Advances in Inorganic Chemistry, Volume 41

Advances in INORGANIC CHEMISTRY Volume 41 ADVISORY BOARD I. Bertini Universita Degli Studi di Firenze Florence, lta...

Author: AG Sykes

35 downloads 2266 Views 26MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Advances in

INORGANIC CHEMISTRY

Volume 41

ADVISORY BOARD I. Bertini Universita Degli Studi di Firenze Florence, ltaly

A. H. Cowley University of Texas Austin, Texas

H. B. Gray California lnstitute of Technology Pasadena, California

M. L. H. Green University of Oxford Oxford, England

D. M. P. Mingos lmperial College of Science, Technology, and Medicine London, England

J. Reedijk Leiden University Leiden, The Netherlands

A. M. Sargeson The Australian National University Canberra, Australia

Y. Sasaki Hokkaido University Sapporo. Japan

0. Kahn

D. F. Shriver

Universite de Paris-Sud Orsay, France

Northwestern University Evanston, lllinois

A. Ludi Universitat Bern Bern, Switzerland

K. Wieghardt Ruhr-Universitat Bochum Bochum, Germany

Advances in

INORGANIC CHEMISTRY EDITED BY A. G. Sykes Department of Chemistry The University Newcastle upon Tyne United Kingdom

VOLUME 41

ACADEMIC PRESS San Diego New York Boston London Sydney Tokyo Toronto

This book is printed on acid-free paper.

@

Copyright 0 1994 by ACADEMIC PRESS, INC. All Rights Reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher.

Academic Press, Inc. A Division of Harcourt Brace & Company

525 B Street, Suite 1900, San Diego, California 92101-4495 United Kingdom Edition published by Academic Press Limited 24-28 Oval Road, London NW 1 7DX

International Standard Serial Number: 0898-8838 lnternational Standard Book Number:

0-12-02364 1-9

PRINTED IN THE UNITED STATES OF AMERICA 94 95 9 6 9 7 98 9 9 E B 9 8 7 6

5

4

3 2

I

CONTENTS The Coordination Chemistry of Technetium

JOHNBALDAS I. Introduction 11. Technetium(-I) 111. Technetium(0)

IV. V. VI. VII. VIII. IX. X. XI.

. . .

Technetium(1) . Technetium(I1) . Technetium(II1) . . Technetium(IV1 . . TechnetiumW . Technetium(V1) . . Technetium(VI1) . . Appendix: Abbreviations References . .

. . . .

. .

.

2 5 5 7 17 27 45 54

. . 80 . 94 . 99 . 101

.

Chemistry of Pentafluorosulfanyl Compounds

R. D. VERMA, ROBERTL. KIRCHMEIER, AND JEAN’NE M. SHREEVE . 126 I. Introduction . . 126 11. Pentafluorosulfanyl Halides . . 130 111. Pentafluorosulfanyl Hypohalites, SFSOX . . . IV. Pentafluorosulfanylalkanes, Alkenes, and Alkynes V. Sulfur Isocyanate Pentafluoride and Sulfur Isothiocyanate Pentafluoride . VI . Sulfur Cyanate Pentafluoride, SF50CN . . VII. Sulfur Cyanide Pentafluoride, SF&N VIII. Sulfur Isocyanide Pentafluoride, SF,NC . . IX. Pentafluorosulfanylamine and Other Derivatives X. Pentafluorosulfanyl N,N-Dichloroamine, SFSNC1z XI. Pentafluorosulfanyl N ,N Difluoramine, SFSNFz XII. Pentafluorosulfanyl Perfluoroalkylamines, SF,N(H)Rf XIII. SF,N(CF,)z . . XIV. SF5N(X)CF3(X=F, C1, Br, I) . . xv. SF,N(Cl)Rf (Rf=C2Fs,n-C3F7,n-C4F9) . XVI. Bis(pentafluorosulfany1)perfluoroalkylamines . XVII. Tris(pentafluorosulfanyl)amine,(SF5)3N . XVIII. Bis(pentafluorosulfanyl)bis(trifluoromethyUhydrazine, SF,(CF,)NN(CFz)SFS . XIX. Tetrakis(pentafluorosulfanyl)hydrazine,(SF&NN(SF& xx. Bis(pentafluorosulfanyl)amine,(SFS),NH . .

.

v

. 132 . 138 . 142

. . .

. .

143 143 144 145 146 146 147 147 149 149 150

. . .

150 150 151

.

. . . . .

vi

CONTENTS

XXI. XXII. XXIII. XXIV.

(SF&NX (X=F, C1) . N-Pentafluorosulfanyl Haloimines, F5SN=CX~ ( X = C1, F) . Pentafluorosulfanyliminodihalosulfanes,SF,N=SX2 ( X = F, C1) Pentafluorosulfanyl-P-Sultonesand Sulfonic Acids . References . .

. 151

. .

.

152 155 . 157 . 161

,

The Hunting of the Gallium Hydrides

ANTHONY J. DOWNSAND COLINR. PULHAM I. Introduction . 11. History and Chemical Background . 111. IV. V. VI. VII.

Conduct of the Hunt: Practical Methods of Attack . Toward Gallane: Preparations for the Hunt . Gallane a t Last! . . Hydrogen-Rich Derivatives of Gallane . Hydrides of the Other Group 13 Metals: Preliminaries and Prospects References .

172 173 177 188 196 211 221 . 228 ,

. . . . .

The Structures of the Group 15 Element(lll) Halides and Halogenoanions

GEORGEA. FISHERAND NICHOLAS c . NORMAN I. 11. 111. IV.

Introduction . Element Trihalides, EX3 Element(II1) Halogenoanions General Comments . References . . Note Added in Proof .

. 233 . 234

.

.

238

. 264 . 268 . 271

lntervalence Charge Transfer and Electro n Exchange Studies of Dinuclear Ruthenium Com plexes

ROBERTJ. CRUTCHLEY I. 11. 111. IV. V.

Introduction . Mixed-Valence Complexes . Electron Exchange . Future Studies . Glossary of Abbreviations and Ligand Structures References .

. . .

273 274 . 304 . 313 , 314 . 319

vii

CONTENTS

Recent Synthetic, Structural, Spectroscopic, and Theoretical Studies on Molecular Phosphorus Oxides and Oxide Sulfides

I. 11. 111. IV. V.

J. CLADE,F. FRICK, AND M. JANSEN Introduction . Molecular Phosphorus Oxides . Molecular Phosphorus Oxide Sulfides . Comparative Considerations . Concluding Remarks . . References .

.

327

. 329

.

364

. 381

. .

383 384

Structure and Reactivity of Transferrins

E. N. BAKER I. 11. 111. IV. V. VI. VII.

Introduction . Biological Roles . Transferrin Structure . Properties of the Metal and Anion Sites Mechanisms of Binding and Release Recombinant DNA Studies . Concluding Remarks . References .

INDEX . CONTENTS OF PREVIOUS VOLUMES .

. 389

.

391

. 393

.

419

. 445 . 452

.

455

. 456

.

. 465 . 477

This Page Intentionally Left Blank

ADVANCES I N INORGANIC CHEMISTRY, VOL.

41

THE COORDINATION CHEMISTRY OF TECHNETIUM JOHN BALDAS Australian Radiation Laboratory. Yallambie, Victoria 3085, Australia

I. Introduction 11. Technetium(-[)

111. Technetium(0) IV. Technetium(1) A. Carbonyl Complexes B. Cyclopentadienyl and Acene Complexes C. Cyano and Isonitrile Complexes D. Dinitrogen, Phosphine, Phosphite, and Related Complexes E. Complexes with Nitrogen Ligands F. Nitrosyl and Thionitrosyl Complexes V. Technetium(I1) A. Organometallic Complexes B. Halide Complexes and Clusters C. Complexes with Nitrogen Ligands D. Phosphine, Arsine, and Related Complexes E. Complexes with Sulfur Ligands F. Nitrosyl and Thionitrosyl Complexes VI. Technetium(II1) A. Carbonyl Complexes B. Cyclopentadienyl Complexes C. Cyano, Isonitrile, and Thiocyanato Complexes D. Aqua, Halide, and Related Dimeric Complexes E. Carboxylato and /3-Diketonato Complexes F. Complexes with Dioximes, Schiff Bases, and Other Nitrogen Ligands G. Complexes with Monodentate Phosphines and Related Ligands H. Complexes with Bidentate Phosphine, Arsine, and Related Ligands I. Complexes with Sulfur Ligands J. Nitrosyl and Thionitrosyl Complexes VII. Technetium(1V) A. Isonitrile and Thiocyanato Complexes B. Halide and Related Complexes C. Complexes with Oxygen Ligands and 0x0-Bridged Complexes D. Complexes with Schiff Base and Other Nitrogen Ligands E. Complexes with Phosphine and Arsine Ligands F. Complexes with Sulfur Ligands 1 Copyright 0 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.

2

JOHN BALDAS

VIII. TechnetiumW A. Mononuclear [TcOI''' Complexes B. Complexes of the trans-[TcO(OH)1" and [TcO,]' Cores C. 0x0-Bridged lTc*0314*and Other Binuclear Complexes D. [TcS13* Complexes E. Nitrido Complexes F. Imido and Hydrazido Complexes G. Complexes Not Containing Multiply Bonded Ligands IX. Technetium(V1) A. 0x0 Complexes B. Nitrido Complexes C. Imido and Hydrazido Complexes D. Dithiolene and Related Complexes X. TechnetiumWII) A. 0x0 and Sulfido Complexes B. Nitrido and Imido Complexes C. Complexes Not Containing Multiply Bonded Ligands XI. Appendix: Abbreviations References

I. Introduction

Technetium, the ekamanganese of Mendeleev and the first of the artificially produced elements, was discovered in 1937 by Perrier and Segre in a molybdenum plate that had been bombarded with deuterons ( I , 2).The name technetium is derived from the Greek word for artificial. Twenty-one isotopes, all radioactive, of mass number 90-110 and several metastable isomers are known (3).Because the half-life of the longest lived isotope, 9 8 T ~is, 4.2 X l o 6 years, primordial technetium has long ceased to exist on earth but minute traces occur in nature (1ng of "Tc in 5.3 kg of pitchblende) as a result of the spontaneous fission of uranium ( 4 ) . The long-lived 99Tc[tuz = 2.11l(12) x lo5 years; p - decay energy = 293.6 keV] (5, 6) is produced in 6% yield from 235Ufission and is isolated in quantity from spent nuclear fuel (7). Technetium-99 is available commercially in gram quantities, usually as ammonium pertechnetate in aqueous solution. This is the only isotope used for macroscopic chemical studies and is here designated simply by the symbol Tc. The ground-state electronic configuration of the Tc atom is [Kr14d55s2with a 6S5/2(2s+1Sj) term symbol (8).Technetium metal dissolves in the oxidizing acids nitric, aqua regia, and concentrated sulfuric and in bromine water. Like rhenium, technetium dissolves in neutral and alkaline solutions of hydrogen peroxide to form the pertechnetate anion. In oxygen the metal burns to form the oxide Tc,O, (7). Apart from radioactivity considerations, the chemistry of

COORDINATION CHEMISTRY OF TECHNETIUM

3

technetium may be investigated by conventional synthetic and spectroscopic methods. Chemically, technetium resembles its third-row congener rhenium, but there are significant differences. In particular, there are the greater ease of reduction of the higher oxidation states of technetium and the greater substitution lability of the lower oxidation states compared with those of the rhenium analogs (9).The organometallic chemistry of technetium, however, rather closely resembles that of rhenium (10).Technetium complexes with the metal in oxidation states from -1 to + 7 are known but, although there is now much research activity in the area, the chemistry of technetium remains relatively undeveloped compared with that of manganese, rhenium, and the neighboring Group 6 and 8 transition metals. The results obtained to date are nonetheless very considerable and show the chemistry of technetium to be among the most varied and interesting of the transition metals. In the last 20 years or so the study of the coordination chemistry of technetium has assumed major practical importance due to the widespread use of the short-lived metastable isomer 9 9 m Tin~ diagnostic compound (radionuclear medicine (11-19). Generally, a 99mTc-labeled pharmaceutical) is injected intravenously into the patient and the in uiuo distribution determined by the use of scintillation techniques, including single photon emission computed tomography (SPECT) (15). The physical properties of 9 9 m Tare ~ near ideal. The gamma ray energy of 140 keV is sufficiently energetic to penetrate deeply seated tissue and is easily externally collimated and detected. The absence of a or /3 emission and the short half-life of 6.01 hr result in a low radiation dose to the patient and activities of up to 1 GBq may be administered. is usually obtained from a Technetium-99m in the form of NaggmTc04 99Mo/99mTc generator based on the decay scheme

Fission-produced 99Mo02- loaded onto an alumina column decays to 99mT~04-, which is conveniently eluted from the column by physiological saline (0.15 M NaCl) while the parent 99Mo0,2-is strongly retained (20,21). The generator eluate contains 9 9 m T~04and a variable quanmainly on the time interval since the previtity of 9 9 T ~ 0 4(depending ous elution) with a total Tc concentration in the range of to M (22,231.This mixture of 99mTc04-/99Tc04is referred to as “no carrier added” and is denoted simply as 99mT~04-. Radiopharmaceuticals are in the presence of a usually prepared by the reduction of 99mT~04-

4

JOHN BALDAS

ligand to give a Y y m complex T~ with the desired physiological behavior. A commonly used reducing agent is stannous tin. The 9 9 m Tradiophar~ maceutical is formed in high yield and radiochemical purity in aqueous solution at near-neutral pH and should be stable in the chemically aggressive in vivo environment at a Tc concentration of the order of 10-”M, which results from dilution by the blood volume (24,251. In a number of cases chromatographic comparisons have shown the structure of 9 9 m Tradiopharmaceutical ~ to be the same as the 99Tccomplex prepared at the macroscopic level but in others the structure and oxidation state are uncertain (19).”“‘Tc radiopharmaceuticals are now available for skeletal, myocardial, renal, hepatobiliary, thyroid, and lung imaging and for a variety of physiological function studies (15).Specific examples are described together with the 99Tcanalogs. The impact of technetium in medical diagnosis may be judged by the 1990 estimate that six t o seven million administrations of 9 9 m Tradio~ pharmaceuticals are performed annually in the United States (17 ) .As a result the study of technetium chemistry has to a degree been driven by the need to understand the chemistry of 9 9 m Tradiopharmaceuticals ~ and to develop new or improved organ-specific agents. Some of this chemistry is now being transferred to rhenium, whose high-energy pemitting lasRe and “‘Re radioisotopes show promise for the development of therapeutic radiopharmaceuticals (9). The aim of this chapter is to provide a fairly comprehensive overview of the status of technetium coordination chemistry up to the latter part of 1993. The term “coordination” is taken to include organometallic compounds. Binary halides are briefly described for the sake of completeness. The material is grouped into oxidation states, with the nitrosyl and thionitrosyl groups being treated as NO+ and NS’, the hydrido ligand as H-, and “noninnocent” ligands such as dithiolenes in the dianionic form. The literature of technetium chemistry consists of two now out-of-date books (26, 2 7 ) and a more recent Russian text ( 2 8 ) together with a comprehensive survey of the literature in two volumes of Gmelin published in 1982 and 1983 (29).Much information is to be found in three conference volumes (30-32) and there are numerous reviews of technetium chemistry (11-13,15,33-36). Specific areas to have been reviewed are crystal structures (37,381, EPR spectroscopy (39-41 1, cluster compounds (42),and analytical chemistry ( 4 3 )and a useful correlation chart of 99Tc NMR chemical shifts and oxidation states of technetium is available (441. Single-crystal X-ray diffraction has been particularly useful. The considerable fraction of technetium complexes to have been characterized by this method may be due to some extent to the difficulty in working with radioactive material but

COORDINATION CHEMISTRY OF TECHNETIUM

5

is no doubt largely due to the recent development of the chemistry and the greater availability of crystallographic structure determination facilities. There is a vast literature developed in the search for potential 9 9 m Tradiopharmaceuticals. ~ In many cases the complexes are poorly, if at all, characterized, although the charge is usually determined by electrophoresis. Such complexes will, in general, be considered here only if there are points of specific chemical interest. “No carrier added” preparations will always be denoted as 99mT~. II. Technetium( - I)

This is the rarest oxidation state for technetium. The IR spectrum of a solution prepared by the addition of N a amalgam to [Tcz(CO)lolin THF showed two v(C0) bands at 1911 and 1865 cm-’, which were assigned to the carbonyl anion [Tc(CO)J by comparison with the spectra of [M(CO),]- (M = Mn, Re). Solutions of Na[Tc(CO),l in THF undergo the expected reactions, including the formation of volatile [HTc(CO),] on treatment with H,PO, (45).The [Tc(CO),]- anion has been used as a nucleophile for the preparation of mixed-metal decacarbonyls (46)by reactions such as

Photolysis of a mixture of [ T C ~ ( C Oand ) ~ ~[Fe(CO)J I in THF is reported to give NEt4[TcFe2(CO)lzl,where Tc(CO)~-replaces Fe(CO), the triangular structure of [Fe3(CO)121 (47). III. Tech netium(0)

Best known, and of great synthetic utility, is the colorless diamagnetic dimer [TC,(CO),~I(m.p., 159-160°C) (48,491, which may be prepared in up to 96% yield by the reaction of NH4Tc04with CO (90 atm initial pressure) in toluene at 200°C with a reaction time of 4 hr (50). The [Mz(CO)lol(M = Mn, Tc, Re) carbonyls are isomorphous (51). The structure of [ T C ~ ( C O (Fig. ) ~ ~ I 1) shows the Tc atoms octahedrally coordinated with a Tc-Tc single bond distance of 3.036(6) A and the equatorial carbonyl groups staggered (approximateDld symmetry) (51). The equatorial carbonyl groups on each Tc are bent away from the axial carbonyl toward the other half of the dimer. The greater n-acceptor character of the axial CO ligands is reflected in C-0 bond distances

6

JOHN BALDAS 05'

L.

c2n. 101'

01 I C

04 '- 04' FIG. 1. The structure of [Tc2(CO),,,](51 1.

0.09 A longer and Tc-C bond distances 0.10 A shorter than those of the equatorial ligands. The vibrational spectra of [M2(CO)lo](M = Mn, Tc, Re) have been extensively investigated and compared (521. For [TC,(CO),~]the equatorial and axial CO stretching force constants of 16.642 and 16.316 mdyn A-', respectively, again demonstrate the greater r-acceptor character of the axial CO ligands. The "Tc NMR spectrum of [Tc,(CO),,] consists of single sharp signal ( A V ~=, ~1.4 Hz) at -2477 ppm relative to Tc0,- (53).The [99"'T~(C0)51'radical is produced in the p- decay of [99Mo(C0)61and reacts with carrier [Mn(CO)J] to form [99"T~(C0)511(541. The heteronuclear carbonyls [M~TC(CO),~] and [TcRe(CO),,l have been prepared by the reaction of a carbonylate anion with a carbonyl halide and characterized by IR and mass spectrometry. The IR spectra of the six possible [M2(CO),,1 (M = Mn, Tc, Re) compounds are closely similar, with the three v(C0) peaks expected in local C,, symmetry (461,a point that emphasizes the general similarity of the structures of Group 7 carbonyls. A mixed cobalt carbonyl [(CO),COTC(CO)~I has also been reported (55)and the CO stretching force and interaction constants have been determined (56).A polymeric [Tc(CO),], , thought to be a trimer, has been claimed but remains inadequately characterized (57). Substitution of the CO ligands in [ T C ~ ( C O by ) ~ the ~ ] strong r-acceptor PF, is achieved either thermally or photolytically. In one study up to eight CO ligands were replaced to give a t least 24 [ T C ~ ( C O ) ~ ~ - ~ ( P F ~ ) , I

COORDINATION CHEMISTRY OF TECHNETIUM

7

isomers, which were assigned on the basis of mass spectra, gas chromatographic retention times, and comparison with the rhenium analogs (58).The monosubstituted ~ ~ ~ - [ T C ~ ( C O )has ~ ( Pbeen F ~ )studied I by 99Tcand 19F NMR (59).Reaction of Tc vapor with PF3 at 77 K gives the volatile [ T c ~ ( P F ~ )~~ ] the basis of IR evidence, the formation (60). On of [ T C ~ ( C O ) ~ ( P Pand ~ , ) ][ T C ~ ( C O ) ~ ( P P has ~ ~been ) ~ I proposed in the reaction of [TC,(CO),~] with PPh, in decalin at 100-150°C (61). Photolysis of [ T C ~ ( C O )in ~ ~the ] presence of butadiene at -20°C gives [TC2(C0)&p-C4H6)],which is isomorphous with the Mn and Re analogs. The trans-butadiene ligand bridges the Tc atoms, which are separated by 3.117(1) A. The Tc-Cbutadiene bond distances (av., 2.389 are markedly longer than Tc-CO (av., 1.945 & (62).A dinitrogen complex originally reported as [Tc(N,)(dppe),I (63) has been shown to be the hydride [HTc’(N2)(dppe),l(641. IV. Technetiurn(1)

A notable feature of this oxidation state is that a considerable number of Tc and 99”’T~ complexes can be prepared in high yields in aqueous media 136). As a consequence the coordination chemistry of Tc(1) has been intensively investigated in the search for 9 9 m Tcationic ~ myocardial imaging agents. Tc(1)complexes have the low-spin d6configuration and are diamagnetic. The 18-electron rule is generally applicable and nicely explains the stability and the prevalence of six-coordinate complexes. A. CARBONYL COMPLEXES Complexes containing cyclopentadienyl and related ligands are considered in Section B.

1. Mononuclear Complexes Complexes containing from one to six carbonyl groups are known and all obey the 18-electron rule. The colorless salt [TC(CO)6]A1C14 is formed by the reaction of [Tc(CO),Cl] with AlCl, under 300 atm CO pressure and is soluble in THF, acetone, and methanol and stable in aqueous solution (65).The carbonyl halides [Tc(CO),X] (X = C1, Br, I) may be prepared by the reaction of the halogen with [TC~(CO),~I. Reaction with chlorine and bromine occurs readily at room temperature but reaction with iodine is extremely slow. The iodide has been prepared by the high-pressure carbonylation of [TC(CO)~II~ (45).An alternative

8

JOHN BALDAS

preparation of the carbonyl halides is by the reaction of K,[TcX,] with CO under pressure a t 230-250°C in the presence of Cu powder (65). The IR spectra of [Tc(CO),Xl (X = C1, Br, I) show the three v(C0) bands (2A, + E ) expected in C,, symmetry in the region 2153-1991 cm-' and a weak 13C0 isotope peak of the intense E mode (45).The ease of halide substitution in [Tc(CO),Xl (usually with the loss of one or more CO groups) makes these compounds key starting materials in technetium carbonyl chemistry (65). Simple substitution of X- occurs in the reaction of CF,COOAg with [Tc(CO),ClI to give [Tc(CO),(OOCCF3)1. The asymmetry introduced by the CF,COO- ligand results in the B , mode becoming IR active and four v(C0) bands are observed (66). Oxidation of [ T C ~ ( C O ) with ~ ~ I NOPF, in MeCN gives [Tc(CO),(CH3CN)]PF6in quantitative yield. This complex is a useful synthetic precursor for the preparation of cationic carbonyl complexes with a variety of ligands (67). The volatile, colorless hydride [HTc(CO),] is produced in only low yield by the reaction of [Tc(CO),]- with &PO, (45). Complexes based on the [Tc(CO),I core are [Tc(CO),(SzCNRz)I(R = Me, Et), the cationic [Tc(CO)4(PPh3)zlA1C14 (651, and [Tc(CO),(acac)l (68). The dithiocarbamato complexes are formed by the reaction of Na(SzCNR2) with [Tc(CO),Cl] in acetone or THF. Grinding of [Tc(CO),ClI with K(P-diketonate) under a layer of CCl, yields the unstable tetracarbonyl P-diketonates (68). A considerable number of complexes containing the [Tc(CO),] core have been prepared and a number of crystal structures have been reported. Cationic complexes are of the type [Tc(CO),L,]X, where L, represents three neutral monodentate ligands, a monodentate and bidentate neutral ligand, or a neutral tridentate ligand. Reaction of [Tc(CO),Brl with AgPF6 in MeCN gives a near quantitative yield of [Tc(C0),(MeCN),1PF6 and [Tc(C0),(MeCN)(PPh3),1PF, and [Tc(CO), (MeCN)(dppe)]PF, may be prepared by ligand exchange (69).Of particular interest in relation to potential 99mTcradiopharmaceuticals is the air-stable, water-soluble [Tc(CO)~(L~)]PF~ (L3 = tan; 1,4,7-trimethyltan; 1,4,7-trithiacyclononane) (67). Only one monoanionic ligand seems to be supported to give neutral complexes of the type [TcX(CO),L,I, where some examples are X = C1, Br, I, 02CR; L = PR, , AsR,, SbR,, P(OR),, py, MeCN, CNR, EtzNH; or Lz = bpy, phen, dppe, en (65,66, 68, 70-72); [Tc(CO),{HB(pz),}] (73). A novel preparative method with CO at atmospheric pressure yields 1(74). Fac and mer isomers may be distinguished by the IR spectrum; two v(C0) bands (A, + E in local CBUsymmetry for the CO groups) are expected for a fuc isomer and three (2A, + B , in local Cz, symmetry), for a mer isomer (67, 72). The "Tc NMR spectra of neutral complexes

9

COORDINATION CHEMISTRY OF TECHNETIUM

:1;

CI

OC-Tc-CO

co 110°C N B u ~ ~ c O C I ~*]

PhMe/MeCN

I’

OG

PPh, (1)

95% yield

show chemical shifts of -940 to - 1820 ppm and those of cationic complexes, -2070 to -3520 ppm against TcO,- (70, 75). The crystal structure of 1 shows almost undistorted octahedral geometry with a P-Tc-P is distorted angle of 174.59(2)”(74),whereas that offac-[T~Br(CO)~(en)] with the Tc-C and C-0 bond distances the same for all three CO groups (76). A “piano stool” structure with CBVsymmetry is found for [Tc(CO),L] [L = HB(pz),, HB(3,5-Me2pz),l, which is isostructural with the Mn and Re analogs (73). An unusual complex is [TcBr(CO),(Ph-Pglup)], prepared by the reaction of [Tc(CO),Brl with a neutral chiral phosphinoglucose derivative (77). Complexes containing the [Tc(CO),] core may be prepared by substitution or carbonylation reactions. The thiolato complexes [Tc(CO),(PPh,),L] [L = SzCNEtz,S2COEt, SzP(OMe),l are formed on heating trans-[T~(CO),Cl(PPh,)~] with the ligand in acetone or THF (78). The } ~ formed ] C ~ O ~by the cis- and trans-isomers of [ T c ( C O ) ~ { P ( O E ~ ) ~ P ~are reaction of [TcClz{P(OEt)zPh}41C104 with CO (1atm) at 50°C (79).The cis-isomer is a distorted octahedron with the two Tc-CO bond distances both 1.90(2) A. mer-[TcX3(PMezPh),] (X = C1, Br) reacts with CO (1atm) in refluxing MeO(CH,),OMe containing added phosphine to ). A variety of [Tc(CO),(PPh3),Ll give only cis-[T~X(CO),(PMe~Ph)~3(72 complexes, where L is a carboxylato, mixed amido, or thiazolato ligand (80),and [Tc(C~)~L{P(OR)~}~]PF, (L = bpy, 4,4-Mezbpy) (67) have been prepared. Crystal structures of the Schiff base complex [Tc(CO),(PPh,), {(C,H,NS)N =CHC,H,O-o}] (81 and the pseudoallyl complexes [ T c ( C O ) , ( P M ~ , P ~ ) ~ ( ~ - M ~ C , H , N ~ N ~ N C , H[TC(CO),(PM~,P~)~ ~M~-~)], ( P h N q ( M e F N P h ) l (821, and [Tc(CO),(PPh,)2{SC(NHPh)S}l(83) show distorted octahedral geometry with the two CO ligands mutually cis and the PPh, ligands trans. Structurally characterized complexes with a tridentate ligand are ci~-[Tc(CO)~(PPh,)(tan)]Cland cis-[Tc(CO),(PPh,){HB(pz),}] (74). Electrochemical oxidation of [ T c ( C O ) ~ C ~ ( P M ~ results ~ P ~ ) , ]in the formation of [Tc~~’(CO)C~(M~CN),

10

JOHN BALDAS

(PMe,Ph),](ClO4), (841, an example of the oxidation of one 18-electron species to another. Reaction of [HTc(N,)(dppe),] with CO in benzene or with methanol in the presence of pyridine gives [HTc(CO)(dppe),I.In the latter reaction methanol serves as the source of CO. On reflux in MeCN, [HTc(CO)(dppe),] is converted to [Tc(CO)(MeCN)(dppe),1PF6(85). 2 . Dimeric and Polynuclear Complexes

The dimers [Tc(CO),X], (X = C1, Br, I) are formed by the reaction of the halogen with [ T C ~ ( C O (45). ) ~ ~ IThe ease of thermal decarbonylation of [Tc(C0I5X] in a n inert solvent or during vacuum sublimation increases in the order I < Br < C1 and decarbonylation proceeds in the sequence [Tc(CO),X] + [Tc(CO),Xl, + [Tc(CO),Xl, (86).Decarbonylation occurs more easily than that for the Mn or Re analogs. The presence of four u ( C 0 )bands and the TcCO bending region in the IR spectra is consistent with the D,, halide-bridged structure ( 2 )for the dimers ( 8 7 ) and has been confirmed crystallographically by the isostructural nature of [M(CO),BrI, (M = Tc, Re) (88).

oc .co

co X

'

3'

I ,.-'cp

'X (2)

(3)

Structure 3 (X = Br), consisting of a cube with F,-B~bridges, was assigned to the tetrameric [Tc(CO),Br], on the basis of X-ray diffraction data (88).This is confirmed by the single-crystal structure determination of 3 (X = Cl), which shows that the tetramer has crystallographic Tdsymmetry with bond distances Tc-C, 1.903(3) A; C-0, 1.128(4) A; and Tc-C1, 2.559(1) A. The Tc-Tc distance of 3.840(1) A shows the absence of a direct Tc-Tc interaction (76).The reaction of [Tc(CO),C1I4 with chlorine is reported to give the trimer [(OC),Tcl(p-C1),Tcl" (p-Cl),T~'(C0)~1 (89). The reaction of thiols, sulfides, diarsines, and Hacac with [Tc(CO),Xl gives the dimers [TC(CO)~(SP~)I, (651, [Tc(C0),C1(EPh2)I2(E = P, S, Se, As) (90,91), [Tc(CO),BrLI, (L = THF,

11

COORDINATION CHEMISTRY OF TECHNETIUM

MeCN; for which the IR spectra are consistent with a centrosymmetric structure) (711, and [Tc(CO),(acac)l, (92).Extensive mass spectral data have been reported (93). Partial carbonylation of NaTcO, in methanol gives the unprece(4) dented cubane-type structure N ~ [ T C ~ ( C O ) ~ ( O M ~ )(Fig. , I 21, with each Tc atom obeying the 18-electron rule. The Na+ cation in 4 forms one corner of the cube with Na-OMe distances of ca. 2.38 and Na-OC interactions (ca. 2.51 A) with adjoining cubes completing the coordination octahedron. In solution, 4 exists as the cubane cluster and not the Na' salt. The [TC,(CO)~(OM~),]group may thus be likened to a n anionic crown ether with a high affinity for Na' (94). The reaction of KTc0, with HCOOH gives [Tc(CO),OH],, which is most likely the cubic tetramer (3) (X = OH) (95). The cubic structure of 3 has been established crystallographically for [M(CO),(p.,-OH)], (M = Mn, Re) (96). Reaction of [TC,(CO),~]with rneso-tetraphenylporphine(H,tpp) or mesoporphyrin IX dimethyl ester (H,mp) gives the unusual dimers [L{Tc(CO),},] (L = mp, tpp) (97, 98). These dimers are also formed by the thermal disproportionation of [(HL)Tc(CO),].The crystal structure of [tpp{Tc(CO),},] shows the two Tc(CO), moieties arranged in a tripod 020

>-

~. 021

i

'p

022

k'

c21

012

'--\

c12

032

010 i

011'

L-c2

030

FIG. 2. The structure of N ~ [ T C ~ ( C O ) ~ ( O M (4)~ )(94). .J

12

JOHN BALDAS

configuration, with one on each side of the porphine ring and each outof-plane Tc atom coordinated to three N atoms. The Tc.-Tc distance of 3.101 is somewhat long to constitute bonding, but is short enough to indicate some metal-metal interaction (97).The dark-red air-stable heteronuclear [mp{(OC),TcRe(CO),}] is formed on heating [Hmp{Re(CO),}] with [Tc,(CO),,l in decalin (98).

B. CYCLOPENTADIENYL AND ARENECOMPLEXES q5-Cyclopentadienyl complexes of the type [Cp’Tc(CO),l may be prepared by the reaction of TcCl,/CO/Cu or [Tc(CO),Xl with NaCp’ or LiCp’ (99,100).Crystal structures of [LTc(CO),I (L = C5Me5,C,Me,Et, indenyl) (101) and [{Me3N(CH,)3C,Me,}Tc(CO)311(100)show the piano stool arrangement ( 5 ) . [CpTc(CO),I undergoes acylation on reaction with PhCOCl to give the PhCOCp derivative (102).

Cp’2Fe+ [Mn(CO),I]

OC‘

-

+ 99mT~04-

oc

co

99m

co (6)

(5)

The 9 9 m Tcomplexes ~ (6) (R = N-methylpiperidine, quinuclidine) may be prepared in 30-90% radiochemical yield by the route shown on heating for 1h r in THF a t 150°C. These esters show high brain uptake in animals (103). Irradiation of [(C,M~,)TC(CO)~] in cyclohexane produces the $-C5Me, carbonyl-bridged dimers 7 and 8. The structure of 7 was established crystallographically and that of 8 was confirmed by spectroscopic comparison with the structurally characterized Re analog (104 1.

Q

COORDINATION CHEMISTRY OF TECHNETIUM

13

The short Tc-Tc bond distance of 2.413(3) A in 7 corresponds to a triple bond and for 8 a Tc-Tc single bond has been proposed. These bond orders are those needed to satisfy the 18-electron rule. As expected, the v(C0) IR absorptions in 7 occur at 1821-1771 cm-', whereas for 8, which contains terminal and bridging CO groups, the range is 2012-1738 cm-'. The [Tc(arene),]PF, (arene = benzene, substituted benzene, aromatic hydrocarbon) complexes are formed by reaction of the arene with TcCl,/AlCl,/Al (105,106).The cations are stable to the air and to acids and bases. A large number of [99"Tc(arene)2]+complexes have been prepared and the structure has been demonstrated by HPLC comparisons with the 99Tccomplexes. The lipophilic 9 9 m Tcomplexes ~ of benzene substituted with four to six carbon atoms show promising myocardial uptake (107). C. CYANO AND ISONITRILE COMPLEXES Olive-green K,[TC(CN)~] has been prepared by the reduction of Tc0,with KICN- and shown to be isostructural with K,[M(CN)6] (M = Mn, Re) (108).The low CN force constant of 14.57 mdyn A-' indicates that cyanide is acting as a relatively strong n-acceptor (109).Following the discovery that the [99"Tc(CNtBu)6]+ cation is concentrated in the human myocardium, this class of complexes has been intensively investigated in the search for improved imaging agents (19, 110). The air- and water-stable [TC(CNR)~]X salts may be prepared by the reaction of [Tc"'(tu),]Cl, with RNC but a more convenient method is the reduction of TcO,- by Na2S204in aqueous ethanol in the presence of the ligand (111,112).The energy of the v(CN) IR absorption is 50-80 cm-' lower than that in the free ligand, consistent with extensive .rr-donation from Tc(1). Reversible one-electron oxidation occurs at 0.82-0.88 V vs SCE for alkyl derivatives, with the phenyl derivative more difficult to oxidize at 1.18 V vs SCE (112).The 99TcNMR spectra show a single signal at about - 1900 ppm relative to TcO,-, with small but significant chemical shift differences due to the substituents (44,113). The crystal structure of [TC(CN%U)~]PF, establishes that the geometry is octahedral with Tc-C bond distances of 2.029(5) A and that the complex is isomorphous with the Re analog (114).Systematic variation of the R group has led to the development of [99"T~(CNR)6]+, where CNR is (2-methoxy-2methylpropyl)isonitrile, as a radiopharmaceutical for myocardial imaging (19).I n uiuo, the methoxy groups are sequentially metabolized to hydroxy groups to give seven products of increasing hydrophilicity and the resulting desired faster blood and lung clearance in comparison

14

JOHN BALDAS

with [99mTc(CN'Bu)6]'(115).At high pH [Tc(CNCMe,COOMe),ICl undergoes random base-catalyzed ester hydrolysis of the coordinated ligands. The nine possible carboxylic acid products have been isolated and identified by HPLC, FABMS, IR, and "Tc NMR (116). Mixedligand complexes of the type [TC(CNR),(CNR')~_,I+ ( n = 0-6) and [Tc(CN'Bu), (PPh,),-,]PF6 ( n = 4, 5) have been prepared by synthesis with a mixture of ligands (117,1181, and truns-[Tc(dppe),(CN'Bu),]PF, (119)and [HTc(CNR)(dppe),l (85) have been prepared by substitution of [HTc(N,)(dppe),]. Photolysis of [TC(CNR)~]PF, in the presence of bpy, phen, or mixed ligand synthesis from Tc0,- gives a series of complexes of the type [Tc(CNR),L]PF,. The crystal structure of [Tc(CN'Bu),(bpy)]PF, shows that one of the isonitrile ligands is considerably bent, with a C=N-C angle of 148",suggesting a "pseudo" internal oxidation of Tc(1) to Tc(II1) (120). Oxidative addition of chlorine or bromine to [TC(CN'BU)~]PF, produces the seven-coordinate [Tc"'(CNLBu),X](PF6), in 75% yield (121 ).

D. DINITROGEN, PHOSPHINE, PHOSPHITE, AND RELATED COMPLEXES Crystallography and 'H NMR have confirmed the formula [HTc(N,)(dppe),] for the product of the reduction of [TcCl,(PPh,),] by Na amalgam under nitrogen in the presence of dppe. The Tc atom is octahedrally coordinated with the hydrido ligand trans to dinitrogen. The Tc-N and N-N distances are 2.05(1)and 0.98(1)8,, respectively, and the Tc-N-N angle is 178(1)" (64). The ease of substitution of dinitrogen and hydride makes this compound a versatile starting material for the preparation of Tc(1) mixed-ligand complexes (85). On UV irradiation [TC(CO),(HB(~,~-M~~~~)~}~ reacts with nitrogen to give the air-stable dinitrogen-bridged dimer [{(HB(3,5-Mezpz)3)Tc(CO)z}~(~-Nz)l. The N-N bond distance of dinitrogen is 1.160(3)A and the Tc-N-N angle is close to linear at 174". In the electronic spectrum a band at 21,552 cm-' ( E = 3175) has been assigned to a Tc + N2(.rr*)MLCT transition (122). Excess ligand serves as the reductant in the preparation of [Tc(dmpe)&F3S03 from Tc04- and dmpe. EXAFS analysis of the fluoride salt has established octahedral geometry with a Tc-P bond distance of 2.40 8, (123). The [TcL,]' (L = dmpe, depe) complexes undergo reversible electrochemical oxidation to [ T C ~ ~ L ~a Ireaction ~', that may be chemically produced by H,02(124).The Tc(II)/Tc(I)couple for depe as ligand is 164 mV more negative than that for dmpe, indicating that [Tc(depe),]' is considerably more easily oxidized than [Tc(dmpe),l+ . Thus, both oxidation states are air-stable for dmpe as ligand, whereas

COORDINATION CHEMISTRY OF TECHNETIUM

15

[Tc(depe)J2+is air-stable but [Tc(depe),I+ must be prepared under airfree conditions. Pulse radiolysis studies show that the oxidation of I ~the + strong oxidant C1,- proceeds at, [Tc(dmpe),] to [ T ~ ( d r n p e ) ~by or near, the diffusion-controlled limit (k = 1 x lo9 M-' sec-') by an outer-sphere mechanism (125). The self-exchange rate of the [T~(dmpe),]+'~+ couple has been calculated by application of the Marcus theory to be 2 x lo6 it-' sec-' (126, 127). The diamagnetic mixed phosphine-phosphite complex [Tc(dppe)(tmp),]PF, has been prepared by substitution of [HTc(N,)(dppe)J and characterized by FABMS and 'H and 99TcNMR (128).A number of homoleptic phosphite, phosphonite, and phosphinite cationic complexes of the type [TcL6]X [L = tmp, PR(OMe)2,PEt,(OMe)] have been prepared, either from TcO,- or by reductive substitution of [Tc"'(tu),]C1,, and characterized by '9'12 and 31PNMR, FABMS, or X-ray photoelectron spectroscopy (129-132).The [99mTc(dmpe)31+ and [99mTc(tmp),]+cations proved disappointing as potential myocardial imaging agents in humans due to slow blood clearance, although the clearance in dogs was fast. This species difference is due to the strong binding of the cations to a plasma component present in human but not in dog blood (24). +

E. COMPLEXES WITH NITROGEN LIGANDS Electrochemical reduction of a mixture of Tc0,- and phen allowed the isolation of the purple crystalline [Tc(phen),]PF6. Conductivity measurements in MeCN confirmed a 1: 1 electrolyte and cerimetric titration confirmed the + 1 oxidation state (133).

F. NITROSYL AND THIONITROSYL COMPLEXES By the reduction of (NH,),[TcCl,] with NH,OH-HCl and addition of ammonia Eakins et al. obtained pink crystals, which they formulated as a hydroxylamine complex (1341,but which were later shown to be the diamagnetic nitrosyl complex trans-[T~(N0)(NH~)~(OH~)lCl~ (135). The ammine ligands are remarkably stable to substitution in acid solution and the nitrosyl group is stable to nucleophilic attack. The crystal structure reveals bond distances of 2.168(4) A for Tc-OH, and 2.164(5) A (av.) for Tc-NH, (136).The short Tc-NO distance of 1.716(4) A and the relatively long N-0 distance of 1.203(6) A together with the low v(N0) IR absorption at 1680 cm-' and the surprisingly acidic water (pK, = 7.3) indicate very strong back-donation from Tc to NO (135).The NO group has been estimated to carry a half-negative charge, which assigns an oxidation state of +2.5 to Tc rather than the + 1 based on the NO+ formalism (136).Oxidation gives the green trans-

16

JOHN BALDAS

[Tc~~(NO)(NH,),(OH,)]C~, with u(N0) at 1830 cm-' and a highly acidic trans water (pK, = 2.0) (135).The novel hydride [Tc(NO)(PP~,)~(H),I is formed by borohydride reduction of a mixture of [ T c ~ ~ ( N O ) ( P P ~ ~ ) ~ C I , I and PPh, (137).The TcH IR absorptions occur at 1733 and 1185 cm-', and v(N0) at 1636 cm-' is shifted to 1659 cm-' in the dideuterio complex, indicating a strong coupling of the nitrosyl and hydrido ligands. Reaction of [Tc(CNtBu),]N03with NOPF, or HN03/HOAcgives a high yield of [Tc(NO)(CN~BU),](PF,)~. The high value of 1865 cm-' for u ( N 0 ) is consistent with NO+ coordination (138). Similarly, [(CSMe5)Tc(CO),]undergoes substitution of CO by the isoelectronic NO+ to produce a good yield of [(C5MeS)Tc(NO)(C0)21PF,,for which v(N0) occurs a t the lower value of 1745 cm-' as a result of the increased back-bonding induced by the anionic Cp' ligand (100). NBu,[Tc"(NOIBr,] is reduced by CN'Bu to the neutral truns-[Tc'(NO)Br2(CNtBu),] [u(NO) a t 1755 cm-'I, for which the stereochemistry has been crystallographically established (138). Electrochemical or hydrazine reduction of the deep-purple (NBu,),[Tc"(NO)(NCS),] results in rustcolored crystals of (NBu,),[Tc(NO)(NCS),l (139).The crystal structure of [Tc(NO)Cl(dppe),]Cl.H,O has been briefly described (140).Reaction of AsPh4[TcVOC1,]with an excess of cyclooctane-1,2-dioxime (codoH,) yields brown crystals, shown by crystallography to be AsPh,[Tc(NO)Cl(codoH),].HC1(9) (141).The nitrosyl group appears to be derived from NH,OH formed by partial hydrolysis of the dioxime. The intense v(N0) IR absorption occurs a t 1701 cm-'.

S PMe,Ph N I ,.' , \ '

0.

'''.H.''

...o

(10)

Although nitrosyl complexes have been long known, the first thionitrosyl complex was reported only in 1974 (142).The Tc-N group shows a marked tendency to abstract sulfur to form Tc(NS) complexes in the +1, +2, and + 3 oxidation states. The Tc'(NS) complex (10) is prepared by the reaction of [TcVNC1,(PMe,Ph),l with 1 eq. of S2C1, (143) and

COORDINATION CHEMISTRY OF TECHNETIUM

17

mer-[Tc(NS)Clz(pic),]by a remarkable reaction in which [TcVINC1,labstracts sulfur from the Sz042-anion (144 1. The 4NS) IR absorptions occur at 1177 [shifted to 1147 cm-' on I5N labeling (14511 and 1173 cm-' for 10 and the picoline complex, respectively. The nearly linear Tc-N-S angles of 177" and 176" are consistent with coordination by NS'. Structure 10 shows a small but distinct shortening of the Tc-C1 bond trans to NS relative to the cis bond, whereas for rner-[Tc(NS)Clz(pic),] the reverse is observed with NSTc-Cl,,,, 2.430(2) A, and trans, 2.443(1) 8, (143, 144).

V. Technetium( II)

Technetium(I1) complexes are paramagnetic with the d5 low-spin configuration. A characteristic feature is the considerable number of mixed-valence halide clusters containing Tc in oxidation states of + 1.5 to +3. This area has been reviewed (42).For convenience, all complexes, except those of [TcZl6+,are treated together here. EPR spectroscopy is particularly useful in both the detection of species in this oxidation state and the study of exchange reactions in solution. The nuclear spin of "Tc (I = $1 results in spectra of 10 lines with superimposed hyperfine splitting. The d5 low-spin system is treated as a d' system in the hole formalism (40). A. ORGANOMETALLIC COMPLEXES Although carbonyl and cyclopentadienyl complexes are well known for Tc(1) and TdIII), none appear to have been reported for Tc(I1). This may be ascribed to the tendency to follow the 18-electron rule, which, due to the odd number of electrons, would require dimer formation for compliance. Similarly, no Tc(I1) cyano or isonitrile complexes appear to have been isolated.

B. HALIDECOMPLEXES AND CLUSTERS 1. Mononuclear and Binuclear Complexes

The only monomeric complex is the tetrahedral [TcBr,I2-, identified crystallographically in the product (11) of the remarkable reaction NBu4[TcV1NBr41 b y

[TcVNBr(bpy)212[TcuBr41.

(11)

18

JOHN BALDAS

The mechanism of formation of 11 is unknown. The Br-Tc-Br bond angles in ITcBr,I2- are approximately tetrahedral, in the range 106.1-112.1", and the Tc-Br bond distances of 2.388-2.417 A are very short (146). From the reduction of KTcO,/HCl by hydrogen (30 atm at 140"C), crystals of the d5-d5 cluster K2[Tc"zC1,].2Hz0 were separated (147, 1481. Structural analysis shows a polymer of Tc2C12-units with strongly distorted D4dsymmetry, linked by bridging C1 in infinite zigzag chains, with the very short Tc-Tc bond distance of 2.044(1) A. On the basis of the short Tc-Tc distance a "quintuple" bond was suggested (148).The crystal structure has been reexamined at 15 and -53°C and Tc-Tc bond distances of 2.047(1) and 2.042(2) 8, found (149). These very short distances are not anomalous for a triple bond with the ~ ~ ~ 7 relectronic ~ 6 ~ t configuration i ~ ~ because, even though the 6 bond order is 0, the low oxidation state of the [Tc214+core strongly enhances the c and 7~ bonding (149).The bromo complex K,[Tc,Br6].2H,0 has also been prepared (147).Both complexes are diamagnetic (150). The turquoise-blue salts (NH,)3[Tc2C1,1~2H,0and Y [TczC1,1.9H,0 of the mixed-valence d4-d5 [Tc,C1,13- anion, with an average oxidation state of +2.5, were first prepared in 1963 by the reduction of [TcCl6l2with Zn/HCl (1341. Improved synthetic methods have been developed and a variety of [Tc2C1,I3- salts, with inorganic or organic cations, is now known (42, 1511. Crystal structures are available for (NH,),[Tc,C1,].2Hz0 (152)and the K+ (153,1541, Y3+(155),and pyH+ (156)salts. Also, the metal cation may be partially replaced by H30+,as in the structurally characterized K',K"3-,(H30), [TczC1,13-nH,0 (157). The [Tc,C~,]~-anion possesses virtual D4,,symmetry with the square-pyramidal end groups in the eclipsed conformation (Fig. 3). The short Tc-Tc distances of 2.117(2) for the K' (154)and 2.1185(5) A for the pyH+ salt (157)and the observed paramagnetism [peff= 1.78(3) BM for the NH,' and Y3+ salts (15811 are consistent with a strong metal-metal bond electronic configuration (1591,a conclusion order of 3.5 and a cz7r4i326*1 supported by the EPR spectra (158) and self-consistent field Xa scat~ ~the first species in tered-wave calculations (160). In fact, [ T c , C ~ , ]is which a bond order of 3.5 was recognized (159).In the electronic spectrum of K3[Tc,C1,], the main component of the 15,700-cm-' (638-nm) band has been assigned to the 6" + 7 ~ transition, * and the band originating at 5900 cm-' in the near IR, to the 6 -+ 6" transition (161). The [ T c , C ~ , ] ~ - / [ T C ~ Ccouple ~ ~ ] ~ - is electrochemically quasireversible in HCl/EtOH (158).In HC1 solution [TczC1813-undergoes hydrolysis, disproportionation, and oxidation by oxygen, with rupture of the Tc-Tc bond (162 1. At 280"C, anhydrous (NH,)3[Tc,C1,1 starts to disproportion-

COORDINATION CHEMISTRY OF TECHNETIUM

19

CL

L3

1 bCL2,

CL33*

FIG.3. The structure of the [ T C ~ C I ~anion ] ~ - in K3[Tc2Cl~l.nH20(154).

ate to (NH4)2[TcC1,]and Tc metal (163).The [Tc2Br8I3-anion is rather less stable than the chloro analog but the gold-colored ( N B U , ) ~ [ T C ~ C ~ ~ ] may be obtained in 70% yield by the reduction of (NBu,),[Tc2Cl8] with BH4- in CH,Cl,. The Cs3[Tc,C18] salt a t 6 K shows a well-resolved vibronic structure of the 6 + 6* transition, with electronic origin a t about 5970 cm-’ (164). Reaction of K,[Tc,X8]~2H20 (X = C1, Br) with glacial acetic acid results in substitution to yield green crystals of [TC~(OAC)~X] (42,1651, and reaction of (NH,)3[Tc2C18]with molten 2-hydroxypyridine (Hhp) gives the dark-green [Tc,(hp),Cl] (166). Crystal structures of these three complexes reveal the familiar “lantern” arrangement, with the bidentate ligands bridging the two Tc atoms of each cluster and axial halide bridging the clusters. Infinite linear chains occur in [Tc,(hp),ClI [Tc-Tc, 2.095(1) A; Tc-C1, 2.679(1) A] and [Tc,(OAC)~B~I [Tc-Tc, 2.112(1) A; Tc-Br, 2.843(1) A] ( 1 6 7 ~complexes ) and zigzag chains with a Tc-C1-Tc angle of 120”for [Tc2(OAc),C11[Tc-Tc, 2.117(1) A1 (I67b).

20

JOHN BALDAS

Also isolated from the acetic acid reaction is K[Tc2(OAc),C1,1, the structure of which shows a distinctly longer Tc-Tc bond distance of 2.1260(5) A, with two axial chlorides at 2.589(1) ( 1 6 7 ~ The ) . effective magnetic moment for the three acetate dimers is 1.78 -+ 0.05 BM and the EPR spectra are consistent with the unpaired electron equally shared by the two Tc centers in the 6" ( b l u )antibonding molecular orbital (168,169). 2. Polynuclear Clusters

The development of this area has been entirely due to the work of Russian chemists (42). Reduction of HTcO, in concentrated HX (X = C1, Br, I) by hydrogen under pressure yields a mixture of products with average oxidation states of +1.5-2.0 for Tc and with varying H20 and H30' contents (42, 170, 171). Crystal structure determinations have identified three basic structural arrangements of the Tc atoms (157, 172,173),the trigonal prismatic ([Tc,X,(p-X),]X,)"- (X = C1, Br; n = 2,3), (NMe,),[Tc,Cl,(p-C1),] (172, 174, 175),the octahedral [Tc,Br6(p3-

0 Br

FIG.4. The structures of technetium bromo clusters. (a)Fragment of the structure of the trigonal prismatic clusters (NEt4)2[Tc6Br6(~-Br)61Brz and (NMe,),[Tc,Br,(p-Br),lBrz (175). (b) Fragment of the structure of the octahedral clusters ~ H 3 0 ~ H 2 0 ~ 3 1 2 ~ T c ~ B r 6 ~ ~ g - B r )equivalent ~l. pg-Br atoms are shown but the posiBr)J and ( N B ~ ~ ) ~ [ T c ~ B r s ( p ~Eight tions are not fully occupied (I72).(c)Fragment of the structure ofthe tetragonal prismatic O ~ zOl ~, T c ~ B r , ~ ~ and - B r ~[H(Hz0)212[TcaBr4(~~lBr, clusters [ T C ~ B ~ ~ ( ~ - B ~ ) ~[ ]HB~PH~z H Br)a]Br2(I 72 1.

COORDINATION CHEMISTRY OF TECHNETIUM

b

21

Y

C

Br1,I2- (1 72, 1761, and the tetragonal prismatic {[Tc,X4(p-X)83X,}l"(X = Br, I; n = 1, x = 1 , O ; n = 2, x = 2; DZhsymmetry) (1 72,177-178). Structural examples are shown in Fig. 4 (172,175). The main structural features of the novel trigonal and tetragonal prismatic clusters are the presence of dimeric Tc-Tc units with strong, localized multiple bonds of the order of 3.0-4.0 and Tc-Tc bond distances in the range 2.16-2.19 A, forming the vertical edges and a system of formally single Tc-Tc bonds (2.51-2.70 A) delocalized along the metal skeleton. In the octahedral (NBU,)~[TC,B~,( P ~ - B ~cluster, ) ~ ] the Tc-Tc bond distances are

22

JOHN BALDAS

2.578(1)-2.609(1) 8, (172). The trigonal prismatic cluster chlorides readily undergo substitution by bromide in HBr at -140°C, with the preservation of the Tc framework. The reaction goes to completion and no mixed-ligand clusters can be isolated. In acetone, [H(H,O),l,[Tc,Br,(p-Br)JBr, undergoes partial substitution with HI at room temperature. Addition of NBu,' gives brown crystals shown by a n X-ray structure determination to be the mixed-ligand tetragonal prismatic cluster (NBu4),[Tc,(Br,,51,,5)4(~-Br~,5p-Io.5~~]Iz (179). X-ray photoelectron spectra (180),magnetic properties, and EPR studies of the Tc clusters have been reported and the mechanism of Tc-Tc bond formation has been discussed (1811. A molecular orbital analysis of the trigonal prismatic [ T C ~ C ~ , ,shows ] ~ - electron-rich c ~ ~ 7 r ~ triple 6 ~ 6 bonding *~ in each dimer unit, single bonding in the triangles, and two electrons in a net antibonding a," orbital (7r* with respect to the dimers) so that the number of framework bonding electrons is 30. This is very different from the magic numbers of 16,24, or 14 known for octahedra and 18 for trigonal prisms (182).Recently, crystal structures of six ternary chalcogenides of the type M4[TcGX,,]( n = 12 or 13; X = S or Se; M = K, Rb, or Cs) have revealed the presence of octahedral Tc clusters with Tc-Tc bond distances of 2.60-2.65 8, (183). c . COMPLEXES WITH NITROGEN LIGANDS The reaction of [ T C " * C ~ , ( C H ~ C N ) ( P P with ~ ~ )bpy, ~ ] phen, and terpy yields the blue-black Tc(I1) complexes [TcLJP+ (L = bpy, phen) and [Tc(terpy),]*+,which may be isolated as the BPh4- or PF,- salts. The crystal structure of [Tc(bpy),](PF,), shows exact D3 symmetry for the cation with all Tc-N bonds distances 2.077(10)A. The cyclic voltammogram of [Tc(bpy),](BPh,), indicates three diffusion-controlled reversible one-electron reduction processes at El,, -0.34, -1.36, and -1.70 V vs SCE, corresponding to successive reduction of Tc" + Tc' +. Tco +. Tc-' (184, 185). An EPR study of [Tc(bpy),](PF,), has shown that the unpaired electron occupies the dxyorbital and that extensive metal-ligand covalent interactions reduce the spin-orbit coupling to about 49% (186).The effective magnetic moment of [Tc(phen),I2+is 1.89 BM, in good agreement with the spin-only value for a Tc(I1) octahedral d5 configuration(1 33 1. A variety of mixed-ligand complexes, including [ T C C ~ , ( P M ~ ~ P(L~=) ~bpy, L I phen) and [TcCl(PMe,Ph),(terpy)]PF,, has been prepared and electrochemically investigated. Crystal structures have established the coordination arrangement for the distorted octahedral (12) and truns(P )-[TcBr(PMe2Ph),(terpy)lS03CF, (187).

23

COORDINATION CHEMISTRY OF TECHNETIUM

PMe-Ph

n

N

N = phen

(13)

(12)

Crystallography, FABMS, and EPR spectroscopy have shown that the deep-purple major product isolated from the reaction of [Tc"'Cl,(PPh,),(MeCN)l with tris(2-aminoethy1)amine and 2-pyridinecarboxaldehyde in methanol is [Tc"(tren-py,)l(PF,>, [tren-py, (1311 (188).The mean Tc-imine and Tc-pyridine nitrogen bond distances are 2.071 and 2.109 A, respectively, and the Tc-+ertiary amine nitrogen distance is 2.933(7) A. The coordination geometry has been described as pseudoseven coordinate capped octahedral. It is arguable whether the long Tc-.Ntert distance constitutes coordination, but a Tc-.N interaction is indicated because the lone pair on this nitrogen is directed toward the Tc atom.

D. PHOSPHINE, ARSINE,AND RELATED COMPLEXES In view of the large number of mixed halide-phosphine complexes of the binuclear [Re,14' and [Re215+cores (1891,it is perhaps surprising that no such complex has been reported for technetium, but this may simply reflect the lack of work in this area. Mononuclear phosphine and arsine complexes are, however, common. The first Tc(I1) complex to be reported was trun~-[Tc(diars)~I~l in 1959 (190),followed by the chloro and bromo analogs in the next year (191). These complexes together with tr~ns-[Tc(dppe)~X~I (X = C1, Br) were prepared by SO2or BH4- reduction of the [Tc1''L2X21Xsalts and characterized by electronic spectroscopy, magnetic measurements (pefl= 2.05, 2.28 BM), and by being shown to be isostructural with the Re analogs (192).These conclusions have been confirmed by a crystal structure determination of truns[Tc(dppe),Cl,I (193).The rather long Tc-C1 bonds (av., 2.424 A) undergo a dramatic shortening of 0.105(2) A on oxidation to truns[Tc"'(dppe),Cl,I +,behavior consistent with a ligand that binds primarily by cr-donation. In contrast, the Tc-P bond lengthens by 0.072(2) A on oxidation from Tc(I1) to Tc(II1) due to r-back-bonding from Tc to P

24

JOHN BALDAS

being less favored in the higher oxidation state. In nonaqueous media truns-[Tc"(dppe),X,] is readily oxidized by a variety of one-equivalent oxidants. The rate of reduction of [(en),Co{S(CH,C,H,Me)CH,CH,NH2}I3+by [Tc(dppe),Cl,] in MeCN a t 25°C is rapid [it, = 3.0(7) x lo4 M -1 sec-'1 (193). Spectroelectrochemical studies of the structurally characterized truns-[T~(dppe),(NCS)~I show that the r-acceptor ability of N-bonded thiocyanate results in a marked stabilization of the lower oxidation state (194).The cationic complexes "I'cL~](PF,)~(L = dmpe, depe) are prepared by HzOz oxidation of [Tc'L,]PF,. The reduction of Tc0,- by excess dmpe gives a mixture of products in the +5, +3, and + 1 oxidation states, but neither truns-[T~~~(dmpe),Cl,I nor [Tc"(drnpe),l2+ appears to be a predominant component (124). Reaction of truns[TcvO2(dmpe),1 with halide may be used to prepare trans-[Tc(dmpe),X,] (1231. A series of cationic dithiocarbamato complexes [ T c ( S , C N R , ) ( ~ ~ ~ ~ )has ~ ] Precently F, been similarly prepared, and the crystal structure of [Tc(S,CNMe,)(depe),lPF,, determined (195).The dithiocarbamate is bidentate and the coordination geometry is octahedral. The E"' values of 0.298 to 0.312 and -0.517 to -0.544 V vs Ag/ AgCl for the Tc(III)/Tc(II)and Tc(II)/Tc(I) couples, respectively, show that dithiocarbamato ligands effectively stabilize the Tc(I1) oxidation state. A crystal structure determination has shown that the room temperature reduction of the tetramethylthiourea complex [TcvO(tmtU),](PF& by dppe in dmf solution is accompanied by a novel rearrangement to give the dithiocarbamato complex [Tc"(dppe),(S2CNMez)1PF6 (196).The mechanism of the rearrangement is uncertain, but the overall reaction is probably described by +

-

2(Me2N),CS+ HzO

Me2NCS2H+ (Me2N)&0 + MezNH

An interesting mixed-ligand complex is [Tc(dppe),(ox)l, prepared by the reduction of Tc04- by dppe in hot ethanol in the presence of oxalic acid, The average Tc-0 bond distances of 2.13(1) and 2.12(1)A in two independent molecules are somewhat longer than those found in other oxalato complexes (197).Reaction of (NH,),[TC'~C~,]with diethyl phenylphosphonite in the presence of BH,- yields the yellow octahedral truns-[Tc"{PhP(OEt),),C1,l, with Tc-P and Tc-Cl bond distances both 2.41(1) A. The low measured magnetic moment of 1.4 BM has been ascribed to partial decomposition (198).

E. COMPLEXES WITH SULFUR LIGANDS The reaction of truns-[TcVO(OH)(dmpe),12fwith excess 4-chlorobenzenethiol yields a mixture of the air-stable black cis- and red truns-

COORDINATION CHEMISTRY OF TECHNETIUM

25

[Tc(SC6H,C1-p),(dmpe),I,which may be separated by fractional crystallization. Crystal structures have been determined for both complexes and show Tc-S-C angles of 114.0(3)" and 123.8(2)' for the cis- and trans-isomers, respectively. In CH2C12,the trans-isomer converts to the more stable cis-isomer with a half-life of about 74 min a t room may be temperature (199).Similarly, trans-[T~~O(OH)(diars)~](PF,), converted to [Tc**(SR),(diars),I(trans, R = Me, Bz; cis and trans, R = Ph), which can then be oxidized to the Tc(II1) complexes. The reversible Tc(III)/Tc(II) couple is in the range -0.32 to -0.47 vs Ag/AgCl. The crystal structure of trans-[Tc(SPh),(diars),] shows a Tc-S-C angle of 119.5(3)" (200). Reduction of NBu,TcO, by SnC1, in the presence of 1,4,74rithiacyclononane yields the dark-brown homoleptic complex [TcL,](BF,),.MeCN, with Tc coordinated by the two thioether ligands in a fac tridentate "sandwich" fashion. The three independent Tc-S bond distances are in the narrow range 2.372(3)-2.381(3) A. Electrochemical oxidation yields the yellow Tc(II1) complex and reduction yields the air-stable cherry-red Tc(1) complex (201). F. NITROSYL AND THIONITROSYL COMPLEXES The first Tc(I1) nitrosyl complex to be identified was the green trans[Tc(NO)(NH3),(OH,)]Cl3, prepared by Ce(1V) oxidation of the pink trans-[Tc1(NO)(NH3),(OH~)]Cl~ (135).From the reaction of Tc02.nH,0 with NO gas in 4 M HBr, red crystals of NBu,[Tc(NO)Br,I may be isolated (139).The NBu,[Tc(NO)X,l (X = C1, I) complexes are prepared by halide substitution (34, 202) and the ink-blue (NBu,),[Tc(NO)(NCS)5]is prepared by substitution with NCS- (139),whereas (ASP~,),[TC(NO)(NCS)~] may be prepared by the reduction of TcO,by NH,OH.HCl in the presence of NCS- (203). The preparation of (NBu4),[Tc(NO)C1,l has also been reported (202). The structure of (NBu,)trans-[Tc(NO)C14(MeOH)] shows a near-linear Tc-N-0 angle of 175.5(10)" and Tc-N and N-0 bond distances of 1.689(11) and 1.171(15) A, respectively. The Tc-O(H)Me distance is fairly short at 2.128(7) A (2041,a reflection of the wacceptor nature of the NO+ ligand. For the halide complexes v(N0) is observed in the IR spectrum a t -1800 cm-'. Reaction of [Tc(NO)Cl,]- with bpy or phen yields [Tc(NO)Cl,LI (205). Black-green crystals of [ T C ( N O ) C ~ , ( P M ~ ~ P ~ ) ~ ] (206) and dark-green [Tc(NO)Cl3(PPh3),1[v(NO) at 1805 cm-'1 (207) are formed by the reaction of NO with [TcCl,(PMe,Ph),] and [TcC1,(PPh3),(MeCN)1,respectively. Red (AsPh4)mer-[Tc(NO)C1,(acac)l (14) is formed when acacH is added to a solution of [Tc(NO)C1,XIn(X = C1, H,O), prepared by the addition of NH,OH.HCl to [TcC1,I2-.

26

JOHN BALDAS

The Tc-N-0 angle is substantially bent at 158.6(33)"and the Tc-O,,,, bond distances cis and trans to the NO+ ligand are not significantly for potendifferent at 2.06(1)and 2.08(1)8, (208,209). [99"T~(NO)C141tial radiopharmaceutical applications is formed in high yield by the addition of NH20H.HC1 to a previously heated 99"Tc0,-/HC1 solution and may be extracted by CH,Cl, after the addition of NBu,Cl (210).

1-

(14)

(15)

The AsPh,[Tc(NS)X,I (X = C1, Br) salts are prepared by the reaction of (ASP~,)~[TCX~I with (NSCl), and AsPh,[Tc(NS)(NCS),I is prepared by ligand exchange. In the case of [TcBr,12-, mixed-ligand complexes [Tc(NS)Cl,Br,-,l- ( n = 1-31 are formed and the addition of HBr is required to effect full substitution and to give red-brown crystals of AsPh,[Tc(NS)Br,]. The d N S ) IR absorptions occur a t 1232-1214 cm-', and for the thiocyanate complex the NCS deformation mode at 501 cm-' indicates that these ligands are N-bonded. In solution, [Tc"(NS)X,] readily loses sulfur to form the nitrido complex [TcV1NX41-(211). Treatment of [TcVNC1,(PMezPh),]with a n excess of S2C1, in CH2Clz under argon has been shown by FABMS and crystallography to yield not [Tc"(NS)C~,(PM~,P~)~I but the phosphine oxide complex (15) (2121,and presumably this also applies to the bromo complex prepared by ligand exchange in HBr (213).The source of oxygen for the oxidation of the phosphine in 15 is not clear. The v(NS) IR band at 1240 cm-' has been confirmed by a n isotope shift to 1206 cm-' on 15N labeling (145).On reaction with excess PMe2Ph, 15 undergoes sulfur abstraction to yield the starting material [TcVNC1,(PMe2Ph),]. The linear Tc-N-S [179.9(1)"]angle and the Tc-N and N-S bond distances of 1.7466)and 1.521(5) 8,, respectively, in 15 are consistent with NS' coordination (212).EPR spectroscopy has confirmed that the product of the reaction of [TcVNL2][L = N-(N"-morpholinylthiocarbonyl)benzamidinate~1-)1 with S2C12is [Tc"(NS)Cl,L] (214). The EPR spectra of low-spin 4d5 Tc(I1) nitrosyl and thionitrosyl complexes have been examined in detail and the results, reviewed (39-41 ). -

COORDINATION CHEMISTRY OF TECHNETIUM

27

Complexes with the most n-bonding between Tc and the equatorial ligands show the largest glland smallest All, as seen by gll = 1.985, and All= 260 x cm-' for [Tc(NO)Cl,l2- andgll= 2.262, andAli= 155 x cm-' for [Tc(NO)I,]- (202, 215). In the case of [Tc(NO)(NH3)4(OH2)13+, for which there is no r-bonding in the equatorial plane, cm-' (216).EPR spectroscopy is particugll= 1.861 and Ail = 297 x larly useful for the study of ligand exchange such as that between [Tc(NO)Br,]- and [Tc(NO)C1,12- or [Tc(NO)141-(217).The g values and 99Tc hyperfine coupling constants are proportional to the spin-orbit coupling constants of the donor atoms and the composition of the mixedhalide coordination sphere may be unambiguously assigned (40). VI. Technetium( III)

The coordination chemistry of this oxidation state is rather more extensive and varied than that of Tc(I1). With appropriate ligands Tc(II1) is water- and air-stable, and cationic complexes with bidentate phosphine and arsine ligands have been extensively studied in the search for myocardial imaging agents. A marked difference between Tc and Re is the absence, at present, of any Tc analog of the extensive chemistry based on the trinuclear [Re3I9+core (189).A notable feature is that nearly all the seven-coordinate Tc complexes are found in this oxidation state. This is understandable in terms of the d4 electronic configuration of Tc(III), which requires seven singly bonded ligands to achieve an 18-electron count. A. CARBONYL COMPLEXES Yellow plates of the seven-coordinate [TcC1,(CO)(PMe2Ph),1~EtOH are formed by passing CO at atmospheric pressure through a refluxing solution of rner-[TcCl3(PMe2Ph),1in ethanol. The structure possesses approximate C,, symmetry and may be described as a distorted capped octahedron with the CO ligand inserted along the C3 axis on the phosphine face [Tc-C-0 angle, 178(2)"1(218).The pentagonal-bipyramidal complex (16) (Fig. 5 ) was unexpectedly prepared by the reduction of Tc04- with formamidinesulfinic acid [NH2(NH)CS02H]in the presence of Na(S2CNEt2).The CO stretch occurs as a n intense band at 1895 cm-' in the IR spectrum. The mechanism of the formation of the coordinated CO is unclear; a scheme that involves initial loss of SO2 from the coordinated sulfinic acid has been proposed. The CO ligand occupies an apical position with a near linear Tc-C-0 angle of 177.8(10)" and

28

JOHN BALDAS

J: w

C15

C

1

3

c12

FIG.5. The structure of [Tc(S,CNEt&CO)I (16)(219).

Tc-C and C-0 bond distances of 1.861(12)and 1.15(1)A, respectively. A comparison with the isostructural [Re(S2CNEt2),(CO)Iindicates that Tc(II1) is a poorer n-donor than is Re(II1) (219). Alkyl dithiocarbamate derivatives of 16 have been prepared by the above method [v(CO) a t 1907-1895 cm-'I and the CO ligand shown to be inert to substitution by EPh, (E = P, As) (220).The CO ligand is, however, readily substituted by the isoelectronic NO to give seven-coordinate [Tc(NO)(S2CNR2),l+cations (2211. A variety of [99mT~(SzCNR2)3(C0)1 complexes has been prepared by reduction of TcO,- with S2042-in the presence of CO and found to behave as hepatobiliary agents when injected into mice (222). A six-coordinate carbonyl complex is truns-mer-[TcCl3(PPh3)2(CO)I,prepared by the bubbling of CO through a solution of rner-[T~Cl,(PPh,)~(MeCN)l. The long Tc-CO bond of 1.985(9) A, the short C-0 bond of 1.12(1) A, and the high IR v ( C 0 ) absorption at 2054 cm-' indicate the absence of significant n-back-bonding in this complex (207). The bubbling of CO through a solution of [Tc(SAr),(MeCN),I (SAr = tmbt) yields orange crystals of [Tc(SA~),(CO)~]. One CO ligand may be displaced to give [Tc(SAr),(CO)(MeCN)]and [Tc(SAr),(CO)(py)l. Crystal structures show trigonal-bipyramidal coordination for both monocarbonyl complexes, with the three S atoms of the sterically hin+

COORDINATION CHEMISTRY OF TECHNETIUM

29

dered thiolates occupying the equatorial plane and the CO and MeCN or pyridine ligands in the axial positions (223). B. CYCLOPENTADIENYL COMPLEXES The product of the BH,- reduction of TcCl,/NaCp in THF is the hydrido complex [HTcCp,], analogous to [HReCp,] and most likely with the same bent structure. The basic nature of [HTcCp,] is shown by the equilibrium

On addition of PF,- the rather insoluble [H,TcCp,IPF, salt precipitates (224).The TcCl,/KCp reaction in THF yields the diamagnetic airstable [TcCp,Cl], which on reaction with KCp gives the red diamagnetic [Tc(q5-Cp),(q'-Cp)], in which one ring is a-bonded (225).The structures are shown in Fig. 6. The oxidation of [(q5-C5Me5)Re1(C0),1by H202 yields the trioxo compound [(q5-C,Me5)ReV"O31,but with [(q5C,Me,)Tc'(CO),] the product has been assigned the polymeric structure (171,with Tc in the +3.5 oxidation state (226).

The planes of the C5Me5rings and the bridging oxygens are exactly parallel but the most striking feature is the exceptionally short Tc-Tc bond distance of 1.867(4) A. The Tc-Tc bonding hat3 been described as ~ r ~ ( , r r t 3 ) ~with 8 * , a net bond of approximate order 2.5 (227). IR absorptions at 909 and 880 cm-' have been assigned to v,,(TcO) and vWm (TcO), respectively (226).The crystal structure determination of 17 has, however, been questioned and it has been suggested that the (228).Treatment product obtained may in fact be [(q5-C5Me5)TcVI1O31 of [Tc1(C5Me5)(CO>,1 with Br,/CF,COOH gives [ T c ( C ~ M ~ ~ ) ( C Oas )~B~~I a mixture of cis- and trans-isomers (100).

30

JOHN BALDAS

c1

FIG.6.

C. CYANO, ISONITRILE, AND THIOCYANATO COMPLEXES The only Tc(II1) cyano complex is the seven-coordinate yellow-orange K,[TC(CN)~].~H~O prepared by the reaction of (NH,)2[Tc'VI,l with KCN in methanol under nitrogen. Raman and IR spectra indicate a pentago-

COORDINATION CHEMISTRY OF TECHNETIUM

31

nal-bipyramidal structure symmetry) both in the solid state and in solution. In aqueous solution K4[Tc(CN),1.2H2Ois oxidized by air to [Tc"O(CN)J- (229).Seven-coordinate isonitrile complexes of the type [Tc(CNR)~X](PF~)~ (X = C1, Br) are formed by the oxidative addition of chlorine or bromine to the six-coordinate [Tc'(C"),]+ (1211. The reaction of (NH,),[TcX,I (X = C1, Br) with NH,NCS produces a mixture of the intensely purple ["I'C'"(NCS)~]~(Amm = 500 nm; E = 76,200) and = 400 nm) anions (230). the air-sensitive, yellow [Tc"'(NCS),I3- ,A( The redox couple [TC(NCS)~]~+ e [TC(NCS)~]~is electrochemically reversible (El,, = 0.18 V vs SCE) and the reduction of [Tc1"(NCS),l2is readily produced chemically by hydrazine. The crystal structure of (NBU,)~[TC(NCS)~] shows near-perfect octahedral geomery and establishes that thiocyanate is N-bonded with a n average Tc-N-C angle of 173(2)".

*

D. AQUA,HALIDE,AND RELATED DIMERIC COMPLEXES No mononuclear Tc(II1) halide is known. Thin-layer spectroelectrochemical techniques show that the [Tc'"X6I2- (X = C1, Br) complexes undergo a reversible one-electron reduction in HX/NaX aqueous media with the loss of 2.7 0.1 chloro ligands and 5.9 0.5 bromo ligands, respectively. These results indicate a low affinity of Tc(II1) for halide and the possibility of preparing [ T c ( O H ~ ) ~in] ~a+weakly coordinating aqueous medium (2311, The stability of the mixed-valence [ T C " ' ~ ' ' ~ C ~ ~ ] ~ and the apparent inability to prepare [ T c " ' ~ C ~ ~were I ~ - a puzzle for a number of years in view of the stability of [Re2C1,12- and the only fleeting existence of [RezC1813-(159). However, in 1980 the brightgreen (NBU,)~[TC~C~,] was prepared by the reduction of [TCCl6I2-with Zn/HC1 and converted to the carmine-red ( N B U ~ ) ~ [ T C ,by B ~ligand ~I by exchange with HBr (232). A recent synthesis of (NBU,)~[TC~C~,] reduction of NBu,[TcOCl,] with NBu,(BH4), followed by carefully controlled air oxidation of the initially formed brown product in CH2C1, in the presence of HC1 gas, gives yields of up to a 85% (164).The crystal structure of (NBu,),[Tc,Cl,] shows that the quadruple Tc-Tc a2n46 bond is, a t 2.147(4) A, about 0.04 A longer than the u2n46%*bond in [Tc,C~,]~-(bond order, 3.5)(233).This unexpected result is due to the greater influence of the change in the oxidation state of Tc than of the weak 6 bonding on the Tc-Tc bond distance (149). A number of complexes of the [Tc,]~' core have been prepared, either by oxidative substitution of [Tc2C1,I3- (234, 235), by substitution of [TC,C~,]~(236,2371, or by the reduction of Tc0,- by hydrogen in the presence of ligands (238). The crystal structure of the red [Tcz-

*

*

32

JOHN BALDAS

c4

FIG.7. The structure of cis-[T~~(OAc)~Cl,(dmaa)~] (237).

(00CCMe3)4C12]reveals a lantern structure, with the four pivalato ligands bridging the two Tc atoms in the eclipsed configuration (D4h symmetry), a Tc-Tc bond distance of 2.192(2) A, and axial chlorides with Tc-Cl bond distances of 2.408(4) A (234).Similar structures are found in [Tc,(OAC)~(TCO~)~I [Tc-Tc, 2.149(1) A1 (239) and K,[Tc, (SO,),(OH,),] [Tc-Tc, 2.155(1) A] (235,157).Reaction of [TczC1,12- with Ac20/HBF4yields czs-[Tcz(OAc)2C1,(OH,)21,in which the weakly bound axial water ligands are easily replaced by donor bases to give the green adducts cis-[Tc,(OAc),Cl,L,] (L = dmf, dmso, OPPh3, py). The structure of the dimethylacetamide (dmaa)adduct [Tc-Tc, 2.1835(7) A; Tc-Odmaa, av., 2.320 A] is shown in Fig. 7 (237). Orange-red [Tcz(OAc),Br,l is prepared by reaction of [Tc2BrsI2-with HOAc/Ac,O (236).A characteristic feature in the electronic spectra is the 6 6" transition at 600-700 nm (232,237).Normal coordinate analyses of [Tcz(OAc),Xzlgive Tc-Tc force constants of 4.08 and 3.99 mdyn for the chloro and bromo complexes, respectively (236). The Tc-Tc IR and Raman stretching frequencies of cis-[T~,(OAc)~Cl,L~l are lowered with increasing donor strength of the axial ligand L (240). Magnetic studies of [Tc2I6+complexes show only temperature-independent paramagnetism (150).

-

E. CARBOXYLATO AND P-DIKETONATO COMPLEXES Technetium(II1) complexes with aminocarboxylato ligands have been reported but none are well characterized (2411. 99mTc-iminodiacetate complexes formed with 2,6-alkylphenyl [ArNHCOCH2N(CH2C00)212ligands are used to image the hepatobiliary system. Studies with "Tc show evidence for [TC"~L,]-in the radiopharmaceutical preparations,

COORDINATION CHEMISTRY OF TECHNETIUM

33

but [TcVOL,]- is also possible (19).A variety of Tc(II1) acac complexes and substituted analogs has been prepared by substitution/reduction of [TC'~X,(PR,),]and [TcrVX6Iz-or by S z 0 2 - reduction of Tc0,- in the presence of the ligand. These include [Tc(acac),l (2421; the dipivaloyl, trifluoro, and hexafluoro analogs (243); and [TcX(acac),(PPh,)I and [TcX,(a~ac)(PPh,)~l (X = C1, Br) (242).The S2O:- reduction method is suitable for the preparation of the neutral lipophilic 9 9 m Ttris ~ complexes, but these show little brain uptake (244).The cationic [Tc(acac),(MeCN),]ClO, is formed by the reaction of [Tc(acac),I with MeCN in the presence of HClO, (245).The crystal structure of [Tc(acac),I shows closely octahedral coordination, with a n average cis 0-Tc-0 angle of 90.2"and Tc-0 distances in the range 2.013(6)-2.030(6) (246).The structures of two crystalline forms of truns-[TcCl(acac),(PPh,)3,which show differences in the IR spectra, have been reported (247,248).The kinetics of ligand exchange of [Tc(acac),] have been studied by the use of 14C-Hacac, and the I, mechanism has been assigned to the rate determining formation of a n intermediate containing one monodentate acac and Hacac ligand (249).The base hydrolysis of [Tc(acac),I is kinetically more complex than that of [Ru(acac>,l (250).A variety of sixcoordinate tris complexes of monothio-P-diketonates has been prepared by substitution of [Tc(tu),]Cl, in refluxing methanol and characterized by IR, electronic, and mass spectrometry; 'H NMR; and, for the phenyl derivative [Tc{SC(Ph)CHC(Ph)O},],a crystal structure determination (251, 252).

F. COMPLEXES WITH DIOXIMES, SCHIFF BASES,AND OTHER NITROGEN LIGANDS An alternative approach to cationic myocardial imaging agents has been the development of neutral seven-coordinate Tc(II1) complexes based on 1,2-dioxime ligands (dioximeH,) with one end capped by a boronic acid derivative (19, 253). These complexes are generally referred to as BATOs (boronic acid udducts of technetium dioximes) and have the general structure [TcX(dioximeH),(dioxime)BR'](X = C1, Br; R' = alkyl) shown in 18. At the uncapped end the three dioxime oxygen atoms are intramolecularly bonded to two bridging protons. BAT0 complexes are prepared by template synthesis from TcO,- and Sn2+or from NBu,[TcOCl,], M,[TC&] (X = C1, Br) in the presence of the dioxime, HX, and the alkylboronic acid (254).The formation of BATOs from TcO,- and Sn2+ proceeds via several intermediates, including an Sn-monocapped [Tc"'(dioximeH),(p-OH)SnC13],which undergoes acid decomposition to give

34

JOHN BALDAS

R

the uncapped [Tc"'X(dioximeH),(dioximeHz)] complex. The uncapped tris complex is then monocapped by the boronic acid (255).Interestingly, although bidboron-capped) clathrochelates [M(dioxime),(BR),] have been known for a number of years for M = Co, Fe, and Ru, the BATOs are the first monocapped examples. Crystal structures have been reported for [TcBr(cdoH),(cdo)BRl (R = Me, Bu) and [TcBr(dmgH),(dmg)BRl (R = Me, Bu) (254);the structure of the n-butyl dimethylglyoxime complex is shown in Fig. 8. The six nitrogen atoms form a distorted trigonal prism monocapped by Br, which causes the two flanking dioximes to be spread away by about 20"toward the third dioxime ligand, thus probably precluding the addition of a second boron cap (2541. The cdoHz derivative [99mTcCl(cdoH)2(cdo)BMel is a radiopharmaceutical for differentiating normal from ischemic and infarcted myocardium (19).The axial chloride is labile to substitution and under physiological conditions is replaced by a hydroxy group with pK, between 7 and 7.4, which indicates that there may be a n equilibrium in uiuo between the neutral hydroxy and cationic aqua forms (24, 256, 257). The lability of the axial chloride is consistent with X-ray photoelectron spectra of "Tc BATOs, which show that the binding energy is between that for covalent and that for ionic bonds (258).The mechanism of chloride-hydroxide exchange has been shown to be S,l-CB, proceeding via a transient neutral six-coordinate complex (256).Electrochemically, chloro and bromo BATOs undergo a n irreversible twoelectron reduction that appears to be biologically inaccessible (259). The S- and N-bonded isomers [TcL(cdoH),(cdo)BMel (L = NCS, SCN) have been prepared. In solution, the S-isomer converts to the N-isomer when exposed to light (257).Cationic BATOs have also been prepared (260).The reaction

-

Tc04-+ 3 dioximeHz + 2 SnClZ

[Tc"'(dioximeH)3(p-OH)Sn1VC13] (19)

COORDINATION CHEMISTRY OF TECHNETIUM

35

FIG.8. The structure of [TcBr(drngH),(dmg)BBu](254).

proceeds to completion. The oxygen bridge between SnIVand Tc"' is most likely in the hydroxyl form. Acid decomposition of 19 yields [TcCI (dioximeH),(dioximeH,~],which may be reconverted to 19 in 97% yield on reaction with SnC1, (255).When the crystal structure of [Tc(dmgH),(p-OH)SnC131~3H,0 was reported in 1976 the oxidation state of Tc was thought to be + 5 (2611,but a + 3 oxidation state is indicated by FABMS and the chemical behavior of 19 (255).The Sn" center is six coordinate, with the three chloro ligands in a fuc arrangement, and, in addition to the hydroxy bridge, two oxime oxygens complete the octahedral coordination sphere (2611. It may be noted that a + 3 oxidation for Tc is also consistent with an 18-electron seven-coordinate species. In the structurally characterized [T~~~~Cl(dmgH),(drngH~)l, the

36

JOHN BALDAS

four protons are shared by the three oxygens on each trigonal face and in the lH NMR spectrum appear as a broad singlet at 15.3 ppm (255). Crystallography has established that a by-product of the reaction of dmgH~/[TcC1,(MeCN)(PPh3),l/EtB(OH), is [TcCl(dmg)(dmgH)(butane2,3-dioneimineoxime)BEtl, where one of the uncapped C=NOH groups has been converted to C=NH and there is only one intramolecularly bound proton (262).A variety of seven-coordinate Re analogs of BATOs, uncapped [Re"'C1(cdoH),(cdoH~~],and monocapped [Re"'Cl(cdo)(cdoH),BR] has been prepared. Yields from Re0,- are low but [ReC13(MeCN)(PPh,),] is a suitable starting material. As with Tc, the biscapped Re complexes could not be prepared. Reaction of Mn(OAc),/ cdoH,/(OH),BPh/MeOH, however, gives a high yield of the biscapped six-coordinate [Mnll(cdo)(cdoH),{B(OMe)Ph}~l, in which each cap is covalently bonded to two oxime oxygen atoms (263). A series of cationic [Tc(L)(PR3),1PF,(PR3 = PEt3, PEtzPh, PEtPh, , PPh3)complexes of tetradentate (acac),en ligands and aromatic derivatives has been prepared by substitution/reduction of [TcOCl,I- (2641. The E for the reversible Tc(III)/Tc(II)couple is sensitive to the nature of the substituents on the Schiff base and the phosphine but is in the range -1.11 to -0.69 V vs Ag/AgCl (265). These cations are thus essentially biologically nonreducible and the 9 9 m Tcomplexes ~ are of interest as potential myocardial imaging agents. All complexes exhibit characteristic MLCT bands in the visible region, the energy of which correlates linearly with the potential of the Tc(IV)/Tc(III)and Tc(III)/ Tc(I1) couples (2641. The crystal structure of trans-[Tc{(acac),en}(PPh3),]PF6shows approximate octahedral coordination with the tetradentate Schiff base in the equatorial plane (264).The thio derivative [Tc{(sacac),en}(PPh,),]PF, has also been prepared from NBu,[TcOCl,I (266).A rather mixed coordination sphere is present in [Tc(quin)(PR,)L] (L = 20; E = 0, S; PR, = PMe,Ph, PEtzPh, PPh,). The crystal structure of [Tc(quin)(PEt,Ph)Ll (L = 20; E = 0)shows approximate octahedral O'

(E = 0, OphsalH,) (E = S, SphsalH,) (20)

P = PMe,Ph

(21)

COORDINATION CHEMISTRY OF TECHNETIUM

37

geometry with the tridentate ligand L spanning three mer positions and the phosphine trans to the quinoline nitrogen (267). Crystallography has shown that the product of the reaction of a Schiff-base dithiocarbazate ester derivative (H,L) with [TcOCl,]-/PPh, is the octahedral cis(CZ)-truns(P)-[TcCl2(HL)(PPh3),1,where HL functions as a bidentate S,N-ligand (268). For [TcClL,(PMe,Ph)] (L = N-phenylsalicylidineiminate),the two bidentate chelate ligands are mutually orthogonal and the chloro and phosphine ligands, cis to each other (269).Novel complexes are [{TcL(PR,),},(p-O)] (L = 20; E = 0, S; PR3 = PMe,Ph, PPh,), which represent the only examples of Tc(II1) (21) the oxo-bridged dimers. In [(T~(Ophsal)(PMe~Ph)~},(p-O)] Tc-O-Tc angle is near-linear at 176.1(14)"and the Tc-Obridgedistances are 1.81(2) and 1.87(2) (270). The complexes rner-[TcC13L31(L = py, pic) and mer-[TcCl,(pic)(PMe,Ph),l are prepared from NBu4[TcOC1,1 dissolved in neat pyridine or picoline by the use of a phosphine as the oxygen acceptor. Linear correlations of reduction potentials in dmf with electrochemical ligand activity parameters are observed for the Tc(IV)/Tc(III), Tc(III)/Tc(II), and Tc(II)/Tc(I) couples. Crystal structures for mer-[T~Cl~(pic)~I and rner-[TcC13(pic)(PMe2Ph),lconfirm the expected octahedral geometry (2711. A variety of Tc(II1) complexes containing polypyridyl ligands has been prepared by substitution of [TcCl3(PR,R'),1 or [TC(tU)&13 or substitution/reduction of [TcC1,(PPh3),l (272, 273), and the electrochemical behavior has been studied (2741. Crystal structures have been reported for [TcCl,(PPh,)(bpy)I (2721, cis(CZ),truns(P)-[TcCl,(PMe,Ph),LlBPh, (L = bpy, phen), and ck(CZ),truns(P)-[TcCl,(PEtPh2)2(bpy)lCF3S03(273). The preparations of [TcCl,L(HB(pz),}] (L = PPh, ,OPPh, ,py), containing the tridentate HB(pz),- ligand, have been reported (275). Reaction of [Tc(tu),I3+ with phen is thought to give [Tc(phen),I(PF,), (276). The organohydrazine chemistry of Tc parallels that of Re (189).The air-stable biddiazenido) [TcC1(NzAr),(PPh3),lcomplexes are formed by the reaction of [TcVOC1,l- or [TcC~,(PP~,)~I with excess ArNHNH, in alcoholic solution or directly from TcO,- (277, 278, 140). The use of p-NO,C,H,NHNH, gives the lime-green monodiazenido complex [ T C C ~ , ( N N C , H , N O ~ - ~ ) ( Pin P ~high ~ ) ~ yield ] (140).Organodiazenido ligands most commonly bond in the singly bent, three-electron-donor mode with the doubly bent, one-electron-donor mode much less com(X = C1, Br) (22) mon. Crystal structures of [TcC1(NNC6H4X-p),(PPh3),l show trigonal-bipyramidal geometry with Tc-N-N angles of 166.2(6)" and 170.7(7)"for X = Br and the same essentially linear arrangement for X = C1 (277, 140).

38

JOHN BALDAS

PPh,

(22)

f

singly bent 3-electron donor

doubly bent l-electron donor

The bond distances indicate extensive delocalization and multiple bonding in the -NNAr moieties together with significant Tc backbonding. The diazenido ligands are thus singly bent, three-electron donors and [TcCl(NNAr),(PPh,),I complexes have a formal valence electron count of 18 (140). The cationic monodiazenido [TcCl(NNAr) (dppe),]+ complexes may be prepared by substitution of [TCCUNNA~)~ (PPh,),] or directly from Tc04- and isolated as the PF6- or BPh,- salts. The crystal structure of trans-[T~Cl(NNPh)(dppe~~IPF~~H~O shows slightly distorted octahedral geometry, a Tc-N-NPh angle of 163(2)",andTc-N and N-NAr bond distances of 1.917(19)and 1.25(4) A, respectively (140). Substitution of [TcC1(NNC6H,C1),(PPh,),1 with Na(S,CNMe,) in methanol yields the dark-orange [Tc(NNC6H4Cl)(S,CNMe2),(PPh,)]. The crystal structure shows distorted octahedral geometry, with the PPh, and diazenido ligands in cis positions and Tc-N-NAr and N-N-Ar angles of 178.6(4)"and 122.5(5)",respectively. The Tc-N and N-NAr bond distances are 1.763(3) and 1.236(4) A, respectively. The trans influence of the diazenido ligand is apparent because the trans Tc-S bond distance is longer [2.537(1) A1 than the other three Tc-S distances [2.412(2)-2.477(2) A] (279).

G . COMPLEXESWITH MONODENTATE PHOSPHINES AND RELATED LIGANDS The products of the reduction of Tc04- by phosphine/HX (X = C1, Br) depend on the nature of the phosphine and the reaction conditions. With PPh, only trans-[T~'~X,(PPh,),] is formed, whereas the more strongly reducing PR,Ph (R = Me, Et) gives the Tc(1V) complex at a Tc0,- : phoshine ratio of 1 :5 and rner-[T~"~X,(PR,Ph)~l at a ratio of 1 : 15 or higher (280). Alternatively, the Tc(II1) complexes may be prepared by the reduction of trans-LT~X,(PR,Ph)~l with excess PR2Ph (2811 or by the reaction of (NH,),[TcCl,] with PRzPh (282).The magnetic

39

COORDINATION CHEMISTRY OF TECHNETIUM

moment of 2.8 BM for the Tc(II1) complexes is consistent with a t2: configuration in an octahedral environment (280).In MeCN solution, mer-[TcCl,(PMe,Ph),] may be electrochemically oxidized to [Tc'"Cl,(PMezPh)21or [Tc'VC13(PMe,Ph),]C104or reduced to Tc(I1) and Tc(1) phosphine complexes (2831. Reduction of NBu4[TcVOC141with PMe, yields mer-[TcC13(PMe3)31(284 and with PPh,/MeCN, [TcCI,(PPh,), (MeCN)] (207), whereas reduction of Tc0,- by PPh,/HCl/dmf yields 23 (L = dmf) (285). The MeCN complex (23) is a useful synthetic intermediate. On reaction with CO or NO only the MeCN ligand is substituted (207) but bpy and phen result in complete substitution, producing [TC"L,]~' salts (185).The crystal structures of mer-[TcCl, (PMe,),].(PhNCO), and mer-[TcCl,(PMe,Ph),] show a marked trans influence of the phosphine ligands with the Tc-C1 bond distances trans to P about 0.08-0.13 A longer than those trans to C1(284,282). Crystal structures have also been reported for 23.2PPh3 (L = dmf) (285)and truns-mer-[TcC1,(MeCN){P(m-MeCsH,),),l (185).The reaction of [Tc (S-tu),](PF,), with PMe, in methanol gives the hydrido complex [Tc(H){r)2-N,-S-NHC(NH2)S}(PMe3)41PFG (241, in which the thiourea ligand has undergone deprotonation and binds in the unusual $-N,-S mode. Structure 24 was established by crystallography, multinuclear NMR, and IR spectroscopy [v(TcH) at 1898 em-'] (284). CI

PMe,

l+

(24)

(L

MeCN, dmf)

(23)

A number of Tc(II1) phosphonite complexes of the type [TcX,{P(OEt)2Ph},]CIO, (X = C1, Br, I) have been prepared from (NH,),[TcX,]. The magnetic moments are in the range 2.3-2.6 BM (79).

H. COMPLEXES WITH BIDENTATE PHOSPHINE, ARSINE,AND RELATED LIGANDS The Tc(II1)complexes trans-[Tc(diars),X,]X (X = C1, Br, I) were first reported in 1959 (190) and the chemistry of the dppe analogs was described in detail later (192).These and related complexes have been

40

JOHN BALDAS

intensely investigated after it was shown by Deutsch et al. that the + 1 cation tr~ns-[~~"Tc(dmpe)~Cl~1+ accumulates in the heart (286). Complexes of the type trans-[TcL,X,IY (L = diars, depe, dmpe, dppe; X = C1, Br, I; Y = PF,, CF3S03, BPh,, BF,) are prepared by the reduction of TcO,-, [TcOX,I-, or [TCX,I2- with excess phosphine or arsine (287). Electrochemical and spectroelectrochemical studies have shown that the E "' value of the reversible Tc(III)/Tc(II)couple depends on the nature of X and of the bidentate ligand, with reduction being easier with the heavier halogen and also easier for dppe than for diars complexes (287,288). These effects result from the stabilization of the Tc(I1) d5 center over the Tc(II1) d4 center by a n increased ligand field. Typical E"' values are in the biologically accessible range of 100 to -250 mV vs Ag/AgCl. Under common laboratory conditions Tc(II1) is the stable state for the chloro and bromo complexes but when X = NCS, E"' is 390 mV and [Tc"(dppe),(NCS),] is the stable state (287). Comparison of tran~-[M"""L~X~]+'~ (M = Tc, Re) couples has shown that the Tc complex is always easier to reduce than the Re analog, with - E o f k )219 2 15mV (289).Thus, the significantly different biological behavior of [99"Tc(dmpe)2C1,1+and [ 1R6Re(dmpe),C12]appears to be due to the in uiuo reduction of the 9 9 m Tbut ~ not the '"Re complex (2901. Interestingly, pulse radiolysis studies have shown that in aqueous anionic surfactant media the [Tc(dmpe),Cl,] cation effectively partitions into the anionic micelles and is there relatively protected from the highly reactive negatively charged strong reductant eaq- and the strong oxidant C1,- (291). The chloro ligands in trans[Tc(dppe),Cl21 are rather unreactive to exchange (287). Reaction of tran~-[Tc(dppe)~Cl~l with LiAlH, yields yellow crystals of [Tc"'(H), (dppe),Cl] [v(TcH) at 1851 and 1775 cm-'1 (292). The electronic spectra of truns-[TcLzXz]+(X = C1, Br; L = diphosphine, diars) exhibit well-defined intense bands in the visible region (-20,000-23,000 cm-') that are 2500 ? 370 cm-' lower in energy than in the corresponding Tc(I1) complex (287). These bands have been assigned to X +. Tc LMCT transitions, and for Tc(II1) Cl/Br pairs the difference is about 1600 cm-'. All complexes are paramagnetic; the magnetic moment of trans-[Tc(dppe),Br,lBr is 2.47 BM (192, 287). FABMS has proven useful in the study of these Tc(II1) cationic complexes (293,294 ). Crystal structures for tr~ns-[Tc(diars)~Cl,]Y (Y = C1, ClO,) (295, 296), tr~ns-[Tc(dppe)~Br~lBF, (287), truns-[T~(dmpe)~Cl~] CF3S03 (I231, and tr~ns-[Tc(dppe)~C1,1NO,~HNO, (193)reveal the expected distorted octahedral geometry. The search for nonreducible Tc(II1)cations has led to the preparation of a variety of thiolato complexes of the type [TC(SR)~L,I+ (L = depe, +

+

+

COORDINATION CHEMISTRY OF TECHNETIUM

41

dmpe, dppe, diars) (297, 298, 200). The geometry is generally trans, but for dmpe complexes with R being an aryl group the cis-isomer is formed (299).These complexes exhibit a reversible Tc(III)/Tc(II)couple for which E "' spans a range of values from - -200 to -600 mV vs Ag/ AgCl and are thus generally more difficult to reduce than the halide complexes (300). Again, the Re complexes are more difficult to reduce than the Tc analogs (301). A general method of synthesis is by the reaction of the thiolate with ~ ~ ~ ~ ~ - [ T C ~ O ( O H(297). ) L ~ FABMS ](PF~)~ (298,302) and spectroelectrochemical studies have been reported (3031. Crystal structures are available for trans-[Tc(SMe),L,]Y (L = depe, Y = PF,; L = dmpe, Y = CF3S03)(297), trans-[T~(SMe)~(diars)~]PF~ (2001, and cis-[Tc(SPh),(dmpe),IPF, (299). In the cis complex a trans influence is evident with the averaged Tc-P distance trans to P, 2.42(1) A, and that trans to S, 2.49(3) A. A crystal structure of the product of the reaction of tran~-[TcO(OH)(dmpe)~](PF,)~ with Na(S2CNEt2)has shown this to be trans-[Tc"*(scp),(dmpe),](PF,),, where scp represents the zwitterionic ligand -SCH,P+Me2(CH,),P(S)Me2.This unusual ligand appears to be formed by nucleophilic attack of a dmpe phosphorus center on the CS, elimination product of the dithiocarbamate, followed by a molecular rearrangement (300).Dithiolene ligands are well known to stabilize trigonal prismatic geometry, and the structure of the 3,4toluenedithiolato complex [Tc(tdt)(dmpe),]PF, (Fig. 9) shows a mean twist angle of 33(3)", which is about midway between the 60" of ideal octahedral geometry and the 0" of the ideal trigonal prism. Electrochemical and spectroelectrochemical studies show reversible Tc(III)/Tc(II) and Tc(II)/Tc(I)couples at -0.600 and -1.217 V, respectively, and a quasireversible Tc(IV)/Tc(III) couple at 0.680 V vs Ag/AgCl (304). A related cationic complex is [Tc(0-SC,H~0)(dmpe)~lBPh,, prepared by the reaction of 2-mercaptophenol with [Tc(dmpe),Cl,]Cl in ethanol. The geometry of the cation is described as distorted octahedral with a dihedral angle of 1 8 . ~ 3 )between " the TcOS plane and the "trans" TcPP plane containing one P atom from each ligand (305).An interesting series of Tc(II1) complexes based on the mixed bidentate ligands 25 and 26 has been prepared by the reduction of Tc0,- by the ligand and complex formation (306-309). Crystal structures have been reported for the octahedral phenolate [Tc(dppo),] (3071, thiolate [Tc(dppbt),] (307, 308), and propionate [Tc(dppp),].2dmso (306). In each case the three P atoms occupy mer positions. For the amine ligand (25) (R = NH,) an acid-base equilibrium is established and either the triply deprotonated [Tc(dppba),I or salts of the doubly deprotonated [Tc(dppba),(dppbaH)]+may be isolated depending on the pH. The crystal structure of [Tc(dppba),(dppbaH)I-

42

JOHN BALDAS

R = NH2 (dppbaH) R=SH (dppbtH) R=OH (dppoH) R = COOH

Ph2P(CH2),COOH n=l n=2 (dpppH)

E(CH2CH2PPh&

(26)

(27)

E=N,P

(25)

C10, again shows a mer arrangement for the phosphorus atoms. The proton is thought not to be delocalized over the three nitrogen atoms but to reside on the single nitrogen which corresponds to the longest of the 1.948(5)-, 1.979(5)-, and 2.04EK5I-A Tc-N bond distances (309).

FIG.9. The structure of the cation in [T~(tdt)(drnpe)~IPF~ (304).

43

COORDINATION CHEMISTRY OF TECHNETIUM

The corresponding Tc-N-C bond angles of 128.3(4)", 126.6(4)", and 129.1(4)",however, seem t o offer little distinction between neutral and anionic nitrogen. Cationic complexes of the type [TcCl2L1 have been prepared by the reaction of [TcCl,(PPh,),] with the tetradentate ligands (27) (310). +

I. COMPLEXES WITH SULFUR LIGANDS Only complexes in which sulfur ligands form the major part of the coordination sphere are discussed here; other sulfur complexes are described under various headings. Most important is the homoleptic orange-red thiourea complex "l"c(tu),]Cl,, which precipitates in high yield from a concentrated HCUethanol solution containing TcO,- and thiourea (111). The thiourea ligands are readily replaced, making this, and related complexes, valuable synthetic precursors for the preparation of Tc(II1) and other low-valent technetium complexes. Thus, for example, reaction of [Tc(~u)~I(PF,), with CNtBu results in reduction, (111).Crystal structures for giving a 62% yield of [TC'(CN~BU)~]PF, [ T c ( ~ u ) , ] C ~ (111 ~ . ~) H ~ ~[TcLG](PF6),(L = N-methylthiourea, N , N ' and dimethylthiourea) (311) establish approximate octahedral geometry with S-bonded thiourea ligands. Another complexes with a n all-sulfur coordination sphere is the cationic [TcL(SR)~]PF,,where L represents a linear tetradentate thioether. The crystal structure of [TcL(SP~)~]PF, (L = 5,8,11,14-tetrathiaoctadecane) shows the two benzenethiolato ligands to occupy cis positions with the thioether wrapped around the remaining four coordination sites to give strongly distorted octahedral geometry (312). Complexes of the type [TcL'L,], where L represents a dithiocarbamato or xanthato ligand, have been prepared by various routes. Crystal structures for [Tc(PMe2Ph)(S2CNEt2),1(313), [Tc(PMe2Ph)(S2COEt),1 (3141, and [Tc(PPh3)(S2COC4H9),1 (315)show pentagonal-bipyramidal geometry, similar to that of [ T c ( S ~ C N E ~ ~ ) ~ (in C OFig. ) ] 5 , with the monodentate ligand in an apical position. Reaction of NH4{S2P(OMe)2} (Me,dtp) with rner-[TcC1,(PMe2Ph),1 yields orange-red crystals, shown by crystallography to be the octahedral truns(CZ)-cis(P)-[TcC12 (PMe,Ph),(Me,dtp)l(314 1, whereas reaction with Na,(mnt) yields PPh4 [T~(PMe,Ph),(mnt)~1(316). The preparation of neutral [TcL,] complexes with bidentate N, N-substituted benzoylthiourea ligands has been reported (317). The reaction of the sterically hindered anions tmbt and 2,4,6-triisopropylbenzenethiolate (SAr) with [TcrVCl6I2-/L/Zndust (L = MeCN, py, PEt,) in the absence of air yields the diamagnetic [Tc(SAr),L21(318, 319). The crystal structure of the tetramethylben-

44

JOHN BALDAS

zenethiolato complex, [Tc(tmbt),(MeCN),], reveals trigonal-bipyramidal geometry, with the two MeCN ligands in the axial positions and the three S atoms in the equatorial plane, with the bulky aryl groups arranged two above and one below this plane. The orientation of the aryl rings observed in the crystal structure is shown by the 'H NMR spectra to persist in solution. The [Tc(tmbt),L,] (L = MeCN, py) complexes can be oxidized to TcW) 0x0 compounds by oxygen atom transfer from dmso and other oxygen donors, and [TcvO(tmbt),(py)] may be reduced to [Tc(tmbt),(PEt,),I by PEt3. In the oxidation of the MeCN complex, an intermediate Tc(II1) complex was isolated and shown by FABMS and crystallography to be [Tc(tmbt),(MeCN)(dmso)l. A catalytic amount of [TcvO(tmbt),(py)l results in the oxidation of PPh, to OPPh, by dmso via a n oxidative and reductive oxo-transfer cycle, with the catalyst still fully active after 500 turnovers (319). Reduction of Tc04- by S2042- in the presence of CN'Pr and the tetradentate "umbrella" ligand P(o-C6H4SH),(H,L) yields the trigonal-bipyramidal 14electron complex [TcL(CNiPr)l,with the isonitrile in a n axial position [Tc-CNR, 2.06" 8,l. In the presence of a large excess of the isonitrile, a sixth ligand is bound, giving the octahedral 16-electron cis-[TcL (CN'Pr),] [PTc-CNR, 2.058(8) 8,; STc-CNR, 2.081(7) A] (320).

J. NITROSYL AND THIONITROSYL COMPLEXES Reaction of NBu,[Tc"(NO)Cl41 with tmbtH yields orange crystals of the neutral [Tc11'(NO)Cl(tmbt)31. The nitrosyl and chloro ligands occupy the axial positions in the trigonal-bipyramidal structure. The Tc-N and N-0 bond distances are 1.767(6) and 1.150(7) A, respectively, and the Tc-N-0 angle is 176.8(6)".The 4 N O ) IR absorption occurs at 1798 cm-' (3211. A variety of seven-coordinate dithiocarbamato complexes [Tc(NO)(S,CNR,),]Y (Y = BF,, PF6,ClO,) is prepared by substitution of [Tc(S2CNR2),(CO)1with NOBF,. These complexes show 4 N O ) a t 1795-1771 cm-' (2211. The seven-coordinate [Tc(NS)X,(S~CNE~,)~] (X = C1, Br) is prepared by sulfur abstraction from S2C1, or SOC1, (X = C1) and SOBr, (X = Br) by [TcN(S2CNEt2),1.Absorptions at 1248 cm-' (X = C1) and 1250 cm-l (X = Br) in the IR spectra have been assigned to u(NS). Crystal structures for both complexes show pentagonal-bipyramidal coordination geometry with the NS ligand, one halide in the axial positions, and Tc-N-S angles of 177(2)" and 174(2)" for the two independent molecules of the chloro complex and 177.2(7)"for the bromo complex. In [Tc(NS)Br,(S2CNEt2),1the SNTc-Br,,,, bond distance of 2.595(1) 8, is lengthened by a small but significant amount over that of Tc-Br,,, [2.564(1) A1 (322, 323).

COORDINATION CHEMISTRY OF TECHNETIUM

45

VII. Technetiurn(1V)

This oxidation state is intermediate between the low oxidation states stabilized by n-acceptor ligands and the high oxidation states stabilized by n-donor ligands. Thus, carbonyl complexes are unknown and, although bridging 0x0 groups are not uncommon, terminal 0x0 groups are at present unknown. The most useful starting material for the preparation of Tc(1V) complexes is [TCX,]~-(X = C1, Br), but TcO,- or [TcVOX,l- may also be used. A. ISONITRILE AND THIOCYANATO COMPLEXES Complexes ofthe type [TcX,L,] (X = C1, Br) are formed by the reaction of MeCN or CNR with TcX,. The IR spectra indicate that the yellow chloro and red bromo crystalline products are the &-isomers (324). The deep red-violet color (A,,, = 500 nm) produced when TcO,- is reduced in the presence of NCS- or by substitution of [TcX,I2- (X = C1, Br) is now known to be due to [TC'~(NCS),]~-. This anion is reduced by hydrazine to the yellow, air-sensitive [TC"'(NCS)~]~(Tc(IV)/Tc(III), 0.18 V vs SCE). The magnetic moment of 4.1 BM for the purple (ASP~,),[TC(NCS)~I is consistent with an octahedral d3 configuration (2301, and the presence of N-bonded thiocyanate is confirmed by the crystal structure of the octahedral (AsPh,) [Tc(NCS),]-CH,C1, . The Tc-N bond distances are 2.00(1) and 2.01(1) and the N-Tc-N angles are exactly 90". Two Tc-NCS groups are linear and for the remaining four the Tc-N-C and N-C-S angles are 175.9(9)" and 175.3(10)", respectively (325).

1

B. HALIDEAND RELATED COMPLEXES The highest binary chloride of technetium is the dark red TcCl,, formed as the major product of the chlorination of Tc metal (26,326). Crystallography reveals a polymeric chain structure of C1-bridged distorted octahedral TCC& units (327).Reaction of TcC1, with Me,SiBr yields "TcBr," (324).Of great importance and synthetic utility are the stable complex halides [TcX,I2-. Salts of the bright yellow chloro complex [TcC1,I2- are best prepared by prolonged reflux of Tc0,- in concentrated HC1 in order to ensure complete reduction of the initially formed [TcVOCl,l-. Concentrated HBr rapidly yields the red [TcBr,I2-, whereas the deep purple [TcI,12- may be prepared by ligand exchange of the chloro and bromo complexes with HI (26, 27, 328). The white fluoro complex KJTcF,] has been prepared by fusion of K,[TcBr,] with KHF,

46

JOHN BALDAS

(329);a convenient high-yield synthesis is by ligand exchange with AgF in 40% HF (330).All eight possible mixed [TcC1,Br6-,,]'- (n = 1-5) complexes have been separated by ion-exchange chromatography. A notable feature of this work is the use of the greater truns effect of Br compared with that of C1 in order to accomplish stereospecific synthesis. Ligand exchange of [TcBr,I2- with HC1 results in the cis/ fuc complexes for n = 2,3, and 4, whereas ligand exchange of [TcCl6I2with HBr yields the trunslmer isomers (3311. From the LMCT spectra an optical electronegativity value of 2.25 for Tc(IV) is indicated, compared with 2.05 for the less oxidizing Re(IV), and lODq is 28,400 cm-' for [TcF,I'- and 32,800 cm-' for [ReF6I2-(332,333).The force constants for all the [TcX,I'- complexes have been determined (334),and the IR and Raman spectra of the 10 [TcCl,,Br,-, 1'- ( n = 0-6) species, including the pure geometrical isomers, at 80 K have been completely assigned and supported by normal coordinate analysis. Due to the C1< Br truns influence, the force constants indicate that in asymmetric C1'-Tc-Br' axes the Tc-Br' bonds are strengthened by on average 6% and the Tc-C1' bonds weakened by 10% relative to symmetric Br-Tc-Br and C1-Tc-C1 axes, respectively (331). Luminescence spectra for mixed ClBr species have been reported (335,336).Recent peflvalues,utilizing diamagnetic corrections, are in the range 3.34-3.80 BM for M2[TcX,] (M = NH,, K; X = C1, Br, I), (NBU,)~[TCCI,],and (NEt4),[Tc16]a t 300 K (150).In general, EPR spectra are observed only at <5 K ( 4 0 ) . A surprising number of [TcX,]'- salts have been studied crystallographically (38).The anion may either be regular octahedral (cubic crystal class) or have lower symmetry. Single-crystal structures are available for M2[TcC1,] [M = NH, (3371, AsPh, (33811, H,[TcC1,I.9H2O (3391, (NH2Mez)(Me2NHCOMe)[TcC1,1-OPPh3 (2851, M,[TcBr,] [M = NH, (340),H,O (34111,and K2[Tc16](342). Radiolabeling studies have shown that ligand exchange of [TcBr,]'occurs at about 170 times the rate for [TcC1,]'- and that these complexes undergo ligand exchange at 20 to 50 times the rate of the Re analogs (343).The [TcF6l2-anion is remarkably inert and is hydrolyzed only by hot, concentrated alkali (329).Spectroelectrochemical studies show that [TCX6I2-(X= C1, Br) undergoes a reversible one-electron reduction in HX/NaX aqueous media (231).The aquation of [TCXG]~(X = C1, Br) is promoted by UV and visible light (344).Spectrophotometric and paper electrophoretic studies in HX, HClO,, and H2S04 have shown the formation of anionic, neutral, and cationic species together with Tc04-, with the proportions depending on the acid (345-349). The anionic species is considered most likely to be [TcX,(OH,)]-. The octahe-

COORDINATION CHEMISTRY OF TECHNETIUM

47

dral [TcCl,(OH,)]- anion has been identified in the crystal structure a product isolated of [(H30)(15-crown-5)1[TcC15(OHz~l~(15-crown-5), from a solution of TcC14/15-crown-5/CHzC12(324). The red K,[TcCl,(OH)I precipitates from solution in the reduction of KTcO, by HCl/I(350) and the yellow Zn[TcCl,(OH)] and La2[TcC1,(OH)13are formed by reduction of the Tc0,- salts with HC1 (351).Unlike for rhenium, there is no evidence for the formation of [TC~OC~,,,]~-. On heating (NEt4)2[TcBr,] in CF,COOH, the dimer NEt, [Br,Tc(p-Br),TcBr31 is formed. The vibrational spectra have been assigned in D3,, symmetry, with TcBr force constants of 1.045 and 0.80 mdyn k' for terminal and bridging bromide (3521. c . COMPLEXES WITH OXYGEN LIGANDS AND 0x0-BRIDGED COMPLEXES Chemical or electrochemical reduction of TcO,- in aqueous solution or hydrolysis of [TcX,I2- (X = C1, Br) with aqueous ammonia results in a brown-black precipitate of Tc0,-nH20 (n = 2). This precipitate is generally regarded as the "thermodynamic sink" of Tc(1V) chemistry when hydrolysis competes favorably with substitution of Tc(1V) cores. It is, however, a useful starting material (34).A Tc(1V) aqua cation is formed when Tc0,- is reduced in solutions of weakly coordinating acids. The structure is unknown; the usual formulation as [TcO(OH)]+ (353) implies the presence of Tc=O and would seem unlikely in the absence of any characterized Tc'"=O complexes. In this respect it may be noted that the structure of crystalline TcO, consists of linked TcO, octahedra in a n infinite three-dimensional network (3541. A spectroelectrochemical study of the reduction of KTcO, in bicarbonate buffer /CF3S03Na at pH 8 indicates the sequential formation of cationic pink Tc(IV) and blue Tc(II1) carbonate species (355). 1 . Mononuclear Complexes

The unexpectedly high solubility of K,[TcBr,] (in contrast to that of the fluoro and chloro complexes) in methanol appears to be due to partial substitution of Br- by MeO-. By the use of KOR, salts such as K,[Tc(OMe),I and K2[Tc(OCH2CH20)31may be isolated. These complexes show a v(Tc0) IR absorption at 450-460 cm-' (356).A crystal structure of 28, with the zwitterionic tripod ligand Me3N+C(CH,0-), , shows octahedral geometry with Tc-0 distances of 1.987(4)-2.005(4) A. Complex 28 is water soluble and stable at pH > 4 for over 24 hr (357).

48

JOHN BALDAS

(28)

(29)

An unusual phosphine/diolato complex is 29, formed from [TcOCl,Iand o-(dipheny1phosphino)benzaldehyde.The Tc-0 bond distances are 1.95 8, (358).Reflux of (PPh,),[TcC1,] in salicylaldehyde yields PPh,[TcCl,(sal)] (per = 3.8 BM), for which the phenolic and aldehyde Tc-0 bond distances are 1.98(2) and 2.04(2) 8,, respectively (359).The pale yellow oxalato complex ( A ~ P ~ , ) , [ T c ( C ~ is O prepared ~)~I by substitution of [TcBr,]'- in oxalic acid solution. The crystal structure shows six oxygen atoms in distorted octahedral coordination with pseudo 0,symmetry and Tc-0 distances of 1.978(5)-2.001(4) 8, (360).The substitution of [TcX,]'- and the reduction of Tc0,- in the presence of carboxylic, hydroxycarboxylic, and aminocarboxylic acids has been extensively studied and it appears that, in general, the Tc(1V) species formed are dimeric (3611. The reaction of acacH with [TcX,l2- or [TcX,(PPh,),] (X = C1, Br) yields products depending on the reaction conditions, and PPh,[TcX4(acac)], [TcX,(acac),], and [TcBr,(acac)(PPh,)l have been isolated (2421. The [TcX,(acac),l complexes are stable to acid but in alkaline solution undergo loss of halide followed by loss of the acac anions (362).The cationic [Tc(acac),]BF, is formed by oxidation of [Tc(acac),] with [Fe(Cp),l+ (363). 2 . Binuclear Complexes

A novel series of p-0x0 complexes is formed when a starting material such as NBu,[TcOX,I, (NBu,),[TcX,I (X = C1, Br), truns-[TcO2(py),1C1, or TcO,-/BH,- reacts with pyridine or alkyl pyridines either in neat solution or in a noncoordinating solvent (364-367). Crystal structures of the picoline derivatives show that these mixed-valence Tc(III)/Tc(IV) complexes are of the asymmetric [X,L3TcOTcX3L21(30)and dissymmetric [XL,TcOTcX,LI (31) type.

COORDINATION CHEMISTRY OF TECHNETIUM

f' ,'.'

L

L,, 0,

L

49

iiiCl 0

I ,..'

.CI

clz-cl L

(30)

L

=

picoline

(31)

The reaction of [TcOCl,]- with hot picoline results first in the formation of (30) and truns-[TcvO,(pic),]+,with the concentration of the latter remaining nearly constant; this species is not an immediate precursor of the dimeric forms. In the later stages of the reaction the asymmetric form (30)is converted to the dissymmetric form (31).The formation of picoline N-oxide indicates that oxygen atom transfer occurs in the reduction process. Both 30 and 31 are stable in organic solvents at room temperature but undergo equilibration on heating. In o-dichlorobenzene, the process is first order in C1- and requires the presence of free picoline to prevent decomposition (365).In the solid state both forms of pox0 pyridine derivatives have small magnetic moments of -0.9-1.3 BM and v,,(TcOTc) at 726-698 cm-' in the IR spectra. In the electronic spectra three relatively narrow intervalence CT bands appear at about 10,000 cm-' for both forms. X-ray photoelectron spectroscopic analysis indicates that the Tc ions differ by no more than a single oxidation state in both forms (366). The complexes [(T~X(bpy)~}~(p-O)lX~.bpy (X = C1, Br) and [{TcCl(phen),),(p0)]C12.4H20have been prepared and the crystal structures, determined (367). For these complexes the Tc-0-Tc bond angles of 171.6(9)"-173.0(3)"show a slight bending, whereas for 30 (asymmetric) the angle is 176.5(2)"(366)and for 31 (dissymmetric) the two independent molecules in the unit cell have angles of 175.7(9)"and 177.1(9)" (364). Reduction of Tc0,- in the presence of aminocarboxylic and carboxylic acids or substitution of [TcX,12- in aqueous solution leads to the formation of bis(p-0x0) Tc(IV/IV) or Tc(III/IV) dimers. Structural and IR data are listed in Table I and the structure of the oxalato complex is shown in Fig. 10. The four-membered Tc(p-O),Tc ring is near planar in all cases and the short Tc-Tc distances are consistent with a multiple

50

JOHN BALDAS

TABLE I STRUCTURAL AND IR DATAFOR T ~ ( p - 0 ) ~ TCOMPLEXES c Com p Iex

Tc-Tc

(A,

Tc-06, !A1 Tc-0-Tc I") iav.) 1av.l

K ~ I { T c ' " ( C ~ O ~ ~ ~ ~ ~ ~ ~ -2.361(11 O)~I~~H 1.913(1) ~O [{Tc'"iedtaH, l),!p-0)21.5H,O 2.331 Na21{Tc~~nta)t2(p-0),1.6H20 2.36312) Ba,I{Tc"' '"(tcta)t2I p-0)21CI04~9H~02.40211) (32)

1.913 1.919(2) 1.936(7)

75.7 75.2 76.0(1) 76.6(3)

i,(TcO,TcI (IR, cm-', asym, syml

Ref.

730, 401 725, 404 715, 410 734"

368 369, 368 370, 368 371

Raman spectrum of sodium salt.

bond. The edtaH, complex is diamagnetic and extended Huckel calculations have suggested the partly antibonding d r 2 6 * 2configuration of a Tc-Tc single bond for the six metal d electrons, rather than the triply bonded a2r26'configuration (369). This suggestion has been questioned, and it has been noted that a Tc-Tc triple bond with the 6 component weakened to such an extent that its contribution to the overall energy of the Tc-Tc bond is close to zero is consistent with the observed long bond distance (42). A characteristic feature of the electronic spectra of the Tc(IV/IV) dimers is an intense visible absorp-

013

01

FIG.10. The structure of the anion in K4[(Tc(C2O4)~}~(p-O)~I~3H~O (368).

COORDINATION CHEMISTRY OF TECHNETIUM

51

tion at about 500 nm. In the IR spectra the asymmetric and symmetric oxygen stretches of the ring system occur at about 725 and 400 cm-', respectively. Treatment of a solution of [ T C ~ O ( O C H , C H ~ O ) ( ~ C ~ ~ ) ] ~ with BH,- yields, on heating, the deep-blue Tc(III/IV) dimeric anion (32)(3711. The EPR spectrum of the solid shows a [{T~(tcta)}~(p-O)~l~broad signal with the hyperfine splitting expected for a single electron coupled equally between two Tc atoms with spin 8. On oxidation by K2S208,blue 32 is converted to the pink Tc(IV/IV) dimer [{Tdtcta)}, (p-0)2]2-and the reaction is reversed on treatment of the pink dimer with hydrazine. For 32,the Tc-Tc bond distance of 2.402(1)8, is distinctly longer than the Tc-Tc distances of the Tc(IV/IV) dimers in Table I. A polynuclear Tc(1V) citrate complex of uncertain structure has been prepared by substitution of [TcBr6I2-(372). 3. Phosphonato Complexes

Of great clinical importance as skeletal imaging agents are the 9 9 m T ~ complexes of the phosphonates CH2(P0,H2)2(mdpH,) and RC(0H)(P03H2)2(R = H, Me), which localize in bone due to the affinity of the coordinated diphosphonate for calcium in actively growing bone. The radiopharmaceutical preparations appear to be a mixture of oligomers ~ but thought and polymers with the oxidation state of 9 9 m Tuncertain to be +4 (12,19). The crystal structure of the polymeric {[Li(OH,)31[Tc(OH)(mdp)l.jH20},, prepared by substitution of (NH,)2[TcBr61with mdpH,, is shown in Fig. 11.The structure consists of infinite polymeric

FIG.11. A portion of the {[Li(OH~)31[Tc(OH)(mdp)l.fH,0), structure (reproduced from Ref. 373 with permission).

52

JOHN BALDAS

chains, with each mdp ligand bridging two symmetry-related Tc atoms and each Tc atom bound to two symmetry-related mdp ligands. The bridging 0x0 ligand appears to be in the hydroxy form, consistent with a Tc(1V) oxidation state (373). An EXAFS study of the Tc-mdp form of the 99mTc-mdp bone seeking complex in solution indicates a Tc(1V) tetrameric structure, with each Tc having 1.5 +- 0.5 Tc neighbors and surrounded by six singly-bonded oxygen atoms from water or the diphosphonato ligands, and the absence of Tc=O groups (374).Raman spectroscopy of Tc-MeC(OH)(PO,), prepared by BH,- reduction, however, indicates the presence of T-0 and O=Tc=O cores (375) and thus of Tc(V) components in this preparation.

D.

COMPLEXES WITH SCHIFF BASEAND OTHER NITROGEN LIGANDS

A number of Schiff base complexes have been prepared by substitution of [TcCl6I2-or [TcCI,(PP~,)~] (376,269). The reaction of TcC1, with bpy yields [TcCl,(bpy)] (377) and thermolysis of (pyH),[TcCl,]] yields [TcCl,(py),], which has been suggested to be the cis-isomer on the basis of the far IR spectrum (378).Orange-colored [ T C C ~ ~ { H B ( ~isZformed )~)I by the reaction of [TcVOC1,]~/KHB(pz),/HC1 and has a magnetic moment of 3.7 BM, consistent with a d3 configuration (379). The reaction of [TcOCl,]- with aromatic amines and dppe in refluxing alcohols gives the purple air-stable imido complexes [TcCl(NAr)(dppe),]+ in good yield. Paramagnetism is evident in the broadened NMR spectra. The shows that the hycrystal structure of tr~ns-[TcCl(NNMe~)(dppe)~IPF~ drazido(2-) ligand is bonded as a linear four-electron donor (278).

E. COMPLEXES WITH PHOSPHINE AND ARSINE LIGANDS The emerald-green air-stable truns-[TcC1,(PPh3),lis readily prepared in high yield by the reduction of TcO,- with HCl/PPh3 (280). If the reaction is performed in acetone at room temperature, then the salts R[TcC1&PPh3)l(R = PPh,H, orange; AsPh, , yellow) are formed (380). In the case of the more highly reducing PMe,Ph and PEt,Ph, truns[TcCl,L,I is formed if the Tc: L ratio is 1: 5, and mer-[Tc"'Cl,L,] is formed at a 1: 15 ratio. On reflux in CCl,, the Tc(II1) complexes are oxidized to [TcCl,L,I (280). A number of bromo analogs and [TcC ~ , ( A S P ~ ,have ) ~ ] been reported (192, 280, 377). In all reactions of TcO,- with monodentate phosphines the intermediate oxidation states Tc(V1) and Tc(V) are not observed, whereas bidentate phosphines, in general, favor reduction to Tc(II1). The magnetic moments of 3.4-3.8 BM for [TC'~X,L~I are consistent with an octahedral d3 environment

COORDINATION CHEMISTRY OF TECHNETIUM

53

(280). Crystal structures have been reported for (Ph3PCMe2CH2COMe)[TcCl,(PPh,)] (3801, (PEt,H)[TcCl,(PEt,)] (381), and trans-[TcC1,L2] [L = PMe, (3821, PMezPh (3831, PMePh,, PEt, (38111. The truns[TcCl,(PPh,),] complex is a useful starting material for Tc(1V) and Tc(II1) chemistry. On heating in coordinating solvents, such as dmso or pyridine, the PPh, ligands are displaced to yield [TcC1,L21 (L = dmso, py), whereas reaction with pyridine/PPh, results in reduction to "I'cCl&py),I (384 1.

F. COMPLEXES WITH SULFUR LIGANDS The reaction of [TcOCl,I- with a dithiocarbamate (S2CNC4H,0-) results in loss of the 0x0 group to give neutral [Tc(S2CNC,H80),].H20 (3851,and reaction with 2-mercaptopyrimidine (mcpH) gives the structurally characterized NBu,[TcCl,(mcp)l (386).A cationic complex is the blue-violet paramagnetic [Tc(S,CNEt,),(PMe2Ph)1PF,, formed by air oxidation of [TC~~'(S,CNE~,),(PM~,P~)] in the presence of HCl(387, 388). Substitution of [TcBr,]'- with Na,(mnt) in ethanol yields (AsPh,), [T~'~(rnnt),l. The crystal structure shows that Tc is coordinated to six S atoms with chelate twist angles of 32.6'-39.0', which are intermediate between the value of 60" for a regular octahedron and 0" for a trigonal prism (3891. Reduction of TcO,- by 1,2-benzenedithiol /HCl yields on standing the wine-red dimer [Tc,(bdt),] (390,391). The crystal structure (Fig. 12) shows each Tc atom coordinated to a trigonal prismatic array

FIG. 12. The structure of [Tcz(bdt)*I(391).

54

JOHN BALDAS

of six S atoms with a shared quadrilateral face and with the eight S atoms delineating a rhombohedral prism. An arrangement in which two bdt ligands span the rhombohedral faces and two span opposite edges is found rather than a "paddle wheel" with the bdt ligands spanning the four vertical edges. The Tc-Tc distance of 2.591(3) A and the d3-d3 configuration would seem to indicate a multiple bond, but any assignment needs to consider the noninnocent nature of the dithiolene ligands (3911. The dark-green dimer [Tc2(edt),(SCHCHS),1has been isolated from the reaction of 1,2-ethanedithiol with [TcC1,I2- (379).The structure is similar to that of [Tc,(bdt),l, with a Tc-Tc distance of 2.610(3) A. A novel feature is the dehydrogenation of (SCH,CH,S)2t o (SCHCHS)2-to give a mixed dithiolate-dithiolene coordination, with each dithiolene ligand coordinated to one Tc atom only and each S atom of the dithiolato ligands coordinated to both Tc atoms. VIII. Technetium(V)

The chemistry of this oxidation state is dominated by complexes containing oxygen and nitrogen multiple bonds. This is a reflection of both the tendency of high oxidation states to induce deprotonation of aqua or amine ligands and the ability of good n-donors such as O2 and N3- to stabilize high oxidation states. The greater ease of reduction of Tc in comparison with Re is seen for Tc in the absence of analogs of the large number of [Re0I3+complexes with monodentate phosphines of the type [ReOX3(PR3),](189). Otherwise, the chemistry of Tc(V) resembles that of Re(V), with the predominance of complexes based on the [Tc0l3', [TcO,]', [ T c , ~ , ] ~ 'and , [TcNl" cores. Tc(V) complexes not containing a multiply bonded oxygen or nitrogen ligand are relatively few. The only binary halide is the yellow TcF, (m.p., 50"C), formed as a by-product of the direct fluorination of Tc metal (392).The complex fluorides M[TcF,] (M = Na, K, Rb, Cs) have been prepared by the reduction of TcF, in the presence of MC1 and IF5 and the rhombohedral unit cell parameters determined (393). In NO[TcF,], the presence of the free NO+ cation results in an IR absorption at 2315 cm-' (394).

A. MONONUCLEAR [Tc013+COMPLEXES The structure and chemistry of the square-pyramidal five-coordinate and pseudo-octahedral six-coordinate [TcV0I3+complexes are dominated by the strong tetragonal distortion induced by the multiply bonded 0x0 ligand. The d orbital energy levels in C,, symmetry are in

COORDINATION CHEMISTRY OF TECHNETIUM

55

the order b2 (dx.y)< e (dx,, dyz)< bl (dx2-y2)< a, (d,d (35,395).The d2 electrons are paired in the low-energy, essentially nonbonding, b2 (dxy) orbital, resulting in complexes with a 'A, ground state, which are either diamagnetic or show only temperature-independent paramagnetism (150).The [TcOI3+core may thus be regarded as a "closed shell" electronic configuration and complexes expected to be relatively kinetically inert to substitution, but this is dependent on the nature of the coordinated ligands (35).The TcO bond is formally triple with one cr and two 7~ (0 px, p,,/Tc d,,, dyz)components, but because of the unfavorable charge distribution in T c - d + , the bonding will be intermediate between triple and double. The strong trans influence of the 0x0 ligand results in the trans ligand being only weakly bound and often absent and the Tc atom being raised above the square basal or equatorial ligand plane. In complexes in which the trans ligand is present, the Tc-Lt,,, distance may be 0.1-0.2 A longer than that for the same ligand in a n equatorial position. An aqua cation of the type [TcO(OH2),13+,or of polymeric forms, is not found because [Tc0l3+is unstable to disproportionation to Tc(1V) and TcO,- (35).When the [TcO13+core is stabilized by suitable ligands, kinetically stable and substitution inert complexes result. A general route to [TcO13+complexes is by substitution of [TcOX,]- (X = C1, Br) (35,396).The NBu4[TcOC1,1salt is conveniently prepared in 99% yield from TcO,-/HCl and is readily soluble in polar organic solvents such as methanol, acetone, or MeCN (397).An alternative method is by the reduction of Tc0,- in the presence of the ligand(s1. A variety of reducing agents has been used, of which sodium dithionite is convenient and popular, but reduction to a lower oxidation state may also occur. The in uitro stability of [TcO13+complexes has been related to the solid angle factor sum of the coordinating atoms (398). 1 . Cyano and Thiocyanato Complexes

Green-yellow K2[TcVO(CN),l.4H20is formed in low yield from the reaction of Tc02.nH20with KCN or from aerobic crystallizations of K,[TC~~'(CN)~].~H,O. In the IR spectrum v(Tc0) occurs at the rather low value of 910 cm-' and three u(CN) absorptions are observed at 2095,2080, and 2035 cm-', which have been assigned to the 2Al + E modes in C,, symmetry. The lilac (NBu4),trans-[TcO(OMe)(CN),3is formed on substitution of NBu,[TcOCl,] with CN- in MeOH [u(TcO) at 932 cm-'I (229),and (NMe4)trans-[TcO(OH2)(CN)4].2H20 has been isolated from the protonation of [TcO2(CN),I3- (399).The strong trans labilizing effect of the 0x0 ligand is apparent in the rapid rate of exchange of the trans water for NCS-, for which the forward rate constant is 22 M-' sec-' at 25°C. The crystal structure of (bpyH)2trans-[TcO-

56

JOHN BALDAS

(NCS)(CN),] shows N-bonded thiocyanate and a short Tc=O bond dis(399). Substitution of [TcOCl,]- with NCS- gives tance of 1.612(8)i% a high yield of the bright-red (ASP~,)~[TCO(NCS)~I. In the presence of NCS- this complex is easily reduced to mixtures of [ T c ~ ~ ( N C S and )~]~[TC"1(NCS)6]3- (400).

2. Halide Complexes When TcO,- is added to concentrated HC1 at room temperature, a ~C formed l ~ l ~ and - , is yellow solution, thought to contain f u ~ - [ T c ~ ~ ~ O is then converted to an olive-green color on reduction to [TcVOC1,1- (35). If the solution is heated, the kinetically controlled product [TcOCl,]undergoes further reduction to the yellow thermodynamic product [TC'~C~,]~-. These steps are described by the equations Tc0,-

+ 3HC1

+ 3HCl + H30' [TcOCIJ + 3HC1-

[Tc03C1,]'-

-

[TcO3C1,I2-t H30+ [TcOCl41-t 3H20 t CI,

[TcClJ-

+ H30' + 112 Clp.

With concentrated HBr as the reductant, the preparation of [TcOBr,lis performed at, or below, 0°C to avoid reduction to [TCBr6I2- (396). The reduction of [TcOB~,,,I-'~-in 8.7 M HBr proceeds by a combination of first- and zero-order reactions (401). The product isolated on addition of cations to solutions of [TcOX,]- (X = C1, Br) in HX is dependent on the nature of the cation. Large cations such as NBu,' result in the small cations such precipitation of the five-coordinate R[TcOX,] (402); as NH,+, K+, or Cs+ result in the six-coordinate M2[TcOX,l (403,4041, and with NEt4+ the trans-aqua complex NEt,[TcO(OH2)Br,l has been isolated (405).These results show that the trans ligand is labile and indicate that crystal packing forces determine the composition of the solid form. In aqueous HX solution the most likely form is [TcO(OH,)X,]- (but is generally written simply as [TcOCl,l-). For [TcOCl,]- in 12 M HC1 the equilibrium

has been demonstrated by Raman spectroscopy and [TcO(OH,)C1,1was found to predominate by a factor of about 60. In CH2Cl2solution the equilibrium constant is ca. 400 times larger, indicating the equilibrium [TcOCl41- + C1-

[TcOC1,12-,

57

COORDINATION CHEMISTRY OF TECHNETIUM

with the trans position in [TcOCl,]- either vacant or containing an only weakly interacting CH2C12molecule (404). In water, [TcOCl,Idisproportionates to Tc02.nH20and TcO,- in the reaction 3Tc(V) + 2Tc(IV) + Tc(VII1, whereas in 1M p-toluenesulfonic acid Cs2[TcOC1,] dissolves to give a brown Tc(1V)cation and TcO,- (35,406).This disproportionation is very slow in >2 M HC1 solutions (35). Salts such as NBu,[TcOX,] (X = C1, Br) may also be prepared directly from NBu4[Tc041and HX (407) and NBu4[TcO141by ligand exchange of NBu4[TcOC1,l with NaI in acetone (408). Structural and v(Tc0) data are listed in Table 11. The first structural characterization of the [TcOCl,I- anion in the [N(PPh3)21+salt showed only approximate C2, symmetry (402).This distortion is a consequence of the presence of the large cation in the crystal because in AsPh,[TcOX,] (X = C1, Br) the anions possess ideal C4,symmetry (409,410) and C,, symmetry for the anions is also indicated by the vibrational spectra of NBu4[TcOX41(X = C1, Br, I) (407,408). In the square-pyramidal five-coordinate [TcOX41complexes, the 0x0 ligand is in the apical position and the Tc=O bond distance is rather short at 1.60-1.62 A. Structurally, the square-pyramidal five-coordinateand octahedral six-coordinate complex anions are dominated by the trans influence of the 0x0 ligand, which results in the displacement of the Tc atom above the square basal or equatorial plane and, in six-coordinate complexes, the weakening of the bond trans to the 0x0 ligand. This trans bond weakening is indicated by the long Tc-OH, bond distance of 2.317(9) A in (NEt,)trans-[TcO(OH2)Br41 (405). In the five-coordinate complexes, the trans influence may be regarded as sufficiently large to prevent the bonding of a trans ligand. The TcO IR stretching frequency is sensitive to the presence and nature of the trans ligand. For [TcOX41- (X = C1, Br, I) this absorption TABLE I1 STRUCTURAL AND IR DATAFOR [Tc0l3+HALIDECOMPLEXES IAI

CampI e x

Tc=O 1b .

(NlPPh312tlTcOC141 AsPh,[TcOCI,l AsPh,lTcOBr,l NEt,[TcO(OH2)Br,l Cs21TcOCI5lb

1.610141 1.593(8) 1.61319) 1.618(9) 1.65

2.305 av. 2.309(21 2.460111 2.507(11 av. 2.36,iS

Cs,lTcOBrJ

1.66

2.50trnns 2.54cm 2.74hIlS

Tc-X

0-Tc-X

(A)

ulTcO) (cm-')

Ref.

("I

GTcX4'

103.2, 110.4

0.66 0.67 0.70

1016 1025

0.37

1000 954

402 409 410 405 404

952

404

106.8 106.6 99.5, 97.6

Displacement of Tc above the square basal or equatorial plane. Bond distances from solid-state EXAFS spectra.

-

58

JOHN BALDAS

occurs at 1025-1000 cm-', whereas the presence of trans halide in Cs,[TcOX,] (X = C1, Br) results in a substantial lowering in energy to 954 cm-' (404). The value of 992 cm-' reported for M2[TcOC1,] (M = NH,, K)in the solid state, however, indicates that the nature of the cation is important (4111. Normal coordinate analysis of NBu,[TcOX41 results in force constants of 8.41, 8.39, and 8.04 mdyn k' for X = C1, Br, and I, respectively, (412). These values, when compared with 8.61 and 8.55 mdyn k' for Ru=N in AsPh,[RuNX,l (X = C1, Br) (4131, indicate considerable triple bond character for [TcO13+.A bond order of 2.55-2.59 has been calculated for M2[TcOC1,1 (414)and the value will be higher for [TcOX41-. The NBu,[TcOX,l (X = C1, Br, I) and M,[TcOCl,] (M = NH,, K)salts are diamagnetic at 80-300 K, which is consistent with an 'A, (b;) electronic ground state (150,395).Three d-d bands in the electronic spectrum of (NH,),[TcOC1,1 in HC1 a t 10,700 ( E = 181, 16,700 (6), and 20,600 (24) cm-' have been assigned to ' E (b,e) t 'A,, 'A, (b,b,) + 'A,, and 'B2(b2al)+ 'A, transitions, respectively (414). Recent L-edge spectra of [MoO13' complexes, however, show that the assignment of 'B, + 'A, for the 20,600-cm-' peak in [TcOCl,12- is most likely incorrect and that this transition is more likely dxy+ dx2-y2in nature (415). Brown, thermally stable TcOC1, and grey-black TcOBr, have been prepared by chlorination or bromination of TcO, .The chloro compound is very readily hydrolyzed by water to Tc02-nHz0and Tc0,- in the ratio 2 : 1 (416). Water-sensitive [TcOX,(bpy)] (X = C1, Br), [TcOC13(phen)l.H,0 and [TcOCl,(OEt)(bpy)l are prepared by substitution of [TcOX,]- in ethanol/HX. The v(Tc0) IR absorptions of 910-850 cm-l for [TcOX,L] and 922 cm-' for the ethoxy complex indicate a somewhat lower TcO bond order than that for [TcOX,l- or [TcOX,12(417). Another example is [TcOCl,(terpy)], for which the terpy ligand is thought to be bidentate (418).Interesting related complexes are (AsPhJmer-[TcOX,(hbt)l (X = C1, Br), prepared by substitution of [TcOXJ [v(TcO)at 945 cm-', X = C1; 940 cm-', X = Brl. The structure

59

COORDINATION CHEMISTRY OF TECHNETIUM

of the chloro complex (33) shows a Tc=O bond distance of 1.650(6)8, and a trans OTc-Ophenolic distance of 1.948(4)8, (419). 3. Complexes Based on the TcO{04}, TcO{S4}, TcO{O4-,,S,}, and TcO{Se4}Cores

Square-pyramidal complexes of the type [TcOL4]-,where L is an 0, S, or Se ligand, are readily prepared either by substitution of [TcOC14]or by the reduction of Tc04- in the presence of the ligand. A large number of complexes have been reported, mainly with bidentate ligands. Structural data and v(Tc0) values are summarized for representative complexes in Table 111. General features are square-pyramidal geometry with the 0x0 ligand in the apical position, e.g., 34, a T-0 bond distance of 1.63-1.67 A, and a considerable displacement of Tc by 0.70-0.88 8, above the square basal plane. The anionic ligands effectively neutralize the positive charge on [Tc0I3+and the position trans to the 0x0 ligand is usually vacant. The TcO{O,} complexes such as M[TCO(OCH~CH~O)~] (M= Na, NBu,) are relatively weak and hydrolyze in the absence of excess diol (420, 433).The catecholate complex [ T C O ( O ~ C ~ His, ~ )however, ~Imore stable and may be prepared by the addition of a stoichiometric amount of TABLE I11 STRUCTURAL AND IR DATAFOR [Tc0I3+C O M P L E X E S WITH 0,s, OR Se LICANDS Complex

Tc-L tAl iav.1

TcO 11%)

TcOiO,\ 1.64861 NBu,lTcOto-O,C,H,I,I N B u ~ 1 T c O l o - O ~ C ~ H ~ N O , ~ ~ I 1.634(4J NBu~[TcOlo-O,C,C1,l~1 1.646i5) iAsPh~~~1TcOiox~~lHox~l~3H~0 1.64016) TcOIS,j NBu,lTcOiSAr),l' AsPh~lTcOiedtl~l NBu,ITcO~SCH~COS)~I AsPh,lTcOtSCOCOSl,l AsPh41TcOlmntlpl NBu,lTcOLzl' A~Ph~lTcOIbdtl~l

1.659i11J 1.64(11 1.672181 1.646(4) 1.655(6) 1.672(61 1.658i5J

TcO{S,O,t AsPh,lTcOtSCH,CH,0121

1.662i5)

TcO{Se,t NEt,lTcOlSe,CCtCN),~~l

1.67(2)

_______~_____

1.957i3l 1.966 1.955 2.016cis

2.380

2.300 2.320(31 2.329ill 2.315(1J 2.316 2.315(2J

0 1.950(41 S 2.291(2) 2 47114) ~

Displacement of Tc above the square basal plane. Ar = 2.4.6-trimethylphenyl. 'L = SCHlCOOMelCH1COOMe)S.

6sbp'

IAJ

0.701 0.695 0.25

utTcOI iIR. cm-'I

Ref.

970 983 969 985

420 421 422 423

940

0.846 0.761 0.791 0.759 0.742 0.78 0.732

950 940 938

424 425 426 427 428 429 430

0.720

948

431

0.88

965

432

950

60

JOHN BALDAS

catechol to [TcOCl,]- (420).Substitution of [TcOCl,]- with oxalic acid yields pale-green crystals of the AsPh,+ salt of an oxalato complex with v(Tc0) at 963 cm-'. Recrystallization from acetone-water containing oxalic acid results in the isolation of emerald-green (AsPh,),[TcO(ox),(Hox)].3H20(35) with v(Tc0) at 985 cm-'. The structure (Fig. 13) is unusual, with a monodentate protonated oxalate coordinated cis to the 0x0 ligand (423).Also unusual is the absence of a significant trans influence of the 0x0 ligand, with the oxalate Tc-Otransdistance of 2.069(6)! Ibeing similar to 2.016 8, (av.)for Tc-OCis.The low susceptibility of oxalate to the trans influence has been noted for [MoO13+complexes but the reason is unclear (434). It is likely that the form in solution is the truns-aqua complex and that the crystallization of 35 is the result of crystal packing effects. The gluconate and heptagluconate complexes are of uncertain structure but thought to be [TcOL,I- from IR and Raman evidence (19).The 9 9 m Tcomplexes, ~ and 99mTc-diethylenetriaminepentaacetate [of unknown structure but the oxidation state is probably Tc(V)I,are useful as kidney and brain imaging agents (19). ,01

Y

c4

FIG. 13. The structure of the anion in (AsPh,),[TcO(ox)2(H~~)]~3Hz0 (36)(423).

COORDINATION CHEMISTRY OF TECHNETIUM

61

The preparation of a variety of [TcO(P-diketonate),Cl] complexes has been reported (435). Pertechnetate is reduced by thiols in the presence of acid in a firstorder process in TcO,- and the thiol (436,437). The kinetic data for a series of p-substituted benzenethiols follow the Hammett relationship with a decrease in rate by more electron withdrawing substituents (438). TcO{S,} and related complexes are generally prepared from TcO,by the use of a reducing agent such as S2042-or by ligand exchange (439-443).There is now a considerable variety of TcO{S,} complexes; structurally characterized examples are listed in Table 111. Four thiolato ligands effectively satisfy the charge on the [TcO13' core to give square-pyramidal complexes that show little or no tendency to bind a sixth trans ligand. In general, these complexes are highly stable and substitution inert; for instance, [TcO(edt),I- is unaffected by PPh, in refluxing MeCN. Electrochemically, there is no tendency to oxidation to Tc(V1) (4411. The magnetic moments of a representative series have been shown to be field-strength dependent and lie in the region 0.1-1.5 BM. The frequency of v(Tc0) is -20 cm-' lower than that in the [Re0I3+ analog, and an LMCT band a t 330-450 nm in the electronic spectra of TcO{S,} complexes is at lower energy than that for Re (441).An interesting complex is NBu,[TcO(SCH2COS)21,prepared by use of commercial HSCH,COOH, indicating the presence of a significant content of HSCH,COSH as an impurity (441). The complex NB~,[TCO(S,MO)~] shows a low v(Tc0) at 895 cm-' and undergoes reduction by PPh, to give a product formulated as [Tc'V(PPh3)2(S4Mo)2(H20)l, but which is possibly a hydrate (444).Reduction of Tc04- by tetramethylthiourea/HCl yields [TcO(trntu),](PF,), , a labile complex useful for liFrom the reaction of [TcO(tmtu),](PF,), gand-exchange reactions (445). with dppe in dmf solution one of the products isolated has been shown by crystallography to be [TcO(~~~U)~(M~~NCS,)I(PF,)~, in which the dithiocarbamato ligand is presumably derived from tetramethylthiouA radiopharmaceutical rea by sulfur transfer and loss of NHMe, (196). for tumor imaging is [99mTcO(dmsa)21(dmsaH, = rneso-dimercaptosucThe 9 9 m Tcomplex ~ has been shown from 'H NMR and cinic acid) (446). chromatographic studies with [TcO(dmsa),l- and the dimethyl ester to be a mixture of three stereoisomers (446,447).The crystal structure of the ester NEt,[TcO{SCH(COOMe)CH(COOMe~S}21 has shown the product isolated in 21% yield to be the syn-endo form (429). The important 99mTc-dmsarenal agent is thought to contain Tc in a lower oxidation state, possibly Tc(III), but the structure is unknown (19, 446). The potential for different chemical behavior at the 9 9 m Tand ~ 99Tc concentration levels is illustrated by the mnt ligand, for which the

62

JOHN BALDAS

product at macroscopic concentration is [TcvO(mnt),1-, whereas the product at the 9 9 m Tlevel ~ is [99mTcrV(mnt)3]2~ (448).The preparations of a variety of [TcO13' complexes with -S(CH,),X(CH,),S- (X = 0, S)/ -SAr ligands (449-4501, of TcO-metallothioneins (451), and of the three possible NEt4[TcO{XYC=C(CN)2}21(X,Y = S or Se) complexes (452) have been reported. The 0x0 ligand in [TcO{S(CH2)20(CH2)2S}(SAr)l is removed by PPh, at room temperature (450). 4 . Complexes Based on TcO{N4},T c O { N 4 ~ , , 0 ,and } , TcO{N$,} Cores

The discovery that the neutral, lipophilic 9 9 m T ~complex 0 36, prepared by the reduction of TcO,- in the presence of the tetradentate propyleneamine oxime ligand, is able to cross the blood-brain barrier in both directions has stimulated much work in this area (19).

(36)

(37)(d, 1 )

A large number of variously substituted analogs were prepared and the d,Z stereoisomer (37) was found to be sufficiently retained in the brain due to transformation t o a more hydrophilic species, which is then unable to diffuse out of the brain. The 99mTccomplex (37) is now an important radiopharmaceutical for cerebral perfusion imaging and the evaluation of stroke (19).Crystal structures of 36 and meso-37 and a variety of analogs have been reported (453-455). The [TcO13+core is sufficiently electron deficient to deprotonate secondary aliphatic amines, and in 36 and analogs neutrality is achieved by loss of both amine protons and one oxime proton, with the remaining oxime proton being intramolecularly hydrogen bonded. Features of the structures are T-0 bond distances in the range 1.670(4)-1.682(5) A and the displacement of Tc above the N, plane, which for 36 is 0.678(1) A. Also, for 36 the distance between Tc and the deprotonated N (imino) atoms of 1.913 A (av.) is considerably shorter than the Tc-N(oxime) distance of 2.090 A (av.) (453).The u(Tc0) IR absorption in the rather

COORDINATION CHEMISTRY OF TECHNETIUM

63

low range of 934-908 cm-' is consistent with the long TcO distances. A study of analogs of 36 with the aliphatic N(CH2)3Nchain replaced by two, four, and five carbons has shown that for the four- and fivecarbon chains both the five-coordinate monooxo and the six-coordinate truns-[TcO,]+ complexes are formed (455). Reduction of Tc04-/l,2diaminobenzene (pdaH,) by S2042- allows the isolation of diamagnetic NBu4[TcO(pda),](456).A single v(NH) confirms the deprotonated form of the ligands and the low v(Tc0) at 891 cm-' is consistent with coordination by four -NH groups with Tc-N 1.98 A. The Tc=O distance is 1.668(7)A and Tc lies 0.67 A above the N4 plane. Another complex of this group is [TcO(octaethylporphyrinate)]OAc (457). A variety of five- and six-coordinate structurally characterized complexes with mixed TcO{N,-, 0,) and related cores have been reported. The only seven-coordinate complex is [TcO(edta)l-, prepared by the reaction of [TcOCl,]- with edtaH, in anhydrous dmso. A crystal structure of the barium salt shows distorted pentagonal-bipyramidal geometry, with the 0x0 group and the two N atoms bound equatorially (458). The six-coordinate Schiff-base complex truns-[TcO(OH,){(acac),en}]X has distorted octahedral geometry, a Tc=O bond distance of 1.648(2) A, and Tc 0.39 A above the N202plane. As expected, the trans influence of the 0x0 ligand results in a long Tc-OH, bond distance of 2.282(2) A. Similarly, in truns-[TcO{(sal),en)C11 Tc-C1 is long at 2.527(4) A and Tc is displaced by -0.27 A (459). A structurally characterized sixcoordinate 8-quinolinolate complex is [TcO(Ophsal)(quin)](460). In the orange phenolic derivative NBu,[TcO(epa)l.H,O [epaH, = N,"-ethylenebis(2-phenoxyacetamide)], both the amide and the phenolic groups are deprotonated and v(Tc0) occurs at 925 cm-'. The average Tc-N distance is 1.977(6) A and Tc lies 0.65 A above the N202plane (4611.

(38)

The novel neutral complex 38 is formed by substitution of [TcO(OCH,CH,O),]- or [TcOCl,]- (462,463). Triple deprotonation of the starting ligand includes the loss of an amine and pyrrolic proton. The 99mTc-38 complex is undergoing clinical evaluation for efficacy in the detection and determination of the severity of stroke and illustrates the degree

64

JOHN BALDAS

of ligand design undertaken t o achieve the desired in vivo behavior. The Tc-Npyrrole bond distance of 1.993(4)8, is, as expected, rather longer than the TC-N,,,,~,,distance of 1.897(4)A. Other structurally characterized six-coordinated complexes are [TcOL] (L = ONNNO Schiff base) (464), the unusual [TcO(apa)l, where apa represents a pentadentate ONNNO Schiff-base ligand derived from dehydroacetic acid (4651, and [TcOL(sal)][L = N-salicylidine-D-glucosamine(2-)] (466).The last complex precipitates from a methanol solution of the glucose derivative and [TcOCl,]-. The presence of salicylaldehyde in the product does not appear to be the result of hydrolysis of the Schiff base prior to coordination. The bidentate coordination of the salicylaldehyde(-1) anion is unusual. The preparation of a variety of complexes with heterocyclic N,O and other ligands (467,468) and of salicylidine Schiff-base complexes with amino acids has been reported (469). Reaction of NBu,TcO, with 25 (R = NHJ under rigorously controlled conditions in ROH, to avoid further reduction to Tc(IIIj, yields [TcVO(OR)Lzl [R = Me, Et; L = 25(1-)].The crystal structure of the methoxy complex shows approximate octahedral geometry with an NNPP equatorial plane, a Tc=O bond distance of 1.700(8)A, a Tc-OMe bond distance of 1.999(8)8,, and an O=Tc-OMe angle of 158.3(3)".The trans alcoxy group accounts for the stability of these complexes and results in low v(Tc0) values of 878 (R = Me) and 857 cm-' (R = Et) (470).

5. Complexes Bused on TcO{N,-,S,} Cores The search for neutral, lipophilic 9 9 m Tcerebral ~ perfusion imaging agents has led t o the intensive investigation of the chemistry of [TcO13+ with bisaminedithiolato (BAT) ligands (17,191. The NzSzcoordination results in highly stable complexes in which neutrality is achieved by deprotonation of one of the amino groups. A common method of preparation is by the reduction of TcO,- with Sz02-in the presence of the 1igand . If one of the amino groups is substituted, then syn/unti isomerism is possible. For 39 (R1 = Me, Rz = HI, crystal structures of both isomers have been determined and the major product has been shown to be the syn form (with the methyl group pointing in the same direction as TcO). The Tc-N bond distance to the anionic nitrogen is 0.288(9) and 0.198(9) 8, shorter in the syn and anti forms, respectively, than the corresponding Tc-NMe bond distance (471). In syn-39 (R1 = Et, R2 = H), the Tc-N and Tc-NEt distances are 1.921(2)and 2.224(2) A, respectively (4721. A large number of variously imaginatively substituted BAT ligands have been synthesized and the "Tc and 9 9 m Tcomplexes, ~ prepared (473-478). In general, 'H and 13C NMR are useful for stereo-

COORDINATION CHEMISTRY OF TECHNETIUM

(39)

65

(40)

chemical assignment of the 99Tccomplex and the stereochemistry is complex important in determining the level of brain uptake of the 99mT~ (477).Complexes include those for which a benzene ring forms part of the ligand skeleton (474)and for which R1 is a steroid moiety (475). An interesting example is 39 (R1= H, R2 = CH2-NC5HgPh),containing a pendant phenylpiperidine group for which the crystal structures of the syn and anti forms (with respect to R,) are available and differences in the brain uptake of the 9 9 m Tcomplexes ~ are found (477).Other crystal structures, including 40 (478),have been reported (474).The complexities of in uiuo behavior are illustrated by 99mTc-40,which is retained in the brain on trapping by enzymatic hydrolysis of one ester group to the free acid and the formation of a charged species. The hydrolysis is stereospecific and only the L,L enantiomer is trapped (19).Additionally, high brain uptake appears limited to humans and primates, presumably due to the high serum and lower brain esterase levels in lower animal species (17).Cationic BAT complexes have been prepared with or without alkylated amine groups and these are of interest as potential myocardial imaging agents (476).A number of crystal structures are available (476,479-481 1. Cationic NzSz complexes such as the six-coordinate tr~ns-[Tc0(0H,){(sacac)~en}]C1 are formed with imine nitrogen ligands. The v(Tc0) absorption in this complex occurs a t 964 cm-' and the Tc-OH, distance is quite long at 2.384(3) A (482). The greater acidity of the amide protons in diamidedithiols results in the loss of two amide protons, producing anionic complexes for which the 9 9 m Tpreparations ~ are of interest as renal agents (19). Yellow salts of the parent complex (41), and derivatives, may be isolated from the in situ hydrolysis of the S-protected ligand and Tc04-/Na2S204(483). Crystal structures of (AsMePh3)(4l)and the PPh4+salt of the butanediamine derivative show the usual square-pyramidal geometry, with the Tc atom displaced 0.771 and 0.67 A, respectively, above the N2S2 plane (484,485).The preparation of a variety of substituted analogs of 41 (and the monoamides) and crystal structures have been reported

66

JOHN BALDAS

(41)

(42)

(486-489). In a novel example the CH7CH7 - - bridge in 41 is replaced by a ribonucleoside (490). Crystallography and NMR have been used to assign stereoisomers (487, 491 1. In the reaction of [TcOCl,]with excess ligand, the blue lantern dimer (A~P~,),[(TCO)~(SCH,CONH(CH2)2NHCOCH2S},],with each Tc coordinated by four S atoms and an intramolecular Tc...Tc distance of 7.175 -81, is formed. In aqueous basic solution the dimer is immediately and quantitatively converted to 2 eq. of (AsPh4)(41)(492). If one of the thiolate groups in 41 is substituted by -CH,CH,-N-(piperidinyl), then a neutral complex is formed. This complex readily undergoes S-CH, bond cleavage in solution, assisted by neighboring group participation of the piperidine nitrogen, to reform 41 (493). The neutral six-coordinate D-penicillaminato (pen) complex (42) contains one bidentate ligand, one tridentate ligand with a Tc-Ocarboxylate distance of 2.214(4) A, and one free carboxyl group (494). The [TcO-(~-pen)(~-pen)lanion is fluxional in solution and racemizes by exchange of bonded and free carboxylate groups trans to the 0x0 ligand. Racemization of the Tc complex is faster than that for the Re analog (495). Other structurally characterized examples with bidentate NS ligands are NBu,[TcO(abt),l (496) and a cationic [TcOL,ICl [L = substituted (thiocarbamoyl)benzamidinate] (497).The preparations of a variety of Schiff-base dithiocarbazate derivative and N-heterocyclic thiolato complexes have been reported (498-500). An N3S complex is [TcO(MAG,)I- (431, for which the negative charge is achieved by deprotonation of the three amide groups of the mercaptoacetyltriglycinato ligand. In uiuo the carboxylic acid group is ionized and [99mT~O(MAG3)12is a n important radiopharmaceutical for the assessment of renal function. The presence of the uncoordinated carboxylate group in the dianion is important for efficient renal clearance (19).The crystal structure of AsPh4[TcO(MAGJ1 shows the carboxylic acid group to be distant from the Tc center, and two crystal forms of the methyl ester AsPh,[TcO(MAG,Me)l differ with the orientation of the carbomethoxy group being approximately parallel and perpendicular to the Tc=O bond (501).Calculations indicate that in solution

67

COORDINATION CHEMISTRY OF TECHNETIUM

(43)

(44)

[TcO(MAG3)I2-is conformationally flexible (5021. A variation of N3S coordination is the inclusion of one pyridine nitrogen in the neutral 44. The Tc=O bond distance is 1.653(4) A, and Tc-Npyridine at 2.102(4) A is substantially longer than the Tc-Namid,distances [1.965(4)A (av.)l (488).A complex with NS3 coordination is [TcO(tmbt),(py)l (319). 6. Complexes Based on Other TcO Mixed Ligand Cores Numerous five- and six-coordinate TcO complexes containing mixed ligand atom coordination are known and many have been structurally characterized. Examples not containing sulfur are [TcO(OR)X2L,l(R = Me, Et; X = C1, Br; L = pyN02 or L2 = bpy, polypyridyl derivative) and [TcOC12(terpy)lTc04 (503-505, 418). For trans(N)-trans(Br)[ T ~ 0 ( 0 E t ) B r ~ ( p y N Othe ~ ) Tc-OEt ~], distance is short at 1.855(6)A and Tc=O is 1.684(6)A. The ethoxy group results in a low v(Tc0) at 938 cm-' (503).Reduction of Tc04- by HX /KBH4in the presence of HB(pz),yields the lipophilic [T~0Cl~(HB(pz)~}l (5061, and the bromo complex may be prepared from [TcOBr41-(396).In the chloro complex, the three N donor atoms span fac positions with OTc-N,, bond distances of 2.086(4) and 2.088(3) A, and OTc-N,,,, is markedly longer at 2.259(4) (506).The neutral six-coordinate [TcOL,Cl] (L = 2-methyl8-quinolinolate), prepared by substitution of [TcOClJ, is the cis-isomer and hence may be regarded as a TcO{N20C1} derivative. The Tc-Oquinolinolat bond distances cis and trans to the 0x0 group are 1.947(3) and 1.994(3) respectively, and Tc-C1 is 2.360(1) A (507).The chloro ligand in this and related complexes undergoes solvolysis in methanol (507,508).Electrochemical studies of [TcOClL,], where L is a bidentate N,O-Schiff base or 8-quinolinolate ligand, have shown reduction to Tc'"0 species (509).In [TcOC1(OCH2CH,0)(phen)l the C1 atom is also cis to the 0x0 group and the OTc-N bond distances are 2.173(4) (cis) and 2.268(4)A (trans) (505). Similar structurally characterized

d,

68

JOHN BALDAS

complexes are [TcOClL,] (L = N-phenylsalicylidineiminate)(510)and [TcOCl,L] (L = NNO Schiff base) (511).A five-coordinate example is [TcOCl(Ophsal)],in which Tc is displaced by 0.67 A above the ONOCl plane (512). Crystallography has shown that a product obtained from the reaction of 20 [E = S (SphsalH,)] with [TcOCl,I- is the octahedral [TcOCl(hbt),] with equatorial ONNCl coordination. The hbt ligand (see 33) is formed by an oxidative intramolecular ring closure (513).The five-coordinate [TcOCl(Sphsal)]has since been prepared by substitution of [TcOClJ, with a stoichiometric amount of the ligand and crystal structures of this complex (514)and the related [TcO(Sphsal)(SPh)l(515)reported. Crystal structures are also available for the dithiocarbazate derivative (45) (2681, the potential brain imaging agent (46) (5161, cis-[Tc0{8hydroxy-3,6-dithiaoctan-l-olate-(O,S,S, )}C121 (517), and the squarepyramidal AsPh,[TcO(MAG,)I [MAG2 = mercaptoacetylglycylglycinate(2-)],with the carboxylate group participating in the ONNS coordination (518).The preparation ofTcO complexes with a variety of dithiocarbazate derivatives (519, 520) and of tridentate Schiff bases with a thiolato coligand (521 has been reported. 0 II

An interesting structurally characterized complex is the distorted square-pyramidal [TcVO(SC,H2iPr,)2(PhNNCON2HPh)l (522),which has also been assigned the Tc(II1) oxidation state on the basis of structural and spectroscopic features (523).In view of the absence of any other Tc=O group in an oxidation state below Tc(V)and Holm's generalization that M=O groups are stabilized at metal centers with an oxidation state of no less than +4 (524),the TcW) assignment would seem preferable.

B. COMPLEXES OF THE tr~ns-[TcO(OH)1~' AND [TcO,l+ CORES Technetium, in common with rhenium (1891,forms a considerable number of cationic complexes containing the truns-[Tc021+core. The poor

COORDINATION CHEMISTRY OF TECHNETIUM

69

ability of neutral cr- or weak 7r-donor equatorial ligands in truns[TcO(OH,)L,]~+t o neutralize the positive charge on the [TcO13' core results in enhanced acidity of the truns water and the following acid-base equilibria (525, 35): [0=Tc-0HJ3+

Ka 1

[O=Tc-OH]2t

__ K.2

[0=T~0]+.

Negatively charged ligands do not favor proton loss but an exception is CN-, which, although a good cr-donor, is also an effective n-acceptor. From neutral solution K,truns-[TcO2(CN),1 [v,,,(Tc02) at 785 cm-'] (PIC,, is isolated and acidification to pH 1yields trans-[TcO(OH2)(CN4)12.90) via the [TcO(OH)(CN),I2-intermediate. The 7r-acceptor nature of the equatorial cyanides in (NMe4)trans-[TcO(OH2)(CN),1~2H20 is apparent in the high value of 1029 cm-' for v(Tc0) (399).The dioxo complex is also formed by the hydrolysis of K2[TcO(CN),1 (229). In general, cationic truns-[TcO,]+ complexes are prepared by the reaction of Tc0,-/Na,S,O, (399) or [TcOX,I- (526, 527) and neutral nitrogen or N2S2cyclic thioether ligands. The truns-[Tc02(py),lC1complex is readily prepared by hydrolysis/oxidation of [TcCl6I2-in neutral conditions or by substitution of [TcOCl,]- in the presence of water and serves as a useful starting material for ligand exchange reactions (528,399). Crystal structures have been reported for [Tc02(cyclam)lC104~H20 (5251,truns-[TcO,(en),]X (X = C1, I) (529),[Tc02L,lC1~nH20 (L = imidazole, n = 2; L = 1-methylimidazole, n = 3) (530),[Tc02(4-tert-butylpyridine),]CF3S03.H20 (5261, [Tc0,(1,4-dithia-8,11-diazacyc1otetradecane)]PF, (5311, and a polymeric {Li[Tc02(1,4,8,11-tetraazaundecane)](CF3S03),},,(532). There is a brief mention of the structure of [TcO2(CN),I3-(35).Characteristic features illustrated by the structure of truns-[TcO,(en),]Cl are Tc=O bond distances of 1.752(1) and 1.741(1)A, a Tc-N bond distance of 2.158 A (av.) and an 0-Tc-0 angle of 178.6(3)" (529). Interestingly, although 36 exists as the monooxo form, an increase in the hydrocarbon chain to N(CH2I6Nresults in the formation of the neutral dioxo complex (no amine nitrogen deprotonated) with Tc=O, 1.745(3)A, and the 0-Tc-0 angle, 170.1(1)"(455). The long Tc=O distances and low asymmetric O=Tc-O IR stretching frequencies in the range 850-750 cm-' are indicative of a lower bond order than that in [Tc0l3' complexes. Group theoretical analysis presymmetry the maximum Tc-0 bond dicts that for truns-[TcO,]+ in order is 2 (533) and this is consistent with the bond order of 2.10 for indicated by the Tc=O stretching force constant of 6.23 mdyn truns-[TcO,(en),]C1(4141. Kinetic studies of pyridine exchange in truns[TcO,(py),]+ indicate a dissociative mechanism and the Tc complex has

70

JOHN BALDAS

been found to undergo exchange at ca. 8000 times the rate of the Re analog (534, 535). The cationic nature of [TcO,I+ complexes has attracted considerable radiopharmaceutical interest. A promising myocardial imaging agent that shows good blood and liver clearance is the diphosphine derivative ["""TcO,L,]+ [L = {CH2P(CH2CH,0Et),},]. The Tc=O bond distance in trans-[Tc02L,l[Cr(SCN),(NH3),] is 1.738(17) A (536). With the dmpe ligand the hydroxo complex trans-[TcO(OH)(dmpe),](CF3S03), has been isolated and Tc=O and Tc-OH bond distances of 1.66 and 1.96 A have been determined by EXAFS (123).The trans-[T~0(0H)(CN),]~species could not, however, be isolated due to rapid dimerization (399),but (NBu,),truns-[TcO(OMe)(CN4)], with a nonionizable methoxy group, may be regarded as a trapped form (229). The isolation of K,[TcO(OH)Cl,], with v(Tc0) at 900 cm-', has been claimed (537),but has not been substantiated. The formation of monoanionic TcO, complexes with Schiff bases has also been reported (538), but these may be the p-0x0 dimers. C. 0x0-BRIDGED[Tc2O3I4+ AND OTHERBINUCLEAR COMPLEXES Protonation of [TcO2(CN),I3- in acidic aqueous solution and rapid dimerization of the initially formed truns-[TcO(OH)(CN),I2~yield the purple p-0x0 dimer [TC,O~(CN)~]~(399). Generally, [Tc2O3I4 complexes are prepared from substitution reactions of [TcOCl,]-, for example, [{TcO(S,CNEt2),},(pL-0)1 (539),or from reduction of TcO,-. The reaction sequence is illustrated by truns-[TcOClL], prepared by substitution of [TcOCl,I- with (sacac),enH, (H,L) in dry solvents. In the presence of moisture, the labile chloride is replaced by water to give the cationic [TcO(OH,)LlCl, which then forms the p-0x0 dimer [{TcOL},(p-O)I by reaction of [TcOClLI with the intermediate hydroxy complex [TcO(OH)L1(540).Crystal structures of [{TcOL},(p-O)], where L represents a variety of tetradentate ONNO aminephenolato ligands, (e.g.,47) (541), the ONNO Schiff base ligands N,Nf-2-hydroxypropane1,3-bis(salicylideneiminate)(542) and N,N'-propane-l,3-diylbis(salicylideneiminate) (5431,and 48 (517 )have shown the presence of either a crystallographically imposed linear or a near-linear (167"- 173") Tc-0-Tc bridge with O=Tc-0 angles of 163"-171", giving an essentially linear [O=Tc-O-Tc=0I4+ core analagous to the [Re,O3l4+ core (189). Other features are Tc-Obridgebond distances of 1.90-1.92 A and the near-eclipsed arrangement of the donor atoms of the two Tc centers as shown in 48, in which the near-eclipsed atoms are C1 and S (517).The occurrence of linear d2-d2 [MV2O3l4+ (M = Tc, Re) cores is explained +

71

COORDINATION CHEMISTRY OF TECHNETIUM

(47)

(48)

by MO theory. For a [O=ML,-0-L,M=O] complex in D4,, symmetry, the T interactions between the two tz9sets from the two metal atoms and the px,p,, sets from the three oxygen atoms give rise to two nonbonding molecular orbitals (b, + b,,), which do not correspond to any oxygen T linear combinations. A d2-d2 configuration corresponding to the occupation of these nonbonding orbitals satisfies the closed-shell electronic appears as a n configuration (544).In the IR spectra v,,,(Tc-0-Tc) intense broad band at 625-680 cm-' but v(Tc=O) is of variable intensity and may not be observed (541,543).A complex formulated as the mixed-valence K3[T~V'1V202C1al has been obtained by reflux of K,[Tc,. Cla].2H,0 in 2-butanone in air. The IR spectrum indicates the presence of both Tc=O (1020 cm-'1 and Tc-0-Tc (680 cm-'1, but the structure is uncertain (42). Formation of the novel dimer 49 on reaction of [TcOCl,I- with 1.5 eq. of edtH, may be viewed as the interaction of the Lewis base [TcO(edt),]- with the Lewis acid [TcO(edt)]+ (Fig. 14) (35, 545). On reaction with excess edtH,, 49 is converted quantitatively to [TcO(edt),]-, but [(TCO)~(S(CH~)~S}~] does not react with further amounts of ligand. Also, although a dimeric intermediate was found to form in the reaction of [TCO(SCH~CH~O)~]with 2 eq. of edtH, , no intermediate was detected in the substitution of [TcO(OCH2CHz0)21to [TcO(edt),]- (546).An interesting binuclear complex is [{TcO(OEt)Cl,},(p-L)], where L is a n N6 heterocyclic nitrogen ligand (547).Addition of NBu4[TcOC1,1 to ( N B U ~ ) ~ [ H ~ P Win ~ , MeCN O ~ ~ I yields purple crystals of the Tc-substituted Keggin polyoxotungstate derivative ( N B U ~ ) ~ [ P W ~ ~ TThe C Osilicon ~ ~ ] . derivative ( N B U ~ ) ~ [ S ~ W ~has ~TCO~~] been prepared by the addition of N~[TCO(OCH~CH,O)~] to a-KaSiWl,039~12Hz0 in sodium acetate buffer. Electrochemically, the Tc center of [PWl,TcO,O]"- appears to exhibit only three accessible oxidation states, Tc(1V) Tc(V) Tc(VI), in contrast to the five oxidation states, Re(II1)-Re(VII), accessible for the analogous Re cluster (548).

72

JOHN BALDAS

56

s4

TC2

c5-

- c3

c2

FIG.14. The structure of [ ( T ~ O ) ~ ( e d(49) t ) ~ l(545).

D. [TcS13' COMPLEXES at 520 The diamagnetic sulfido complex AsPh,[TcS(edt),] [ 4T-S) cm-'1 is formed from the reaction of [TcC1,12- with 1,2-ethanedithiol and [TCSC~,{HB(~~)~)I from the 0x0 complex by S atom transfer from B2S3 (379,549).The [TcS13' core is less stable than [Tc013+and readily undergoes replacement of the sulfido ligand by 0x0 in solution and under aerobic conditions. E. NITRIDOCOMPLEXES The nitrido ligand (N3-) is isoelectronic with the 0x0 ligand (02-) and is a powerful velectron donor that effectively stabilizes technetium in the + 5 to +7 oxidation states. The first complexes containing the [TcVNI2 core, [TcNCl2(PPh3),1and ITcN(S,CNEt,),I, were prepared from the reaction of TcO,- /ligand with NH2NH2.HC1as the reducing agent and source of the nitrido ligand (550,551), but this method is of limited applicability. Two general methods for the synthesis of [TcNI2+ complexes are by ligand exchange of [TcNCl2(PPh,),1 and by reduction/ exchange of [TcV'NX4]-(X = C1, Br) (551,552,220). A characteristic feature is the generally sharp v(TcN)IR absorption at 1100-1028 cm-', +

COORDINATION CHEMISTRY OF TECHNETIUM

73

which is shifted by -30 cm-' on 15N labeling (145).The TcN bond is formally triple with one (T and two n components and quite short in the range 1.60-1.64A. The lower charge on [TcN12+in comparison with the isoelectronic [Tc0I3+results in little tendency to deprotonation of coordinated amine ligands. The TcN bond is very resistant to hydrolysis or removal by other reactions but readily reacts with active sulfur sources such as S2C1, to yield thionitrosyl complexes (322).In general, [TcNI2+complexes are not readily reduced and require agents such as chlorine for oxidation (553,554).Structurally, the strong trans influence of the nitrido ligand results in either five-coordinatesquare-pyramidal complexes or six-coordinate complexes with the trans ligand only weakly bound. The anion in (AsPh4),trans-[TcN(OH2)(CN),1~5H20 has distorted octahedral geometry with a TEN bond distance of 1.60(1)A, a very long NTc-OH, distance of 2.559(9)A, and the Tc atom 0.35 8, above the equatorial plane. A high v(TcN) at 1100 cm-' results from the n-acceptor property of the equatorial cyanides. The pK,, value of the coordinated water has not been determined but the long bond distance indicates very low acidity and high kinetic lability (555).In the case of [Re(E)(OH,)(CN),I"-(E = 0, N) the strong trans effect of the nitrido ligand is apparent in the pK,, values of 1.4 and 11.7 for the 0x0 and nitrido complexes, respectively, and a reaction rate constant some 9 x lo5times greater for the nitrido complex (556).The structurally characterized Cs2Na trans-[TcN(N3)(CN),I.2H2O is obtained by ligand exchange with N3- (5571, and Cs,K[TcN(CN),] may be isolated in the presence of cyanide (558).The IR spectrum of the latter complex shows three well-defined v(CN) absorptions (2A1+ E ) consistent with ClU symmetry. Reaction of NCS- with [TcNC1,(PPh3),1yields yellow [TcN(NCS),(PPh,),], which on reflux in MeCN is converted to orange-red crystals of trans,trans-[T~N(NCS)~(PPh~)~(MeCN)]4MeCN. The TEN bond distance is 1.629(4)A and the weak binding of the MeCN ligand is apparent in the long Tc-NCMe distance of 2.491(4)8, and the formation of the five-coordinate complex on dissolution in CHC13(559).The thiocyanato ligands areN-bonded, as is also the case for the structurally characterized (NEt,),[TcN(NCS),(MeCN)I, prepared by reaction of [TcNCl,]- with NCS- and crystallization from MeCN (560).The preparation of (AsPh,),[TcN(NCS),] has been reported (561) but crystallography shows the product crystallized from MeCN/EtOH is the trans-aqua complex (AsPh,)2[TcN(OH2)(NCS),l~EtOH (557).Spectroelectrochemical studies at -60°C show a reversible one-electron reduction of [Tc"'NX,]- (X = C1, Br) to [TcNX4I2-,but the colorless reduced species have not been isolated (562).

74

JOHN BALDAS

No TcvN(04) complex has yet been identified, but five-coordinate TcVN{S4}complexes are readily prepared by ligand exchange of [TCNC~,(PP~,)~] or reduction/substitution of [TcNClJ (552,5631. Key structural data are summarized in Table IV. Of particular interest is a comparison of the trans influence of the nitrido and 0x0 ligands in the same coordination environment. The three TcN/TcO pairs in Table IV show that, although the nitrido ligand exerts the greater trans influence, the 0x0 ligand exerts the greater structural steric effect (566). This is seen in the greater displacement of Tc above the square basal plane in the five-coordinate square-pyramidal complexes in Table IV and the correspondingly greater OTcL angles. The Tc-N bonds are shorter than the Tc=O bonds and it has been suggested that this may be largely accounted for by a-electron effects, with nitrogen utilizing an sp hybrid orbital and oxygen, an sp2 orbital in the TcN/O bond. The Tc-L distances are longer in the nitrido complexes, but this effect largely may be due to the lesser core charge on [TcNI2' (566). Reaction of [ T c N C ~ ~ ( P P ~with , ) ~ I K(S2COEt) and treatment with aqueous ethanol yield the dithiocarbonato complex K2[TcN(S2C0)21 on hydrolysis of the intermediate xanthate (564). The mixed-ligand complex AsPh4[TcN(S2CNEt2)(SCOCOS)I is prepared in a controlled fashion by the reduction/substitution of [TcV'NC1,(S2CNEt2)] with dithiooxalic acid (565).Reaction of thiourea with [TcNC14]-yields the orange [TcN(tu),Cl]Cl, which is a useful starting material for ligand exchange in aqueous solution (568).Mass spectrometry has shown that [TcN(SSeCNEtJ,] undergoes a thermally induced scrambling to give the S4, S,Se, and SSe, species (569).Chromatographic studies have shown the formation of [99"T~NL2]2(L = mnt, dto) on reaction of TABLE IV STRUCTURAL DATAFOR TcVN{S4}COMPLEXES AND TcVN{S,/Se4}/TcVO{S4/Se4} COMPLEXES WITH THE S A M E COORDINATION ENVIRONMENT hbp' Complex lTCNlS2CNEt2)2l K21TcN(S2CO)ZI.2H2O AsPh4[TcNC32CNEt211dto11 (AsPh4)p[TcN(dto121 AsPh,[TcO(dto),I tAsPh412[TcN(mnt)21 AsPh4[TcOlmnt121

INBu,)21TcN[Se2CC(CN12}21 NEt4[TcO{Se~CClCN12)~l

thl

TcL

1,604161 1.621(61 1.5412) 1.613(41 1.646141 1.5911) 1.655163 1.6111) 1.67(21

2.401 av. 2.390 av.

108.1 av. 107.3 av.

2.393 av.

106.0av.

2.378(21-2.391121 2.327(1)-2.330(1) 2.367(4)-2.41914) 2.31012)-2.320(21 2.508(21-2.52812) 2.463141-2.476141

105.4(2l-106.113) 108.6(21-109.9(2) 101.818)-106.8(81 107.4121-109,8121 106.1(4l-109.5(41 108.1(6l-112.416)

TcNlO

Displacement of Tc above the square basal plane.

th)

NTcLiOTcL ("I

(A)

Ref.

0.741(51 0.71 0.66 0.65 0.76 0.59 0.74 0.768(1) 0.88

550 564 565 427 427 566 428 567 432

75

COORDINATION CHEMISTRY OF TECHNETIUM

[99mT~V1NC1,]with the ligands (448).Exchange of [99mT~N(dto)212(and the 99Tccomplex) with mnt occurs via an intermediate, presumably the mixed complex [99mTcN(dto)(mnt)12(570). The neutral [99mTcN{S2CNEt(OEt)},] is a promising neutral myocardial imaging agent. Cyclic voltammetry of the 99Tccomplex has shown no oxidation or reduction in the interval + 1.225 to - 1.75 V vs SCE, indicating that this complex should be stable in uiuo (571). A variety of TcN{N,} complexes has been prepared from [TcNCl,(PPh3)21or from [TcNCl,]- (usually in the presence of a n auxiliary reducing agent such as KBH4or PPh,) (572,573).The cationic ethyleneis typical, with v(TcN) a t diamine complex trans-[T~N(en)~Cl]BPh, 1085 cm-', distorted octahedral geometry, a TEN bond distance of 1.603(3)A, and the very marked trans influence of the nitrido ligand, resulting in a n NTc-Cl bond distance of 2.7320(8)8,. In trans-[TcN(tad)ClIBPh, (tad = 1,5,8,12-tetraazadodecane) the Tc-C1 distance is 2.663(2)A (572).Crystallography has shown that the product formed from [TcNC1,(PPhJ2] with excess diethylenetriamine in benzenelethano1 under aerobic conditions is the novel dicationic 50(BPh4),. The mechanism of formation of the zwitterionic NH, +CH2CH2NHCOO-carbamato ligand and the cleavage of the triamine to ethylenediamine is not clear. Under anhydrous conditions in an inert atmosphere, [TcNCl,(PPhJ2] is recovered unchanged, but 50 is readily isolated when a stream of C02is passed into the reaction mixture. The crystal structure shows that the zwitterionic ligand lies in a peculiar "transient state" and is stabilized by strong intramolecular hydrogen bonding (5741. H,

(50)

12+

(51)

Other related complexes are [TcN(cyclam)C1lC1and 51 (X = 01, for which neutrality is achieved by deprotonation of the two amide groups, with the resultant Tc-N bond distance of 2.051 8, significantly shorter than the Tc-NH distance of 2.126 8, (573).In the cationic 51C1 (X = H2)the NTc-OH2 bond is very long at 2.560(2) A (575).Crystallography

76

JOHN BALDAS

has shown the product of the reaction of [TcNBr,]- with bpy in ethanol to be {cis-[TcVNBr(bpy),l}z[TcllBr,]. The formation of the previously unknown [TcBr,],- under mild conditions is unprecedented. With methanol as the solvent the product is cis-[TcNBr(bpy),]BPh, (146,576).In cis-[TcNBrz((pyCHz)2NCHzCMezSBz}l, the Tc-N bond distance of the tertiary amine N atom coordinated trans to the nitrido ligand is 2.47(1) A and that of the pyridine N atoms coordinated cis is 2.141 A (av.). In solution there is a n equilibrium between the dibromo form and one in which a bromide ion is expelled, and the thioether sulfur is coordinated (577). Other structurally characterized complexes are cis-[TcNCl(phen),]PF, (which exhibits a pseudo-twofold symmetry axis that gives rise to reproducible enantiomeric disorder) and cis-[TcNCl(phen)zlC1.HzO(578). The preparations of [TcNL41Cl2(L = py,imidazole) (579) and [TcN(phthalocyanine)l (580) have been reported. The pyridine ligands in [TcN(OH)(py),]BPh, are labile and undergo exchange with pyridine in solution. The [TcN(OH)]+core shows similar "Tc NMR shifts to [TcO,I+. Reaction with tmbtH gives a quantitative yield of trans-[TcN(tmbt),(py),l (581). The most important phosphine complex is the synthetic intermediate [TcNCl2(PPh3),1,which may be prepared by a variety of routes, including hydrazine.HC1 reduction of Tc0,- in the presence of PPh, (5511, reductionlsubstitution of [TcNCl,]- (5521, and substitution of [TcN(tu),Cl]Cl (568). Reaction of PPh3 or AsPh, with [TcNX,]- gives the five-coordinate [TcNX,(EPh,),] (X = C1, Br; E = P, As) in high yield and with the sterically less demanding PMe,Ph, the six-coordinate cis[TCNX,(PM~,P~)~]. The presence of the trans halide ligand in the sixcoordinate complexes results in u(TcN) a t 1048 cm-' (X = C1) and 1028 cm-' (X = Br) compared with 1095-1090 cm-' for the five-coordinate complexes. All these complexes readily undergo ligand-exchange reactions (582). Tri(cyanoethy1)phosphine yields the anionic NBu,[TcNX3L] (583).The reduction of 9 9 m T ~ 0 in 4 - the presence of NH,-NR-C(=S)SMe/PPh,/HCl and addition of the ligand have been developed for the preparation of 99mT~N radiopharmaceuticals (584). The Tc=N bond distances in [ T C N C ~ ~ ( E P(E ~ , )=~ P, ] As) are 1.602(8) and 1.601(5) A, respectively, and the geometry may be regarded as intermediate between square-pyramidal and trigonal-bipyramidal, as shown in Fig. 15 for the arsine complex (585,586). The six-coordinate cis-mer4TcNC1,(PMe2Ph),1 shows NTc-Cl bond distances of 2.441(1) A (cis) and 2.665(1) A (trans) and a T c z N distance of 1.624(4)A (587). For the octahedral trans-[TcNCl(dmpe),]BPh, , a Tc=N bond distance of 1.853(6)A has been reported, but problems in the refinement were noted (576).This distance seems unreasonably long and is likely due

77

COORDINATION CHEMISTRY OF TECHNETIUM

FIG.15. The structure of [TcNClz(AsPh&J (586)

to disorder between the trans nitrido and chloro ligands, giving rise to the crystallographic artifact of "distortional isomerism". Cationic [99mTcN(dppe)zC11+ and related complexes undergo in uiuo reduction and are then washed out of the myocardium (588).Cyclic voltammetry has shown that [TcNCl(dmpe),]+ undergoes reversible reduction with unexpected ease at -0.02 V vs SCE when compared with the irreversible reduction of [ReNCl(dppe),l+ at -1.8 V vs SCE (184). Trigonal-bipyramidal geometry is observed for 52,with Tc=N, 1.601(4) A, and a near-linear P-Tc-P angle of 176.5'. The ether oxygens cannot be regarded as coordinated with Tc-0 contact distances of 3.190(2) A. A related [TcNCl,Ll complex with L containing a tertiary amine brid e has square-pyramidal geometry and a Tc-.Naminedistance of 2.70(1) , indicative of incipient coordination (589).

1

COOEt

78

JOHN BALDAS

A variety of complexes containing 25 (R = SH), 27, or PPh, in a mixed coordination sphere has been reported, indicating the versatility of the [TcN],' core (3O8,310,590). The binuclear [{Tc,N2C1,L21(L = zPr2PCH,CH2PPr2)is thought to contain C1 bridges (576). Structurally characterized examples are 53 (5911, [TcNL(PPh,)l (L = tridentate S-methyl dithiocarbazate) (5921, and [TcNCl(PPh,){PhN=C(OEt)S}] (593).The TcN{02S2}core is found in thio-P-diketonato complexes (5941. Structurally characterized TcN{N,S,} complexes are [TcN(tox),] (tox = 8-quinolinethiolate) (5521, [TcN{(~acac)~en}] (540), and [TcN(Me, CNNC(S)SMe),] (592). A novel complex is [TcN(tmbt),L2], where L represents what is generally regarded as a "noncoordinating" proton sponge, 1,1,2,2-tetramethylganidine. The T e N bond distance and v(TcN) are unexceptional at 1.615(6) A and 1057 cm-l (581). Neutral bisaminedithiolato and N-(N"-morpholinylthiocarbony1)-N'-phenylbenzamidinato complexes have been prepared (481,595).

F. IMIDO AND HYDRAZIDO COMPLEXES The reaction of [TcOXJ (X = C1, Br) with ArNCO in toluene yields the moisture-sensitive blue-black [Tc(NAr)X,]- in high yield (278). Imido complexes containing phosphine ligands are formed from the reaction of [TcOC141~/organohydrazine or aromatic amine/phosphine (278, 523). Alternatively, TcO,- may be used as in the reaction with PPh,/PhNHNHCOMe in methanol containing a minimal amount of HC1 to give a good yield of the yellow-green octahedral imido complex [TcCl3(NPh)(PPh3),1, with the phosphine ligands in trans positions. The Tc=N bond distance of 1.704(4) A is longer than that in [Tc0I3+ or [TcN],' complexes, but the Tc=N-C bond angle of 171.8(4)"confirms that the imido(2-) ligand is in the linear triply bonded form. This bonding mode is also consistent with the v(TcN) IR absorption a t 1090 cm-'. Reaction with py/MeOH gives the mixed-ligand complex [TcC13(NPh)(PPh,)(py)] (523, 596). Similarly, in fuc-[TcCl,(NPh)(dppe)], v(TcN) occurs at 1110 cm-I and the Tc=N-C bond angle is 175.7(9)" (523). The hydrazido(2-) complex [TcCl3(NNMePh)(PPh3),1 is formed by reaction of [TcOC1,1-/NH2N(Me)Ph/PPh, in refluxing methanol. With less bulky phosphines, [TcCl2(NNMePh)(PMe2Ph),I+and the structurally characterized truns-[T~Cl(NNMe,)(dppe)~]PF, cations are formed (278). The hydrazido(2-) ligand in the dppe complex is coordinated in the linear (four-electron donor) mode. The reaction of MePhNNH,/dppe/[TcOCl,]- in methanol, however, yields the cationic oxoimido complex truns-[TcO(NH)(dppe),l , with marked asymmetry of +

COORDINATION CHEMISTRY OF TECHNETIUM

79

imido complex tr~ns-[TcO(NH)(dppe)~] +,with marked asymmetry of the two axial ligands shown by crystallography (278). The structurally characterized trigonal prismatic diazene complex [Tc(HNNCSPh),(S,CPh)] (597) and the octahedral hydralazino complex [TcCl,(CaH5N4)(PPh3),l(598)were assigned the Tc(V) oxidation state but a n alternative assignment, that of Tc(1) and Tc(II1) species, respectively, has been proposed (523).The deep-green thiobenzoyldiazene complex NBu,[Tc(HNNCSPh),] is also likely to be trigonal prismatic (597). Addition of HC1 to [ T c ~ ~ ~ C ~ ( N N A ~ ) , ( Pyields P ~ , ) ,the ] neutral [TcCl,(NNAr)(NNHAr)(PPh,),] and addition of HBr yields the cationic doubly protonated [TcBr,(NNAr)(NHNHAr)(PPh,),]Br (277).These complexes and the diazene complex [TC(C~H,N,N=NH)~IBP~, have been assigned the Tc(1) oxidation state, but the 99TcNMR chemical shifts fall in the established TdV) region (277, 599). G. COMPLEXES NOTCONTAINING MULTIPLYBONDED LIGANDS Treatment of the 16-electron [T~~~'(diars),Cl,]Cl with chlorine results in oxidative addition, producing the brown 18-electron [Tcv(diars),Cl,]Cl with a magnetic moment of 0.9 BM (191). The crystal structure of [Tc(diars),Cl,]PF, shows DZddodecahedra1 eight-coordination geometry, with Tc-As bond distances of 2.578(2) and Tc-Cl bond distances of 2.442(4) (295).Reaction of NBu4[TcVO(abt),]with 12 M HC1 yields the blue NBu4[TcVCl,(abt)]by removal of the 0x0 ligand in a formally nonredox process. The crystal structure and the magnetic moment of 2.86 BM establish the presence of Tc(V) and thus that the abt ligand is in the doubly deprotonated form (600). The Tc=O bond in NBu4[Tc"O(abt),] is abnormally long [1.73(2) A] (496) and appears to be susceptible to protonation as indicated by 'H NMR evidence of an equilibrium between the anionic and the neutral species in wet CDC1, (600).Removal of the nitrido ligand in AsPh4[TcV1NC1,1by 1,2-benzenedithiol and reduction gives a low yield of AsPh,[TcV(bdt),]. The structure of the anion shows only small distortions from ideal trigonal prismatic geometry, with chelate twist angles for the three dithiolene ligands of 1.1",16.3", and 5.8", compared with the ideal value of 0" (360).This complex is more conveniently prepared in quantitative yield by the reduction of Tc0,- by bdtH, in refluxing EtOH/H20/HC1(558). The intermediate [TcvO(bdt),l- is the kinetically controlled product formed at room temperature. The thermodynamic product [Tc(bdt),]is then formed by removal of the 0x0 ligand in a formally nonredox process. A related complex is [Tc(abt),]- (601).

80

JOHN BALDAS

IX. Tech net i urn(V1)

The [Tcv1014+core is highly susceptible to hydrolysis and disproportionation and unlike [TcV0I3+is not readily stabilized by coordination. The nitrido ligand is, however, very effective in stabilizing Tc(V1) as [TcNI3'. A characteristic feature is the formation of dimeric [NTcOTcNI4+and [ N T C ( ~ - O ) ~ T Ccomplexes N ] ~ ~ that have no analogs for any other transition metal. Complexes not containing an 0x0, nitrido, or imido ligand are relatively few and confined to fluorides and complexes with dithiolene and other noninnocent ligands. Monomeric Tc(V1) (dl) is easily and reliably detected by EPR spectroscopy (401, but the dimeric species are EPR silent due to spin pairing (602).The only binary halide, and the highest fluoride for Tc, is the golden-yellow TcF,, prepared by the reaction of fluorine gas on the metal powder (603, 392). Reaction of TcF, with NOF and N02F yields (NO),[TcF,I and N02[TcF71,respectively. Magnetic moments of 1.72 and 1.67 BM confirm the +6 oxidation state (394).

A. 0x0 COMPLEXES The Tc0,2- and Re0,2- anions are rather less stable than Mn0,2-. Pulse radiolysis and cyclic voltammetry have shown that in alkaline aqueous solution Tc04,- has a lifetime of the order of milliseconds (604). In neutral solution Tc0:decays by a second-order process, about 100-foldmore slowly than Re0:- (605,606).The pK,, of H,TcO, is estimated to be 2-0.5 (607).The paramagnetic, violet (NMe4)2[T~041 [ p e K= 1.60 BM; v(Tc0) a t 780 cm-ll has been prepared by electrochemical reduction of Tc04- in MeCN with rigorous exclusion of air and water. The salt is extremely sensitive to air and atmospheric moisture, which cause rapid oxidation and disproportionation (608,609).Fluorination of Tc yields the blue [TcOF,] (m.p., 134°C)as a by-product (392). The blue monoclinic form is isostructural with ReOF,, the structure of which consists of infinite chains of F-bridged octahedra (610). A minor product, the green hexagonal form, is the F-bridged cyclic trimer with a Tc=O bond distance of 1.66(3) (611). The purple, light-sensitive [TcOCl,] has been prepared from the chlorination of Tc metal (326). Reduction of Tc04- by HC1 in concentrated HZSO4gives a deep-blue solution (Amax = 572 nm), shown by the EPR spectrum to be a Tc(V1) species, most likely [TcOClJ, although the presence of [TcOCl,] cannot be totally excluded. The EPR parameters are gll = 2.057, g, = 1.938, All = 230 x and A, = 96 x lop4 cm-'. After 1 hr the blue color vanishes and the EPR signal decreases (612).Unstable deep-blue

COORDINATION CHEMISTRY OF TECHNETIUM

81

solutions containing Tc(V1)oxochloro complexes are also formed by the reduction of AsPh,TcO, in SOCl, or POCl, (613).The reaction of Tc207 with SnMe, gives the sublimable organometallic bis(p-oxo) dimer (54) (228).Coordination about each Tc atom is distorted square-pyramidal with Tc=O bond distances of 1.666(2) and 1.647(2)8, and Tc-Obridge distances of 1.900(2)-1.925(2) 8,. The dimer has been reported to be paramagnetic on the basis of the absence of "Tc NMR signals but the Tc-Tc distance of 2.5617(3) 8, would seem to indicate a single bond and consequent diamagnetism.

(54)

B. NITRIDOCOMPLEXES 1. Monomeric [TcN13+Complexes

The reaction of TcO,-/NaN, in refluxing HX (X = C1, Br) gives high yields of orange-red R[TcNCl,I and intensely blue R[TcNBr,] (R = AsPh,, NBu,] on precipitation with the organic cations (614). The structure of the square-pyramidal [TcNCl,I- is shown in Fig. 16. These

FIG.16. The structure of the [TcNCl41- anion in AsPh4[TcNCl41(614).

82

JOHN BALDAS

salts are air-stable and the remarkable resistance of the TcVINbond to acid hydrolysis is apparent from the method of preparation. Evaporation of the MeCN extract of the dried TcO,-/NaN,/HCl reaction mixture and dissolution of the residue in concentrated HC1 yield an orange-red solution that probably contains HTcNC1,. Addition of CsCl to this solution gives red crystals of the six-coordinate Cs2[TcNC1,] (5521, whereas NEt,Cl gives orange crystals of (NEt,)trans-[TcN(OH,)Cl,] (615). The aqua complex (NEt4)truns-[TcN(OH,)Br41may be prepared in high yield directly from the TcO,- /NaN3/HBr reaction. In concentrated HX solution the major species is most likely [TcN(OH2)X,](615). The [TcNClJ anion is also formed by the oxidation of TcVN species (554, 586, 616 ) and substitution/oxidation of [TcOCl,] with azide (617). The reaction of NH,0S03H with TcO,-/HCl also yields [TcNCl,]- and shows that a single amine nitrogen attached to a good leaving group may serve as an N3- precursor, but the product is contaminated with nitrosyl species and [TcCl6l2-(558).The structural features observed for the isoelectronic [TcVNI2+ /[TcV0I3+pairs are again apparent in a comparison of the [ T c ~ ' N I ~ ' / [ T c ~pairs ~I~+ in Table V. The nitrido ligand exerts the greater trans influence, in terms of the transNTc-OH, bond distance in the aqua complexes, but the OTcX angles are greater than the NTcX angles and the displacement of Tc above the square basal or equatorial plane is consequently greater for the 0x0 complexes. Also, the Tc-X bond distances are significantly greater

TABLE V STRUCTURAL AND IR DATAFOR [TcV1NI3'HALIDECOMPLEXESAND SOME [ T c " ~ ] ~ANALOGS ' Complex

Tc-NI=O iAi

AsPh4[TcNCI41 AsPh4lTcOCl41 AsPh4[TcNBrll AsPh,[TcOBr41 NEt41TcN10H21Br41

1.561i5l 1.593i6l 1.59616 ) 1.613i9) 1.559(9)

NEt4[TcOtOH21Br41

1.616191

[Rb(15-crown-5121 ITcNiOH,)C141 Cs?lTcNCI61

1.600i3) 1600'

Tc-X iA) 2.3220i9) 2.309i2) 2.4616i5) 2.46011) 2.510ill 2.518(11 2.50511I 2.506i 1I 2.320i2) 2.373(5),,, 2.740i511,,,

Tc-OH,

A)

2.44317, 2.317i9l 2.43i4)

OIN-Tc-X ("1 103.34(3) 106.6i1) 103.04(21 106.59(3) 97.212) 98.0i2l 97.6i2l 99.5i3) 94.5i2) 99.73i6l

6'

tA) 0.54

0.67 0.56 0.70 0.33 0.37

0.401

UiTcNIiiTcO) 1cm-l)

Ref.

1063

614 409 618 410 615

1000

405

1074

619

1027

615

1076 1025 1074

Displacement of Tc above the square basal or equatorial plane. The [TcYX,]- iY = N , 01 anions have ideal C,,, symmetry. Value fixed in the refinement due to the statistical disorder of the ligands in the cubic space group.

COORDINATION CHEMISTRY OF TECHNETIUM

83

for the nitrido complexes. The v(TcN) IR absorption occurs at higher energy than that of v(Tc"0) but this difference may be accounted for, either partially or entirely, by the greater mass of the l60atom. Thus, the difference betwen v(Tc14N)at 1076 cm-' for AsPh4[TcNC1,1 and v(Tcl6O)at 1025 cm-' for AsPh,[TcOCl,I is less than the 61 cm-' calculated by the simple diatomic oscillator model. The presence of trans halide in Cs2[TcNX51(X = C1, Br) results in the decrease of z4TcN) to 1027 (X = C1) and 1028 cm-' (X = Br) from the values of 1076 and 1074 cm-' for AsPh,[TcNX,I (552, 614, 620). The NEt4[TcN(OH,)C1,1 complex undergoes complete dehydration to NEt,[TcNCl,I under vacuum and the aqua complex is reformed on exposure to atmospheric moisture. The small change in v(TcN) from 1065 to 1070 cm-' on removal of the trans water is indicative of very weak binding and of a Tc-OH, bond distance in the aquachloro complex that is longer than that in NEt,[TcN(OH,)Br,l, which does not undergo dehydration under the same conditions (615). The [Rb(15-crown-5),1[TcN(OH2~C141 salt contains [Rb(l5-crown-5),]+sandwich cations and isolated anions, with a n NTc-OH, bond distance of 2.43(4)A and v(TcN) at 1074 cm-'. Because this salt is prepared from SOC1, solution, the coordinated water presumably arises from atmospheric moisture (619). The 4d' (S = 1) configuration of Tc(V1) results in readily observed EPR spectra at temperatures of >77 K (40,41).The spectra of [TcNX,I(X = C1, Br) have been examined in detail (620,621 ), including singlecrystal EPR, electron nuclear double resonance (ENDOR), and electron spin echo envelope modulation (ESEEM)studies of 15N-enrichedAsPh,[TcNCl,] doped into the diamagnetic AsPh,[TcOC141 host (622, 623 1, as well as single-crystal EPR and 15N powder ENDOR studies of NBu,[TcNBr,]/[TcOBr4] (624). The molecular orbital of the unpaired electron is a combination of the Tc dx, and equatorial ligand p orbitals. Analysis of the hyperfine data for AsPh,[TcNCl,] indicates 20% of the spin density is localized in the 3p orbitals of the C1 atoms (622),but a polarized neutron diffraction study has shown exceptionally high covalence of the Tc-Cl bonds, with 46(5)%of the spin density located on the C1 atoms (625).Interestingly, AsPh4(CF3SO3)may also serve as a host lattice for AsPh4[TcNC141,giving an extremely well-resolved spectrum at 130 K (619). EPR spectroscopy has proven particularly useful for the identification of [TcNI3+species in solution and the monitoring of ligand-exchange reactions. Mixed-ligand [TcNBr,-, C1,I- ( n = 1-31 species have been identified in mixtures of [TcNCl,]- and [TcNBr,l- and equilibrium constants, determined (579, 626, 627). Mixed species are readily assigned because the EPR parameters are nearly linearly dependent on the spin-orbit coupling constants of the equato-

84

JOHN BALDAS

rial donor ligands (40,579,626,628). A 0.002 M solution of Cs,[TcNCl,] in 28.6 M HF shows the presence of the five [TcNF,-,Cl,I- (n = 0-4) species, presumably due to the low activity of fluoride ion in the solution. Partial removal of C1- by the addition of 1 eq. of AgF results in the disappearance of signals due to [TcNCl,I- and [TcNFCl,]-, and after the addition of 3 eq. of AgF only signals due to [TcNF,Cl]- and [TcNF,]- remain. The [TcNF,]- species may be prepared in solution by the dissolution of “TcN(OH),” in 50% HF but has not been isolated. The EPR parameters for [TcNF,]- are gll = 1.895, g , = 1.990, A,, = 377 x lo-,, and A, = 179 x lo-, cm-l(628). EPR studies of AsPh,[TcNX,] (X = C1, br) in organic solvents in the presence of a large molar ratio of X- and of Cs,[TcNX,] (X = C1, Br) in HX solution show no evidence for the equilibrium (620) [TcNXJ-

+ X-

[TcNXS]’-

The mixed-ligand species [TcNCl,(CN)I- and [ T c N C ~ ~ ( C N have ) ~ ] - been identified by EPR in the reaction of [ ( T C N ( C N ) ~ } ~ ( ~ - with O ) ~ IHCl ~(5651, [TcNBr,(NCS)]- and [TcNBr,(NCS),l- in the reaction of [TcNBr,]- with NCS- (5791, and [TcNX,(N,),-,,I- (X = C1, Br; n = 1-4) in the reaction of NBu,[TcNX,] with azide in acetone (629).Also, a variety of [TcNI3+ species such as [TcN(HSO,),I- and [TcN(H,PO,),I- have been identified in concentrated acid solution (630). The oxidation of [TcVNC1,(EPh3),l (E = P, As) to [TcNCl,I- by S0C12 has been shown by EPR to proceed via the Tc(V1) species [TcNCl,(EPh,)l (586). Substitution of R[TcNX,I in organic solvents occurs readily but generally results in reduction and the [TcVNI2+substituted product. Thus, reaction with PPh, , KNCS, Na(S2CNEt2),and 8-quinolinethiol yields [TcNCl,(PPh,),], (NEt,),[TcN(NCS),(MeCN)I, [TCN(S~CNE~,)~], and [TcN(C9H6NS),],respectively (552).Reduction also occurs upon substitution by nonreducing ligands such as bpy or phen (146,578). Substitution reactions in which the Tc(V1) oxidation state is retained are the reaction of AsPh,[TcNC1,3 with LiBr in acetone to give AsPh,[TcNBr,I and of NBu,[TcNBr,] with (sal)enH, to form [TcN{(sal)en}]C1(552,631). Attempts to prepare R[TcNI,l by ligand exchange with LiI in acetone result in oxidation of iodide to iodine. The ease of reduction of [TcNX,1(X = C1, Br) and the inability t o prepare [TcNI,I- may be understood in terms of the lowest energy n X + Tc LMCT transitions at 18,975 cm-’ (X = C1) and 13,100 cm-l (X = Br) for NBu4[TcNX41in MeCN (632). According to the theory of charge-transfer spectra and the optical electronegativity difference of 0.5 between C1- (n1 and I- (n), substitution of iodide for chloride is expected to result in a red shift of -15,000

COORDINATION CHEMISTRY OF TECHNETIUM

85

cm-' for the lowest energy Laporte-allowed LMCT transition (633).An LMCT transition may be regarded as the transfer of an electron from a predominantly ligand orbital to a predominantly metal orbital and, if the energy difference is less than about 10,000 cm-', then, commonly, there will be total electron transfer, resulting in reduction of the metal and oxidation of the ligand (634). For [TcNI,l- in organic solvents the first LMCT band is calculated to be at -4000 cm-l and a facile redox reaction is apparent. The [TcNL,]- (L = HSO,, H2P04)species are colorless and show only an intense absorption at 33,800 and 37,600 cm-', respectively, which has been assigned to a TN += Tc LMCT transition (630). has The formation of [99"'TcNC1,]- on reaction of 99mTc0,-/NaN3/HC1 been established (635). After removal of HC1 the residue is stable t o oxidation under acidic conditions but undergoes oxidation to 99m T~04at pH > 4 (636). Addition of ligand solutions results in the formation of 99mTcNcomplexes with biological distributions different from those of 99mTccomplexes prepared from 99mT~04and the ligand by use of complexes are Sn2+or other reducing agents. In general, the 99mT~N cleared from the blood more slowly, indicating greater in uiuo reactivity and the exchange of 99mT~N with serum proteins. Although reaction of [99"'TcNC1,]- with thiolato ligands leads to reduction to Tc(V), the oxidation state with ligands such as gluconate or phosphonates is unclear (635). 2. Dimeric and Polymeric [TcN13+Complexes The very moisture-sensitive neutral TcNC1, may be prepared by the reaction of TcC1, with IN, or NBu,[TcNCl,] with GaCl, . The IR spectrum indicates a polymeric structure with TcNTc and TcC1,Tc bridges. TcNC1, is insoluble in CH,C1, but dissolves on addition of AsPh,Cl due to the formation of AsPh4[TcNC1,1 (637). Addition of 18-crown-6 to a suspension of Cs,[TcNCl,] in SOC1, results in the formation of an orange-red solution, which on slow evaporation of the solvent yields crystals of [Cs(18-crown-6)][TcNC14](619, 638). The structure consists of the unprecedented "infinite sandwich" M+/crown ether configuration with ordered and disordered infinite chains of [TcNClJ anions arranged in an antiparallel fashion (Fig. 17a). In the ordered TEN... Tc=N.-chain the TEN and N-Tc bond distances are 1.561(36) and 2.714(36)A, respectively. An unusual aspect of the structure is that the nearest neighbors of each Cs+cation are two Cs+cations at 4.275 A, whereas the Tc atoms of the four nearest anions are at 7.95 h; and the nearest Cs+...Clcontacts are 6.4-6.6 h;. Also, although each Cs+cation has neighbors a t 4.275,8.55, and 12.825 h; along each vertical column,

COORDINATION CHEMISTRY OF TECHNETIUM

87

FIG.17b. A portion of the structure of [ C s ~ l 8 - c r o w n - 6 ~ 1 , [ ~ T c N C l ~ ~(639). ~(OH~~~]

in the horizontal plane the nearest Cs+ neighbors are at 11.23 A. The IR spectrum shows a single u(TcN) absorption at 1041 cm-', but partial 15N labeling results in complex behavior due to the coupling of TcN oscillators in the infinite [TcNCl,]- chains and is diagnostic for this arrangement. At a 12.5% 15N content the spectrum shows essentially two peaks, a t 1042 and 1025 cm-', with the latter peak being predominantly due to ''NTc.-15NTc.-'4NTc groupings. With 50% 15N content, the major "NTc absorption at 1015 cm-' is strong but the 14NTcabsorptions are reduced to two weak shoulders a t 1053 and 1045 cm-'. The EPR spectra over the temperature range 130-290 K indicate the presence of exchange interactions along the -.TcN.-TcN-- chains (619). Recrystallization of the infinite sandwich [Cs(18-crown-6)][TcNC14] from MeCN, acetone, or ethanol, or the reaction of Cs2[TcNCl51with with 18-crown-6 in 6 M HC1, yields [Cs(18-crown-6)14[(TcNC14),(OH2),l two u(TcN) absorptions at 1055 and 1046.5 cm-', which are shifted to 1023 and 1016 cm-' on 15N labeling. Ether diffusion into a n MeCN solution of the aqua complex may result in the crystallization of either the infinite sandwich or a mixture of the infinite sandwich and the

88

JOHN BALDAS

aqua complex. The crystal structure of the aqua complex shows dimeric [N&I’cC~,-.N+I’C(OH~)C~,]~units inside a square cage formed by four [Cs(18-crown-6)]+cations, with two monomeric [TcN(OH2)C1,1- units present in the lattice (Fig. 17b) (639).The [(H30)(18-crown-6)1,[(TcNC14)2(OHz)] complex has also been isolated and shown by crystallography to contain only dimeric [ N 9 c C 1 , ~ ~ ~ N 9 c ( O H 2 ) C l 4units I 2 - (639). Reaction of NBu4[TcNCl41with ( N B U ~ ) ~ [ H ~ P W in ~MeCN ~ O ~results ~I in the incorporation of TcN to give dark crystals of the Keggin polyoxotungstate derivative (NBu4)4[PWllTcN03gl,for which the Tc(V1) oxidation state is thought to be retained (548). A characteristic feature of the chemistry of [TcN13+is the formation of dimeric complexes based on the [NTc-O-TcNI4+and [NTC(~.-O),TCN]~+ cores (423,565, 630,640).Analogous nitrido complexes are not known for any other transition metal but the chemistry and structural aspects of the TcV*Ndimers parallel those of the well-known isoelectronic [OMoV-O-MoV014+ and [OMoV(pO)2MoV012+ dimers (640,641). Hydrolysis of Cs,[TcNCl,] in an ample quantity ofwater gives a brown precipitate of “TcN(OH),,”which has been formulated as the bis(p-0x0) dimer [(TCN(OH)(OH~)}~(~-O)~] (55) on the basis of its reactions and the presence of v(Tc0Tc) absorptions in the IR spectrum (628,640).This precipitate is the isoelectronic analog of “MoO(OH),,”a compound of unknown structure (641). Solutions of 55 in 7.5 MCF3S03H(avery weakly coordinating medium) are orange (A,,, = 474 nm) and EPR silent, showing the absence of monomeric species. The monomeric aqua cation [TcN(OH,),I3+ is thus not a viable species even in strongly acid solution and appears to spontaneously dimerize to the p-0x0 aqua cation [{TcN(OH2)4}2(p-0)14+ (56) (630). 14+

(56)

(57)

Solutions of 55 in 1 M p-toluenesulfonic acid, CF,S03H, or MeS03H are pale yellow and shown by paper electrophoresis to contain a single cationic species (630, 640).Also, dilution of a solution of 56 in 7.5 M CF3S03Hleads to the slow formation of the yellow species. That this species is the bis(p-oxo) aquanitrido cation [{TcN(OH,)~}~(~-O),]~+ (57) is indicated by the similarity of the electronic spectrum to that of the well-established [{MoO(OH2)3}2(p-0)212+ cation, the absence of EPR signals, and the isolation of [{TcN(S2CNEt2)},(p-O),1on reaction with

COORDINATION CHEMISTRY OF TECHNETIUM

89

Na(S2CNEt2)(602). Structure 57 has been confirmed in solution by EXAFS studies, but the actual number of coordinated water molecules is uncertain (642). Addition of ethanol to a solution of 55 in aqueous CsOH precipitates the yellow Cs2[{TcN(OH)2}2(p-O)21 with u(TcN) at 1046 cm-' and v(Tc0Tc) at 734 cm-'. On treatment with HC1 this salt is converted to [TcNClJ (640). The only p-0x0 complex to have been isolated and structurally characterized is the cyclic tetramer (AsPh4)4[Tc4N4(0)2(ox)61 prepared by the reaction of AsPh,[TcNCl,] with oxalic acid in aqueous acetone (423). The centrosymmetric structure consists of two [(ox)TcN-0-TcN(ox)l units joined by two tetradentate oxalates (Fig. 18). The TEN bond distances are 1.639(17) and 1.606(17) A and the Tc-0-Tc bridges are only approximately linear,with angles of 150.4(8)"and Tc-Obridgedistances of 1.840(13) and 1.869(13) A. A marked asymmetry due to the trans influence of the nitrido ligand is apparent in the Tc-0 bond distances of the bridging oxalates, with NTc-Ot,,, , 2.410(11) and 2.369(12) A, and NTc-O,,,, 2.076(11)and 2.061(11) A. The p-0x0 structure of the oxalato complex with the nitrido ligands cis to the oxygen

FIG.18. The structure of the anion in (AsP~~)~[Tc~N~(O)~(OX)~] (423).

90

JOHN BALDAS

bridge may be contrasted to the linear (or near linear) [OTcV-0-T~V014td2-d2 dimers, in which the 0x0 ligands are trans to the oxygen bridge. This difference in geometry is explained in terms of closed-shell electronic structures. In the syn conformer (Czvsymmetry) shown in 56 and 60 [and also for an anti conformer (C,,, symmetry)], only one molecular orbital is available for the metal d electrons and the Tc(V1)d'-d' configuration will thus satisfy the closed-shell requirement (544). Addition of AsPh,Cl and then HCl to a solution of Cs,[TcNCl,] in water with sufficient MeS0,H added to dissolve the initial precipitate gives a high yield of the yellow (AsPh4)z[(TcNC12),(p-O)21, and the bromo complex may be similarly prepared (6431. Dithiocarbamato, cyano, and ethanedithiolato complexes have been prepared by the addition of the ligand to solutions of Cs2[TcNCl51in aqueous Na4P2O7(565). The structure of the [ ( T ~ N C l ~ ) ~ ( p - 0anion ) ~ 1 ~(643) in Fig. 19 shows the features of the Tc0,Tc ring system and structural data are given in Table VI. The geometry of the [Tc2N2O2I2+ complexes may be described as two square pyramids sharing the bridging oxygens to give a bent TczOzring. Each Tc atom is displaced above the plane of the four basal donor atoms by 0.50-0.67 A. A comparison of the isostructural [{TcN~S2CNEtz~}z~p-0~21 and [ ( M O O ( S ~ C N E ~ ~ ) } ~(644) (~-O)~I again shows the greater effect of the 0x0 ligand, with the Mo atoms displaced by 0.73 A above the square basal planes compared with 0.65 A for the Tc atoms (565).The Tc-Tc distances of 2.542(2)-2.591(1) A correspond to a d,..'-d~,,l single bond and account for the absence of EPR spectra. The two nitrogen atoms in complexes are N

N

FIG. 19. The structure of the [ ( T C N C I ~ ) ~ ( ~ -anion O ) ~ ](643). ~-

91

COORDINATION CHEMISTRY OF TECHNETIUM

TABLE VI STRUCTURAL DATAFOR Complex

T-N

th

[NTC(~-O),TCN]~' DIMERS

Tc-Tc

(A)

T d b &

"4)

6-sbp0 ~

AI

Ref.

~~

I AsPh,)zl{TcNClzjzlp-OlZl

1.648(81 1.65018) 1.6101131

2.5791 11

1.896(71-1.953(5)

2.57511I

IAsPh,121{TcN~CN12}2~p-OIZl 1.70111 I(TcNIS2CNEt2it2( p-0121

tAsPh4121{TcNBr2t21 p-0l21

I{TcN1S~CNC,H,1)21/~-O121

1.62314) 1.624(4) 1.6512) 1.59121

632

1.9%11-1.95( 1)

0.54(11 0.5211) 0.59(11

2.560121

1.921(91-1.924191

0.50

557

2.543(11

1.935131-1.942(3)

565

2.542121

1.934(131-1.9471121

0.65111 0.65(1) 0.65111 0.67111

632

565

Displacement of Tc atom above the square basal plane

bent back from each other to a n N.-N contact distance of 3.3-3.5 A. In the absence of this bending the N-N distance would be the same as the Tc-Tc bond distance and rather shorter than the van der Waals contact distance of about 3.1 A (632).The Tc0,Tc ring system is readily detected in the IR spectrum by the presence of a strong asymmetric stretching mode a t 710-700 cm-' and a weaker symmetric mode a t 515-450 cm-'. These assignments have been confirmed by l80labeling (565, 632). All dimers have the syn stereochemistry shown in Fig. 19 and show two v(TcN) absorptions as a result of the in-phase and outof-phase vibration of the coupled TcN oscillators. For (AsPh,), [(TcNX2)2(p-0)2] these absorptions occur a t 1063 and 1054 cm-' (X = C1) and 1059 and 1051 cm-' (X = Br) (632).Surprisingly, (AsPh,),[{TcN(CN)2}2(p-O)21 does not show significant 4 C N ) IR absorptions (565). The crystal structure, however, shows the CN bond distances to be normal, at 1.12-1.18 A (643). The formation of bis(p-0x0) dimers greatly reduces the susceptibility with of TcVINto reduction. Thus, reaction of (AsPh,),[(T~NCl~)~(p-0)~1 in good yield, Na(S2CNEt2) in MeCN gives [{TcN(S~CNE~~)},(~-O)~I whereas reaction of [TcNCl,]- in the same solvent gives only the reduced [ T C ~ N ( S ~ C N E ~ ~ ) The ~ I Tc02Tc bridge is readily cleaved (565). by HCl in organic solvents. This reaction allows the preparation of Tc(V1) species such as [TcNC12(S2CNEt2)1, which are not accessible by partial substitution of [TcNClJ. The electronic spectra of the bis(p0x0) dimers do not show pronounced visible absorptions (565, 632). The interconversions and equilibria of [TcNI3' species in solutions of inorganic and organic acids have been studied by UV-visible and EPR spectroscopy (630,643,645)and are described by Scheme 1.Monomeric species are identified by their EPR spectra. The pox0 dimers

92

JOHN BALDAS

2[TcNL,

SCHEME 1. L

=

1-

monoanionic ligand.

are readily distinguished from bis(p-oxo) dimers by the intense visible absorption at 470-580 nm, which arises from a transition in the threecenter Tc"O2Tc .rr-bond system. High acidity and the presence of coordinating anions such as C1- favor the monomeric species. The reaction sequence may occur in either direction, depending on whether 55 or Cs,[TcNCl,] is dissolved in the acid. Solutions of Cs2[TcNC1,] in 3.33 M HC1 show only the presence of [TcNCl,]- whereas in 0.5 M HC1 a pink species (A,, = 538 nm) is formed. If Cs,[TcNCl,] is first hydrolyzed and then HC1 added to 3.33 M ,an intensely blue species (Ama = 566 nm) that is converted to [TcNCl,]- by first-order kinetics is formed. The pink and blue species have been formulated as 59 and 60 (L = Cl), respectively (632, 643). At low acid concentrations the interconversions are slow and paper electrophoresis is a useful technique for the separation of species. Thus, when Cs,[TcNCl,] is dissolved in 0.5 M H2S04an orange anionic species and a colorless slow moving cationic species (with some S02- coordination) may be separated (630). The presence of pox0 and bis(p-oxo)dimers in HC1 and H2S04solutions has been confirmed by EXAFS studies (642). C. IMIDOAND HYDRAZIDO COMPLEXES The reaction of [ T c ( N A ~ )with ~ I ] Na/THF at room temperature yields the homoleptic diamagnetic dimer [Tc,(NAr),I (Ar = 2,6-diisopropylphenyl). The dimer is air-stable in solution and adopts an unprecedented "ethane-like" structure (61),with the six imido ligands symmetry equivalent in a staggered arrangement.

COORDINATION CHEMISTRY OF TECHNETIUM

(61)

93

(62)

The unsupported Tc-Tc single bond distance is 2.744(1) 8, and the Tc=N bond distances are 1.758(2) 8, (646).Reduction of [Tc(NAr),II (Ar = 2,6-dimethylphenyl) with 1 eq. of Na gives the dimer [Tc,(NAr),(p-NAr),I, which on reaction with MeMgCl undergoes the unprecedented substitution of imido by methyl groups to give successively [TcM~~(NA~)(~-NA~)~Tc(NA~),~ and tetramethyl derivative (62) (647). The Tc-Tc bond distances in the dimethyl and tetramethyl derivatives of 2.673(2) and 2.733(1) 8,, respectively, are similar to the bond distance in 61 and consistent with a d'-d' single bond. For the tetramethyl derivative only the "Z-type" isomer (62) is observed in the solid state. Reaction of the tetrachlorocatecholate complex NBU,[TCO(C~C~,O~)~] with NHzNPhzin CH2C12followed by addition of methanol yields purple crystals of the unusual paramagnetic Tc(V)/(VI) mixed-valence complex N B ~ , [ T C ~ ( N N P ~ ~ ) ~ ( C ~ C ~ ~ ~The ~ )crystal ~I~CH~C~~ structure of the dimeric anion shows the presence of bridging hydrazido(2-) ligands and a Tc-Tc bond distance of 2.612(2) 8,. The complex is EPR silent in various solvents to 196 K but at lower temperatures shows a broad line centered at g = 2.015. In the electronic spectrum a weak absorption a t 12,000 cm-' is consistent with an intervalence charge-transfer band (4221.

D. DITHIOLENE AND RELATED COMPLEXES Green [Tc(tdt),] is formed in about 5% yield from the reaction of TcO,- with Zn(tdt) in 7.5 M H2S04. Electrochemically, [Tdtdt),] is oxidized to [TcV**(tdt),1+ and readily reduced to [Tcv(tdt),1- and with more difficulty to [ T ~ ' ~ ( t d t ) , ](648). ~The deep-green [Tdbdt),] may be prepared quantitatively by the oxidation of [Tc"(bdt),]- with iodine (558).The green [Tc(abt),I complex is formed on allowing a mixture of TcO4-/2-aminobenzenethiol/HC1 to stand overnight (649). The coordination sphere is a tapered trigonal prism, with the three N and S atoms occupying the triangular faces and Tc-N and Tc-S bond distances of 1.982(9)-2.004(8) 8, and 2.339(3)-2.359(3) A, respectively (650). The EPR spectrum of [Tc(abt),] has been analyzed (649, 651)

94

JOHN BALDAS

and the effect of concentration and solvent composition on the spectra of frozen solutions, interpreted in terms of the breakdown of molecular aggregates to the monomeric species (651). Dark-blue [Tc(dbcat),I (A,, = 594 nm; E = 19,000) is formed in high yield from NH4Tc04and 3,5-di-tert-butylcatechol (dbcatH,) in methanol. A well-resolved 10-line EPR spectrum is observed in solution at room temperature. Reversible electrochemical oxidation yields the Tc(VI1) species with surprising ease and there are two reversible reductions to Tc(V)and Tc(1V)species. The coordination geometry of [Tc(dbcat),l is approximately octahedral with the twist angle of 41.7" much closer to the ideal octahedral value of 60" than to the ideal trigonal primatic value of 0". The Tc-0 bond distances are in the range 1.945(6)-1.974(6) A (652). X. Technetium(VI1)

The aqueous solution chemistry of Tc(VI1)is dominated by the stability of the TcO,- anion. Strong oxidizing agents such as HN03 or H202 ultimately, but at varying rates, oxidize all technetium compounds to TcO,- (12).Technetium, unlike rhenium, does not form a heptafluoride (7). The coordination chemistry of Tc(VI1)has been regarded as rather limited but recent results show it to be potentially extensive and novel.

A. 0x0 AND SULFIDO COMPLEXES The only product formed when Tc metal is burned in an excess of oxygen at 500°C is the volatile, crystalline, yellow Tc207(m.p. 119.5"C) (7).In the solid state the structure of Tcz07consists of isolated centrosymmetric molecules with tetrahedral coordination about Tc and a linear Tc-O-Tc bridge with Tc-Obridgebond distances of 1.840 A (653). The oxide dissolves in water t o give a colorless solution of the strong acid HTcO, . Concentrated solutions of the acid are red and on evaporation dark-red crystals of hygroscopic anhydrous HTcO, are obtained (7). The Tc0,- anion absorbs strongly in the UV region a t 244 and 287.5 nm (654) and the red color of HTcO, is thought to be due to a disturbance from tetrahedral symmetry, resulting in the movement of the edge of the 287-nm absorption band into the visible region (27). The alkali metal salts are very stable; KTcO, sublimes at about 1000°C without decomposition (7). A considerable number of Tc0,- salts have been prepared and structurally characterized (655-659). In KTcO the TcO,- anion is tetrahedral with Tc-0 bond distances of 1.711(3) (or 1.724 A if corrected for librational oscillation) (659)but in NMe,TcO,

COORDINATION CHEMISTRY OF TECHNETIUM

95

the anion has approximate CSusymmetry with Tc-0 bond distances of 1.589(11)A and 1.696(6)-1.719(9)A (656). The TcO,- anion shows no tendency to form polyanions and, unlike ReO,- (6601,appears to have little tendency to act as a ligand, although it may be present as a counteranion. The E" values for the M04-/M02couples of 1.695,0.738, and 0.510 V for Mn, Tc, and Re, respectively, show that TcO,- is a stronger oxidizing agent than ReO,-, but very much weaker than MnO,- (654). Many technetium complexes in lower oxidation states may be prepared directly from TcO,- in the presence of the ligand and a suitable reducing agent. Key starting materials such as [TcOCl,]and [TcC1612-are readily prepared by the reduction of TcO,- by 12 M HCl in the cold and under reflux, respectively, and [TcNClJ by HCl reduction in the presence of azide (35, 614). A kinetic study of the reduction of Tc0,- by HBr shows that the first-step Tc0,- + [TcVOBr,]is a pseudo-first-order process and the second-step [TcOBr,]- + [TcBr612-is a combination of a first-order with a zero-order process (661). Pertechnetate is an effective catalyst in the oxidation of hydrazine by NO3- or ClO,- (662). A unique property of TcO,- is the remarkable inhibition of the corrosion of soft iron or carbon steels at concentrations as low as 5 x M.The ReO,- anion is inactive in this respect (654). Brown-black Tc2S7may be prepared by H2Sprecipitation from 2-4 M HC1 or H2S04 (7). There is some evidence for the presence of the [TcO,S]- anion in solution (663) but the formation of thiopertechnetates needs further investigation. Reaction of Tc207 with SnMe, yields MeTcO,, the dimer [(Me2(54), TCO)~(~-O ) ~ ]and the polymeric ester {Me3SnOTc0,}, (228, 664). The structure of the polymer consists of infinite zigzag chains with Tc=O bond distances of 1.655(13) and 1.676(15) A and a Tc-Obridge distance of 1.72(1)A (664).The oxide MeTcO, is a much stronger Lewis acid than MeReO, and reacts with olefins such as cyclohexene to form a Tc(V) glycolato complex, which decomposes in the presence of water and acids to stereospecifically produce the cis-diol and the disproportionation products Tc02.nH,0 and TcO,- (228). The ester [TcO,(OSiMe,)] is a useful synthetic intermediate (665, 666). Yellow crystals of [Tc03Fl(m.p., 18.3"C)are formed from the reaction of fluorine with TcO, at 150°C (667)or by the dissolution of NH,TcO, in anhydrous HF (668).In the presence of water, [TcO,F] hydrolyzes to TcO,- and HF (667). The pale-yellow liquid [Tc03C11(b.p., 25°C) is formed quantitatively on heating TcC1, in oxygen at 450°C (326).The vibrational spectra of [TcO,X] (X = F, C1) have been assigned in CSu symmetry (669, 670). The equilibrium Tc03F + HF Tc03+ + HF2has been demonstrated by 99Tcand 170NMR and confirmed by the

*

96

JOHN BALDAS

addition of AsF, to a solution of Tc0,- in HF. NMR has also identified the species [Tc2O5F41 and [TcO,F,I in the reaction of XeF, with [TcO,F]/ HF (671).Pure [Tc02F31(m.p., 200 + l°C) has been isolated from the reaction of Tc,O,/HF/XeF, and consists of open chains of F-bridged cis-Tc02F4octahedral units, with Tc=O bond distances of 1.646(9) A, a Tc-F terminal bond distance of 1.834(7) L$, and a bridging bond distance of 2.080(5) L$ (672). The yellow transitory intermediate formed on addition of TcO,- to 12 M HC1 is thought, by analogy with the ~1~ attempts to isolate the reaction of Re04-, to be f a c - [ T ~ ~ " O ~ C 1but NBu,+ salt result in reduction to [TcOCl,J- or hydrolysis to Tc04- (35). The presence of choline chloride appears to stabilize [TcO3Cl3I2-and the solution remains bright yellow for several hours (673).In the presence of bpy or phen the reaction of TcO,- with ethanolic 12 M HC1 yields a yellow precipitate of [TcO,ClLI and with HBr yields orange [TcO,Br(bpy)l. These complexes are hydrolyzed by water to TcO,- and are reduced by reflux in ethanolic HX to [TcVOX3Ll.In the IR spectra there are three v(Tc0) bands in the range 910-850 cm-' (417). Slurries of [TcO,ClLI (63) (L = phen, bpy, Me4-phen, NO2-phen) in acetone or CH2C12cleanly oxidize olefins at 22°C to give high yields (>70%) of the stereospecific TcVOdiolato complexes (64).

Hydrolysis of 64 with concentrated HCl yields [TcOCl,L] and the stereospecific diol. Thus, reaction of cis-4-octene with [TcO,Cl(phen)l and hydrolysis gives only the meso-diol, whereas trans-4-octene gives 80% of the DL and 20% of the meso isomers, indicating some racemization (L = during the hydrolysis process (674).The binuclear [(TCO,X)~(~.-L)I polynitrogen heterocycle; X = C1, OR) has been prepared from Tc04or [TcOClJ (547). Reaction of TcO,-, the tripodal ligand [(~5-Cp)Co{PO(OR)2},l-, and concentrated HNO, gives [LTc03]in 97%yield. This complex may also be prepared, but in low yield, from the oxidation of [LTcOCl,] (675). The structure of the rhenium analog indicates that in [LTcO,] the geometry is distorted octahedral with coordination by three facial oxygens from the tripodal ligand and three technetyl 0x0 ligands (676). Similarly, HNO, oxidation of [TcVOC12{HB(pz),}]yields [TcO,{HB-

97

COORDINATION CHEMISTRY OF TECHNETIUM

(pz),}], which may also be prepared from TcO,-/HB(pz),- in ethanol containing concentrated HzS04.Bubbling ethylene through a CH2Clz solution of [Tc03{HB(pz)3}lyields the glycolato complex [TcVO(OCH, CHzO){HB(pz)3}l.Notably, [ReO,{HB(pz),}I does not react with ethylene due to the greater difficulty of reducing Re(VI1) to Re(V) (677). B. NITRIDOAND IMIDO COMPLEXES In view of the stability of the Tc=N bond and the preparation of

Kz[Re(N)O,], it would seem likely that nitridotechnetic(VI1) acid [Tc(N)O3Hz1,or its salts, could be prepared from the reaction of Tc207 with liquid ammonia or NHz-/NH3,but these reactions have not been attempted (640).Slow evaporation of a solution of Cs,[TcNCl,] in 10% HzOzyields yellow-orange crystals of the explosive nitridoperoxo complex C ~ [ T C N ( O ~ )(Fig. ~ C ~20). I The coordination geometry is a distorted pentagonal pyramid with the nitrido ligand in the apical position [Tc=N, 1.63(2)A] and q 2 peroxo ligands with 0-0 bond distances of 1.41(2) and 1.46(2)A (678).The AsP~,[TcN(O,)~X] (X = C1, Br) complexes are prepared from AsPh,[TcNX41/Hz02 and are thermally more stable. Addition of bpy, phen, or oxalic acid to the pale-yellow solution of “TcN(OH),” in 10% H202yields [TcN(O,),LI (L = bpy, phen) and (679). The crystal structure of the dimeric (AsPh4)z[{TcN(0,)2}2(ox)l the oxalate dimer shows the anion to consist of two TcN(02), units bridged by a tetradentate sideways-bound oxalate with distorted penN

CL

FIG.20. The structure of the anion in CS[TCN(O~)~CI] (678).

98

JOHN BALDAS

tagonal-bipyramidal geometry about each Tc atom (680). In the IR spectra of the nitridoperoxo complexes, 4TcN) occurs a t 1069-1035 cm-', 40-0) at 912-894 cm-'; and u,,(Tc02) at 665-647 cm-' (679). These complexes are the only examples of nitridoperoxo complexes and rare examples of peroxo complexes of a metal in the + 7 oxidation state. The [TcN(O,),] core is isoelectronic with the well-known [MO(O,),] (M = Cr, Mo, W) cores and emphasizes the analogy between isoelectronic [MOO]and [TcNl complexes noted for [TcV'NI3+dimeric species. Oxidation of [TcOCl,]- by H202gives TcO,-, with no evidence for the formation of transitory peroxo species (678). The reaction of NBu,[TcOCl,] with Ph2NNH2and 2,4,6-triisopropylbenzenethiol yields yellow crystals of the novel nitrido-hydrazido(2-), formally Tc(VII), binuclear complex 65.0.5Et20. The geometry about each Tc atom is distorted square-pyramidal with long Tc= NNPh, bond distances of 1.88(1)A, Tc-N bond distances of 1.64(1) A, and Tc-N-NPh2 angles of 140.2(11)"and 141.7(11)".The Tc-S (bridging) distance of 2.470(7) A is significantly longer than the average Tc-S (terminal) distance of 2.379(6) A. The nitrido ligands result from N-N bond cleavage of the organohydrazine (6811.

Ar

H Tc \ ArN411 'NAr

A rS

SAr (65)

N ..

Ar

(66)

The reaction of ArNCO (Ar = 2,6-dimethyl- or 2,6-diisopropylphenyl) with [TcO,(OSiMe,)] yields the imido complex [Tc(NAr),(OSiMe,)l. Tetrahydrofuran solutions of [Tc(NAr),(OSiMe,)] react readily with Grignard reagents to form deep blue-green [Tc(NAr),R] (R = Me, Et, qlallyl) and with F - to give the oxoimido complex [(Ph,P),NI[TcO(NAr),l. Reaction with ISiMe, in toluene yields [Tc(NAr),I]. Crystal structures of [Tc(NAr)&OSiMe,)l and [Tc(NAr),I] (Ar = 2,6-diisopropylphenyl) show approximate tetrahedral geometry and, for the iodo complex, Tc=N bond distances of 1.740(7)-1.763(6) A and Tc-N-C bond angles of 164.8(6)"-169.4(6)". The presence of three imido ligands imparts a high degree of stability to the Tc(VI1) center. Electrochemical studies show that the complexes are difficult to reduce and are also moderately air stable. NMR spectra indicate free rotation about the N-C(Ar) bonds (666).The reaction of KCp with [Tc(NAr),I] rapidly forms the green,

COORDINATION CHEMISTRY OF TECHNETIUM

99

air- and water-stable r)'-Cp complex (66).Two of the Tc=N bond distances in 66 are similar, a t 1.748(2) and 1.753(2) A, but the third is significantly longer, a t 1.761(2) A. With a n excess of KCp, the airsensitive K[Cp,Tc(NAr),] is formed and a symmetrical structure is indicated by the 'H NMR spectrum (682). C. COMPLEXES NOTCONTAINING MULTIPLY BONDEDLIGANDS Treatment of NH,TcO, with K/en/EtOH yields the classic hydrido complex K,[TcH,I, isostructural with K,[ReH,l (683).The structure of the [TcH,I2- anion is thus a trigonal prism capped on the three rectangular faces (684).The chemical behavior of [ReHglZpand [TcHg12is similar but the Tc complex is more reactive. In solution, [TcHg12has been shown by 'H and "Tc NMR to be stereochemically nonrigid (6711. The preparation of [TcH,(PEt,Ph),J has been reported (685).The green [Tc(pda),]TcO, is formed on reflux of a solution of Tc0,- and 1,2 diaminobenzene (pdaH,) in methanol. The geometry of the [Tc(pda),l cation is trigonal prismatic with the pda ligands in the paddle wheel arrangement and the six Tc-NH bond distances in the range 1.98(1)-2.03(2) A. The presence of a single dNH) IR absorption at 3235 cm-' confirms that the ligands are in the deprotonated dianionic form (456). +

XI. Appendix: Abbreviations

abtH acacH (acac),enH, AcO atm av. bdtH, bPY Bu 'Bu Bz cdoH, CP 15-crown-5 18-crown-6 cyclam depe

2-aminobenzenethiol acetylacetone N , N'-ethylenebis(acety1acetoneimine) acetate atmosphere average value l,2-benzenedithiol 2,2'-bipyridine n-butyl tert-butyl benzyl cyclohexane-1,2-dioxime cyclopentadienyl 1,4,7,10,13-pentaoxacyclopentadecane 1,4,7,10,13,16-hexaoxacyclooctadecane 1,4,8,11-tetraazacyclotetradecane 1,2-bis(diethylphosphino)ethane

100 diars dmf dmgH2 dmpe dmso dPPe dtoH, &

edtH, edtaH, en

EPR Et EXAFS FABMS HB(pz), hbt HPLC LMCT Me MLCT mntH, ntaH, OphsalH, ox Ph phen pic Pr PY quinH (sacac),enH, salH (sal),enH, SphsalH, tan tctaH, tdtH terPY THF tmbtH tmP tmtu tu

JOHN BALDAS

1,2-phenylenebis(dimethylarsine) dimethylformamide dimethylglyoxime 1,2-bis(dimethylphosphino)ethane dimethylsulfoxide 1,2-bis(diphenylphosphino)ethane dithiooxalic acid molar extinction coefficient (M-’ cm-’) 1,2-ethanediethiol ethylenediaminetetraacetic acid 1,2-ethanediamine electron paramagnetic resonance ethyl extended X-ray absorption fine structure fast atom bombardment mass spectrometry hydrotris(pyrazo1-1-yllborate(1-1 2-(2-hydroxyphenyl)benzothiazolate( 1-) high-performance liquid chromatography ligand-to-metal charge transfer methyl metal-to-ligand charge transfer maleonitriledithiol nitrilotriacetic acid N -(2-hydroxyphenyl)salicylideneimine oxalate(2-) phenyl 1,lO-phenanthroline 4-methylpyridine is0 - propyl pyridine 8-hydroxyquinoline N,” -ethylenebis( thioacetylacetoneimine) salicylaldehyde bis(salicy1idine)ethylenediamine N - (2-sulfidophenyl)salicylideneimine 1,4,7-triazacyclononane 1,4,7-triazacyclononane-N,”,N’’-triacetic acid 3,4-toluenedithiol 2,2 : 6’,2”-terpyridine tetrahydrofur an 2,3,5,64etramethylbenzenethiol trimethylphosphite tetramethylthiourea thiourea

COORDINATION CHEMISTRY OF TECHNETIUM

101

ACKNOWLEDGMENTS

I thank Dr. S. F. Colmanet for preparing the structural figures.

REFERENCES 1. Perrier, C., and Segre, E., Rend. Accad. N u . Lincei 25,723 (1937). 2. Perrier, C., and Segre, E., Nature 140, 193 (1937). 3. Schwochau, K., Radiochim. Actu 32, 139 (1983). 4. Kenna, B. T., and Kuroda, P. K., J. Inorg. Nucl. Chem. Lett. 23, 142 (1961). 5. Coursey, B. M., Gibson, J . A. B., Heitzmann, M. W., and Leak, J . C. Int. J. Appl. Radmt. Isot. 35, 1103 (1984). 6. Wapstra, A. H., and Audi, G., Nucl. Phys. A 432, 1 (1985). 7. Kotegov, K. V., Pavlov, 0. N., and Shvedov, V. P., Adu. Inorg. Chem. Radiochem. 11, li1968). 8. Rard, J. A,, Rand, M. H., Thornback, J. R., and Wanner, H., J. Chem. Phys. 94, 6336 (1991). 9. Deutsch, E.,Libson, K., Vanderheyden, J.-L., Ketring, A. R., and Maxon, H. R., Nucl. Med. Biol. 13,465 (1986). 10. Boag, N. M., and Kaesz, H. D. in Wilkinson, G., (Ed.), “Comprehensive Organometallic Chemistry” (G. Wilkinson, Ed.), Vol. 4,p. 161.Pergamon, New York, 1982. 11. Clarke, M.J., and Fackler, P. H., Struct. Bonding 50, 57 (1982). 12. Deutsch, E., Libson, K., Jurisson, S., and Lindoy, L. F., Prog. Inorg. Chem. 30, 75 (1983). 13. Deutsch, E., and Libson, K., Comments Inorg. Chem. 3,83 (1984). 14. Saha, G. B., “Fundamentals of Nuclear Pharmacy,” 3rd ed. Springer-Verlag, New York, 1992. 15. Clarke, M. J . , and Podbielski, L., Coord. Chem. Reu. 78,253 (1987). 16. Johannsen, B., and Spies, H., Isotopenpraris 24,449 (1988). 17. Dewanjee, M.K., Semin. Nucl. Med. 20, 5 (1990).

18. Steigman, J., and Eckelman, W. C., “The Chemistry of Technetium in Medicine.” National Academy Press, Washington, D.C., 1992. 19. Jurisson, S., Berning, D., Jia, W., and Ma, D., Chem. Rev. 93, 1137 (1993). 20. Richards, P., Tucker, W. D., and Srivastava, S. C . , Int. J . Appl. Radiat. Isot. 33,

793 (1982). 21. Boyd, R. E., Int. J. Appl. Radiat. h o t . 33, 801 (1982). 22. Deutsch, E.,Heineman, W. R., Zodda, J. P., Gilbert, T. W., and Williams, C. C.

Int. J . Appl. Radiat. Isot. 33, 843 (1982). 23. Bonnyman, J., Int. J. Appl. Radiat. Isot. 34,901 (1983). 24. Deutsch, E.,and Hirth, W., J. Nucl. Med. 28, 1491 (1987). 25. Eckelman, W. C., and Volkert, W. A., Int. J . Appl. Radiat. h o t . 33,945 (1982). 26. Colton, R., “The Chemistry of Rhenium and Technetium.” Interscience Publishers, London, 1965. 27. Peacock, R. D., “The Chemistry of Technetium and Rhenium.” Elsevier, Amsterdam, 1966. 28. Spitsyn, V. I., and Kuzina, A. F., “Technetium.” Nauka Publishers, Moscow, 1981 (in Russian).

102

JOHN BALDAS

29. “Gmelin Handbook of Inorganic Chemistry; Tc, Technetium,” 8th ed., Suppl. Vol.

1. Springer-Verlag, Berlin, 1982;Suppl. Vol. 2. Springer-Verlag, Berlin, 1983. 30. Deutsch, E., Nicolini, M., and Wagner, H. N., Jr. (Eds.),“Technetium in Chemistry and Nuclear Medicine” Cortina International, Verona/Raven Press, New York, 1983. 31. Nicolini, M., Bandoli, G . , and Mazzi, U. (Eds.), “Technetium in Chemistry and Nuclear Medicine 2” Cortina International, VeronaiRaven Press, New York, 1986. 32. Nicolini, M., Bandoli, G., and Mazzi, U. (Eds.),“Technetium and Rhenium in Chemistry and Nuclear Medicine 3” Cortina International, Verona/Raven Press, New York, 1990. 33. Peacock, R. D., in “Comprehensive Inorganic Chemistry” (Bailar, J . C., Emeleus, H. J., Nyholm, R., and Trotman-Dickenson, A. F., Eds., Vol. 3,p. 877.Pergamon, Oxford, 1973. 34. Jones, A. G . , and Davison, A., Znt. J . Appl. Radiat. Isot. 33, 867 (1982). 35. Davison, A., and Jones, A. G., Znt. J . Appl. Rudiut. Zsot. 33, 875 (1982). 36. Davison, A., in “Technetium in Chemistry and Nuclear Medicine” (Deutsch, E., Nicolini, M., and Wagner, H. N., Eds.), p. 3. Cortina International, VeronaiRaven Press, New York, 1983. 37. Bandoli, G., Mazzi, U., Roncari, E., and Deutsch, E., Coord. Chem. Rev.44, 191 (1982). 38. Melnik, M., and Van Lier, J . E., Coord. Chem. Rev. 77, 275 (1987). 39. Abram, U., and Kirmse, R., J . Radioanal. Nucl. Chern., Articles 122,311 (1988). 40. Kirmse, R., and Abram, U., Zsotopenpraxis 26, 151 (1990). 41. Raynor, J. B., Kemp, T. J., and Thyer, A. M., Znorg. Chirn. Actu 193, 191 (1992). 42. Spitsyn, V. I., Kuzina, A. F., Oblova, A. A., and Kryuchkov, S. V., Russ. Chern. Rev. (Engl. Transl.) 54, 373 (1985). 43. Schwochau, K., Fortschr. Chem. Forsch. 96,109 (1981). 44. O’Connell, L. A., Pearlstein, R. M., Davison, A,, Thornback, J . R., Kronauge, J. F., and Jones, A. G., Znorg. Chim. Acta 161, 39 (1989). 45. Hileman, J . C., Huggins, D. K., and Kaesz, H. D., Znorg. Chern. 4,933 (1962). 46. Michels, G.D., and Svec, H. J., Znorg. Chern. 20, 3445 (1981). 47. Lindauer, M. W., Evans, G. O., and Sheline, R. K., Inorg. Chem. 7, 1249 (1968). 48. Hileman, J . C., Huggins, D. K., andKaesz, H. D., J . A m . Chem. Soc. 83,2953 (1961). 49. Hieber, W., and Herget, C., Angew. Chern. 73, 579 (1961). 50. Calderazzo, F.,Mazzi, U., Pampaloni, G., Poli, R., Tisato, F., and Zanazzi, P. F., Gatz. Chirn. Ztul. 119,241 (1989). 51. Bailey, M. F., and Dahl, L. F., Znorg. Chern. 4, 1140 (1965). 52. Bor, G., and Sbrignadello, G., J . Chem. Soc., Dalton Trans., 440 (1974). 53. Findeisen, M., Kaden, L., Lorenz, B., Rummel, S., and Wahren, M., Znorg. Chirn. Acta 128, L15 (1987). 54. Muto, T., and Ebihara, H., J . Znorg. Nucl. Chem. 43,2617 (1981). 55. Sbrignadello, G.,Tomat, G., Magon, L., and Bor, G., Inorg. Nucl. Chem. Lett. 9, 1073 (1973). 56. Sbrignadello, G., Znorg. Chim. Actu 48, 237 (1981). 57. Kaesz, H. D., and Huggins, D. K., Can. J . Chem. 41, 1250 (1963). 58. Grimm, C. C., and Clark, R. J., Orgunornetullics 9, 1123 (1990). 59. Grimm, C. C., Clark, R. J., and Rosanske, R., Znorg. Chirn. Actu 159, 137 (1989). 60. Klimov, V. D., Mamchenko, A. V., and Babichev, A. P., Radiokhimiya 33(5),33 (1991).

COORDINATION CHEMISTRY OF TECHNETIUM

103

61. Fawcett, J. P., and Poe, A,, J . Chem. Soc., Dalton Trans. 2039 (1976). 62. Kanellakopulos, B., Nuber, B., Raptis, K., and Ziegler, M. L., 2. Naturforsch. B

46,55 (1991). 63. Kaden, L., Lorenz, B., Schmidt, K., Sprinz, H., and Wahren, M., Z . Chem. 19,

305 (1979). 64. Struchkov, Yu. T., Bazanov, A. S., Kaden, L., Lorenz, B., Wahren, M., and Meyer,

H.,Z.Anorg. Allg. Chem. 494,91 (1982). 65. Hieber, W., Lux, F., and Herget, C., 2. Naturforsch. B 20, 1159 (1965). 66. Borisova, I. V., Miroslavov, A. E., Sidorenko, G. V., and Suglobov, D. N.,Radiokhim-

iya 33(6),l(1991). 67. Knight Castro, H. H., Hissink, C. E., Teuben, J. H., Vaalburg, W., and Panek, K.,

Rec. Trau. Chim. Pays-Bas 111, 105 (1992). I. V., Miroslavov, A. E., Sidorenko, G. V., Suglobov, D. N., and Shcherbakova, L. L., Radiokhimiya 33(4),27 (1991). 69. Kaden, L., Lorenz, B., Rummel, S., Schmidt, K., and Wahren, M., Inorg. Chim. Acta 142, l(1988). 70. Lorenz, B., Findeisen, M., Olk, B., and Schmidt, K., 2.Anorg. Allg. Chem. 566, 160 (1988). 71. Borisova, I. V., Miroslavov, A. E., Sidorenko, G. V., and Suglobov, D. N., Radiokhimiya 33(3),1 (1991). 72. Mazzi, U., Bismondo, A,, Kotsev, N., and Clemente, D. A., J . Organomet. Chem. 135, 177 (1977). 73. Joachim, J . E., Apostolidis, C., Kanellakopulos, B., Maier, R., Marques, N., Meyer, D., Muller, J . , Pires de Matos, A., Nuber, B., Rebizant, J., and Ziegler, M. L., J . Organomet. Chem. 448, 119 (1993). 74. Alberto, R., Herrmann, W. A., Kiprof, P., and Baumgartner, F., Znorg. Chem. 31, 895 (1992). 75. Findeisen, M., Kaden, L., Lorenz, B., and Wahren, M., Znorg. Chim. Acta 142, 3 (1988). 76. Grigoriev, M. S., and Kryutchkov, S. V., Radiochim. Acta 63, 187 (1993). and Utilization of Technetium ’93,Sendai, Japan,” p. 23.[Abstr.] 77. Kaden, L., Findeisen, M., Lorenz, B., Schmidt, K., and Wahren, M., Isotopenprazis 27, 265 (1991). 78. Lorenz, B.,Findeisen, M., and Schmidt, K., Zsotopenprazis 27, 266 (1991). 79. Biagini Cingi, M., Clemente, D. A., Magon, L., and Mazzi, U., Znorg. Chim. Acta 13,47 (1975). 80. Marchi, A., Rossi, R., Duatti, A., Zucchini, G. L., and Magon, L., Transition Met. Chem. 11, 164 (1986). 81. Rossi, R., Marchi, A., Magon, L., Duatti, A., Casellato, U., and Graziani, R., Inorg. Chim. Acta 160,23 (1989). 82. Marchi, A,, Rossi, R., Duatti, A., Magon, L., Bertolasi, V., Ferretti, V., and Gilli, G., Inorg. Chem. 24, 4744 (1985). 83. Rossi, R., Marchi, A,, Magon, L., Casellato, U., and Graziani, R., J . Chem. Res. ( S ) ,78,(1990). 84. Mazzi, U., Roncari, E., Seeber, R., and Mazzocchin, G. A., Znorg. Chim. Acta 41, 95 (1980). 85. Kaden, L., Findeisen, M., Lorenz, B., Schmidt, K., and Wahren, M., Znorg. Chim. Acta 193,213 (1992). 86. Miroslavov, A. E., Sidorenko, G. V., Borisova, I. V., Legin, E. K., Lychev, A. A., and Suglobov, D. N., Radiokhimiya 32(6),14 (1990). 68. Borisova,

104

JOHN BALDAS

87. El-Sayed, M. A., and Kaesz, H. D., Znorg. Chem. 2, 158 (1963). 88. Miroslavov, A. E., Sidorenko, G. V., Borisova, I. V., Legin, E. K., Lychev, A. A,, Suglobov, D. N., and Adamov, V. M., Radiokhimiya 32(4), 6 (1990). 89. Miroslavov, A. E., Borisova, I. V., Sidorenko, G. V., and Suglobov, D. N., Radiokhimiya 33(6), 14 (1991). 90. Hieber, W., Opavsky, W., and Rohm, W., Chem. Ber. 101,2244 (1968). 91. Hieber, W., and Opavsky, W., Chem. Ber. 101, 2966 (1968). 92. Borisova, I. V., Miroslavov, A. E., Sidorenko, G. V., and Suglobov, D. N., in “Technetium and Rhenium in Chemistry and Nuclear Medicine 3” (Nicolini, M., Bandoli, G., and Mazzi, U., Eds.), p. 337. Cortina International VeronaiRaven Press, New York, 1990. 93. Adamov, V. M., Belyaev, B. N., Borisova, I. V., Miroslavov, A. E., Sidorenko, G. V., and Suglobov, D. N., Radiokhimiya 33(4),38 (1991). 94. Herrmann, W. A,, Alberto, R., Bryan, J . C., and Sattelberger, A. P., Chem. Ber. 124, 1107 (1991). 95. Miroslavov, A. E., Sidorenko, G. V., Borisova, I. V., Legin, I. K., Lychev, A. A,, and Suglobov, D. N., Radiokhimiya 31(6),33 (1989). 96. Copp, S. B., Subramanian, S., and Zaworotko, M. J . , Angew, Chem. Znt. Ed. Engl. 32, 706 (1993). 97. Tsutsui, M., Hrung, C. P., Ostfeld, D., Srivastava, T. S., Cullen, D. L., and Meyer, E. F., J . Am. Chem. SOC.97, 3952 (1975). 98. Tsutsui, M., and Hrung, C. P., J. A m . Chem. SOC.95, 5777 (1973). 99. Palm, C., Fischer, E. O., and Baumgartner, F., Naturwiss. 49, 279 (1962). 100. Knight Castro, H. H., Meetsma, A,, Teuben, J . H., Vaalburg, W., Panek, K., and Ensing, G., J . Organomet. Chem. 410, 63 (1991). 101. Raptis, K., Dornberger, E., Kanellakopulos, B., Nuber, B., and Ziegler, M. L., J . Organomet. Chem. 408, 61 (1991). 102. Fischer, E. O., and Fellmann, W., J . Organomet. Chem. 1, 191 (1963). 103. Wenzel, M., and Saidi, M., J . Labelled Comp. Radiopharm. 33, 77 (1993). 104. Raptis, K., Kanellakopulos, B., Nuber, B., and Ziegler, M. L., J . Organomet. Chem. 405, 323 (1991). 105. Palm, C., and Fischer, E. O., Tetrahedron Lett., 253 (1962). 106. Fischer, E. O., and Schmidt, M. W., Chern. Ber. 102, 1954 (1969). 107. Wester, D. W., Coveney, J . R., Nosco, D. L., Robbins, M. S., and Dean, R. T., J . Med. Chem. 34, 3284 (1991). 108. Schwochau, K., and Herr, W., Z. Anorg. Allg. Chem. 319, 148 (1962). 109. Krasser, W., Bohres, E. W., and Schwochau, K., 2. Nuturforsch. A 27, 1193 (1972). 110. Holman, B. L., Jones, A. G., ListerJames, J., Davison, A., Abrams, M. J., Kirshenbaum, J. M., Tumeh, S. S., and English, R. J., J . Nucl. Med. 25, 1350 (1984). 111. Abrams, M. J . , Davison, A., Faggiani, R., Jones, A. G . , and Lock, C. J. L., Znorg. Chem. 23,3284 (1984). 112. Abrams, M. J . , Davison, A., Jones, A. G., Costello, C. E., and Pang, H., Znorg. Chem. 22, 2798 (1983). 113. Findeisen, M., Lorenz, B., and Abram, U., Z. Chem. 29, 29 (1989). 114. Tulip, T. H., Calabrese, J., Kronauge, J. F., Davison, A,, and Jones, A. G., in “Technetium in Chemistry and Nuclear Medicine 2” (Nicolini, M., Bandoli, G., and Mazzi, U., Eds.), p. 119. Cortina International, Verona/Raven Press, New York, 1986. 115. Kronauge, J . F., Kawamura, M., Lepisto, E., Holman, B. L., Davison, A,, Jones, A. G., Costello, C. E., and Zeng, C.-H., in “Technetium and Rhenium i n Chemistry

COORDINATION CHEMISTRY OF TECHNETIUM

105

and Nuclear Medicine 3” (Nicolini, M., Bandoli, G., and Mazzi, U., Eds.), p. 677. Cortina International, Verona/Raven Press, New York, 1990. 116. Kronauge, J . F., Davison, A., Roseberry, A. M., Costello, C. E., Maleknia, S., and Jones, A. G., Znorg. Chem. 30, 4265 (1991). 117. Abram, U., Beyer, R., Munze, R., Findeisen, M., and Lorenz, B., Znorg. Chim. Acta 160, 139 (1989). 118. O’Connell, L. A,, and Davison, A., Znorg. Chim. Acta 176, 7 (1990). 119. Abram, U., Abram, S., Beyer, R., Muenze, R., Kaden, L., Lorenz, B., and Findeisen, M., Znorg. Chim. Acta 148, 141 (1988). 120. O’Connell, L. A., Dewan, J., Jones, A. G., and Davison, A., Znorg. Chem. 29, 3539 (1990). 121. Farr, J . P., Abrams, M. J., Costello, C. E., Davison, A., Lippard, S. J., and Jones, A. G., Organometallics 4, 139 (1985). 122. Joachim, J. E., Apostolidis, C . , Kanellakopulos, B., Maier, R., Meyer, D., Rebizant, J., and Ziegler, M. L., J . Organomet. Chem. 455, 137 (1993). 123. Vanderheyden, J.-L., Ketring, A. R., Libson, K., Heeg, M. J., Roecker, L., Motz, P., Whittle, R., Elder, R. C., and Deutsch, E., Znorg. Chem. 23, 3184 (1984). 124. Ichimura, A,, Heineman, W. R., Vanderheyden, J.-L., and Deutsch, E.,Znorg. Chem. 23, 1272 (1984). 125. Roodt, A,, Sullivan, J. C., Meisel, D., and Deutsch, E., Znorg. Chem. 30,4545 (1991). 126. Doyle, M. N., Libson, K., Woods, M., Sullivan, J . C., and Deutsch, E., Inorg. Chem. 25, 3367 (1986). 127. Libson, K., Woods, M., Sullivan, J. C., Watkins, J. W., Elder, R. C., and Deutsch, E., Znorg. Chem. 27, 999 (1988). 128. Abram, U., Beyer, R., Munze, R., Stach, J., Kaden, L., Lorenz, B., and Findeisen, M., Polyhedron 8, 1201 (1989). 129. Wester, D. W., White, D. H., Miller, F. W., Dean, R. T.,Znorg. Chem. 23,1501 (1984). 130. Abram, U., and Abram, S., 2. Chem. 26, 181 (1986). 131. Wester, D. W., White, D. H., Miller, F. W., Dean, R. T., Schreifels, J. A., and Hunt, J. E., Znorg. Chim. Acta 131, 163 (1987). 132. Davison, A,, Kronauge, J. F., Jones, A. G., Pearlstein, R. M., and Thornback, J. R., Inorg. Chem. 27,3245 (1988). 133. Linse, K. H., Kremer, C., Su, Z. F., and Schwochau, K., in “Topical Symposium on the Behavior and Utilization of Technetium ’93, Sendai, Japan,” p. 59 [Abstr.] 134. Eakins, J . D., Humphreys, D. G., and Mellish, C. E., J. Chem. Soc., 6012 (1963). 135. Armstrong, R. A., and Taube, H., Znorg. Chem. 15, 1904 (1976). 136. Radonovich, L. J., and Hoard, J. L., J . Phys. Chem. 88,6711 (1984). 137. Roseberry, A. M., Davison, A., and Jones, A. G., Znorg. Chim. Acta 176,179 (1990). 138. Linder, K. E., Davison, A., Dewan, J . C., Costello, C. E., and Maleknia, S., Znorg. Chem. 25, 2085 (1986). 139. Orvig, C., Davison, A., and Jones, A. G., J . Labelled Comp. Radiopharm. 18, 148 (1981). 140. Archer, C. M., Dilworth, J. R., Jobanputra, P., Thompson, R. M., McPartlin, M., and Hiller, W., J . Chem. Soc., Dalton, Trans. 897 (1993). 141. Baldas, J., Colmanet, S. F., Raptis, R. G., and Williams, G. A., unpublished results. 142. Chatt, J., and Dilworth, J. R., J . Chem. Soc., Chem. Comm., 508 (1974). 143. Hiller, W., Hubener, R., Lorenz, B., Kaden, L., Findeisen, M., Stach, J., and Abram, U., Znorg. Chim. Acta 181, 161 (1991). 144. Lu, J., and Clarke, M. J., Znorg. Chem. 29,4123 (1990). 145. Lorenz, B., Schmidt, K., Kaden, L., and Wahren, M., Zsotopenpraxis 22,444 (1986).

106

JOHN BALDAS

146. Archer, C. M., Dilworth, J. R., Kelly, J. D., and McPartlin, M., J . Chem. SOC., Chem. Comm., 375 (1989). 147. Kryuchkov, S. V., Kuzina, A. F., and Spitsyn, V. I., Dokl. Akad. Nauk SSSR 266, 127 [Engl. 3041 (1982). 148. Kryuchkov, S. V., Grigor’ev, M. S., Kuzina, A. F., Gulev, B. F., and Spitsyn, V. I., Dokl. Akud. Nauk SSSR 288,389 [Engl. 147J(1986). 149. Cotton, F. A,, Daniels, L. M., Falvello, L. R., Grigoriev, M. S., and Kryuchkov, S. V., Inorg. Chin. Acta 189, 53 (1991). 150. Kryuchkov, S. V., Rakitin, Yu. V., Kazin, P. E., Zhirov, A. I., and Konstantinov, N. Yu., Koord. Khim. 16, 1230 [Engl. 6601 (1990). 151. Cotton, F. A., and Bratton, W. K., J.A m . Chem. SOC.87, 921 (1965). 152. Bratton, K. W., and Cotton, F. A., Inorg. Chem. 9, 789 (1970). 153. Koz’min, P. A., and Novitskaya, G. N., Russ. J . Inorg. Chem. (Engl. Trunsl.) 17, 1652 (1972). 154. Cotton, F. A,, and Shive, L. W., Inorg. Chem. 14, 2032 (1975). 155. Cotton, F. A., Davison, A., Day, V. W., Fredrich, M. F., Orvig, C., and Swanson, R., Inorg. Chem. 21, 1211 (1982). 156. Grigor’ev, M. S., Kryuchkov, S. V., Struchkov, Yu. T., and Yanovskii, A. I., Koord. Khim. 16, 90 (1990). 157. Kryutchkov, S. V., Grigoryev, M. S., and German, K. E., in “Technetium and Rhenium in Chemistry and Nuclear Medicine 3” (Nicolini, M., Bandoli, G., and Mazzi, U., Eds.), p. 253. Cortina International, VeronaiRaven Press, New York, 1990. 158. Cotton, F. A., and Pedersen, E., Znorg. Chem. 14, 383 (1975). 159. Cotton, F. A., and Walton, R. A,, “Multiple Bonds between Metal Atoms.” Wiley, New York, 1982. 160. Cotton, F. A., and Kalbacher, B. J., Inorg. Chem. 16,2386 (1977). 161. Cotton, F. A,, Fanwick, P. E., Gage, L. D., Kalbacher, B., and Martin, D. S., J . A m . Chem. SOC.99, 5642 (1977). 162. Kryuchkov, S . V., Kuzina, A. F., and Spitsyn, V. I., Russ. J . Inorg. Chem. (Engl. Transl.)28, 1124 (1983). 163. Spitsyn, V. I., Kuzina, A. F., Kryuchkov, S. V., and Simonov, A. E., Russ. J . Inorg. Chem. (Engl. Transl.)32, 1278 (1987). 164. Peters, G., Skowronek, J., and Preetz, W., 2. Naturforsch. A 47, 591 (1992). 165. Spitsyn, V. I., Baierl, B., Kryuchkov, S. V., Kuzina, A. F., and Varen, M., Dokl. Akud. Nauk SSSR 256,608 [Engl. 301 (1981). 166. Cotton, F . A., Fanwick, P. E., and Gage, L. D., J . A m . Chem. SOC.102, 1570 (1980). 167. Koz’min, P. A,, Larina, T. B., and Surazhskaya, M. D. Koord. Khim. (a) 9, 1114 (1983); (b) 7, 1719 (1981); (c) 8, 851 (1982). 168. Rakitin Yu. V., and Nefedov, V. I., Russ. J . Inorg. Chem. (Engl. Trunsl.)29, 294 (1984). 169. Rakitin, Yu. V., Kryuchkov, S. V., Aleksandrov, A. I., Kuzina, A. F., Nemtsev, N. V., Ershov, B. G.; and Spitsyn, V. I. Dokl. Akad. Nauk SSSR 269, 1123 [Engl. 2531 (1983). 170. Kryuchkov, S. V., Kuzina, A. F., and Spitsyn, V. I., Dokl. Akad. Nauk SSSR 287, 1400 [Engl. 1121 (1986). 171. German, K. E., Kryuchkov, S. V., Kuzina, A. F., and Spitsyn, V. I. Dokl. Akad. Nauk SSSR 288,381 [Engl. 1391 (1986). 172. Spitzin, V. I., Kryutchkov, S. V., Grigoriev, M. S., and Kuzina, A. F., 2. Anorg. Allg. Chem. 563, 136 (1988).

COORDINATION CHEMISTRY OF TECHNETIUM

107

173. Kryutchkov, S. V., Kuzina, A. F., and German, K. E., in “Technetium and Rhenium

in Chemistry and Nuclear Medicine 3” (M. Nicolini, G. Bandoli, and U. Mazzi, Eds.), p. 275. Cortina International, Verona/Raven Press, New York, 1990. 174. Koz’min, P. A., Surazhskaya, M. D., and Larina, T. B.,Koord.Khim. 11,1559 (1985). 175. Kryuchkov, S. V., Grigor’ev, M. S., Yanovskii, A. I., Struchkov, Yu. T., and Spitsyn, V. I., Dokl. Akad. Nauk SSSR 297,867 [Engl. 5201 (1987). 176. Kryuchkov, S. V., Grigor’ev, M. S., Kuzina, A. F., Gulev, B. F., and Spitsyn, V. I., Dokl. Akad. Nauk SSSR 290,866 [Engl. 3961 (1986). 177. Koz’min, P. A., Surazhskaya, M. D., and Larina, T. B. Dokl. Akad. Nauk SSSR 265, 1420 [Engl. 6561 (1982). 178. Kryuchkov, S. V., Grigor’ev, M. S., Kuzina, A. F., Gulev, B. F., and Spitsyn, V. I., Dokl. Akad. Nauk SSSR 288, 893 [Engl. 1721 (1986). 179. Kryuchkov, S. V., Grigor’ev, M. S., Yanovskii, A. I., Struchkov, Yu. T., and Spitsyn, V. I., Dokl. Akad. Nauk SSSR 301, 618 [Engl. 2191 (1988). 180. Gerasimov, V. N., Kryutchkov, S. V., German, K. E., Kulakov, V. M., and Kuzina, A. F. in “Technetium and Rhenium in Chemistry and Nuclear Medicine 3” (M. Nicolini, G. Bandoli, and U. Mazzi, Eds.), p. 231. Cortina International, Verona/ Raven Press, New York, 1990. 181. Kryutchkov, S. V., Kuzina, A. F., and Spitzin, V. I. 2.Anorg. Allg. Chem. 563, 153 (1988). 182. Wheeler, R. A., and Hoffmann, R., J . Am. Chem. SOC.108,6605 (1986). 183. Bronger, W., Kanert, M., Loevenich, M., Schmitz, D., and Schwochau, K., Angew.

Chem. Int. Ed. Engl. 32, 576 (1993). 184. Archer, C. M., Dilworth, J . R., Kelly, J. D., and McPartlin, M., Polyhedron 8,

1879 (1989). 185. Archer, C. M., Dilworth, J. R., Thompson, R. M., McPartlin, M., Povey, D. C., and 186. 187. 188. 189.

190. 191. 192. 193. 194. 195. 196. 197.

Kelly, J. D., J . Chem. Soc., Dalton Trans., 461 (1993). Kirmse R., and Abram, U., 2.Anorg. Allg. Chem. 608, 184 (1992). Wilcox, B. E., Ho, D. M., and Deutsch, E., Inorg. Chem. 28, 3917 (1989). Thomas, J. A., Davison, A., and Jones, A. G., Inorg. Chim. Actu 184, 99 (1991). Conner, K. A., and Walton, R. A., in “Comprehensive Coordination Chemistry” (G. Wilkinson, R. D. Gillard, and J. A. McCleverty, Eds.), Vol. 4, p. 125. Pergamon Press, Oxford, 1987. Fergusson, J . E., and Nyholm, R. S., Nature 183, 1039 (1959). Fergusson, J. E., and Nyholm, R. S., Chem. Znd., 347 (1960). Fergusson, J. E., and Hickford, J. H., Aust. J . Chem. 23,453 (1970). Libson, K., Doyle, M. N., Thomas, R. W., Nelesnik, T., Woods, M., Sullivan, J . C., Elder, R. C., and Deutsch, E., Znorg. Chem. 27,3614 (1988). Bandoli, G., Mazzi, U., Ichimura, A., Libson, K., Heineman, W. R., and Deutsch, E., Inorg. Chem. 23,2898 (1984). Okamoto, K.-I., Chen, B., Kirchhoff, J. R., Ho, D. M., Elder, R. C., Heineman, W. R., and Deutsch, E., Polyhedron 12, 1559 (1993). Rochon, F. D., Melanson, R., and Kong, P.-C., Inorg. Chim. Actu 194,43 (1992). Seifert, S., Muenze, R., Leibnitz, P., Reck, G., and Stach, J., Znorg. Chim. Acta 193, 167 (1992).

198. Mazzi, U., Clemente, D. A., Bandoli, G., Magon, L., and Orio, A. A., Inorg. Chem.

16, 1042 (1977). 199. Konno, T., Heeg, M. J., and Deutsch, E., Znorg. Chem. 28, 1694 (1989). 200. Konno, T., Heeg, M. J., Stuckey, J . A., Kirchhoff, J. R., Heineman, W. R., and

Deutsch, E., Inorg. Chem. 31, 1173 (1992).

108

JOHN BALDAS

201. White, D. J . , Kuppers, H.-J., Edwards, A. J., Watkin, D. J . , and Cooper, S. R., Inorg. Chem. 31, 5351 (1992). 202. Kirmse, R., Stach, J . , Abram, U., and Marov, I. N., 2. Anorg. Allg. Chem. 518, 210 (1984). 203. Baldas, J., Boas, J . F., Bonnyman, J., and Williams, G. A,, J . Chem. Soc., Dalton Trans., 827 (1984). 204. Brown, D. S., Newman, J. L., Thornback, J. R., and Davison, A,, Acta Cryst. C 43, 1692 (1987). 205. Newman, J. L., Thornback, J. R., and Nowotnik, D. P., J . Labelled Comp. Radiopharm. 26,24 (1989). 206. Kirmse, R., Lorenz, B., and Schmidt, K., Polyhedron 2, 935 (1983). 207. Pearlstein, R. M., Davis, W. M., Jones, A. G., and Davison, A., Inorg. Chem. 28, 3332 (1989). 208. Brown, D. S., Newman, J . L., Thornback, J. R., Pearlstein, R. M., Davison, A,, and Lawson, A,, Inorg. Chim. Acta 150, 193 (1988). 209. Brown, D. S., Newman, J. L., and Thornback, J . R., Acta Cryst. C 44, 973 (1988). 210. Cheah, C. T., Newman, J . L., Nowotnik, D. P., and Thornback, J. R., Nucl. Med. Biol. 14, 573 !1987). 21 1. Abram, U.,Hubener, R., Wollert, R., Kirmse, R., and Hiller, W., Inorg. Chim. Acta 206, 9 (1993). 212. Kaden, L., Lorenz, B., Kirmse, R., Stach, J.,Behm, H., Beurskens, P. T., and Abram, U.,Znorg. Chim. Acta 169, 43 (1990). 213. Abram, U., Kirmse, R., Kohler, K., Lorenz, B., and Kaden, L., Inorg. Chim. Acta 129, 15 (1987). 214. Abram, U., Hartung, J., Beyer, L., Kirmse, R., and Kohler, K., 2. Chem. 27, 101 (1987). 215. Kirmse, R., Stach, J . , and Abram, U.,Polyhedron 4, 1275 (1985). 216. Yang, G. C., Heitzmann, M. W., Ford, L. A., and Benson, W. R., Inorg. Chem. 21, 3242 (1982). 217. Kirmse, R., and Abram, U.,Z . Anorg. Allg. Chem. 573, 63 (1989). 218. Bandoli, G., Clemente, D. A,, and Mazzi, U., J . Chem. SOC.,Dalton Trans., 373 (1978). 219. Baldas, J., Bonnyman, J.,Pojer, P. M., Williams, G. A,, and Mackay, M. F.,J. Chem. SOC.,Dalton Trans., 451 (1982). 220. Abram, U.,and Spies, H., Inorg. Chim. Acta 94, L3 (1984). 221, Bonnyman, J., personal communication. 222. Baldas, J., and Bonnyman, J., Nucl. Med. Biol. 19, 741 (1992). 223. de Vries, N., Dewan, J. C., Jones, A. G., and Davison, A,, Inorg. Chem. 27, 1574 (1988). 224. Fischer, E. O., and Schmidt, M. W., Angew. Chem. Int. Ed. Engl. 6,93 (1967). 225. Apostolidis, C., Kanellakopulos, B., Maier, R., Rebizant, J., and Ziegler, M. L., J . Organomet. Chem. 396,315 (1990); 411, 171 (1991). 226. Kanellakopulos, B., Nuber, B., Raptis, K., and Ziegler, M. L., Angew. Chem. Znt. Ed. Engl. 28, 1055 (1989). 227. Chan, A. W. E., Hoffmann, R., and Alvarez, S., Inorg. Chem. 30, 1086 (1991). 228. Herrmann, W. A,, Alberto, R., Kiprof, P., and Baumgartner, F., Angew. Chem. Znt. Ed. Engl. 29, 189 (1990). 229. Trop, H. S., Jones, A. G., and Davison, A., Inorg. Chem. 19, 1993 (1980). 230. Trop, H. S., Davison, A,, Jones, A. G., Davis, M. A., Szalda, D. J . , and Lippard, S. J., Inorg. Chem. 19, 1105 (1980).

COORDINATION CHEMISTRY OF TECHNETIUM

231. 232. 233. 234. 235. 236. 237. 238. 239. 240. 241. 242. 243. 244. 245. 246. 247. 248. 249. 250. 251. 252. 253.

254. 255. 256. 257.

258. 259.

109

Huber, E. W., Heineman, W. R., and Deutsch, E.,Inorg. Chem. 26,3718 (1987). Preetz, W., and Peters, G., 2.Naturforsch. B . 35, 797 (1980). Cotton, F. A., Daniels, L., Davison, A., and Orvig, C., Inorg. Chem. 20,3051 (1981). Cotton, F. A,, and Gage, L. D., Nouu. J . Chern. 1, 441 (1977). Kryuchkov, S.V., and Simonov, A. E., Koord. Khim. 16,339 (1990). Skowronek, J., and Preetz, W., 2.Naturforsch. B 47,482 (1992). Skowronek, J., Preetz, W., and Jessen, S. M., 2.Naturforsch. B 46, 1305 (1991). Zaitseva, L. L., Kotel’nikova, A. S., and Rezvov, A. A., Russ. J . Inorg. Chem. (Engl. Transl.)25, 1449 (1980). Baturin, N. A., German, K. E., Grigor’ev, M. S., and Kryuchkov, S. V., Koord. Khim. 17, 1375 (1991). Skowronek, J., and Preetz, W., 2.Anorg. Allg. Chem. 615, 73 (1992). Hashimoto, M., Wada, H., Omori, T., and Yoshihara, K., J . Radioanal. Nucl. Chern., Lett. 153, 283 (1991). Mazzi, U., Roncari, E., Bandoli, G., and Magon, L., Transition Met. Chem. 4, 151 (1979). Patterson, G.S., Davison, A., Jones, A. G . , Costello, C. E., and Maleknia, S. D., Inorg. Chim. Acta 114, 141 (1986). Leesmeister, K., and Schwochau, K., Nucl. Med. Biol. 19, 73 (1992). Mutalib, A., Omori, T., and Yoshihara, K., J . Radioanal. Nucl. Chem., Lett. 165, 19 (1992). Hashimoto, K., Kabuto, C., Omori, T., and Yoshihara, K., Chern. Lett., 1379 (1988). Bandoli, G., Clemente, D. A., and Mazzi, U., J . Chem. SOC.,Dalton Trans., 1837 (1977). Bandoli, G., Clemente, D. A., Mazzi, U., and Roncari, E., Acta Cryst. B 34, 3359 (1978). Kido, H., and Hatakeyama, Y., Inorg. Chem. 27, 3623 (1988). Mutalib, A., Omori, T., and Yoshihara, K., J . Radioanal. Nucl. Chem., Articles 170, 67 (1993). Spies, H., Abram, U., Uhlemann, E., and Ludwig, E., Inorg. Chim. Acta 109, L3 ( 1985). Bandoli, G., Mazzi, U., Spies, H., Miinze, R., Ludwig, E., Ulhemann, E., Scheller, D., Inorg. Chim. Acta 132, 177 (1987). Linder, K. E., Jurisson, S., Francesconi, L., Nowotnik, D. P., Eckelman, W. C., and Nunn, A. D., in “Technetium and Rhenium in Chemistry and Nuclear Medicine 3” (M. Nicolini, G. Bandoli, and U. Mazzi, Eds.), p. 195. Cortina International, VeronalRaven Press, New York, 1990. Treher, E. N., Francesconi, L. C., Gougoutas, J . Z., Malley, M. F., and Nunn, A. D., Inorg. Chern. 28,3411 (1989). Linder, K.E., Malley, M. F., Gougoutas, J . Z., Unger, S. E., and Nunn, A. D., Inorg. Chem. 29,2428 (1990). Jurisson, S.S., Hirth, W., Linder, K. E., Di Rocco, R. J., Narra, R. K., Nowotnik, D. P., and Nunn, A. D., Nucl. Med. Biol. 18, 735 (1991). Jurisson, S., Linder, K. E., Hirth, W., Feld, T., Nowotnik, D. P., Nunn, A. D., and Eckelman, W. C., in “Technetium and Rhenium i n Chemistry and Nuclear Medicine 3” (M. Nicolini, G. Bandoli, and U. Mazzi, Eds.), p. 189. Cortina International, VeronaiRaven Press, New York, 1990. Thompson, M., Nunn, A. D., and Treher, E. N.,Anal. Chem. 58,3100 (1986). Cyr, J . E., Linder, K. E., and Nowotnik, D. P., Inorg. Chim. Acta 206, 97 (1993).

110

JOHN BALDAS

260. Su, Z. F., Linse, K. H., Steinmetz, H. J., and Schwochau, K., J . Labelled Comp. Radiopharm. 31, 61 (1992). 261. Deutsch, E., Elder, R. C . , Lange, B. A., Vaal, M. J . , and Lay, D. G., Proc. Natl. Acad. Sci. USA 73, 4287 (1976). 262. Linder, K. E., Nowotnik, D. P., Malley, M. F., Gougoutas, J . Z., and Nunn, A. D., Inorg. Chim. Acta 190, 249 (1991). 263. Jurisson, S., Francesconi, L., Linder, K. E., Treher, E., Malley, M. F., Gougoutas, J. Z., and Nunn, A. D., Znorg. Chem. 30, 1820 (1991). 264. Jurisson, S.S., Dancey, K., McPartlin, M., Tasker, P. A., and Deutsch, E., Znorg. Chem. 23,4734 (1984). 265. Ichimura, A,, Heineman, W. R., and Deutsch, E., Inorg. Chem. 24, 2134 (1985). 266. Vavouraki, H., Papadopoulos, M., Mastrostamatis, S., Varvarigou, A. D., Psilla, M., and Chiotellis, E., J . Labelled Comp. Radiopharm. 33, 249 (1993). 267. Mazzi, U., Refosco, F., Tisato, F., Bandoli, G., and Nicolini, M., J . Chem. Soc., Dalton Trans., 847 (1988). 268. Marchi, A., Rossi, R., Magon, L., Duatti, A., Pasqualini, R., Ferretti, V., and Bertolasi, V., J . Chem. Soc., Dalton Trans., 1411 (1990). 269. Duatti, A,, Marchi, A., Luna, S.A., Bandoli, G., Mazzi, U., and Tisato, F., J . Chem. Soc., Dalton Trans., 867 (1987). 270. Tisato, F., Refosco, F., Mazzi, U., Bandoli, G., and Nicolini, M., Inorg. Chcm. Acta 157, 227 (1989). 271. Lu, J., Yarnano, A., and Clarke, M. J., Inorg. Chem. 29, 3483 (1990). 272. Breikss, A. I., Nicholson, T., Jones, A. G., and Davison, A., Inorg. Chem. 29, 640 (1990). 273. Wilcox, B. E., Ho, D. M., and Deutsch, E., Inorg. Chem. 28, 1743 (1989). 274. Wilcox, B. E., and Deutsch, E., Znorg. Chem. 30, 688 (1991). 275. Abrams, M. J., Davison, A., and Jones, A. G., Znorg. Chim. Acta 82, 125 (1984). 276. Kremer, C., and Kremer, E., J . Radioanal. Nucl. Chem., Lett. 175,445 (1993). 277. Nicholson, T., de Vries, N., Davison, A., and Jones, A. G., Inorg. Chem. 28, 3813 (1989). 278. Archer, C. M., Dilworth, J. R., Jobanputra, P., Thompson, R. M., McPartlin, M., Povey, D. C., Smith, G. W., and Kelly, J. D., Polyhedron 9, 1497 (1990). 279. Dilworth, J .R., Jobanputra, P., Thompson, R. M., Archer, C . M., Kelly, J . D., and Hiller, W., Z. Naturforsch. B 46, 449 (1991). 280. Mazzi, U., De Paoli, G., Di Bernardo, P., and Magon, L., J . Znorg. Nucl. Chem. 38, 721 (1976). 281. Fergusson, J. E., and Heveldt, P. F., J . Inorg. Nucl. Chem. 38, 2231 (1976). 282. Bandoli, G., Clemente, D. A., and Mazzi, U., J . Chem. SOC.,Dalton Trans., 125 (1976). 283. Mazzocchin, G. A,, Seeber, R., Mazzi, U., and Roncari, E., Znorg. Chim. Acta 29, 1 (1978);5 (1978). 284. Watson, P. L., Albanese, J. A., Calabrese, J. C., Ovenall, D. W., and Smith, R. G., Znorg. Chem. 30,4638 (1991). 285. Rochon, F. D., Melanson, R., and Kong, P.-C., Can. J . Chem. 69, 397 (1991). 286. Deutsch, E., Bushong, W., Glavan, K. A., Elder, R. C., Sodd, V. J., Scholz, K. L., Fortman, D. L., and Lukes, S. J . , Science 214, 85 (1981). 287. Libson, K., Barnett, B. L., and Deutsch, E., Inorg. Chem. 22, 1695 (1983). 288. Hurst, R. W., Heineman, W. R., and Deutsch, E., Znorg. Chem. 20,3298 (1981). 289. Kirchhoff, J. R., Heineman, W. R., and Deutsch, E., Znorg. Chem. 26, 3108 (1987). 290. Vanderheyden, J.-L., Heeg, M. J., and Deutsch, E., Znorg. Chem. 24, 1666 (1985).

COORDINATION CHEMISTRY OF TECHNETIUM

111

291. Roodt, A., Sullivan, J. C., Meisel, D., and Deutsch, E.,Inorg. Chem. 30,4545 (1991). 292. Kaden, L., Lorenz, B., Schmidt, K., Sprinz, H., and Wahren, M., 2. Chem. 21, 232 (1981). 293. Cohen, A. I., Glavan, K. A,, and Kronauge, J. F., Biomed. Mass Spectrom. 10, 287 (1983). 294. Unger, S. E., Anal. Chem. 56, 363 (1984). 295. Glavan, K. A., Whittle, R., Johnson, J . F., Elder, R. C., and Deutsch, E., J . Am. Chern. SOC.102, 2103 (1980). 296. Elder, R. C., Whittle, R., Glavan, K. A., Johnson, J .F., and Deutsch, E., Acta Cryst. B 36, 1662 (1980). 297. Konno, T., Heeg, M. J., and Deutsch, E., Inorg. Chem. 27,4113 (1988). 298. Abram, U., Beyer, R., Mading, P., Hoffmann, I., Munze, R., and Stach, J.,2. Anorg. Allg. Chern. 578, 229 (1989). 299. Konno, T., Seeber, R., Kirchhoff, J. R., Heineman, W. R., Deutsch, E., and Heeg, M. J., Transition Met. Chem. 18, 209 (1993). 300. Okamoto, K.-I., Kirchhoff, J . R., Heineman, W. R., and Deutsch, E., Polyhedron 12, 749 (1993). 301. Chang, L., Deutsch, E., and Heeg, M. J., Transition Met. Chem. 18, 335 (1993). 302. Stach, J., Abram, U., and Munze, R., 2. Chem. 29,249 (1989). 303. Konno, T., Kirchhoff, J. R., Heineman, R. W., and Deutsch, E., Inorg. Chem. 28, 1174 (1989). 304. Konno, T., Kirchhoff, J. R., Heeg, M. J., Heineman, W. R., and Deutsch, E., J. Chem. Soc., Dalton Trans., 3069 (1992). 305. Munze, R., Abram, U., Stach, J.,and Hiller, W., Inorg. Chim. Acta 196,151 (1991). 306. Refosco, F., Bandoli, G., Deutsch, E., Duatti, A., Mazzi, U., Moresco, A., Nicolini, M., and Tisato, F., in “Technetium and Rhenium in Chemistry and Nuclear Medicine 3” (M. Nicolini, G. Bandoli, and U. Mazzi, Eds.), p. 221. Cortina International, Verona/ Raven Press, New York, 1990. 307. Bolzati, C., Refosco, F., Tisato, F., Bandoli, G., and Dolmella, A., Inorg. Chim. Acta 201, 7 (1992). 308. Dilworth, J. R., Hutson, A. J., Morton, S., Harman, M., Hursthouse, M. B., Zubieta, J., Archer, C. M., and Kelly, J. D., Polyhedron 11, 2151 (1992). 309. Refosco, F., Bolzati, C., Moresco, A., Bandoli, G., Dolmella, A., Mazzi, U., and Nicolini, M., J. Chem. SOC.,Dalton Trans., 3043 (1991). 310. Dilworth, J. R., Griffiths, D. V., Hughes, J. M., Morton, S., Archer, C. M., and Kelly, J. D., Inorg. Chim. Acta 196, 145 (1992). 311. Rochon, F. D., Melanson, R., and Kong, P. C., Acta Cryst. C 46, 571 (1990). 312. Pietzsch, H . J . , Spies, H., Leibnitz, P., Reck, G., Beger, J.,and Jacobi, R.,Polyhedron 11, 1623 (1992). 313. Batsanov, A. S., Struchkov, Yu. T., Lorenz, B., and Wahren, M., 2. Anorg. Allg. Chem. 510, 117 (1984). 314. Lorenz, B., Schmidt, K., Hiller, W., Abram, U., and Hubener, R., Inorg. Chim. Acta 208, 195 (1993). 315. Nicholson, T., Thornback, J., OConnell, L., Morgan, G., Davison, A., and Jones, A. G. Inorg. Chem. 29, 89 (1990). 316. Lorenz, B., and Schmidt, K., 2. Chem. 23,423 (1983). 31 7 . Abram, U., and Abram, S., 2. Chem. 23,228 (1983). 318. Davison, A., de Vries, N., Dewan, J., and Jones, A., Inorg. Chim. Acta 120, L15 (1986). 319. de Vries, N., Jones, A. G., and Davison, A,, Inorg. Chem. 28, 3728 (1989).

112

JOHN BALDAS

320. de Vries, N., Cook, J.,Jones, A. G., and Davison, A,, Inorg. Chem. 30,2662 (1991). 321. de Vries, N., Cook, J . , Davison, A., Nicholson, T., and Jones A. G., Inorg. Chem. 29, 1062 (1990). 322. Baldas, J., Bonnyman, J.,Mackay, M. F., and Williams, G. A., Aust. J . Chem. 37, 751 (1984). 323. Baldas, J., Colmanet, S. F., and Williams, G. A., Aust. J . Chem. 44, 1125 (1991). 324. Abram, U., Wollert, R., and Hiller, W., Radiochim. Acta 63, 145 (1993). and Utilization of Technetium ’93, Sendai, Japan,” p. 58. [Abstr.] 325. Williams, G. A,, Bonnyman, J . , and Baldas, J., Aust. J . Chem. 40, 27 (1987). 326. Guest, A., and Lock, C. J . L., Can.J . Chem. 50, 1807 (1972). 327. Elder, M., and Penfold, B. R., Inorg. Chem. 5, 1197 (1966). 328. Dalziel, J.,Gill, N. S., Nyholm, R. S., and Peacock, R. D., J . Chem. Soc., 4012 (1958). 329. Schwochau, K., and Herr, W., Angew. Chem. 75,95 (1963). 330. Alberto, R., and Anderegg, G., Polyhedron 4, 1067 (1985). 331. Preetz, W., and Wendt, A,, 2. Naturforsch. B 46, 1496 (1991). 332. Jgrgensen, C. K., and Schwochau, K., 2. Naturforsch. A 20, 65 (1965). 333. Schenk, H. J., and Schwochau, K., 2. Naturforsch. A 28,89 (1973). 334. Schwochau, K., and Krasser, W., 2. Naturforsch. A 24, 403 (1969). 335. Flint, C. D., and Lang, P. F., J. Chem. Soc., Dalton Trans., 921 (1986). 336. Wendt, A,, and Preetz, W . ,2. Naturforsch. A 47, 882 (1992). 337. Elder, R. C., Estes, G. W., and Deutsch, E., Acta. Cryst. B 35, 136 (1979). 338. Baldas, J., Bonnyman, J . , Samuels, D. L., and Williams, G. A,, Acta Cryst. C 40, 1343 (1984). 339. Koz’min, N. A,, Novitskaya, G. N., and Kuzina, A. F., J . Struct. Chem. (Engl. Transl.) 13, 878 (1972). 340. Shepel’kov, S. V., Konarev, M. I., Chebotarev, N. T., Zaitseva, L. L., and Vinogradov, I. V., Russ. J . Inorg. Chem. (Engl. Transl.)20, 1828 (1975). 341. Kryuchkov, S. V., Grigor’ev, M. S., Kuzina, A. F., and Spitsyn, V. I., Russ. J . Inorg. Chem. (Engl. Transl.)32, 1714 (1987). 342. Kryuchkov, S. V., Grigor’ev, M. S., Kuzina, A. F., and Spitsyn, V. I., Russ. J . Inorg. Chem. (Engl. Transl.) 32, 1708 (1987). 343. Schwochau, K., 2. Naturforsch. A 20, 1286 (1965). 344. Koyama, M., Kanchiku, Y.,and Fujinaga, T., Coord. Chem. Reu. 3, 285 (1968). 345. Kanchiku, Y., Bull. Chem. SOC. Japan 42, 2831 (1969). 346. Kawashima, M., Koyama, M., andFujinaga, T., J .Inorg. Nucl. Chem. 38,819 (1976). 347. Ianovici, E., Kosinski, M., Lerch, P., and Maddock, A. G., J . Radioanal. Chem. 64, 315 (1981). 348. Colin, M., Ianovici, E., Lerch, P., and Maddock, A. G., J . Radioanal. Nucl. Chem., Articles 92, 283 (1985). 349. Colin, M., Ianovici, E., and Lerch, P., J . Radioanal. Nucl. Chem., Articles 120, 175 (1988). 350. Elder, M., Fergusson, J. E., Gainsford, G. J., Hickford, J. H., and Penfold, B. R. J . Chem. SOC.A , 1423 (1967). 351. Zaitseva, L. L., Kruglov, A. A., Romanov, A. V., Samsonov, V. E., and Teterin, E. G., Russ. J . Inorg. Chem. (Engl. Transl.)32, 1716 (1987); 33, 1452 (1988). 352. Wendt, A., and Preetz, W., 2. Anorg. Allg. Chem., 619, 1669 (1993). 353. Sundrehagen, E., Int. J . Appl. Radiat. Isot. 30, 739 (1979). 354. Magneli, A., and Andersson, G., Acta Chem. Scand. 9, 1378 (1955). 355. Paquette, J., and Lawrence, W. E., Can. J . Chem. 63,2369 (1985). 356. Alberto, R., Anderegg, G., and May K., Polyhedron 5, 2107 (1986).

COORDINATION CHEMISTRY OF TECHNETIUM

113

357. Alberto, R., Albinati, A.,Anderegg,G.,andHuber,G.,Inorg.Chem.30,3568(1991). 358. Refosco, F., Bandoli, G., Mazzi, U., Tisato, F., Dolmella, A., and Nicolini, M., Znorg. Chem.29,2179(1990). 359. Mazzi, U., Roncari, E., Bandoli, G., and Clemente, D. A,, Transition Met.Chem. 7, 163 (1982). 360. Colmanet, S . F., Williams, G. A., and Mackay, M. F., J.Chem.Soc., Dalton Trans., 2305 (1987). 361. Kennedy, C. M., and Pinkerton, T. C., Appl. Radiat. Isot. 39, 1159 (1988). 362. Omori, T., Yamada, Y., and Yoshihara, K., Znorg. Chim.Acta 130,99 (1987). 363. Abrams, M. J., Davison, A., Jones, A. G., and Costello, C. E., Inorg. Chim.Actu 77,L235 (1983). 364. Kastner, M. E., Fackler, P. H., Podbielski, L., Charkoudian, J., and Clarke, M. J., Znorg. Chirn.Acta 114,L11 (1986). 365. Lu, J . , and Clarke, M. J., Znorg. Chem.27,4761 (1988). 366. Clarke, M. J., Kastner, M. E., Podbielski, L. A., Fackler, P. H., Schreifels, J., 110, 1818 (1988). Meinken, G., and Srivastava, S. C., J.A m . Chem.SOC. 367. Lu, J.,Hiller, C. D., and Clarke, M. J.,Znorg. Chem.32, 1417 (1993). 368. Alberto, R., Anderegg, G., and Albinati, A., Znorg. Chim.Acta 178,125 (1990). 369. Burgi, H. B., Anderegg, G., and Blauenstein, P., Inorg. Chem.20, 3829 (1981). 370. Anderegg, G., Muller, E., Zollinger, K., and Burgi, H.-B., Helu. Chim.Acta 66, 1593 (1983). 371. Linder, K. E., Dewan, J . C., and Davison, A., Znorg. Chem.28,3820 (1989). 372. Muenze, R., Rudiochern.Radioanal. Lett.48,281 (1981). 373. Libson, K., Deutsch, E., and Barnett, B. L., J.A m . Chem.SOC. 102,2476 (1980). 374. Martin, J. L., Yuan, J., Lunte, C. E., Elder, R. C., Heineman, W. R., and Deutsch, E., Znorg. Chem.28,2899 (1989). 375. Mikelsons, M. V., and Pinkerton, T. C., Appl. Radiat. Zsot. 38,569 (1987). 376. Hunter, G., and Kilcullen, N., J . Chem.Soc., Dalton Trans., 2115 (1989). 377. Fergusson, J . E., and Hickford, J . H., J. Znorg. Nucl. Chem.28,2293 (1966). 378. Levanda, 0.Yu., Oblova, A. A., Kuzina, A. F., Belyaeva, L. I., and Spitsyn, V. I., RUSS. J . Znorg. Chem.(Engl. Transl.) 30,522 (1985). 379. Tisato, F., Bolzati, C., Duatti, A., Bandoli, G., and Refosco, F., Znorg. Chem. 32, 2042 (1993). 380. Bandoli, G., Clemente, D. A., Mazzi, U., and Roncari, E., J. Chem.Soc., Dalton Trans., 1381 (1982). 381. Rochon, F. D., Melanson, R., and Kong, P.-C., Acta Cryst. C 47,732 (1991). 382. Rochon, F. D., and Melanson, R., Acta Cryst. C 49, 1259 (1993). 383. Cotton, F. A., Day, C. S., Diebold, M. P., and Roth, W. J., Acta Cryst. C 46, 1623 (1990). 384. Breikss, A. I., Davison, A., and Jones, A. G., Znorg. Chim.Actu 170,75 (1990). 385. du Preez, J. G. H., Gerber, T. I. A., and Knoesen, O., Inorg. Chim. Actu 109, L17 (1985). 386. Bush, C. D., Hamor, T. A., Hussain, W., Jones, C. J., McCleverty, J. A., and Rothin, A. S., Acta Cryst. C 43,2088 (1987). 387. Lorenz, B., and Schmidt, K., 2.Chem.24,443 (1984). 388. Findeisen, M., Lorenz, B., and Olk, B., 2.Chem. 27,447 (1987). 389. Colmanet, S.F., and Mackay, M. F., Aust. J. Chem.41, 1127 (1988). 390. Colmanet, S.F., and Mackay, M. F., J. Chem.SOC.,Chem.Comm., 705 (1987). 391. Colmanet, S.F., and Mackay, M. F., Aust. J.Chem.41,269 (1988). 392. Edwards, A. J., Hugill, D., and Peacock, R. D., Nature 200,672 (1963).

114

JOHN BALDAS

393. Hugill, D., and Peacock, R. D., J . Chem. SOC.A 1339 (1966). 394. Holloway, J . H., and Selig, H., J . Znorg. Nucl. Chem. 30, 473 (1968). 395. Jezowska-Trzebiatowska, B., and Natkaniec, L., J . Struct. Chem. (Engl. Transl.)

8, 460 (1967). 396. Thomas, R. W., Davison, A,, Trop, H. S., and Deutsch, E., Znorg. Chem. 19, 2840

(1980). 397. Davison, A., Trop, H. S., DePamphilis, B. V., and Jones, A. G., Znorg. Synth. 21,

160 (1982). 398. Kung, H. F., Liu, B.-L., Wei, Y., and Pan, S., Appl. Radiat. Zsot. 41, 773 (1990). 399. Roodt, A., Leipoldt, J. G., Deutsch, E. A., and Sullivan, J . C., Znorg. Chem. 31,

1080 (1992). 400. Davison, A,, Jones, A. G., Miiller, L., Tatz, R., and Trop, H. S., Znorg. Chem. 20,

1160 (1981). 401. Caron, S., Ianovici, E., Lerch, P., and Maddock, A. G., Znorg. Chim. Acta 109,

209 (1985). 402. Cotton, F. A., Davison, A., Day, V. W., Gage, L. D., and Trop, H. S., Inorg. Chem.

18, 3024 (1979). 403. Jeiowska-Trzebiatowska, B., and BaTuka, M., Bull. Acad. Polon. Sci., Ser. Chim.

13, l(1965). 404. Thomas, R. W., Heeg, M. J., Elder, R. C., and Deutsch, E., Znorg. Chem. 24,

1472 (1985). 405. Mantegazzi, D., Ianoz, E., Lerch, P., and Tatsumi, K., Znorg. Chim. Acta 167,

195 (1990). 406. Baldas, J., Boas, J. F., Bonnyman, J., and Colmanet, S. F., in “Technetium and

407. 408. 409. 410. 411. 412. 413. 414. 415. 416. 417. 418. 419. 420. 421. 422. 423.

Rhenium in Chemistry and Nuclear Medicine 3” (M. Nicolini, G. Bandoli, and U. Mazzi, Eds.), p. 63. Cortina International, Verona/Raven Press, New York, 1990. Preetz, W., and Peters, G., 2.Naturforsch. B 35, 1355 (1980). Peters, G., and Preetz, W., Z . Naturforsch. B 36, 138 (1981). Baldas, J., and Colmanet, S. F., Aust. J . Chem. 42, 1155 (1989). Hiibener, R., and Abram, U., Z . Anorg. Allg. Chem. 617,96 (1992). Hanuza, J., B a h k a , M., Machner, W., and Jeiowska-Trzebiatowska,B., Acta Phys. Polon. A 71, 91 (1987). Baran, E. J., and Cabello, C. I., 2. Naturforsch. A 38, 563 (1983). Collin, R. J., Griffith, W.P., and Pawson, D., J . Mol. Struct. 19, 531 (1973). B a h k a , M., Hanuza, J., and Jezowska-Trzebiatowska, B., Bull. Acad. Polon. Sci., Ser. Chim. 20, 271 (1972). George, G. N., Cleland, W. E., Enemark, J . H., Smith, B. E., Kipke, C. A,, Roberts, S. A,, and Cramer, S. P., J . Am. Chem. SOC.112, 2541 (1990). Colton, R., and Tomkins, I. B., Aust. J . Chem. 21, 1981 (1968). Davison, A,, Jones, A. G., and Abrams, M. J., Znorg. Chem. 20,4300 (1981). Du Preez, J. G. H., Gerber, T. I. A., and Jacobs, R., J . Coord. Chem. 26,259 (1992). Duatti, A,, Marchi, A., Rossi, R., Magon, L., Deutsch, E., Bertolasi, V., and Bellucci, F., Znorg. Chem. 27,4208 (1988). Davison, A., DePamphilis, B. V., Jones, A. G., Franklin, K. J., and Lock, C. J . L., Znorg. Chim. Acta 128, 161 (1987). Rochon, F. D., Melanson, R., and Kong, P.-C., Acta Cryst. C 48, 785 (1992). Abrams, M. J., Larsen, S. K., and Zubieta, J., Znorg. Chem. 30, 2031 (1991). Baldas, J., Colmanet, S. F., and Mackay, M. F., J . Chem. SOC.,Dalton Trans., 1725 (1988).

COORDINATION CHEMISTRY OF TECHNETIUM

115

424. Hamor, T. A,, Hussain, W., Jones, C. J., McCleverty, J. A., and Rothin, A. S.,Znorg.

Chim. Actu 146,181 (1988). 425. Smith, J. E., Byrne, E. F., Cotton, F. A., and Sekutowski, J. C., J . A m . Chem. SOC.

100,5571 (1978). 426. DePamphilis, B. V., Jones, A. G., Davis, M. A., and Davison, A., J . A m . Chem. SOC.100,5570 (1978). 427. Colmanet, S. F., and Mackay, M. F., Znorg. Chim. Acta 147, 173 (1988). 428. Colmanet, S. F., and Mackay, M. F., Aust. J . Chem. 41, 151 (1988). 429. Bandoli, G., Nicolini, M., Mazzi, U.,Spies, H., and Munze, R., Transition Met.

Chem. 9, 127 (1984). 430. Colrnanet, S. F., and Mackay, M. F., Aust. J . Chem. 40, 1301 (1987). 431. Jones, A. G., DePamphilis, B. V., and Davison, A.,Inorg. Chem. 20, 1617 (1981). 432. Bandoli, G., Mazzi, U.,Abram, U., Spies, H., and Munze, R., Polyhedron 6, 1547

11987). 433. Huber, G., Anderegg, G., and May K., Polyhedron 6, 1707 (1987). 434. Sykes, A. G., in “Comprehensive Coordination Chemistry” (G. Wilkinson, R. D. Gillard, and J . A. McCleverty, Eds.), Vol. 3,p. 1250.Pergamon Press, Oxford, 1987. 435. Hashimoto, K.,Yamada, Y., Omori, T., and Yoshihara, K., J . Radioanal. Nucl.

Chem., Lett. 135, 187 (1989). 436. Grases, F.,Arnat, E., Genestar, C., and Palou, J.,J . Radioanal. Nucl. Chem.,Articles 437. 438. 439. 440. 441. 442. 443. 444. 445. 446. 447. 448. 449.

450. 451. 452. 453. 454. 455. 456.

116,459 (1987). Spies, H., and Johannsen, B., React. Kinet. Catal. Lett. 10, 153 (1979). Spies, H., React. Kinet. Catal. Lett. 10, 131 (1979). Spies, H., and Johannsen, B., Znorg. Chim. Actu 33, L113 (1979). Byrne, E.F., and Smith, J. E., Znorg. Chem. 18, 1832 (1979). Davison, A., Orvig, C., Trop, H. S., Sohn, M., DePamphilis, B. V., and Jones, A. G., Znorg. Chem. 19, 1988 (1980). Jones, A. G., Orvig, C., Trop, H. S., Davison, A., and Davis, M. A., J . Nucl. Med. 21,279 (1980). Spies, H., and Johannsen, B., Inorg. Chim. Acta 48,255 (1981). Du Preez, J. G. H., and Gerber, T. I. A., Inorg. Chim. Acta 110,59 (1985). Abrams, M. J., Brenner, D., Davison, A., and Jones, A. G., Znorg. Chim. Acta 77, L127 (1983). Blower, P. J., Singh, J., and Clarke, S. E. M., J . Nucl. Med. 32, 845 (1991). Spies, H., and Scheller, D., Znorg. Chim. Acta 116, 1 (1986). Baldas, J., Bonnyman, J., and Ivanov, Z., J . Nucl. Med. 30, 1240 (1989). Pietzsch, H. J., Spies, H., Hoffmann, S., and Stach, J., Inorg. Chim. Acta 161, 15 (1989). Pietzsch, H . J . , Spies, H., and Hoffmann, S., Znorg. Chim. Acta 165, 163 (1989); 168,7 (1990). Morelock, M. M., Cormier, T. A., and Tolman, G. L.,Inorg. Chem. 27,3137 (1988). Abram, U.,and Spies, H., 2.Chem. 24, 74 (1984). Fair, C. K., Troutner, D. E., Schlemper, E. O., Murmann, R. K., and Hoppe, M. L., Acta Cryst. C 40, 1544 (1984). Jurisson, S . , Schlemper, E. O., Troutner, D. E., Canning, L. R., Nowotnik, D. P., and Neirinckx, R. D., Inorg. Chem. 26,543 (1986). Jurisson, S., Aston, K., Fair, C. K., Schlemper, E. O., Sharp, P. R., and Troutner, D. E., Inorg. Chem. 26,3576 (1987). Gerber, T. I. A., Kemp, H. J., du Preez, J. G. H., Bandoli, G., and Dolmella, A., Inorg. Chim. Actu 202, 191 (1992).

116

JOHN BALDAS

457. Lawrence, A. J., Thornback, J. R., Zanelli, G. D., and Lawson, A,, Znorg. Chim. Acta 141, 165 (1988). 458. Deutsch, E. A., Elder, R. C., and Packard, A., et ul., cited in Coord. Chem. Rev. 44, 191 (1982). 459. Jurisson, S., Lindoy, L. F., Dancey, K. P., McPartlin, M., Tasker, P. A,, Uppal, D. K., and Deutsch, E., Inorg. Chem. 23, 227 (1984). 460. Mazzi, U., Refosco, F., Tisato, F., Bandoli, G., and Nicolini, M., J . Chem. Soc., Dalton Trans., 1623 (1986). 461. Abrams, M. J., Shaik, S. N., and Zubieta, J., Znorg. Chim. Acta 186, 87 (1991). 462. Morgan, G. F., Abram, U., Evrard, G., Durant, F., Deblaton, M., Clemens, P., Van den Broeck, P., and Thornback, J. R., J . Chem. Soc., Chem. Comm., 1772 (1990). 463. Morgan, G. F., Deblaton, M., Hussein, W., Thornback, J. R., Evrard, G., Durant, F., Stach, J., Abram, U., and Abram, S., Znorg. Chim. Acta 190, 257 (1991). 464. Refosco, F., Tisato, F., Mazzi, U., Bandoli, G., and Nicolini, M., J . Chem. Soc., Dalton Trans., 611 (1988). 465. Liu, S., Rettig, S. J., and Orvig, C . , Znorg. Chem. 30, 4915 (1991). 466. Duatti, A., Marchi, A., Magon, L., Deutsch, E., Bertolasi, V., and Gilli, G., Znorg. Chem. 26,2182 (1987). 467. DuPreez, J . G. H., Gerber,T. I. A,, andGibson, M. L., J . Coord. Chem. 22,159 (1990). 468. Du Preez, J. G. H., Gerber, T. I. A., and Kemp, H. J.,J . Coord. Chem. 26,177 (1992). 469. Du Preez, J. G. H., Gerber, T. I. A., Fourie, P. J., and Van Wyk, A. J., J . Coord. Chem. 13, 173 (1984). 470. Refosco, F., Tisato, F., Bandoli, G., Bolzati, C., Dolmella, A,, Moresco, A,, and Nicolini, M., J . Chem. Soc., Dalton Trans., 605 (1993). 471. Lever, S. Z., Baidoo, K. E., and Mahmood, A., Znorg. Chim. Acta 176, 183 (1990). 472. Mahmood, A., Halpin, W.A., Baidoo, K. E., Sweigart, D. A., and Lever, S. Z., Acta Cryst. C 47, 254 (1991). 473. Kung, H. F., Guo, Y.-Z., Yu, C.-C., Billings, J., Subramanyam, V., and Calabrese, J . C., J. Med. Chem. 32,433 (1989). 474. John, C. S., Kung, M.-P., Billings, J., and Kung, H. F., Nucl. Med. Biol. 18, 551 (1991). 475. DiZio, J. P., Anderson, C. J., Davison, J. A., Ehrhardt, G. J., Carlson, K. E., Welch, M. J . , and Katzenellenbogen, J. A., J . Nucl. Med. 33, 558 (1992). 476. Ohmomo, Y . , Francesconi, L., Kung, M.-P., and Kung, H. F., J . Med. Chem. 35, 157 (1992). 477. Francesconi, L. C., Graczyk, G., Wehrli, S., Shaik, S. N., McClinton, D., Liu, S., Zubieta, J., and Kung, H. F., Inorg. Chem. 32, 3114 (1993). 478. Edwards, D. S., Cheesman, E. H., Watson, M. W., Maheu, L. J., Nguyen, S. A., Dimitre, L., Nason, T., Watson, A. D., and Walovitch, R., in “Technetium and Rhenium in Chemistry and Nuclear Medicine 3” (M. Nicolini, G. Bandoli, and U. Mazzi, Eds.), p. 433. Cortina International, Verona/Raven Press, New York, 1990. 479. Faggiani, R., Lock, C. J. L., Epps, L. A., Kramer, A. V., and Brune, D., Actu Cryst. C 44, 777 (1988). 480. Faggiani, R., Lock, C. J. L., Epps, L. A., Kramer, A. V., and Brune, H. D., Actu Cryst. C 46, 2324 (1990). 481. Marchi, A., Marvelli, L., Rossi, R., Magon, L., Bertolasi, V., Ferretti, V., and Gilli, P., J . Chem. SOC.,Dalton Trans., 1485 (1992). 482. Stassinopoulou, C. I., Mastrostamatis, S., Papadopoulos, M., Vavouraki, H., Terzis, A,, Hountas, A., and Chiotellis, E., Znorg. Chim. Acta 189, 219 (1991).

COORDINATION CHEMISTRY OF TECHNETIUM

117

483. Davison, A,, Jones, A. G., Orvig, C., and Sohn, M., Inorg. Chem. 20, 1629 (1981). 484. Jones, A. G., Davison, A,, LaTegola, M. R., Brodack, J. W., Orvig, C., Sohn, M.,

485. 486. 487. 488. 489. 490. 491. 492. 493. 494. 495. 496. 497. 498. 499. 500. 501.

502. 503. 504. 505. 506. 507. 508. 509. 510.

Toothaker, A. K., Lock, C. J. L., Franklin, K. J.,Costello, C. E., Carr, S. A., Biemann, K., and Kaplan, M. L., J . Nucl. Med. 23, 801 (1982). Chen, B., Heeg, M. J., and Deutsch, E., Znorg. Chem. 31,4683 (1992). Brenner, D., Davison, A,, ListerJames, J., and Jones, A. G., Inorg. Chem. 23, 3793 (1984). Rao, T. N., Adhikesavalu, D., Camerman, A,, and Fritzberg, A. R., J . Am. Chem. SOC.112,5798 (1990). Bryson, N., Lister-James, J., Jones, A. G., Davis, W. M., and Davison, A., Inorg. Chem. 29, 2948 (1990). Canney, D. J., Billings, J., Francesconi, L. C., Guo, Y.-Z.,Haggerty, B. S.,Rheingold, A. L., and Kung, H. F., J . Med. Chem. 36, 1032 (1993). Chen, Y.-C. J., and Janda, K. D., J . Am. Chem. SOC. 114, 1488 (1992). Rao, T. N., Brixner, D. I., Srinivasan, A., Kasina, S., Vanderheyden, J.-L., Wester, D. W., and Fritzberg, A. R., Appl. Radiat. h o t . 42, 525 (1991). Bryson, N. J., Brenner, D., Lister-James,J., Jones, A. G., Dewan, J. C., and Davison, A., Znorg. Chem. 28, 3825 (1989). Bryson, N., Dewan, J. C., ListerJames, J., Jones, A. G., and Davison, A., Znorg. Chem. 27,2154 (1988). Franklin, K. J.,Howard-Lock,H. E., and Lock, C. J. L.,Inorg. Chem.21,1941 (1982). Johnson D. E., Fritzberg, A. R., Hawkins, B. L., Kasina, S., and Eshima, D., Znorg. Chem. 23,4204 (1984). Bandoli, G., and Gerber, T. I. A., Inorg. Chim. Acta 126,205 (1987). Abram, U.,Muenze, R., Hartung, J., Beyer, L., Kirmse, R., Koehler, K., Stach, J., Behm, H., and Beurskens, P. T., Inorg. Chem. 28,834 (1989). Du Preez, J . G. H., Gerber, T. I., and Knoesen, O., Inorg. Chim. Acta 133,3 (1987). Du Preez, J. G. H., Gerber, T. I. A., and Knoesen, O., Inorg. Chim. Acta 132, 241 (1987). Du Preez, J. G. H., Gerber, T. I. A., and Gibson, M. L., J . Coord. Chem.22,321(1991). Nosco, D. L., Tofe, A. J., Dunn, T. J., Lyle, L. R., Wolfangel, R. G., Bushman, M. J., Grummon, G. D., Helling, D. E., Marmion, M. E., Miller, K. M., Pipes, D. W., Strubel, T. W., and Wester, D. W., in “Technetium and Rhenium in Chemistry and Nuclear Medicine 3” (M. Nicolini, G. Bandoli, and U. Mazzi, Eds.), p. 381.Cortina International, Verona/Raven Press, New York, 1990. Hansen, R., Cini, R., Taylor, A., and Marzilli, L. G., Inorg. Chem. 31,2801 (1992). Fackler, P. H., Kastner, M. E., and Clarke, M. J., Inorg. Chem. 23,3968 (1984). Du Preez, J. G. H., Gerber, T. I. A., Gibson, M. L., and Geyser, R., J . Coord. Chem. 22,249 (1990). Pearlstein, R. M., Lock, C. J. L., Faggiani, R., Costello, C. E., Zeng, C.-H., Jones, A. G., and Davison, A.,Inorg. Chem. 27,2409 (1988). Thomas, R. W., Estes, G. W., Elder, R. C., and Deutsch, E., J . Am. Chem. SOC.101, 4581 (1979). Wilcox, B. E., Heeg, M. J., and Deutsch, E., Inorg. Chem. 23, 2962 (1984). Omori, T., Yamada, Y., Iino, S., and Yoshihara, K., J . Radioanal. Nucl. Chem., Lett. 119,223 (1987). Refosco, F., Mazzi, U., Deutsch, E., Kirchhoff, J. R., Heineman, W. R., and Seeber, R., Znorg. Chern. 27,4121 (1988). Bandoli, G., Mazzi, U., Clemente, D. A., and Roncari, E., J . Chem. Soc., Dalton Trans., 2455 (1982).

118

JOHN BALDAS

511. Tisato, F., Refosco, F., Moresco, A., Bandoli, G., Mazzi, U., and Nicolini, M., J . Chem. Soc., Dalton Trans., 2225 (1990). 512. Bandoli, G., Mazzi, U., Wilcox, B. E., Jurisson, S., and Deutsch, E., Znorg. Chim. Actu 95, 217 (1984). 513. Wilcox, B. E., Cooper, J. N., Elder, R. C., and Deutsch, E., Znorg. Chim. Actu 142, 55 (1988). 514. Tisato, F., Refosco, F., Mazzi, U., Bandoli, G., and Nicolini, M., J . Chem. Soc., Dalton Trans., 1693 (1987). 515. Bandoli, G., Mazzi, U., Pietzsch, H . J . , and Spies, H., Acta Cryst. C 48, 1422 (1992). 516. Abram, U., Abram, S., Hiller, W., Morgan, G. F., Thornback, J. R., Deblaton, M., and Stach, J., 2. Nuturforsch. B 46, 453 (1991). 51 7. Pietzsch, H.J . , Spies, H., Leibnitz, P., Reck, G . ,Beger, J., and Jacobi, R., Polyhedron 12, 187 (1993). 518. Johannsen, B., Noll, B., Leibnitz, P., Reck, G., Noll, S., and Spies, H., Znorg. Chim. Actu 210, 209 (1993). 519. Du Preez, J. G. H., Gerber, T. I. A,, and Knoesen, O.,Znorg. Chim. Actu 130,9 (1987). 520. Du Preez, J . G. H., Gerber, T. I. A., and Knoesen, O., J . Coord. Chem. 16,285 (1987). 521. Pietzsch, H.J., Spies, H., Hoffmann, S., and Scheller, D., Appl. Rudiat. hot. 41, 185 (1990). 522. Nicholson, T., Davison, A., and Jones, A. G., Znorg. Chim. Acta 168, 227 (1990). 523. Nicholson, T., Storm, S. L., Davis, W. M., Davison, A,, and Jones, A. G., Znorg. Chim. Actu 196, 27 (1992). 524. Holm, R. H., Chem. Rev. 87, 1401 (1987). 525. Zuckman, S. A., Freeman, G. M., Troutner, D. E., Volkert, W. A., Holmes, R. A., Van Derveer, D. G., and Barefield, E. K., Znorg. Chem. 20, 2386 (1981). 526. Kastner, M. E., Fackler, P. H., Clarke, M. J., and Deutsch, E., Znorg. Chem. 23, 4683 (1984). 527. Truffer, S., Ianoz, E., Lerch, P., and Kosinski, M.,Znorg. Chim. Actu 149,217 (1988). 528. Kuzina, A. F., Oblova, A. A., and Spitsyn, V. I., Russ. J . Znorg. Chem. (Engl. Transl.) 17, 1377 (1972). 529. Kastner, M. E., Lindsay, M. J., and Clarke, M. J., Znorg. Chem. 21, 2037 (1982). 530. Fackler, P. H., Lindsay, M. J., Clarke, M. J., and Kastner, M. E., Znorg. Chim. Actu 109, 39 (1985). 531. Ianoz, E., Mantegazzi, D., Lerch, P., Nicolb, F., and Chapuis, G., Znorg. Chdm. Acta 156, 235 (1989). 532. Mantegazzi, D., Ianoz, E., Lerch, P., Nicolb, F., and Chapuis, G., Znorg. Chim. Acta 176, 99 (1990). 533. Lin, Z., and Hall, M. B., C O O FChem. ~ . Reu. 123, 149 (1993). 534. Lu, J., and Clarke, M. J., Inorg. Chem. 28, 2315 (1989). 535. Helm, L., Deutsch, K., Deutsch, E. A., and Merbach, A. E., Helv. Chim. Acta 75, 210 (1992). 536. Kelly, J. D., Forster, A. M., Higley, B., Archer, C. M., Booker, F. S., Canning, L. R., Chui, K. W., Edwards, B., Gill, H. K., McPartlin, M., Nagle, K. R., Latham, I. A., Pickett, R. D., Storey, A. E., and Webbon, P. M., J . Nucl. Med. 34, 222 (1993). 537. Spitsyn, V. I., Glinkina, M. I., and Kuzina, A. F., Dokl. Akud. Nuuk SSSR 200, 1372 [Engl. 8751 (1971). 538. Du Preez, J. G. H., Gerber, T. I. A., Fourie, P. J., and Van Wyk, A. J., Znorg. Chim. Actu 82, 201 (1984). 539. Trop, H. S., cited in Znt. J . Appl. Radiat. Zsot. 33, 875 (1982).

COORDINATION CHEMISTRY OF TECHNETIUM

119

540. Tisato, F., Mazzi, U., Bandoli, G., Cros, G., Darbieu, M.-H., Coulais, Y .,and Guiraud, R., J . Chern. Soc., Dalton Trans., 1301 (1991). 541. Pillai, M. R. A., John, C. S., Lo, J . M., Schlemper, E. O., and Troutner, D. E., Inorg.

Chem. 29, 1850 (1990). 542. Tisato, F., Refosco, F., Mazzi, U., Bandoli, G., and Dolmella, A., Inorg. Chim. Actu

164,127 (1989). 543. Bandoli, G., Nicolini, M., Mazzi, U., and Refosco, F., J . Chem. Soc., Dalton Trans.,

2505 (1984). 544. Lin, Z., and Hall, M. B., Znorg. Chem. 30,3817 (1991). 545. Davison, A., DePamphilis, B. V., Faggiani, R., Jones, A. G., Lock, C. J . L., and

Orvig, C., Can. J . Chem. 63,319 (1985). 546. DePamphilis, B. V., Jones, A. G., and Davison, A.,Inorg. Chem. 22, 2292 (1983). 547. Du Preez, J. G. H., Gerber, T. I. A., and Gibson, M. L., J . Coord. Chem. 22,33 (1990). 548. Abrams, M. J., Costello, C. E., Shaik, S. N., and Zubieta, J., Inorg. Chim. Actu

180,9 (1991). 549. Duatti, A,, Tisato, F., Refosco, F., Mazzi, U., and Nicolini, M., Inorg. Chem. 28,

4564 (1989). 550. Baldas, J., Bonnyman, J., Pojer, P. M., Williams, G. A., and Mackay, M. F., J. Chem. Soc., Dalton Trans., 1798 (1981). 551. Kaden, L., Lorenz, B., Schmidt, K., Sprinz, H., and Wahren, M., Zsotopenpraris 17, 174 (1981). 552. Baldas, J., Bonnyman, J., and Williams, G. A., Znorg. Chem. 25, 150 (1986). 553. Abram, U., Kirmse, R., Stach, J., and Lorenz, B., Z. Chem. 25, 153 (1985). 554. Kirmse, R., Stach, J., and Abram, U., Polyhedron 4, 1403 (1985). 555. Baldas, J., Boas, J. F., Colmanet, S. F., and Mackay, M. F., Inorg. Chim. Acta 170, 233 (1990). 556. Leipoldt, J. G., Basson, S. S., Roodt, A., and Purcell, W.,Polyhedron 11,2277(1992). 557. Baldas, J., Colmanet, S. F., and Williams, G. A., unpublished results. 558. Baldas, J., and Ivanov, Z., unpublished results. 559. Baldas, J., Bonnyman, J., and Williams, G . A., J . Chem. SOC.,Dalton Trans., 833 (1984). 560. Williams, G. A,, and Baldas, J., J . Nucl. Med. Allied Sci. 33, 327 (1989). 561. Abram, U.,Spies, H., Abram, S., Kirmse, R., and Stach, J.,2. Chem. 26,140 (1986). 562. Baldas, J., Heath, G. N., Raptis, R. G., and Williams, G. A., unpublished results. 563. Abram, U., Spies, H., Gorner, W., Kirmse, R., and Stach, J.,Inorg. Chim. Acta 109, L9 (1985). 564. Rossi, R., Marchi, A., Magon, L., Casellato, U., and Graziani, R., J . Chem. Soc., Dalton Trans., 2923 (1990). 565. Baldas, J., Boas, J . F., Colmanet, S. F., and Williams, G. A., J.Chem. SOC.,Dalton Trans., 2845 (1992). 566. Williams, G. A., and Baldas, J., Aust. J. Chem. 42,875 (1989). 567. Abram, U., Abram, S., Stach, J., Dietzsch, W., and Hiller, W., 2.Naturforsch. B 46,1183 (1991). 568. Baldas, J., and Bonnyman, J., Inorg. Chim. Acta 141, 153 (1988). 569. Stach, J., Dietzsch, W., and Abram, U., Z. Chem. 29,295 (1989). 570. Baldas, J., and Bonnyman, J., in “Technetium and Rhenium in Chemistry and Nuclear Medicine 3” (M. Nicolini, G. Bandoli, and U. Mazzi, Eds.), p. 429.Cortina International, Verona/Raven Books, New York, 1990. 571. Pasqualini, R., and Duatti, A., J.Chem. SOC.,Chem. Comm., 1354 (1992).

120

JOHN BALDAS

572. Marchi, A., Garuti, P., Duatti, A., Magon, L., Rossi, R., Ferretti, V., and Bertolasi, V., Inorg. Chem. 29,2091 (1990). 573. Marchi, A., Rossi, R., Magon, L., Duatti, A,, Casellato, U., Graziani, R., Vidal, M., and Riche, F., J . Chem. SOC.,Dalton Trans., 1935 (1990). 574. Duatti, A., Marchi, A., Bertolasi, V., and Ferretti, V., J . A m . Chem. SOC.113, 9680 (1991). 575. Bertolasi, V., Ferretti, V., Gilli, P., Marchi, A., and Marvelli, L., Acta Cryst. C 47, 2535 (1991). 576. Archer, C. M., Dilworth, J. R., Griffiths, D. V., McPartlin, M., and Kelly, J . D., J . Chern. SOC.,Dalton Trans., 183 (1992). 577. Dilworth, J. R., Griffiths, D. V., Hughes, J. M., Morton, S., Hiller, W., Archer, C. M., Kelly, J. D., and Walton, G., Inorg. Chim. Acta 192, 59 (1992). 578. Clarke, M. J., and Lu, J . , Znorg. Chem. 31,2476 (1992). 579. Abram, U., Abram, S., Spies, H., Kirmse, R., Stach, J., and Kohler, K., Z. Anorg. Allg. Chem. 544, 167 (1987). 580. Rummel, S., Hermann, M., and Schmidt, K.,Z. Chem. 25, 152 (1985). 581. de Vries, N., Costello, C. E., Jones, A. G., and Davison, A,, Inorg. Chem. 29, 1348 (1990). 582. Abram, U., Lorenz, B., Kaden, L., and Scheller, D., Polyhedron 7,285 (1988). 583. Abram, U., Mading, P., Kirmse, R., and Kohler, K.,Z. Chem. 29, 183 (1989). 584. Duatti, A., Marchi, A., and Pasqualini, R.,J. Chem. Soc.,Dalton Trans.,3729 (1990). 585. Abrams, M. J., Larsen, S. K., Shaikh, S. N., and Zubieta, J . , Znorg. Chim. Acta 185,7 (1991). 586. Baldas, J., Boas, J. F., Colmanet, S. F., and Williams, G. A., J . Chem. SOC.,Dalton Trans., 2441 (1991). 587. Batsanov, A. S., Struchkov, Yu. T., Lorenz, B., and Olk, B., 2. Anorg. Allg. Chem. 564, 129 (1988). 588. Dilworth, J. R., Archer, C. M., Latham, I. A,, Kelly, J . D., Griffiths, D. V., York, D. C.,Mahoney, P. M., and Higley, B., Nucl. Med. Biol. 18,547 (1991). 589. Marchi, A., Marvelli, L., Rossi, R., Magon, L., Uccelli, L., Bertolasi, V., Ferretti, V., and Zanobini, F., J . Chem. SOC.,Dalton Trans., 1281 (1993). 590. Abram, U., Munze, R., Jager, E.-G., Stach, J., and Kirmse, R., Inorg. Chim. Acta 162, 171 (1989). 591. Abram, U., Abram, S., Munze, R., Jager, E.-G., Stach, J., Kirmse, R., Admiraal, G., and Beurskens, P. T., Inorg. Chim. Acta 182,233 (1991). 592. Marchi, A., Duatti, A., Rossi, R., Magon, L., Pasqualini, R., Bertolasi, V., Ferretti, V., and Gilli, G . , J . Chem. Soc., Dalton Trans., 1743 (1988). 593. Rossi, R., Marchi, A., Aggio, S., Magon, L., Duatti, A., Casellato, U., and Graziani, R., J . Chem. SOC.,Dalton Trans., 477 (1990). 594. Uhlemann, E., Spies, H., Pietzsch, H.J . , and Herzschuh, R., Z . Naturforsch. B 47, 1441 (1992). 595. Abram, U., Hartung, J., Beyer, L., Stach, J., and Kirmse, R.,Z. Chem. 30,180(1990). 596. Nicholson, T., Davison, A., and Jones, A. G., Inorg. Chim. Acta 187,51 (1991). 597. Abrams, M. J., Shaikh, S. N., and Zubieta, J., Znorg. Chim. Acta 171, 133 (1990). 598. Abrams, M. J., Larsen, S. K., and Zubieta, J., Znorg. Chim. Acta 173, 133 (1990). 599. Nicholson, T.,Mahmood, A., Morgan, G., Jones, A. G., and Davison, A., Inorg. Chim. Acta 179,53 (1991). 600. Cook, J., Davis, W. M., Davison, A., and Jones, A. G., Znorg. Chem. 30,1773(1991). 601. Spies, H.,Pietzsch, H . J . , and Hoffmann, I., Znorg. Chim. Acta 161, 17 (1989).

COORDINATION CHEMISTRY OF TECHNETIUM

121

602. Baldas, J., Boas, J. F., Bonnyman, J., Colmanet, S. F., and Williams, G. A., J.

Chem. SOC.,Chem. Comm., 1163 (1990). 603. Selig, H., Chernik, C. L., and Malm, J . G., J. Znorg. Nucl. Chem. 19, 377 (1961). 604. Deutsch, E., Heineman, W. R., Hurst, R., Sullivan, J. C., Mulac, W. A,, and Gordon,

S., J . Chem. SOC.,Chem. Comm., 1038 (1978). 605. Libson, K., Sullivan, J . C., Mulac, W. A., Gordon, S., and Deutsch, E., Inorg. Chem.

28, 375 (1989). 606. Lawrance, G . A., and Sangster, D. F., Polyhedron 4, 1095 (1985). 607. Kryuchkov, S. V., Pikaev, A. K., Kuzina, A. F., and Spitsyn, V. I., Dokl. Akad.

Nauk SSSR 247, 1187 [Engl. 6901 (1979). 608. Astheimer, L., Hauck, J., Schenk, H. J., and Schwochau, K., J . Chem. Phys. 63, 1988 (1975). 609. Astheimer, L., and Schwochau, K., J. Inorg. Nucl. Chem. 38, 1131 (1976). 610. Edwards, A. J.,Jones, G. R., and Steventon, B. R., J . Chem. Soc., Chem. Comm., 462 (1967). 611. Edwards, A. J.,Jones, G. R., and Sills, R. J. C., J. Chem. SOC.A , 2521 (1970). 612. Kirmse, R., Stach, J., and Abram, U., Znorg. Chem. 24,2196 (1985). 613. Abram, U., Abram, S., Stach, J., and Kirmse, R., J. Radioanal. Nucl. Chem., Articles 100, 325 (1986). 614. Baldas, J., Boas, J. F., Bonnyman, J . , and Williams, G. A., J. Chem. SOC.,Dalton Trans., 2395 (1984). 615. Baldas, J.,Colmanet, S. F., and Williams, G. A., Inorg. Chim. Acta 179,189 (1991). 616. Abram, U., Munze, R., Kirmse, R., Kohler, K., Dietzsch, W., and GoliE, L., Znorg. Chim. Acta 169,49 (1990). 61 7. Lorenz, B., Isotopenpraris 26, 452 (1990). 618. Baldas, J., Bonnyman, J., and Williams, G. A., Aust. J. Chem. 38, 215 (1985). 619. Baldas, J., Boas, J. F., Colmanet, S. F., Rae, A. D., and Williams, G. A., Proc. R . SOC.London A 442, 437 (1993). 620. Baldas, J., Boas, J . F.,and Bonnyman, J., J.Chem. Soc., Dalton Trans., 1721 (1987). 621. Kohler, K., Kirmse, R., and Abram, U., 2. Chem. 26, 339 (1986). 622. Kirmse, R., Kohler, K., Abram, U., Bottcher, R., Golit, L., and de Boer, E., Chem. Phys. 143, 75 (1990). 623. Kohler, K., Kirmse, R., Bottcher, R., Abram, U., Gribnau, M. C. M., Keijzers, C. P., and de Boer, E., Chem. Phys. 143, 83 (1990). 624. Kohler, K., Kirmse, R., Bottcher, R., and Abram, U., Chem. Phys. 160,281 (1991). 625. Figgis, B. N., Reynolds, P. A., and Cable, J . W., J. Chem. Phys. 98, 7743 (1993). 626. Kirmse, R., Stach, J., and Abram, U., Inorg. Chim. Acta 117, 117 (1986). 627. Kohler, K., Kirmse, R., and Abram, U., 2.Anorg. Allg. Chem. 600, 83 (1991). 628. Baldas, J., Boas, J. F., and Bonnyman, J., Aust. J. Chem. 42, 639 (1989). 629. Abram, U., Kohler, K., Kirmse, R., Kalinichenko, N. B., and Marov, I. N., Inorg. Chim. Acta 176, 139 (1990). 630. Baldas, J., Boas, J. F., Ivanov, Z., and James, B. D., Inorg. Chim. Acta 204, 199 (1993). 631, Pietzsch, H.J.,Abram, U., Kirmse, R., and Kohler, K., 2.Chem. 27, 265 (1987). 632. Baldas, J., Colmanet, S. F., Ivanov, Z., Williams, G. A., and James, B. D., unpublished. 633. J~irgensen,C. K., Prog. Znorg. Chem. 12, 101 (1970). 634. Lever, A. B. P., J. Chem. Ed. 51,612 (1974). 635. Baldas, J., and Bonnyman, J., Int. J.Appl. Radiat. Zsot. 36,133 (1985);919 (1985).

122

JOHN BALDAS

636. Alagui, A., Apparu, M., du Moulinet d’Hardemare, A., Riche, F., and Vidal, M.,

Appl. Radiat. Zsot. 40,813 (1989). 637. Abram, U., and Wollert, R., Radiochim. Acta 63, 149 (1993). 638. Baldas, J., Colmanet, S. F., and Williams, G. A., J . Chem. Soc., Chern. Comm.,

954 (1991). 639. Baldas, J., Colmanet, S. F., Craig, D. C., Rae, A. D., and Williams, G. A,, unpub-

lished results. 640. Baldas, J., Boas, J. F., Bonnyman, J., Colmanet, S. F., and Williams, G. A., Znorg.

Chcm. Acta 179, 151 (1991). 641. Stiefel, E. I., Prog. Inorg. Chem. 22, l(1977). 642. Williams, G. A., and Martin, L. J., personal communication. 643. Baldas, J., Boas, J. F., Colmanet, S. F., Ivanov, Z., and Williams, G. A.,Radiochirn.

Acta, 63,111 (1993). Ricard, L., Martin, C., Wiest, R., and Weiss, R., Inorg. Chem. 14,2300 (1975). Baldas, J.,and Boas, J. F., J.Chem. SOC.,Dalton Trans., 2585 (1988). Burrell, A. K., and Bryan, J. C., Angew. Chem. Int. Ed. Engl. 32,94 (1993). Burrell, A. K., and Bryan, J. C., Organometallics 12,2426 (1993). Kawashima, M., Koyama, M., and Fujinaga, T.,J.Znorg.Nucl. Chem. 38,801(1976). Kirmse, R.,Stach, J., and Spies, H., Inorg. Chim. Acta 45, L251 (1980). Baldas, J., Boas, J., Bonnyman, J., Mackay, M. F., and Williams, G. A,, Aust. J . Chem. 35, 2413 (1982). 651. Baldas, J., Boas, J. F., Bonnyman, J., Pilbrow, J. R., and Williams, G. A,, J.A m . Chem. SOC.107, 1886 (1985). 652. delearie, L. A., Haltiwanger, R. C., and Pierpont, C. G., J . A m . Chem. SOC.111, 4324 (1989). 653. Krebs, B., Z . Anorg. Allg. Chem. 380,146 (1971). 654. Boyd, G. E., J. Chem. Ed. 36,3 (1959). 655. Faggiani, R., Lock, C. J. L., and Poce, J., Acta Cryst. B 36, 231 (1980). 656. German, K. E., Grigor’ev, M. S., Kuzina, A. F., and Spitsyn, V. I., Russ. J.Inorg. Chem. (Engl. Transl.)32,667 (1987). 657. Rochon, F. D., Kong, P. C., and Melanson, R., Acta Cryst. C 46, 8 (1990). 658. Keller, C.,and Kanellakopulos, B., Radiochim. Acta 1, 107 (1963). 659. Krebs, B., and Hasse, K.-D., Actu Cryst. B 32, 1334 (1976). 660. Chakravorti, M. C., Coord. Chem. Rev. 106,205(1990). 661. Truffer-Caron, S.,Ianoz, E., and Lerch, P., Znorg. Chim. Actu 149, 119 (1988). 662. Kemp, T. J., Thyer, A. M., and Wilson, P. D., J . Chem. SOC.,Dalton Trans., 2601 (1993);2607 (1993). 663. Miiller, A., Krebs, B., and Diemann, E., 2. Anorg. Allg. Chem. 353, 259 (1967). 664. Kanellakopulos, B., Raptis, K., Nuber, B., and Ziegler, M. L.,2.Naturforsch. B 46, 15 (1991). 665. Nugent, W. A.,Inorg. Chem. 22, 965 (1983). 666. Bryan, J. C., Burrell, A. K., Miller, M. M., Smith, W. H., Burns, C. J., and Sattelberger, A. P., Polyhedron 12,1769 (1993). 667. Selig, H., and Malm, J. G., J.Znorg. Nucl. Chem. 25,349 (1963). 668. Binenboyn, J.,El-Gad, U., and Selig, H., Znorg. Chem. 13,319 (1974). 669. Guest, A,, Howard-Lock, H. E., and Lock, C. J. L., J . Mol. Spectrosc. 43,273 (1972). 670. Baran, E.J., Spectrosc. Lett. 8, 599 (1975). 671. Franklin, K. J., Lock, C. J. L.,Sayer, B. G.,and Schrobilgen, G. J.,J . A m . Chem. SOC.104,5303 (1982).

644. 645. 646. 647. 648. 649. 650.

COORDINATION CHEMISTRY OF TECHNETIUM

123

672. Mercier, H. P. A,, and Schrobilgen, G. J., Znorg. Chem. 32,145 (1993). 673. Aronson, F. L., Hwang, L. L.-Y., Ronca, N., Solomon, N. A., and Steigman, J., Znorg. Chem. 24, 541 (1985). 674. Pearlstein, R. M., and Davison, A., Polyhedron 7, 1981 (1988). 675. Thomas, J. A., and Davison, A.,Znorg. Chem. 31, 1976 (1992). 676. Banbery, H. J., Hussain, W., Evans, I. G., Hamor, T. A., Jones, C. J., McCleverty, J . A,, Schulte, H.-J., Engles, B., and Klaui, W., Polyhedron 9,2549 (1990). 677. Thomas, J. A., and Davison, A., Znorg. Chim. Acta 190,231 (1991). 678. Baldas, J., Colmanet, S. F., and Mackay, M. F., J . C h e n . Soc., Chem. Comm., 1890 (1989). 679. Baldas, J., and Colmanet, S. F., Znorg. Chim.Acta 176, 1 (1990). 680. Baldas, J., Colmanet, S. F., and Williams, G. A., J . Chem. Soc., Dalton Trans., 1631 (1991). 681. Abrams, M. J., Chen, Q . , Shaikh, S. N., and Zubieta, J . , Znorg. Chim. Acta 176, 11 (1990). 682. Burrell, A. K., and Bryan, J . C., Organometallics 11, 3501 (1992). 683. Ginsberg, A. P., Znorg. Chem. 3, 567 (1964). 684. Abrahams, S.C., Ginsberg, A. P., and Knox, K., Znorg. Chem. 3,558 (1964). 685. Kaden, L., Lorenz, B., and Wahren, M., Zsotopenpraxis 18,400(1982).

This Page Intentionally Left Blank

ADVANCES IN INORGANIC

CHEMISTRY,VOL. 41

CHEMISTRY OF PENTAFLUOROSULFANYL COMPOUNDS R. D. VERMA,’ ROBERT L. KIRCHMEIER, AND JEAN’NE M. SHREEVE Department of Chemistry, University of Idaho, Moscow, Idaho 83843

I. Introduction 11. Pentafluorosulfanyl Halides 111. Pentafluorosulfanyl Hypohalites, SFSOX A. SFSOF B. SF50CI IV. Pentafluorosulfanylalkanes, Alkenes, and Alkynes V. Sulfur Isocyanate Pentafluoride and Sulfur Isothiocyanate Pentafluoride VI. Sulfur Cyanate Pentafluoride, SF50CN VII. Sulfur Cyanide Pentafluoride, SF5CN VIII. Sulfur Isocyanide Pentafluoride, SF5NC IX. Pentafluorosulfanylamine and Other Derivatives X. Pentafluorosulfanyl N,N-Dichloroamine, SF5NCI2 XI. Pentafluorosulfanyl N,N-Difluoramine, SF5NF2 XII. Pentafluorosulfanyl Perfluoroalkylamines, SF5N(H)& XIII. SFSN(CF3)Z XIV. SF5N(X)CF3(X = F, C1, Br, I) x v . SF5N(CI)Rf(R,= CzF5, n-C3F7,n-C4F9) XVI. Bis(pentafluorosu1fanyl)perfluoroalkylamines XVII. Tris(pentafluorosulfanyl)amine,(SF5I3N SF5(CF3)NN(CF3)SF5 XVIII. Bis(pentafluorosulfanyl)bis(trifluoromethyl)hydrazine, XIX. Tetrakis(pentafluorosulfanyl)hydrazine,(SF&NN(SF& x x . Bis(pentafluorosulfanyl)amine,(SF&NH XXI. (SF5)zNX(X = F, C1) XXII. N-Pentafluorosulfanyl Haloimines, F5SN=CX2 (X = C1, F) A. SF5N=CCI2 B. SF5N=CFz XXIII. Pentafluorosulfanyliminodihalosulfanes,SF5N=SX2 (X = F, C1) A. SF,N=SFz B. SF,N=SClz XXIV. Pentafluorosulfanyl-P-sultones and Sulfonic Acids A. ~ ~ - 0 2

B. F50-2 References Visiting Professor (1992), Punjab University, Chandigarh, India. 125 Copyright 0 1994 by Academic Press, Inc. All rights of reprcduction in any form reserved.

126

VERMA. KIRCHMEIER, AND SHREEVE

I. Introduction

Fluorinated compoundscontaining five- and six-coordinatesulfur are of considerable interest. They include those with sulfur as the central atom surrounded by five or six ligands as well as many with six-coordinate sulfur as a functional group, such as pentafluorosulfanyl, SF,. Compoundsin which this group is present are of special interest because they often possess the advantageous properties of the parent compound, SF6,among which are a high group electronegativity, large steric bulk, a nonfunctional hexacoordinate stereochemistry, and high thermal and hydrolytic stability. These properties are manifested in various potential applications such as their use as solvents for polymers, as perfluorinated blood substitutes, as surface active agents, as fumigants, and as thermally and chemically stable systems (11. This chapter gives the reader a broad picture of the synthesis and chemistry of the various classes of pentafluorosulfanyl compounds, many of which are the subject of much ongoing research. I I . Pentafluorosulfanyl Halides

The two compounds of this class that are known, SF,Cl and to a lesser extent SF,Br, are important intermediates in the preparation of derivatives that contain the SF, group. The chloro derivative, SF,Cl, was first prepared in 1960 (2)as a minor product ofthe reaction between SC1, and HF-free F2diluted with nitrogen at - 10°C. Other methods of preparation include chlorination of S2FIo(3),electrolysis of an SC12/HF mixture (41, reaction of C1F with SF, ( 5 , 6 )and KSCN (7), and reaction of chlorine, CsF, and SF, (8). High yields are obtained as shown below (9, 10): SF,

+ CIF

CsF

SF5C1(92%).

25'C. 0.5 hr

A mixture of S2FIoand Br, at 140-150°C results in the formation of SF5Br (11, 12). An alternate method is the reaction of SF, with BrF (or a mixture of BrF, and Br,) at 90-100°C (13). It is reported that SF5C1 (m.p., -64°C; b.p., -21°C) is stable up to ca. 400°C in inert vessels, but decomposes at substantially lower temperature in the presence of Cu/Hg or in ultraviolet light. It is not hydrolyzed by water or aqueous acids, but is rapidly decomposed by aqueous alkali (14): SF5Cl + 7 0 H -

-

+ 5F- + C1- + H' + 3H20.

PENTAFLUOROSULFANYL COMPOUNDS

127

The bromo derivative, SF,Br (m.p., -78°C; b.p., 31"C), is less stable thermally, with decomposition starting at 150°C. The products of the thermal decomposition of SF,X (X = C1, Br) are SF6,SF,, and X, , The products of UV photolysis of SF5Br are SF,, SF,, S2Flo,and Br, (15). The vacuum-UV photolysis of SF,Br in a n Ar matrix a t 8 K provides a convenient method of generating BrF (16). The vapor-phase Raman spectrum of SF5Cl (17 ) ,the argon-matrix Raman and infrared spectra of SF,Cl and SF,Br (181, and the vaporphase infrared and liquid-phase Raman spectra of SF,Br (191,as well as photoelectron diffraction (20)and microwave spectra of SF5C1 (21) and SF,Br (22) have been reported. The ionization potential of SF, (9.65 eV) has been measured by photoionization mass spectrometry of SF,Cl (23). Phosphorus(II1) compounds are oxidatively fluorinated by SF,C1; e.g., C6H,PC1,, (C6H5),PC1, and CH3PC12 form the fluorophosphoranes C,H,PF,, (C,H,),PF,, and CH,PF,, respectively (24). Reaction of SF5C1 with methylamine yields CH,N=SF, (25). The first known metal-pentafluorosulfanyi complex, {PtC1(SF,)[P(C,H,),]2} is synthesized by the reaction of trans-stilbenebis(tripheny1phosphine)platinum(0) with SF,C1 in benzene (26). Under photochemical conditions SF5Cl can be reacted with some simple substrates. For example, with H2 it gives S2F1,(27);with 0 2 , the products are SFBOOSF,and SF,OSF, (28); and with N2F4, SF5NF, is obtained (29).The ability of SF5X to form the stable SF5. radical is a n important feature of its chemistry. In some instances, fluorination takes place with breakdown of the SF, group. Fluorination occurs in the gas-phase photochemical reaction between SF5C1 and SO,, which gives SOF,, S02F2, S02C1,, S2FI0, SF,OSO2F, and SF50SF5(30).Similarly, the photochemical reaction with CO yields COClF, COF,, SF,, S2FI0,SF,, and COS (31). When SF5C1 reacts with trimethylsilylcyanide, (CH3),SiCN, at -1O"C, a white sublimable solid, S(CN), (32),is formed. On reaction of SF5C1with (CF3I2NCNunder photolytic conditions, (CF,),NC(Cl)= NSF, is obtained (33),and with (CF3),NCF2N(C1)CF,, (CF,),NCF,N(SF5)CF3is produced in 50% yield. This product is characterized by spectral (IR, NMR, mass) data (34). The fluorine atoms of SF5X are substituted in reactions with nucleophiles such as dimethylaminotrimethylsilane,(CH3),NSi(CH3)3,and lithium hexafluoropropylidenimine, LiN=C(CF3), ; e.g., SF,X with (CH,),NSi(CH,), at -78°C gives (CH,),NSF,X (35, 36). The equivalence of four fluorine atoms in equatorial positions is supported by 19F NMR of the products (35).The replacement of F rather than C1 or Br in SF5X is probably favored because of the higher SiF bond energy.

128

VERMA, KIRCHMEIER, AND SHREEVE

The reaction of LiN=C(CF,), with SF,X leads to stepwise replacement of fluorine with concomitant fluoride ion migration as shown (37-39 ): SFSX + LiN=C(CF&-

SF3X[=NCF(CF&1 + LiF

SF3X[=NCF(CF3)z] + LiN=C(CF&-

SFX[=NCF(CF3)zlz + LiF.

Both SF,Cl and SF,Br react with unsaturated organic compounds. There are numerous examples of the addition of an SF, group and a halogen atom across a C=C, C=O, C=C, or C=N bond, although under certain conditions fluorination also occurs. The photochemical reactions of SF5C1 with olefins have been studied in detail and are believed to involve a chain reaction as shown below (40).

Similar reactions take place with other olefins (propene, cyclohexene, butadiene, and vinyl chloride). These reactions may also be initiated thermally (90-100°C). Some polymerization is observed. In fact, it is found that with isobutene and styrene only polymerization occurs (41 1. With less reactive fluoroolefins, addition of SF,Cl takes place only under photolytic conditions or by using a radical initiator such as benzoyl peroxide. Trifluoro-, tetrafluoro-, and chlorotrifluoroethylene as well as hexafluoropropylene give mainly 2-chloropolyfluoroalkylsulfur pentafluoride. With excess fluoroolefin, telomers of the type SF,(CF,),Cl are obtained ( 4 2 4 4 ) .Acrylic acid derivatives, CH,=CHR (R = CO,Me, CN), react with SF5C1 in Freon-113 at 120°C to give 32-42% SF,CH,CHClR (45). Addition of SF5X (X = C1, Br) to olefins gives rise to a variety of pentafluorosulfanylalkanes ( 4 6 ) .

PENTAFLUOROSULFANYL COMPOUNDS

129

CH,0CF=CF2 G O ) , CHF=CHF (511, (CH3)3SiCH=CH2 (52), CH,=CFCI, CH2=CHCF3, CHFECHCl, CHF=CFCI (53), CF,=CHCI (54),

X

= C1(40,43,48,49,50,52), X = Br (47,51,52,54)

As shown, the reactions of SF5Br with olefins are the same as those described for SF5C1,but, in general, take place under milder conditions in keeping with the greater reactivity of the bromo compound. Thus, C2H4 and SF,Br give SF,CH,CH,Br at room temperature and addition to the halogenated olefins, CH2=CHF, CH,=CF, , CHF=CF, , and CClF=CF2, occurs without irradiation or use of a catalyst (47, 51 ). Fluorinated polymers are formed when SF,Br reacts with the appropriate fluoroolefins at 90 ? 5°C and autogenous pressures of up to 90 atm for periods of 4 days to 2 weeks (53).

+ CH,=CHF SF,Br

+ CHZ=CFz + CFH=CF2

+ CF2=CF2

(excess)

-

SF5(CH2CHF),,Br

SF5(CH2CF2IsBr SF5(CFHCF2)30Bro.7 SF5(CF2CF2)16Bro,3

With C2H,, no polymer is obtained. In addition to the above polymers, the adducts SF,RBr (R = CH,CHF, CH2CF2, CFHCF,, CFzCFz, CH,CFCl, CH2CHCF3,CHFCHCI, CHFCFCI, CF2CC1H) are also formed (53,54).The monomeric adducts, with the exception of SF,CF2CF,Br, are reported (47).Polymers and telomers containing SF, are found in the patent literature (55)and elsewhere (56). The reaction between SF5X and acetylene is similar to that with olefins (57,58).The addition product SF,CH=CHX can be converted to SF,C=CH [X = C1 (11%)(57), X = Br (-50%) (58)l.Reactions between CH,C=CH or CF3C-CH and SF,Br give SF,CH=C(CH,)Br in 30%yield (b.p., 109 ? l°C) and SF5CH=C(CF3)Brin 58% yield (b.p., 93 ? l"C), respectively. The products are clear, colorless, hydrolytically stable liquids and are identified from their IR, NMR, and mass spectra (59).The reaction of SF5C1with CH3C=COR gives an 85% yield of ethylidenesulfur tetrafluoride, CH3CH=SF4 (60).A trigonal bipyramidal structure is proposed based on NMR studies. The addition of SF5C1to a C=O group occurs less readily than addition to C=C bonds. With ketene the product of reaction at 25°C in a pressure vessel is SF,CH,C(O)Cl. This acid chloride is a useful precursor to the strong acid SF,CH,COOH (45).

130

VERMA, KIRCHMEIER, AND SHREEVE

Addition of SF,C1 to the -C=N functionality is a reaction of considerable preparative significance (8). t ClCN

-

SFSN=CClp

SFSC1 t CFQCN__* SFSN=C(CF3)CI

+ C3FTCN

SF,N=C(C3F,)Cl

It is possible to fluorinate the products using NaF in tetramethylene sulfone. The compound thus obtained from SF5N=CC1,, i.e., SF5N= CF, , forms a mercurial, Hg[NCF3(SF6)I2,with HgF,, and on heating isomerizes to SF4=NCF3. With (C2N2)two molecules of SF5C1add to give SF,N=C(Cl)C(Cl)=NSF,. With methylamine SF,C1 gives CH3N=SF, (61). Irradiation of a mixture of SF&l and (CF3),NC1gives SF,N(CF3)2(62). When SF,Br reacts with pentafluorosulfanyl(fluorosulfuryllketene F5S(S02F)C=C=0, only BrF addition takes place, to give F,SCBr(SO,F)C(O)F (63).Sulfur tetrafluoride is the other product. Pentafluorothiophenyllithium readily attacks SF,Br at - 78°C to form bis(pentafluoropheny1)trisulfane and bis(pentafluoropheny1)disulfane (64).However, in its reaction with SF5C1,C6F,SC1 is produced in addition to the di- and trisulfanes (64).Regardless of the molar ratio of the reactants, these are the only products obtained. Similar behavior is observed in the nucleophilic reactions of trifluoromethylthiolithium. Reaction of SF5X(X= C1, Br) with C6F5Liforms an unstable intermediate, (C,F,),SXF, which is hydrolyzed t o (C6F,),SO. The mechanism for the formation of the compounds isolated is suggested (64). 111. Pentafluorosulfanyl Hypohalites, SF,OX

A. SF50F The first of these compounds to be isolated was SF,OF, formed in low yield by fluorination of SO, or SOF, with F, at 200°C using AgF, as a catalyst (65). It is also obtained in >90% yield by the reaction of F2 with SOF4 at 25°C in the presence of CsF in a static system (66, 67). The salt CsOSF, is believed to be an intermediate, analogous to the report that the salt CsOCF, is an intermediate in the reaction between F2 and COF2in the presence of CsF (68).Pentafluorosulfanyl hypofluorite is also obtained in the reaction between CsOSF, and FS020F at 100°C (69). Although SF50F(b.p., -55.1"C) is thermally stable to about 200"C, at higher temperatures it decomposes to SF6 and 0,(70). Photolysis

PENTAFLUOROSULFANYL COMPOUNDS

131

gives a low yield of the peroxide, (SF&O2, which is consistent with cleavage of the O-F bond on irradiation. That this bond also breaks upon thermolysis is indicated by the formation of SF,0NF2 when the compound is heated with N2F4 (71-73). When SF50Fis reacted with SO2 in the liquid phase, SF,, S02F2,SO,, and SF,0S02F are formed (74). With SF, it gives SF,OSF,, SF500SF5,and SF50SF40SF5.A similar reaction in the presence of O2 gives SF,OSF,OOSF, and SF,OSF,OOSF,OSF, as additional products. These new products react with benzene to give C6H,0SF,0SF,. The reaction of CF,OF with SF, gives CF30SF, as the only product. Trifluoromethyl hypofluorite, sulfur(1V) fluoride, and oxygen react to give CF,OSF,OSF, , CF,OSF,OOSF, , and a compound believed to be CF30SF,00SF,0CF,. A reaction mechanism is proposed (74). Other reactions of SF50Fthat have been investigated include reactions with CO a t 165°C to form COF2 and SOF, and those with CCl, in UV light, giving COF,, C12, and SF,OSF, (75). With NO2, SOF, and N02F are formed (75). Photolytic reaction of SF,OF with oxalyl chloride gives F,SOC(O)F (76). Pentafluorosulfanyl hypofluorite adds readily to a number of alkenes to give only one product, containing the components SF,O and F (77, 78). Because of its facile synthesis, some use is made of SF,OF in electrophilic fluorination (79). The gas-phase structure of SF50F is reported (80). B. SF,OCl This hypochlorite is considerably less stable than the hypofluorite. The hypobromite and hypoiodite have not been prepared. Pentafluorosulfanyl hypochlorite is synthesized from C1F and SOF, in the presence of CsF at -20°C (81, 82). Pentafluorosulfanyl hypochlorite (b.p., 8.9"C) is thermally stable up to about 20"C, a t which temperature it decomposes to SOF, and C1F. Upon photolysis, the peroxide (SF,)2O2 is formed in high yield (83-85) via cleavage of the 0-C1 bond. The yield of the peroxide is much greater than that observed when SF50F is irradiated. Photolysis of SF,OCl with N2F4 (84) and CO (86) gives SF,0NF2 and SF,OC(O)Cl, respectively. Because of the partial positive charge associated with chlorine in the hypochlorite, SF,OCl reacts readily with molecules containing negative chlorine. Seppelt exploited this property (87) and reacted SF,OCl with HC1 at - 95°C to obtain the unstable pentafluoroorthosulfuric acid, SF,OH, which decomposes via elimination of HF at -60°C. Al-

132

VERMA, KIRCHMEIER, AND SHREEVE

though the decomposition of CF,OH is thermodynamically more favored, its greater stability is attributed to a longer intramolecular H-F distance compared with that in SF50H (88,89). A low-temperature addition reaction occurs between SF50C1 and symmetric fluoroolefins to form pentafluorosulfanylalkyl ethers in nearly quantitative yields (78,90-92). It is found that with unsymmetrical olefins the chlorine atom of the hypochlorite most often bonds to the olefinic carbon atom with higher electron density. The reactions of SF50C1and SF,OF with fluorinated ethylenes are used to prepare new SF,O-substituted fluorocarbons in 44-77% yield (93). SFSOC1 t CF,=CFH

+SFSOCF2CFHCI

(44%)

SF5OCI t CF,=CFClSF,OCl

+ CF,=CFBr

SFSOCFClCF2Cl (50%)

-

SF,OCFBrCF,CI

+ SF50CF2CFBrC1 (77%)

The compounds are characterized by "F and 'H NMR, IR, and Raman spectral studies. Compounds of the type SF50Rf are thermally very stable. They do not yield the perfluorovinyl derivative, SF50CF=CF2, on dehalogenation as is the case when CF,OCF=CF, is formed from CF30CFC1CF2C1(94). At room temperature, SF,OCl adds quantitatively to C3F7NCto give C,F7N=C(C1)OSF5. The IR, NMR, and mass spectra are compatible with the structure (95). IV. Pentafluorosulfanylalkanes, Alkenes, and Alkynes

Several perfluoroalkylpentafluorosulfur(V1) compounds, &SF, (R,= CF,, C2F5, n-C3F7,i-C3F7,n-C4Fg,sec.-C4Fg,n-C5F11, i-C5Fll), are reported. When CF,ClSCl is passed over AgF, at 60-70°C with a residence time of 10 min, a 30% yield of SF,CF,, together with CF2C1SF3 (40%) and CF,SF, (30961, is obtained (96). Fluorination of CF,SSCF, with COF, provides SF5CF3in high yield (97). The reaction of CzF5SSC2F5with ClF a t 25°C (10 hr) yields SF5C2F5(17.1%). When RfSCl (Rf = C2F5,n-C3F7,n-C,Fg) is treated with C1F at 25°C (10 hr) small amounts of RfSF5are formed (98).The physical properties and spectral data (IR, "F NMR, and mass) are reported (96,98,99), as is the gasphase electron diffraction structure of SF,CF, (100). Electrofluorination of alkane thiols in AHF leads to variable yields of perfluoroalkylpentafluorosulfur(V1)compounds (101, 102). Dithiols

133

PENTAFLUOROSULFANYL COMPOUNDS

and sulfides also yield SF5Rf. Dithiols give both cyclic and acyclic sulfur(V1) derivatives (1031,e.g., HS(CH2),SH-

ECF

SF5(CF2),,SF5+ ( m F 4

n=4,5

Electrochemical fluorination of sulfur dissolved in CS2in the presence of chlorine gives several products, including SF5CF3,SF5CF2SF5,and CF2C1SF,(104).The reaction of elemental fluorine with branched alkyl mercaptans or sulfides gives SF, organofluorine compounds (105). FZ/He

(CH,)3CCHzSH

- 120°C to rwm temperature

FZ/He

(CH,),CSH

* SFSCF2C(CF3)2F

- 120°C to rmm temperature

Fz/He

(CH,),CSC(CH,),

'SFSCF2C(CF3)3

- 120°C to rwm temperature

SFSCF2C(CF3)2F + C(CF3)3F

F,/He

(CH,)*HCSCH(CH3)2

SFSCF2CFzCF3 - 120°C to rwm temperature

Direct fluorination of CS2with F2diluted with He at - 120°C (3 days), followed by warming to -80°C (3 days), gives a 15% yield of (SF5),CF2 (106,107).The IR spectrum of this product is in agreement with that provided earlier (108). The gas-phase structure of (SF5l2CF2is reported (109). The compound SF,CH, can be obtained from SF5CH2COOAgby the following sequence of reactions. SF5CH2COOAg+ X

2

A SFSCHZX

X2 = Br2 (110,111), 12 (112) SF5CH2X

SFSCH3

Similarly, SF,CH2F (113)and SF5CHF2(114) are also formed. The parent acid is obtained by the reaction of SF5C1with the ketene, CH2= C=O, followed by hydrolysis of the resulting acid chloride (110-112). SFSCI + CHZ=C=O-

SF5CH2C(O)Cl-

HZ0

SFSCHZCOOH

.

134

VERMA, KIRCHMEIER, AND SHREEVE

The silver salt, SF5CH2COOAg,is formed by reacting the acid with silver carbonate (50, 111, 112). When SF5CH2Bris lithiated at -llo"C, LiF is lost on warming to leave the remarkably stable methylene sulfur tetrafluoride, CH2=SF4 (111).It is also synthesized by the reaction of O=C(C1)CH,SF5 with [Mn(CO),]- (115,116). Pentafluorosulfanylalkanesare also obtained by saturation of SF5containing olefins. Dehydrohalogenationof SF,-alkanes results in the formation of SF,-alkenes, which can be converted to additional SF,alkanes. SFSCHFCF2X-

SF&HFCHFBr

KOH

SF,CF=CFz ( 4 3 , 4 7 )

reflux

KOH

(X = Br, CI)

SF,CF=CHF (51)

retlux

A variety of materials add across the double bond in the resulting vinyl-SF, compounds (117), e.g., SF5CH=CH2

70°C

+ SFSBr

SFSCHBrCHzSF5.

8 hr

Attempts to dehydrobrominatethis bis-SF, compound result in the loss of SF, and HF to give SF5CBr=CH2. Indirect "HF" addition (via KF-formamide) to pentafluorosulfur olefins yields hydrylpentafluorosulfur-F-alkanes. SFSCX=CFZ

+ HF

-

SFSCH(XKF3

X = H, F, CF3 (118)

Reaction of SF,CF=CF, with a mixture of I, and IF, gives SF,CFICF, (119).The presence of H in the hydrylpentafluorosulfur-F-alkanesprovides a site to introduce additional functional groups (120).Reaction of SF5CHXCF3with S206F,gives the corresponding fluorosulfate (118, 121).Similarly (CF3),CHSF5with S206Fzat 80°C (25 hr) forms (CF3)ZC(OSO,F)SF, (66%yield) (118).Both (CF3),N0 and C1, add as well. SFSCH=CHX X

=

H, Y

+Y

-

SFSCHYCHXY

(CF3)ZNO;X = CH3, Y

= C1z (121)

PENTAFLUOROSULFANYL COMPOUNDS

135

In some cases, cyclization occurs (122).

SF5CF=CF2

+ CHz=CHCH=CH,

-

CHZ=CHdHCF&H2(!XF)SFS (56)

Trifluoroethenylpentafluorosulfur(VI), SF,CF=CF, , is obtained in 80% yield by debromination of SF,CFBrCF,Br with Cu at 188°C (0.5 hr, 3 Torr). Irradiation of SF,CFBrCF,Br in a silica tube in the presence of mercury gives SF5CF=CF2 in 67% yield (56).Trifluoroethenylpentafluorosulfur(V1) is stable in air up to 300°C. It decomposes at 380°C, giving SF, and unidentified fluorocarbons (56).In perfluorobutylamine SF5CF=CF2 reacts with 0,- or 0, at 25°C (12 hr) to give, respectively, 72 or 52% yields of FC(0)CF20SF5,as well as CF3C1, SOF,, S02F,, and (SF5)20.The IR and 19FNMR spectra of FC(0)CF20SF5are reported (123). When SF5CF=CF2 reacts with NaOCH, in CH,OH a t 60°C (20 hr), SF,CHFCF20CH3 is formed (90% yield; b.p., 73.4"C) (56). In aqueous solution in the presence of K,S,O,-Na,SO,, copolymerization of CF2=CFSF5 with CH2=CF2 at 85°C (20 hr) gives [(CH2CF2)xCF,CF(SF,)], [68% yield; v SF = 867 (vs, br) cm-'I (56).With KF and formamide, CF3CHFSF5 is obtained (86% yield; b.p., 26.9"C) (118). Under similar conditions SF5C(CF3)=CF2 gives a 91% yield of (CF3),CHSF, (b.p., 52.2"C) (118). The former is obtained in -60% yield by shaking CF,CH(SF,)CF,Cl with excess KOH at 20°C for 1hr. The SF, monomer, 2-chloro-3-(pentafluorosulfur)tetrafluoropropene[SF5CF2C(Cl)CF,], is prepared in 20% yield by the addition of SF5Cl to CF2= C=CF, in a metal vessel at 100°C (124). A mixture of AgF in acetonitrile reacts with excess SF5CF=CF2 in 2 2 : 1 molar ratio at 25°C and with stirring to yield AgCF(CF,)SF, that is isolated as the acetonitrile solvate. The solvate is stable up to 50°C. Thermolysis at 80-90°C gives CF,CF(SF,)CF(SF,)CF, as the major product. Other products include C2F5CF(CF3)SF5,SF,, S2F1,, SOF,, CF3CF=CFCF3, and Ag. The solvate reacts with HX (X = C1, Br, OH) at 20°C (1hr) to give CF3CHFSF5and traces of CF3CF(SF5)CF(SF,)CF,. With CH,I at 20°C (12 hr), SOF,, SiF,, and CF3CHFSF, are formed (125). Bromination of AgCF(CF3)SF5 with dry Br, over the temperature range from - 196 to 20°C (12 hr) gives CF,CFBrSF, (45% yield; b.p., 57.2"C) and (CF,CFSF,),. The latter compound is unstable and decomposes to S2F1,and CF3CF=CFCF3 a t 100°C (125).The IR, 19F NMR, and mass spectral data of CF,CF(SF,)CF(SF,)CF, are reported (125). A mixture of SF5CF=CF2 and Br, a t 20°C (20 days) in the dark

136

VERMA, KIRCHMEIER, AND SHREEVE

gives only traces of CF,BrCFBrSF,. Yields of -91% of CF,BrCFBrSF, are obtained on UV irradiation of the same mixture for 20 hr in a silica tube. The photochemical reaction between SF5CF=CF2 and HBr also give CF,BrCFBrSF, in 17.5% yield. An equilibrium mixture of SF,CF=CF2, C1, , and Br, in CH,C1, gives a mixture of CF,BrCFClSF, (9.5%), CF2C1CFC1SF, (l%), CF,ClCFBrSF, (6.5%), and CF,BrCFBrSF, (62%).The same reaction at 80°C (22 hr) gives CF2BrCFC1SF, (13%), CF2C1CFC1SF5(l%), CF2C1CFBrSF, (15%), and CF,BrCFBrSF, (51%) (56). Spectral data (19FNMR) for CF,BrCFBrSF,, CF,BrCFClSF,, and CF,ClCFBrSF, are available (99). Irradiation (UV) of a mixture of SF5CF=CF2 and CF,I in a sealed silica tube gives a 9% yield of CF,CF,CFISF, (56).An equimolar mixture of S2Floand ICF2CF21when pressurized with CF,=CF2 (150 psi) and heated to 150°C for 4 hr yields ICF,CF,SF, (126). Heating the mixture to 150°C with intermittent injection of CF2=CF2 leads to the formation of a product mixture, which on distillation gives three main fractions identified as SF5CF2CF21,SF5(CF2CF2)21, and SF,(CF,CF,),I. The mass spectrum of the residue shows parent ions corresponding to SF5(CF2CF2),I,n = 4-9. With slow warming, a mixture of S2FI0,12, and C2F, (pressurized) (from 20 to 150°C) gives a violet liquid with a n estimated molecular composition of SF,CF2CF21(34%), SF,(CF2CF2)21 (35%),SF5(CF2CF2)31 (14%),SF5(CF2CF2)41 (5%), and SF,(CF,CF,),I (2%) (125, 126). An interesting reaction involving iodo-SF, alkanes is oxidative fluorination by ClF3 (127).

With NaOC1, SF,CF=CF, ,in Freon-113 and in the presence of a phase0

I\

transfer catalyst [N(n-C8H1,),CH3+C1-1, forms the epoxide SF,CFCF2 (128).The epoxide is decomposed by ether and reacts with CsF to give primarily SF4and CF,C(O)F. Carbonyl fluoride reacts with SF5CF=CF2 in the presence of CsF and acetonitrile to give SF,CF(CF,)C(O)F, which reacts further with NH,, CH,OH, and H 2 0 to give, respectively, SF,CF(CF,)C(O)NH,, SF5CF(CF3)C(0)OCH3,and SF5CF(CF3)C(0)OH (129).On dehydration with P4010,the amide, SF,CF(CF,)C(O)NH,, gives the nitrile SF,CF(CF,)CN. All of these compounds, with the exception of the amide, are colorless, stable liquids. The amide is a stable, white solid (m.p., 32-34°C). Trialkylphosphites react with F-alkenes, e.g., CF,CR’=CF,, to give the phosphonates, CF,CR’=CFP(O)(OR), (R = Et, ‘Pr; R’ = F, CF,),

137

PENTAFLUOROSULFANYL COMPOUNDS

and alkyl fluorides (RF) via an Arbuzov-type reaction (130).The use of trimethylsilylphosphites, (RO),POSiMe, (R = Et, SiMe,), is advantageous because of their greater nucleophilicity and the ease of formation of trimethylsilylfluoride (131). Trimethylsilylphosphites react with SF,CF=CF, to give alkenylphosphonates SF5CF=CFP(0)(OR)2 (R = Et, SiMe,) (132).Only (E) isomers are formed. These compounds are colorless, moisture-sensitive liquids. The SF,-substituted iodoperfluoroalkene, SF,CF=CFI(E), is obtained in 24% yield (133)via the reaction oftrimethylphosphine, iodine, and SF,CF=CF,. It is a colorless liquid and is characterized by spectral data (133). The substituted acetylene, SF5C=CH, is obtained in 11%yield from the reaction of SF,C1 with acetylene (134).It is also formed in -9% yield in a four-step synthesis from the reaction of SF,Br with C2H, (135). SF5Br + C H E C H

-

SF5CH=CHBr

SF5CH=CHBr + Br,

SF,CHBrCHBr

F5SCBrCHBr

SF,CHBrCHBr,

%CO, 25°C

Zddiglyme

+

/"=%H Br

/"=%Br

Br

* F,SC E C H

This SF,-acetylene can also be obtained in -50% yield by dehydrobromination of SF,CH=CHBr (135).The gas-phase electron diffraction structure of SF,C=CH is reported (136).In the 'H NMR spectrum, the acetylenic proton resonates at a more shielded position and appears as a pentet with JF-H = 3 Hz (134). This suggests that it couples significantly only with the four equatorial S-F atoms and not the axial S-F atom. Of particular interest are comparative 19F NMR spectral studies of F,SC=CH and other saturated hydrocarbons/fluorocarbons containing the SF, group (137). Pentafluorosulfanyl acetylene is a useful starting reagent for the synthesis of a variety of SF, derivatives of saturated ethers, vinyl ethers, pyrazoles, cyclic alkenes, and alkyl-substituted phenyl-sulfur pentafluorides (134).It is also used for the preparation of a number of F,S-containing alkenes and alkynes (138,139).

+ SF5C=CH SF5Br

+ RC=CH

110°C

3 hr

(R = CH3, CF3)

-

SFSCBr=C(H)SF5

F,SCH=C(Br)R-

KOH

KOH

FSSC~CSFS

F5SC=CR

138

VERMA, KIRCHMEIER, AND SHREEVE

Mono- and bis(pentafluorosulfur)diacetylenes, F,SC=C-C=CH and F,SC=C-C=CSF,, are obtained by the addition of F,SBr to diacetylene followed by dehydrobromination. These monomers yield interesting polymers (140). A variety of uses are proposed for several of these pentafluorosulfur-containing alkanes and alkenes, e.g., as dielectric insulators (122, 141 ), elastomer precursors (142), blood substitutes (143), fumigants (144), and insecticides (145). V. Sulfur isocyanate Pentafluoride and Sulfur lsothiocyanate Pentafiuoride

Pentafluorosulfanyl isocyanate, SF5NC0 and SF5NCS,were first reported in 1964 by Tullock et al. (8).They are obtained from the reaction of pentafluorosulfanyl(trifluoromethy1)aminewith benzoic acid and thiobenzoic acid, respectively. An alternative preparative method of F,SNCS involves thiolysis of F,SN=CC12 with H2S in the presence of NaF.

Hydrolysis of SF,N=CCl, gives only traces of SF,NCO. Another route for the preparation of SF,NCO is found in the reaction of NSF,, COF2, and AHF (146). Reaction of N,N'-bis(pentafluorosulfanyl)urea,(SF,NH),CO, with a slight excess of COF, at 100°C (12 hr) gives essentially pure SF,NCO (147).When COC1, is used instead of COF2,the reaction proceeds much less cleanly. Nevertheless, the infrared spectrum of the product mixture suggests that the reaction proceeds through the formation of a n intermediate cyclic compound, which is not isolated.

With N,N' -bis(trifluoromethyl)urea, the corresponding cyclic intermediate has been isolated and characterized (148). The preparation and purification of SF,NCO are greatly simplified by first preparing SF,NHC(0)F from a n equimolar reaction mixture of NSF,, COF,, and AHF (149), followed by dehydrofluorination (150, 151 >. Although reaction of SF,NH, with COF, gives SF,NCO in good yield

PENTAFLUOROSULFANYL COMPOUNDS

139

(146), the analogous reaction of SF5NH2with either CSClF or CSCl, fails to produce more than 1-2% of SF,NCS (150).Two methods that often produce isothiocyanates in high yield involve the reactions of iminodichloromethanes with either Na2S or P2S5 (152).Reaction of dichloro(pentafluorosulfany1imino)methanewith P2S5 in refluxing toluene gives SF,NCS in 70% yield (150).It is also obtained in -40% yield when SF,N=CCl2 reacts with triphenylphosphine in benzonitrile for 1day (153). Both F,SN=C=O (b.p., 5-55°C) and SF,N=C=S (b.p., 47-48°C) are well characterized by their elemental analyses and IR, 19F, and 13C NMR spectral studies (8, 146, 150). The gas-phase structure of SF,NCO obtained by electron diffraction and microwave spectroscopy is reported (154). The isocyanate, SF,NCO, is easily hydrolyzed to SF,NH, and COP, whereas SF5NCS is hydrolytically very stable. Both compounds undergo addition reactions with substrates containing easily replaceable hydrogen atoms, i.e., alcohols, thiols, and amines. With alcohols (thiols) the isocyanate and isothiocyanate give urethanes and thiourethanes, respectively. SFSNCO + ROH

-

SFSNHC(0)OR

R = CH3, CH2CH2C(O)C(O)CNHSFS,CsH5, I-C6H,C(O)C(O)CNHSF,, (8)~ 4-C6H,OH (15O),C ~ H S C H

SFSNCS + CH30H SFSNCO + RSH SFSNCS + CHSSH

SFSNHC(S)OCH,

SFSNHC(0)SR R

=

CH3, CsH5 (150)

SFSNHC(S)SCHz

The urethanes are stable indefinitely in aqueous solution, but are decomposed by aqueous alkali. The dithiourethanes are, however, unstable and decompose readily at room temperature. With PCl, the urethane SF5NHC(0)CH3gives SF5NC0 as the major product (1551, whereas the fluorosulfonylurethanesgive the corresponding sulfonylchloroimines (156).The thiourethanes, SF,NHC(O)SR, give SF,NCO as well as the imine (155).The crystal structure of SF5NHC(0)SCH3 is reported (157). On treatment of SF,NCO with polynitroalcohols, carbamates are produced. They are potentially energetic and have densities of -2 g/cm3(158). SFSNCO

+ R(N02),CH20H

-

SFSNHC02CH2R(N02),

140

VERMA, KIRCHMEIER, AND SHREEVE

Such carbamates can be further nitrated, e.g., N-pentafluorosulfanyl3,3,3-trinitropropyl carbamate is nitrated with trifluoroacetyl nitrate [a mixture of (CF3CO)20+ 100% HNO31to produce N-pentafluorosulfanyl-N-nitro-3,3,3-trinitropropyl carbamate in 43% yield (159). iCF,COI,O + HNO,

SFSN(NOZ)COzCHZCH2C(N02)3

SF,NHCO,CHzCHzC(N02)3

These nitro compounds, which have a pentafluorosulfanyl group attached, exhibit increased density, decreased shock sensitivity, and good thermal stability and release considerable energy upon detonation (160). Pentafluorosulfanyl isocyanate reacts with ammonia and primary, secondary, and tertiary amines to form a variety of substituted ureas (150). + RNH,

-

SF,NHC ( 0 )NHR

R = H, CH, CH,CH,NHC ( 0 )NHSF,, C,H,,

SF,NCO

+ RR NH

-

4-C6H,CH,C,H,NHC

SF,NHC (0)N R R

R =R

+ R,N

( 0 )NHSF,

= C,H,,C,H,

to)N+R,

SF,N-C

N+R, = +NQ '

, N+H:,

The thioisocyanate, SF,NCS, reacts with aniline to give SF,NHC(S)NHCGH,. The substituted ureas undergo thermal decomposition. The zwitterionic derivatives, SF,N-C(O)N+R, , are far less stable thermally than the analogous fluorosulfonyl derivatives (1611. The reaction of SF,NCO with triphenylphosphine gives some evidence for a zwitterionic compound, but the compound could not be isolated (1611. On the other hand, fluorosulfonyl isocyanate does react with tertiary phosphines, producing the corresponding adducts in high yield (162).The IR, NMR, and mass spectral data for these substituted ureas are reported (150). Dimethylsulfoxide, aldehydes, and formamides react with SF,NCO to give imines and amidines (147,150).

+ (CH,),SO

+ RC(O)H SFSNCO

-co,

SFSN=S(CH,)z

-coy

SFSN=C(H)R

(R = CsH5,4-CGH4CH,, 4-CsH4OCH3)

141

PENTAFLUOROSULFANYL COMPOUNDS

-co,

+ RR'NC(0)H

SFSN=C(H)NRR'

(R= R'

SFSNCO

+ (CH3)2NC(O)CH,

= CH3; R = CH3, R' = CsH5)

-co,

SF,N=C(CH3)N(CH3)2

Certain electron-deficient isocyanates are known to react with organic carbonyls and sulfoxides to yield the corresponding imines (163-165). With acetylacetone, SF,NCO gives the N-(pentafluorosulfanyl) amide of diacetoacetic acid, which enolizes to give two enolic products in CDC13. In (CD,),SO, however, only one enolic form and a keto form is observed. The structures of these tautomers are supported by 'H, 19F,and 13C NMR spectroscopy (150). Analogous reactions with both chloro- and fluorosulfonyl isocyanates are also reported (166).Reactions of SF,NCO with trimethylorthoformate, HC(OMe), , give SF,NHC(0)C(OMe)3 and SF,N(CH,)C(O)OMe (150).The formation of the latter probably takes place via the same mechanism as has been proposed for a similar reaction with chlorosulfonyl isocyanate (167). When SF,N(SiMe,)C(O)OMe is reacted with CsF in the presence of 18-crown-6,Cs(18-crown-6),[SF5NC(0)0Me] is formed. Decomposition occurs to give [Cs(18-crown-6)z][SF,l,which is characterized by singlecrystal X-ray analysis. The SF, anion is described as naked SF,; i.e., the distance between Cs+ and SF,-F is >7 A (167b). At 60-80"C, SF,NCO reacts with PCl, to give SF,N=CCl, (147). The product is identified by comparison of spectral data obtained with those reported earlier (8). Amides containing SF, form when SF5NC0 reacts with carboxylic acids (148). The reaction is believed to pass through a mixed acid anhydride intermediate that loses COz to give the corresponding N-pentafluorosulfanyl amide in good yield. No reaction is observed with carboxylic acids having electron-deficient carboxylate groups. SFSNCO t RCOOH

-

-co,

[SF,NHC(O)OC(O)R]

SFSNHC(0)R

R = CH3 (98%); CH=CH2 (35%)

With (dimethylamino)triethylsilane,fluorine loss occurs via formation of (CH3),SiF (168).

iaopentane

SF5NC0 + (CH3)3SiN(CZH5)2

-196 to -3o'C

SF4=NC(0)N(C2H5)2

142

VERMA, KIRCHMEIER, AND SHREEVE

On reaction with trimethyl(methoxy)silane, cis-CH30SF,NC0 is formed (169). The reaction appears to proceed as shown below, with subsequent intramolecular migration of the methoxy group. SF,NCO

.

F

I NCO I .OCH,

F

S

F'

F,S,

+ (CH,),SiOCH,

,C

( 0 )OCH,

N

I

-(CH,),SiF

TCH3),

F

On further reaction with methanol, the methoxy compound gives a urethane derivative. cZS-CH,OSF~NCOt CH30H

-

c~s-CH~OSF~NHC(O)OCH~

VI. Sulfur Cyanate Pentafluoride, SF,OCN

Seppelt et al. (170)reported the first synthesis of SF50CN by the sequence of reactions below. SF,OCI

+ Cl,C=NCl

Freon-114 -120 to -70°C

-

(16%)

F,SO Hg, -20°C 4 hr stirring

SF50CC12NC1'2

-c1,

SF,OCC1,NCl2

\

c1I"="\c1 (i) (60-90%)

(i)

Hg, 25°C ultrasound

F,SO

+

\

F'

/C=N CI (ii) (10-409)

SF,OCN (10%)

The identifi ation of isomers i and ii is made on the basis of NMR data (170). Only isomer i undergoes further chlorine elimination in the presence of mercury to afford the desired product in 10% yield. Isomer ii under similar conditions decomposes to give SOF, and ClCN. The IR and 14N NMR spectra (170, 172) are used to differentiate between SF5 cyanate and SF, isocyanate.

PENTAFLUOROSULFANYL COMPOUNDS

143

The cyanate, SF,OCN (b.p., 5°C; m.p., -SOT), undergoes rearrangement at high temperature to give SF,N=C=O. The IR, Raman, 19F and 14NNMR, and mass spectral data for SF,OCN are reported (170). Its geometric structure has been determined by gas-phase electron diffraction spectroscopy (173)and the results have been compared with Ab initio calculations (173) the gas-phase structure of SF,NCO (154). are consistent with the experimental geometries of both isomers.

VII. Sulfur Cyanide Pentafluoride, SF,CN

The synthesis of SF,CN by fluorination of methyl thiocyanate, CH,SCN, was first claimed in 1959 (174),but the claim was in error. The actual product isolated was the isomer CF3N=SF2 (175).Various other attempts to synthesize SF,CN, such as by reaction of (FCN), with SF, in the presence of CsF (176),the photolytic reaction between S,F~,, and (CN)2(111, and metathesis between CsSF, and BrCN (177), failed. It was presumed that SF,CN, if formed, isomerized to CF,N=SF, (I 77). The first successful synthesis of SF&N (5%yield) by fluorination of (SCN), in FC1,CCF2C1 with elemental fluorine diluted with N, (1: 10) at -20°C is reported by Losking and Willner (178).It is a stable, colorless gas at room temperature (b.p., -25°C). It does not isomerize as suggested in the earlier literature (177).It does not decompose even on pyrolysis at 350°C. The IR and 19FNMR spectral data and molecular weight data are consistent with the structure (178).The molecular structures obtained by gas-phase electron diffraction and microwave spectroscopy concur (179).

VIII. Sulfur lsocyanide Pentafluoride, SF,NC

The preparation of SF,NC in about 5% yield by the following sequence of reactions is reported by Thrasher (180). SFhN=CClz

+ 3HF

-

SFSNHCF, + 2HC1

SF5NHCF3+ BBr3

SF5N(CF3)BBr2

-HBr

SF,N(CF3)BBrz SFSN=CBr2

SF5N=CBr2

-BF3

MgiTHF

-MgBr,

SF5NC

144

VERMA, KIRCHMEIER, AND SHREEVE

A similar sequence of reactions gives CF,NC in 65-90% yield (1811. Reaction of SF5N=CBr, with lithium alkyls/aryls also produces SF5NC in low yield (-5%) (180).However, SF5N=CC1, under similar reaction conditions produces only traces of the isocyanide. Pentafluorosulfanyl isocyanide is a colorless gas and slowly isomerizes to SF5CN at ambient temperature. The IR and lgFNMR spectral data of SF5NC are reported (180). IX. Pentafluorosulfanylarnine and Other Derivatives

Pentafluorosulfanylamine is prepared in -34% yield by the addition of AHF to NSF3 (182). NSF3

+ 2HF

-

FSSNH2

It is a volatile white solid (m.p., 43°C). The vapor pressure is given by the expression log P,, = -209612' + 9.145. It is soluble in ether even at -78°C. Dissociation into NSF, and HF increases rapidly with temperature and the presence of moisture. It is stable when stored at -78°C and can be handled in a dry glass vacuum system. In aqueous base, hydrolysis occurs. SFSNHz + 6 0 H -

-

SO3NH-

+ 5F- + 3Hz0

Some important reactions of SF5NH, are summarized below.

+ SFI + SOFI + Fz

SFSNHZ + COF2

+ SCI, + PC15

+ SOCI, + Clp-

SFSN=SF*

(183,184)

SFSN=SOFz (185) SFSNFz

(186)

SF5NC0

(146)

SFSN=SClp (187,188) SFSN=PClS (147) SFSN=SC12 (189)

With BF3 and PF,, SF5NH, forms 1:1 adducts (183).Reaction occurs at room temperature between SF5NH2and various acid chlorides and fluorides containing electron-deficient carbonyl groups to produce N-pentafluorosulfanyl amides, F5SNHC(0)R(R = F, CF3, CH,) (149). The reaction of SF5NC0with certain carboxylic acids at room temperature provides a n alternate route for the preparation of amides, SF,-

PENTAFLUOROSULFANYL COMPOUNDS

145

NHC(0)R (149). Malonic acid reacts with SF,NCO a t 60°C to give both the amide acid, SF,NHC(O)CH,COOH, and the diamide, SF,NHC(O)CH2C(0)NHSF5(149). This diamide is also obtained from the reaction of SF,NH, with carbon suboxide (149). 2SF5NHz + C3O2

-

SFSNHC(O)CH2C(O)NHSF,

In water, SF,NHC(O)F gives the urea derivative,. (SF,NH),CO (190). Reaction of the amide SF,NHC(O)R with PCl, gives the corresponding pentafluorosulfanylimines.

-

SFSNHC(0)R+ PC15

60-100°C

CCI,

SFSN=C(Cl)R + POCl3 + HCI

Similar reactions are employed for the synthesis of chlorimines from amides (191 and N-fluorosulfonylimines (192, 193). Acylation of SF,NH, with oxalyl chloride produces the corresponding diamide; SF,NHC(O)C(O)NHSF,, in 78% yield, whereas acylation by fluorosuccinyl chloride yields both the diamide, [SF,NHC(O)CF,I,, and the cyclic succinimide, SF5k(O)CF2CF26O.The identity of all these compounds is confirmed by IR, NMR, and mass spectrometry (149). Nucleophiles such as H20, NH,, and CH,OH open the ring of the cyclic imide to give products such as SF,NHC(O)CF,CF,C(O)X (X = OH, NH,, OCH,) (190).Mono- and disubstituted products, SF,NHC(O)(CF,),,,C(O)F and SF,NHC(0)(CF2)3,4C(O)NHSF5, are formed by the reaction of SF5NH2 with perfluoroglutaryl chloride and perfluoroadipoyl fluoride, respectively (190). The amide acid fluorides are hydrolyzed by atmospheric moisture to the amide acid, SF5NHC(0)(CF2)3,,C(O)OH.The amide, SF,NHC(O)NHSF, (1461, reacts with PCl, to produce the carbodiimide, SF,N=C=NSF, (149). The latter is also obtained by the reaction of SF,NH, with SF,N=CCl2 (147). X. Pentafluorosulfanyl N,N-Dichloroamine, SF,NCI,

Chlorine monofluoride reacts with NSF, at -78°C to give SF,NC12 in 25-32% yield (194, 195). It is also prepared by the reaction of C1, with NSF, in the presence of HgF, (196). NSFB + ClF NSF3 + 2C12

-78°C HgFz

FSSNC12 FSSNC12

146

VERMA, KIRCHMEIER, AND SHREEVE

Pentafluorosulfanyl N,N-dichloroamine is a light yellow volatile liquid [b.p., 64°C (extrapolated; m.p., -120°C]. It is sensitive to mechanical shock and is thermally unstable at 80"C, giving SF,Cl, N,, and Cl,. It is hydrolyzed slowly to give SF,N(H)Cl and finally SF5NH2.It reacts slowly with mercury, producing NSF, . It reacts with PC13 and Se2C12 or Se at low temperature to give SF5N=PC13 and SF5N=SeC1,, respectively, in >80% yield (197). On warming to room temperature, SF,N=PCl, decomposes, giving PF, , Cl, , and (NSCl), .By comparison, SF5N=SeC12 decomposes to give NSF3, SeF, , and SeC1,. The dichloroamine reacts with SC1, or S&l, to give SF,N=SCl,, which is also unstable. With SF5N=SC12, SF5NC12reacts to give SF,N=S=NSF, (198).

With HC1, SF5NC12forms a n adduct, SF,NH,.HCl, which decomposes giving NSF, , HF, and HC1. Analogous reactions of anhydrous HC1 with RfNC1, are well known (199-201 ).

XI. Pentafluorosulfanyl N,N-Difluorarnine, SF,NF,

In 1963 three groups reported the synthesis of SF5NF2(29,202,203). The best preparative method is the UV irradiation of SF, or SF,Cl with N,F,. It is a colorless gas (b.p., -17.5"C) and can be stored in steel cylinders at room temperature. It slowly decomposes on heating to SF, and NF,. The IR, NMR, and mass spectral data (29,202) and the gas-phase structure are reported (204).

XII. Pentafluorosulfanyl Perfluoroalkylarnines, SF,N(H)R,

A 75%yield of SF,N(H)CF, is obtained by the reaction of AHF with perfluoroazomethine, SF,N=CF, (8).Reaction of SF,N=CCl2 and HF also gives the same product. SF,N=CF2

+ HF

-

SFSN(HEF3

PENTAFLUOROSULFANYL COMPOUNDS

147

This amine is a thermally stable liquid (b.p., 28.5-31°C). Although it does not attack glass, it is completely hydrolyzed by aqueous alkali. Its IR, NMR, and mass spectral data are reported (8). The higher homologue, SF5N(H)C2F,,is obtained as one of the products during the reaction of HF with SF5N=C(C1)CF3 (8).It is a stable liquid (b.p., 45.5-47°C). The reaction of HF with SF5N=C(C1)C3F, does not give SF5N(H)C4Fg,but rather SF,N=CFC3F7 (62% yield), which apparently results from the rapid loss of HF by SF5N(H)C4Fg.

XIII. SF,N(CF,),

Dobbie (208) reported the preparation of SF,N(CF3)2 (-10% yield) by prolonged irradiation of a mixture of SF5C1/SF4and (CF3I2NC1.It is a stable compound (b.p., 33°C) and is unaffected by acid or alkali at room temperature. The compound SF5N(CF3)CzF5is also reported (2091.

XIV. SF,N(X)CF, (X = F, CI, Br, I)

The fluorination of sulfur difluoride imides gives SF5N(F)Rf.

When Rf = CF, or CzF6, CsF is used as a catalyst (2051,but when Rf = S02F,the presence of CsF is not necessary (206). Roberts (207) suggests CF3N(F)SF5as the probable structure for the product of fluorination of methyl thiocyanate with elemental fluorine, although the actual workers (174)proposed the alternate structure, SF5CF2NF2. The reaction between SF5N=CC12 and HgFz gives the mercurial, Hg[N(CF3)SF512,in almost quantitative yield (8,210).

Reactions of the mercurial with halogens or interhalogens lead to the formation of a series of pentafluorosulfanyl N-halo(trifluoromethy1) amines, SF,N(X)CF, (X = F, Br, C1, I) (210).

148

VERMA, KIRCHMEIER, AND SHREEVE

2FZ

2SF,N(F)CF,

-HgF2

-

(54%)

PClz

PSF,N(Cl)CF,

-HgCI,

(98%)

H~[N(CF,)SFSI~

2Br2

-HgBr, 21CI

2SF5N(Br)CF3 (80%)

2SFSN(I)CF, (unstable)

When a mixture of SF5NHCF3,AgF,, and C1, is heated, SF,N(C1)CF3 is obtained (8).The N-fluoro derivative is also reported from the direct fluorination of both CF3N=SF2 (205)and SF5N=CF2 (222 ). Halogens do not react with the mercurial Hg[N(SF5),I2(212)or Hg[N(SO2CF3),1 (213). Instead, these N-haloamines are formed either by alternate methods (214,215)or with polar halogenides such as BrOS0,F (213). On the other hand, all of the N-halobis(trifluoromethy1)amineswith the exception of N-fluoro derivatives are obtained from the reaction of Hg[N(CF3),I2with halogens (216).Mews (212) has attributed the lack of reactivity of Hg[N(SF5),l2with halogens to greater N-X bond polarity in the N-haloamines, which would result from the greater group electronegativity of N(SF,), (3.2-3.45) compared with (CF,),N (2.85-3.0). It is also possible that the greater steric bulk of the SF, groups lowers the reactivity of this mercurial compound relative to Hg[N(CF3)2]2.The spectroscopic data [IR, NMR, mass] of the N-halo derivatives are reported (210). Methyl iodide reacts with the mercurial Hg[N(CF3)SF5I2to give a n N-methyl derivative in 33% yield. Hg[N(CFs)SFS]+ 2CH3I

-

2SFSN(CHS)CF3+ HgIz

It is a clear liquid and is identified by spectral analysis (210). The N-bromoamine, SF,N(Br)CF3, adds to the alkenes C2H4 and C,F, to give SF5N(CF3)CH2CH2Br(85%) and the mixture of isomers SF5N(CF3)CF2CF(Br)CF3 and SF,N(CF,)CF(CF,)CF,Br (53%),respectively. These formulations are supported by spectral analysis (210). Similarly, the N-chloroamine, SF,N(Cl)CF,, also reacts with C2H4 and C2F4, giving SF,N(CF,)CH,CH,Cl (88%yield; b.p., 100°C) and SF5N(CF,)CF,CF,Cl (52%).Both are colorless liquids and are characterized from their spectroscopic (IR, NMR, and mass) data (217). The preparation and characterization of CF,N(SF,)TeF, are also reported (218).

PENTAFLUOROSULFANYL COMPOUNDS

149

XV. SF,N(CI)R, (R, = C,F,, n-C,F,, n-C,F,)

The pentafluorosulfanyl N-chloroperfluoroalkylamines, SF5N(Cl)Rf (RF = C2F5, n-C3F7, n-C4F9), are prepared (217) by the reaction of chlorine monofluoride with fluorimines, SF,N=C(F)Rf. The latter are obtained by the reaction of SF5Cl with nitriles (8,217). SF,N=C(F)&

+ CIF-

SF,N(CI)CF,&

& = CzF5 (87%);C3F7(89%);n-C4F9(64%)

TheN-chloroamines, SF,N(C1)C2F5and SF5N(C1)C4Fg, react with C2H4 to give SF,N(C,F5)CH2CH2C1and SF,N(C,Fg)CH2CH,C1, respectively.

XVI. Bis(pentafluorosulfanyl)perfluoroalkylamines

The tertiary amine bis(pentafluorosulfanyl)trifluoromethylamine, (SF&NCF,, is formed in over 90% yield from the gas-phase UV photolysis of SF,N(C1)CF3 (210).

A small amount of hydrazine, SF,(CF3)NN(CF3)SF5,is also produced The tertiary amine, (SF,),NCF,, is a liquid (b.p., in the reaction (8). 72-74°C) and is characterized by IR, NMR, and mass spectroscopy (210).

The tertiary amines bis(pentafluorosulfany1)perfluoroethylamine and bis(pentafluorosulfany1)perfluoropropylamineare prepared by the following sequence of reactions (217).

The yields of (SF5l2NC2F5and (SF,),NC,F, are 58 and 8%,respectively (217).The overall yield decreases with the increased chain length of

150

VERMA, KIRCHMEIER, AND SHREEVE

the perfluoroalkyl group. This decreased yield is attributed to p-elimination of a perfluoroalkyl radical from the perfluoroalkyl chain

Analogous p-eliminations during photolysis of perfluoro-N-chloramines have been observed by Shreeve et al. (219,220). Addition of C1F to unsaturated systems such as N=S (221,222) and N=C (219, 223, 224) is known to produce highly fluorinated N-chloramines. The spectroscopic data (IR, NMR, and mass) of (SF,),NC,F, and (SF,),NC,F, are reported (217).

XVII. Tris(pentafluorosuIfanyl)arnine,(SF,),N

(SF,)3N is obtained in over 90% yield by the UV photolysis of (SF5I2NC1(217,225). These SF,-containing tertiary amines are likely to find commercial applications.

XVIII. Bis(pentafluorosuIfanyl)bis(trifluoromethyl)hydrazine, SF, (CF,)NN(CF,)SF,

The hydrazine SF,(CF,)NN(CF,)SF, is obtained in 62% yield by the reaction of AgF, with F,SN(H)CF, a t 100°C (8). BSF,N(H)CF3 + 2AgFz

-

SFS(CF3)NN(CF3)SFS+ 2AgF

+ 2HF

A similar reaction with (CF3),NH is described (226).A small amount of SF5(CF3)NN(CF3)SF5 is obtained during the photolysis of SF,N(Cl)CF, (210). A hydrolytically stable liquid, SF5(CF3)NN(CF3)SF5,boils a t 103-104°C. It is not attacked by aqueous alkali at 100°C. It reacts with chlorine to give the chloramine ClN(SF,)(CF,), which adds to perfluoropropylene in a fashion similar to that of BrN(CF,), (227).

XIX. Tetrakis(pentafIuorosuIfanyl)hydrazine, (SF,),NN(SF,),

The preparation of (SF,),NN(SF,), (a white solid; m.p., 44.5-46°C) is claimed from the UV photolysis of SF5NClzand SF,CI (228). However, it was later demonstrated that the white solid is (SF,),N (225).

PENTAFLUOROSULFANYL COMPOUNDS

151

XX. Bis(pentafluorosulfanyl)amine, (SF,),NH

When NSFBis fluorinated at low temperature with elemental fluorine, N-(pentafluorosulfany1)iminosulfur tetrafluoride is formed in -50% yield (229). 3NSF3 + 3F2

- 196°C to

SFSN=SF,

-20°C

+ SF, + N2

It is a stable, colorless liquid (b.p., 49°C; m.p., <-1Oo"C). It can be heated without decomposition in metal or Kel-F vessels. The NSF4 group in SF6NSF4functions both as a fluoride ion donor and as a fluoride ion acceptor, e.g., SF5NSF, + AsF5-

SOZCIF

F5SNSFStAsF6- 5 O b -20°C

The anion [N(SF,)2]- is a precursor to bis(pentafluorosulfany1)amines. Bis(pentafluorosulfany1)amine is prepared in quantitative yield by the addition of hydrogen fluoride to SF,NSF, (214). It is a colorless liquid that boils at 60.4"C, and it is stable in glass at room temperature. At elevated temperatures in the presence of alkali metal fluorides, it loses HF and forms SF5NSF4.It is a weak acid, which with large cations forms stable salts such as Ph4P+N(SF5)2-(229).

XXI. (SF,),NX (X = F, CI)

N-fluoro- and N-chlorobis(pentafluorosu1fanyl)aminesare obtained in 91 and 87% yields from the corresponding cesium salt (214).

The N-fluoroamine boils at 51.6"C, and C1N(SF,I2 boils at 78°C. They are both stable, colorless liquids at room temperature. The N-chloroamine reacts readily with substrates having negative halide ions (Cl-, Br-). Because the two SF, groups are magnetically nonequivalent, the 19FNMR spectra of these amines are quite complex. The structure below, consisting of a planar S2NXframework with large SNS angles

152

VERMA, KIRCHMEIER, AND SHREEVE

*

(134.8" k 1" and 138.3" 1") is proposed for (SF5)2NHand FN(SF,I2 based on electron diffraction studies (214).The SN distances increase with increasing electronegativity of the substituents (230). A

XXII. N-Pentafluorosulfanyl Haloimines, F,SN=CX,

(X = CI, F)

A. SF,N=CCl, N-Pentafluorosulfanyl chloroimine, SF5N=CC12, was first synthesized in 30%yield by Tullock et al. ( 8 ) from the photolytically induced reaction of SF,C1 with ClCN. Other derivatives are obtained by the irradiation of SF5C1and RfCN. SF5Cl+ RfCN --%SF,N=C(Cl)Rf

Rf = C1, CFB, CzFS The reaction of SF,NCO with PCl, a t 60-80°C also gives SF,N=CCl, (b.p., 86-88°C) (1471,which can be stored in glass at room temperature without decomposition and is moderately resistant to hydrolysis. Aqueous alkali degrades SF5N=CC12 quickly. Both carbon-chlorine bonds in SF,N=CCl2 are readily cleaved when this compound is reacted with excess sodium methoxide (231). SF,N=CCl,

-

+ 2NaOCHS

SF5N=C(OCH,), (88%)

When SF5N=CC12 and NaOCH, react in a 1 : 1 molar ratio, a mixture of SF,N=C(Cl)OCH, and SF6N=C(OCH3)2 is obtained. The analogous N-fluorosulfuryl derivative, FSO2N=C(C1)OCH,, is obtained by the reaction of FS02NHC(0)OCH3with PC15 (232).The reaction of SF,N= CC1, with NaOC6H5 carried out in a 1 : 2 molar ratio for 48 h r gives SF5N=C(C1)OC6H5 in 17%yield and SF5N=C(OC6H5), in 31%yield. These reactions are representative of a general method for the synthesis of imidates (233,234).

PENTAFLUOROSULFANYL COMPOUNDS

153

On treatment with two molar equivalents of diethylamine (the second molar equivalent being used as a n HC1 scavenger), SF,N=CCl, gives the monosubstituted derivative SF5N=C(Cl)N(C2H,), in 88% yield. This chloroformamidine reacts only slowly with additional diethylamine or other nucleophilic reagents (231).

-

+ 2HN(CzHs)z 11days SF5N=C(C1)N(C2H5)2+ NaOCH,

+ C6H5Li

SF5N=C"(C&)212 (12%)

3 daya

SF5N=C(OCH3)N(C2H5)2 (19%)

12 day8

SF5N=C(CBH5)N(C2H& (41%)

The compounds SF,N=C(CF3)N(C2H5)2 and SF,N=C(CH,)N(C,H,), are also synthesized by reacting excess diethylamine with the corresponding chlorimines (2311. Many N-fluorosulfonyl derivatives have been reported by Roesky and coworkers (192,235-237). The IR, NMR (lH, 19F,W ) , and mass spectral data for the pentafluorosulfanylimine derivatives are recorded (2311. The compounds SF,N=C(Cl)N(C,H,), and SF5N=C(C1)OCH3 are very stable toward hydrolysis. In contrast, the N-fluorosulfonyl derivatives are prone to hydrolysis (235). When SF,N=CCl2 is reacted with P2S5, SF,N=C=S is obtained in 70% yield (150).With excess HgF, , SF,N=CCl, forms the mercurial Hg[N(CF,)(SF,)I2 (210).With a 1: 1 molar ratio of HgF, at 150°C, SF,N=C(Cl)Rf is fluorinated to SF,N=C(F)R, (218).This mercurial is a useful precursor for the preparation of SF,-containing tertiary amines.

At room temperature SF,N=CCl, reacts with TMSNMe, to give SF5N=C(NMe2I2 in good yield (238).With Me,SiNMe, , SF,N=CCl, gives a 66% yield of SF,N=C[N(CH,),], at room temperature (239). The metathesis reactions of SF,N=C(C1)Rf with NaN, produce azido compounds that are valuable precursors to a family of asymmetric carbodiimides (2401, viz.,

-

SF5N=C(CI)Rf + NaN3

S F 5 N = C ( N 3 ) R f ASF,N=C=NRf.

Rf = CF3, CzF5, CFzNF2

154

VERMA, KIRCHMEIER, AND SHREEVE

In the presence of CsF and with CH,CN as a solvent, SF5N= C(C1)C2F5reacts with TMSC6F5and TMSCF3 a t room temperature to give SF5N=C(C6F5)C2F5and SF5N=C(CF3)C2F5, respectively (238). Reaction of SF5N=C(C1)C2F5 with CH31 in the presence of AgF and CH3CN at room temperature gives SF5N(CH3)C3F,along with SF,N= C(F)C2F5(238). No solvent or CsF is necessary for the reaction of TMSNMe, with SF5N=C(CI)C2F5 to give SF5N=C(NMe2)C2F5. B. SF5N=CF2 N-Pentafluorosulfanyl fluoroimine is obtained in 80% yield when SF5N=CC12 is reacted with NaF in tetramethylene sulfone ( 8 ) . SF5N=CC12

+ 2NaF

-

SF5N=CF2

This fluoroimine (b.p., ll-13°C) dimerizes in the presence of pyridine below room temperature to SF5N=CFN(CF3)SF5. The dimer is also formed when SF5N=CF2 is heated to 225°C in the presence of KF. In addition to the monomer and the dimer, a n isomer of SF5N=CF2, i.e., F,S=NCF3, formed on heating the amine, SF5NHCF3 (obtained in 75% yield by the addition of AHF to SF5N=CF,), and KF to 225°C. The isomer is believed to be a slightly distorted trigonal bipyramid with the NCF3 moiety a t an equatorial position ( 8 ) . On treatment with CF,OOH, SF5N=CF2 gives the peroxyamine, SF5N(H)CF200CF3, in 79% yield (b.p., 77.3’0 (241). Dehydrofluorination of this amine with NaF gives the perfluorooxaziridine,

F,= (241), in 83%yield. A minor product of the reaction, SF5N (CF3)C(0)F,is obtained in -8% yield. SFSN=CFZ

+ CFSOOH

SF5N(H)CF,00CF3 + NaF

A glass is formed by SF&%$

-

-

SFSN(H)CF200CF3 S F , N m + COF,

+ NaF.HF

(b.p., 14.9”C) a t -195°C. In the case

of C F 3 W , nucleophiles (e.g., F-) readily and exclusively attack nitrogen (242), whereas nucleophiles attack S F d K b less readily and at both the carbon and nitrogen atoms (241). With CF3CH20Li,SF5N=CF2 gives SF,N=C(OCH2CF3), in 75% yield (34).With CH31in the presence of CsF in CH,CN at room temperature, SF5N(CH3)CF3is synthesized (238).

155

PENTAFLUOROSULFANYL COMPOUNDS

XXIII. Pentafluorosulfanylirninodihalosulfanes, SF,N=SX,

(X = F, CI)

A. SF5N=SF2 Pentafluorosulfanyliminodifluorosulfane,SF5N=SF2 , was synthesized in 1965 from the fluorination of S4N4(243) and also from the reaction of SF, and NSF, using a BF, catalyst (184).However, the best yields of SF5N=SF2 are obtained from the reaction of SF5NH2with SF, in the presence of anhydrous hydrofluoric acid (183). It is a clear, colorless liquid that boils at 38°C. The chemical shifts of the fluorine resonances in the 19FNMR spectrum in various solvents and the effect of the solvent polarity on the chemical shifts are described (244). It is not easily hydrolyzed by water at room temperature; however, alkaline hydrolysis is facile (184).

Reacting elemental fluorine with SF5N=SF2 gives (SF,N=)2SF2 in 16% yield (183).No reaction occurs between AgF2 and SF5N=SF2. Photolytic fluorination of SF,N=SF, takes place to form SF,N(F)SF,N (FISF, (245). Reactions of SF,N=SF2 with sodium alkoxides and aryloxides produce both the mono- and the disubstituted derivatives, SF,N=S(F)OR and SF,N=S(OR), (246). SF5N=SFz

R

=

+ RONa

-

SF,N=S(F)OR-

RONa

SF5N=S(OR),

CH3, CH2CH=CH2, CsH5,p-CsH4NO2,p-C6H4Br,p-C6H4CN

Similarly, Abe and Shreeve (247) react C,F,N=SF, with NaOCH,, but isolate only the disubstituted derivative, C,F,N=S(OCH,), . Silylamines react readily with SF,N=SF2 to give both SF5N=S(F)NR2 and SF,N=S(NR,), (188). B. SF,N=SCl2 Pentafluorosulfanyliminodichlorosulfane, SF5N=SC12, was first synthesized in 90% yield (248)by reacting SF,NH2 with SC12.However, all subsequent attempts to repeat this reaction result in less than a 10% yield of SF,N=SC12. When C1, is present, SF5NH2reacts with SC1, at room temperature over 6 hr to give a 40% yield of SF,N=SC12 (189).This compound is extremely hygroscopic. The reaction between SF,N=SF, and PCl, also gives SF5N=SC12 in 75% yield (147, 188).

156

VERMA, KIRCHMEIER, AND SHREEVE

Chlorination of [(perfluoroalkyl)imino]difluorosulfanes with various chlorinating agents, e.g., AlC13 (2491, PCl, (2501, and SiC1, (2511, is reported in the literature. The reaction between TiC1, and SF,N=SF2 is almost instantaneous and gives better than a n 89% yield of SF5N=SC1,. A small quantity of the mixed halide derivative, SF,N=SClF, is also isolated (189).The reaction between SnC1, and SF,N=SF, is slower than the reaction with TiC1, ,but the same product is formed. The reaction of SnC1, with SF,N=SF, in the presence of trimethylsilane enhances the yield of SF,NSCl, . Antimony(V1 chloride reacts very slowly at 100°C with SF,N=SF2 to give SF,N=SCl,. A small quantity of SF,NSFCl is obtained in this reaction also (189). In the reaction of SF,N=SF, with PCl, (147,1881,AlC13(147),SiC1, (1471, SbCl,, SnCl,, or TiC1, (1891, there appears to be a correlation between the yield and the ability of the chlorinating agent to expand its coordination sphere as well as the relative Lewis acidity of the chlorinating agent. Titanium(1V) chloride has available inner d-orbitals with which it can expand its coordination sphere, as well as a relatively high Lewis acidity. In addition, the Ti-F bond strength is very high (139.7 Kcal/mol). These factors may account for the high yield of SF,N=SCl, from the reaction of SF,N=SF2 with TiC1,. The synthesis of SF,N=S(F)NSF, is reported (252)from the reaction of SF,N=SCl, with Hg(NSF,),. This pentafluorosulfanyliminosulfane derivative loses NSF, presumably through a n intramolecular fluoride ion transfer, to give SF,N=SF, . The reaction of SF5N=SC1, with AgNCO produces the diisocyanate, SF,N=S(NCO), , in 28% yield (246).This diisocyanate is also obtained from the reaction of SF,N=SCl, with KOCN in liquid SO, (246).The diisocyanate is a yellow, nonvolatile liquid that slowly deepens in color on standing at room temperature and becomes more viscous and less volatile, suggestive of polymerization. The diisocyanate, C,F,N= S(NCO),, synthesized by Abe and Shreeve (2471 is also reported to be unstable a t 25°C and is characterized only by IR and 19F NMR spectroscopy. The IR, NMR, and mass spectra of SF6N=S(NC0), are reported (246). The low-temperature reaction of SF,N=SCl, and Ag,O in nitrobenzene gives SF,N=S=O in -15% yield (246). Formation of the same product by the same reactants in MeNO, also occurs (198).However, SF,N=S=O is reported to be converted to SF,N=S=NSF, , which is also obtained by the reaction of SF5N=SC1, with SF,NH2 (147).The compound SF,N=S=O has previously been proposed as an intermediate in both the reaction of SF,N=SF, with water, leading to SF,NH,

PENTAFLUOROSULFANYL COMPOUNDS

157

and SOz (147, 183), and the reaction of SF,NH2 with SOCl,, leading to the formation of SF,N=SClz and SF,N=SClF (189).In the presence of CsF, SF,N=S=O reacts with C1, to give SF,N=S(O)ClF. With CsF or PCl,, SF,N=S(O)ClF gives SF,N=S(O)F, or SF5N=S(O)Cl2, gives respectively (198). With (MezN),S+Me,SiFz-, SF,N=S=O TAS+SF,NS(O)F- (253).The anion is identified by its "F NMR spectrum. Formation of other similar compounds (253)are represented by the general equation,

The fluorosulfuryl analog, FSO,N=S=O, (2531.

gives TAS +FSO,NS(O)F-

XXIV. Pentafluorosulfanyl-p-sultones and Sulfonic Acids

The family of fluoroalkyl sulfonic acids are some of the strongest protonic acids known. These acids or their derivatives are known to have wide chemical applications. There are several methods available for their preparation (254-258).The utility of any of these methods in producing a n SF,-containing sulfonic acid or its derivatives has not been demonstrated. Sulfur trioxide reacts with fluoroolefins, producing p-sultones (256). At least two p-sultones containing the SF, group are well characterized and will be described here, along with their utility in preparing the desired sulfonic acids or their derivatives. The reader is referred to a review by Gard (259) on p-fluorosultones.

A. F,&FCFzO$Oz The pentafluorosulfur p-sultone, 2-hydroxyl-l-(pentafluoro-h6-sulfanyl)-1,2,2-trifluoroethanesulfonic acid sultone (I), is prepared in 58% yield by reacting SF,CF=CFz with distilled SO, at 100°C (260). SF,CF=CF,

+ SO,

-

F,S-CF-CF,

I 1

0,s-0 (I)

158

VERMA, KIRCHMEIER, AND SHREEVE

It is a colorless, stable liquid that boils at 88°C. On treatment with a base, either Ca(OH), or NaOH, the corresponding calcium or sodium hydryl(pentafluoro-A6-sulfanyl)fluoromethanesulfonate is produced. I

-

+ Ca(OH),+ NaOH

Ca(SF5CFHS03)2+ CaFz + CaC03 + HzO Na(SF5CFHS03)+ NaF + Na2COi1+ H 2 0

The calcium or sodium salt with 100%H2S0, gives over a 60% yield of a SF,-containing sulfonic acid, .hydryl(pentafluorosulfur-h6-sulfanyl)trifluoromethanesulfonic acid, SF5CFHS03H.It is a colorless, stable liquid, boiling at 89-90°C. Treatment of I with a catalytic amount of triethylamine causes a quantitative rearrangement to 2-(fluorosulfonyl)-2-(pentafluoro-~~sulfanyl)-2-fluoroacetyl fluoride. A mechanism for the formation and rearrangement reaction is described (260).

On hydrolysis, I gives hydryl(pentafluor0-h6-sulfany1)fluoromethanesulfonyl fluoride. I

+ H20

-

FSSCFHSOZF + COZ + HF

Spectral data (IR, 19F NMR, and mass) for the compounds described are recorded (259). Static fluorination of F,SCFHSO,F under mild conditions and in the presence of sodium fluoride gives F,SCF,SO,F (b.p., 51 t 1°C) in 57% yield (261). Basic hydrolysis of F,SCF,SO,F gives the corresponding stable sulfonate salt, which, on distillation with concentrated sulfuric acid, gives the corresponding sulfonic acid in 39%yield (b.p., 35 ? 1°C).

F5SCF,S02F with CF,CH,OLi gives the sulfonate ester in 68% yield.

-

F5SCFzSOzF+ CH3CH20Li

F5SCF2S020CH2CF3+ LiF

159

PENTAFLUOROSULFANYL COMPOUNDS

B. F,dHCF,0b02 The fluorosultone F5&HCF,080, is prepared by the reaction of F,SCH=CF, with SO3 under autogeneous pressure at 100°C in 58% yield (262). It is a white crystalline solid [m.p., 47-48°C; b.p., 108-111°C (600 mm)l that exhibits a vapor pressure of 9 Torr at 22°C. The rearranged isomer, F,SCH(SO,F)COF, is obtained as a clear liquid in 58% yield (b.p., 115-117°C) when the fluorosultone is heated to -60°C for 4 days. F,SCH-CF,

I

0,s-0

NaF

I

60 "C, 4 d

-

0

II

F,SCHCF \

FSO,

In the presence of water, F5Sd H C F , d 0 2 undergoes rearrangement, followed by a concerted hydrolysis-decarboxylation reaction to give F5SCHZSO2Fin 51%yield (b.p., 110-111°C). The F,S-sulfonyl fluoride with aqueous sodium hydroxide gives the corresponding sodium sulfonate salt in solution, which, when treated with gaseous hydrogen chloride, produces the white solid, SF,-containing sulfonic acid, F,SCH,SO,H.H,O, in 22% yield (m.p., 97 ? 1°C) (262). Pentafluorosulfanyl esters, i.e., F,SCX(SO,F)C(O)OR [R = CH,CH, , (CF3),CH (262),(CH3),CH (263),CH,=CHCH,; X = F/H) (264),are obtained by using the corresponding sultone with the respective alcohol in the presence of sodium fluoride. F , d X C F , O ~ O z+ ROH

+ NaF

-

F5SCX(S02F)C(0)OR+ NaHF,

Fluorinated monomers and polymers containing a fluorosulfonyl group are prepared and characterized via the reaction of the p-fluorosultones F 5 m 0 , and F 5 m 0 2 with ally1 alcohol (264).A polyester containing both SF, and FSOz groups is obtained by

UV photolysis of the monomer. F5-02

+ CH2=CHCH20H

NaF

F,SCX(S02F)C(0)OCH2CH=CH2 CFCI,

F5SCX(SO2F)C(O)OCH2CH=CH2 (X = F, H)

[F5SCX(SO2F)C(O)OCH2CHln

TH2

160

VERMA, KIRCHMEIER, AND SHREEVE

The monomer is obtained in 70% yield and is a stable water-like liquid (b.p., 103 ? 1'0. The diester [F5SCF(SO2F)C(O)OCH2CF2I2CF2 is prepared [79% yield; b.p., 152°C (15 mm)l from the reaction between H0CH2(CF2),(265).Mechanistically, the CH20H and the sultone F5S =O2 sodium fluoride, the alcohol, or both serve as a catalyst for the rearrangement of the sultone (265,266).The esters formed in this fashion are stable in the presence of F- ion at ambient or higher temperatures. The spectral data (IR; 'H, "F, and I3C NMR; and mass) for these

esters and the crystal structure of the sultone F5SdHCF20b02are given (267). Heating of the ester F,SCH(S02F)C(0)OCH(CH,), with P,O,, results in the formation of a ketene (263).

The same ketene is also prepared by NaF-catalyzed rearrangement of the sultone, F,S d H C F 2 0 ~ 0 2 followed , by treatment with BF,.NEt, (2591. F-0,2

+ F,B.NEt,

-

+ F5SC(S02F)=C=0 + HNET,.BF,

F,S=dii&iO,

Several additional ,f3-fluorosultone derivatives are obtained from F,S=O2

(268).

+ CIF (CF,OCI) F4S==O2

+ HF

-

F5=O2

+ (COF2)

-

FS=O2

+ HCl

SF4CIbHCF,0~02

With CsF or CsOCF,, the stable cesium salt Cs+[SF5C(SO2F)C0F]is formed (259). The reactions of the ketene, F5SC(S02F)=C=0, with electrophilic (SO3)and nucleophilic (MF, NaN, , [(CH3)2N12C=0)reagents are also described (63).

-

FSSC(SO*F)=C=O

+ SOB-

SFsC(SO,F)=C=O

+ CSF

F,S=C(SOzF)z + CO2 CS+[FSSC(SO~F)C(O)FI

The formation of the following compounds via halogenation of the

PENTAFLUOROSULFANYL COMPOUNDS

161

cesium salt is also reported (63): F,SC(SO,F)XC(O)F (X = Br, Cl), F,SC(SO,F)ClC(O)Cl, and F5SCBr,S02F. Reaction of F,SC(SO,F)XC(0)F (X = Br, C1) with water gives F5SCHXS02F.With ethylene, F,SC(SO,F)BrC(O)Fyields F,SCH(S02F)CH2CH2Br(63). The SF,-containing ketene, FSSCH=C=O, is obtained in 70%yield from the dehydration of pentafluoro-h6-sulfanylaceticacid (269). P4OlO

FSSCHZCOOH__* FSSCH=C=O

The ketene is a colorless liquid that boils at 47°C. A molecular ion peak is observed in the mass spectrum. The IR and NMR ("F and 'H) spectra for this compound are reported (18). The ketene dimerizes on heating and isomerizes at reduced pressure in the presence of glass. FSSCH=C=O

glass

F,S=CHC(O)F

270-290°C

The F5S-containingketene undergoes typical additionlelimination reactions with HC1, HzO, EtOH, Br,, etc. (269). Treatment of SF,CXHCY,Br (X,Y = H,F) with sodium sulfite affords the salt SF,CHXCYzS03Na,which when reacted with aqueous hydrochloric acid gives the acid/acid hydrate (270). Two acid hydrates, SF,CHFCF2S03H.H20and SF,CH2CF,S03H.H,0, as well as anhydrous SF,CH,CH,S03H, are prepared by this method.

ACKNOWLEDGMENT

The support of the National Science Foundation (CHE-9003509 and INT 8821849) and the Air Force Office of Scientific Research (91-0189) is gratefully acknowledged.

REFERENCES 1. Gard, G. L., and Woolf, C. W., U.S. Patent No. 3,448,121 (1969); Gard, G. L., Bach,

2. 3. 4. 5.

J., and Woolf, C. W., British Patent No. 1,167,112 (1969);Gilbert, E. E., and Gard, G. L., U.S.Patent No. 3,475,453 (1969);Banks, R. E., and Haszeldine, R. N., British Patent No. 1,145,263 (1969); Sheppard, W. A., U.S.Patent No. 3,219,690 (1965); Michimasa, Y., Chem. Abst. 82, 1752556 (1975). Roberts, H. L., and Ray, N. H., J . Chem. SOC., 665 (1960). George, J., and Cotton, F. A., Proc. Chem. Soc., 317 (1959). Mutterties, E. L., US.Patent No. 2,937,123 (1960). Nyman, F., and Roberts, H. L., J . Chem. SOC.,3180 (1962).

162

VERMA, KIRCHMEIER, AND SHREEVE

6. Nyman, F., Roberts, H. L., and Seaton, T., Znorg. Synth. 8, 160 (1966). 7. Moldavskii, D. D., and Temchenko, V. G., Zh. Org. Khim. 6, 1339 (1970); Chem.

Abstr. 73, 65947 (1970). 8. Tullock, C. W., Coffman, D. D., and Mutterties, E. L., J . A m . Chem. SOC.86, 357 (1964). 9. Schack, C. J . , Wilson, R. D., and Warner, M. G., Chem. Commun., 1110 (1969); Inorg. Synth. 24, 8 (1986). 10. Schack, C. J., U S . Patent No. 3,649,222 (1972);Chem. Abstr. 77, 22415 (1972). 11. Cohen, B., and MacDiarmid, A. G., Znorg. Chem. 4, 1782 (1965). 12. Kovacina, T., Berry, A. D., and Fox, W. B., J . Fluorine Chem. 7,430 (1976). 13. Christe, K. O., Curtis, E. C., Schack, C. J., and Roland, A., Spectrochim. Actu A 33, 19 (1977). 14. Roberts, H. L., Quart. Revs. Chem. SOC.15, 30 (1961). 15. Merrill, C. I., U S . Department of Commerce, Office of Technical Services Publication, AD28205 (1962);Chem. Abstr. 60,6463 (1964). 16. Smardzewski, R. R., and Fox, W. B., J . Fluorine Chem. 7,453 (1976). 17. Noftle, R. E., Smardzewski, R. R., and Fox, W. B., Znorg. Chem. 16, 3380 (1977). 18. Smardzewski, R. R., Noftle, R. E., and Fox, W. B., J . Mol. Spectrosc. 62,449 (1976). 19. Christe, K. O., Curtis, E. C., and Schack, C. J., Spectrochim. Acta A 33, 69 (1977). 20. Dekock, R. L., Higginson, B. R., and Lloyd, R. D., Faraday Discuss. Chem. SOC.54, 84 (1972). 21. Marsden, C. J.,and Bartell, L. S.,Znorg. Chem. 15,3004 (1976);J . Mol. Struc. 75, 271 (1981);Jurek, R., Poinsot, A., and Goulet, P., J . Phys. 47, 645 (1986);Chem. Abstr. 104, 1958305 (1986); Verry, C., Jurek, R., Goulet, P., and Lamalle, B., J . Chem. Phys. 42, 223 (1981);Chem. Abstr. 94, 1118265 (1981). 22. Christen, D., Mack, H. G., and Oberhammer, H., J . Chem. Phys. 87, 2001 (1987). 23. Baumgaertel, H., Jochims, H. W., Ruehl, E., Losking, O., and Willner, H., Z . Naturforsch. 44, 21 (1989). 24. Fraser, G. J . W., McCarty, M. E., Sharp, D. W. A., and Winfield, J. M., J . Inorg. Nucl. Chem. 34, 1455 (1972). 25. Cohen, B., and MacDiarmid, A. G., J . Chem. SOC.A, 1780 (1966). 26. Kemmitt, R. D. W., Peacock, R. D., and Stocks, J . , Chem. Commun., 554 (1969). 27. Roberts, H. L., J . Chem. SOC.,3183 (1963). 28. Roberts, H. L., J . Chem. SOC.,2774 (1960). 29. Logothetis, A., L., Sausen, G. N., and Shozda, M., Znorg. Chem. 2, 173 (1963). 30. Emelbus, H. J., and Packer, K. J., J . Chem. SOC.,771 (1962). 31. Massey, A. G., and Packer, K. J.,J . Chem. Soc., 5554 (1966). 32. Kumar, R. C., and Shreeve, J. M., Z . Naturforsch. B 36, 1407 (1981). 33. Kirchmeier, R. L., Patel, N., and Shreeve, J. M., private communication. 34. Patel, N., Kirchmeier, R. L., and Shreeve, J. M., J . Fluorine Chem. 48,395 (1990). 35. Kitazume, T., and Shreeve, J. M., J . A m . Chem. SOC.99, 3690 (1977); J . Chem. Soc., Chem. Comm., 982 (1976). 36. von Halasz, S. P., and Glemser, O., Chem. Ber. 103, 594 (1970). 37. Swindell, R. F., and Shreeve, J . M., J . A m . Chem. SOC.94,5713 (1972). 38. Swindell, R. F., Babb, D. P., Ouellette, T. J., and Shreeve, J . M., Znorg. Chem. 11, 242 (1972). 39. Gibson, J. A., and Schmutzler, R., Z . Anorg. Allg. Chem. 416, 222 (1975). 40. Sidebottom, H. W., Tedder, J . M., and Walton, J. C., Trans. Faraduy SOC.65, 2103 (1969). 41. Case, J. R., Ray, N. H., and Roberts, H. L., J . Chem. SOC.,2066 (1966).

PENTAFLUOROSULFANYL COMPOUNDS

163

42. Case, J. R., Ray, N. H., and Roberts, H. L., J . Chem. SOC.,2070 (1966). 43. Banks, R. E., Haszeldine, R. N., and Morton, W. D., J . Chem. SOC.C, 1947 (1969). 44. Coffman, D. D., and Tullock, C. W., US.Patent No. 3,102,903(1963);Chem. Abstr.

60,1599 (1964). 45. Studnev, Yu N., Stolyarov, V. P., Prigorelov, G. A., and Fokin, A. V., Izv. Akad.

Nauk SSSR, Ser. Khim. 1, 197 (1989);Chem. Abstr. 111, 38856g (1989). 46. Shreeve, J . M., in “Sulfur in Organic and Inorganic Chemistry” (A. Senning, Ed.),

Vol. 4,pp. 131-191. Dekker, New York, 1982,and references therein. 47. Steward, J . , Kegley, L., White, H. F., and Gard, G. L., J . Org. Chem. 34,760(1969).

48. Sidebottom, H. W., Tedder, J. M., and Walton, J. C., J . Chem. SOC.D,253 (1970). 49. Sidebottom, H. W., Tedder, J. M., and Walton, J. C., Trans. Faraday SOC.66, 2038 (1970). 50. Bekker, R. A., Dayatkin, B. L., and Knunyants, I. L., Izv. Akad. Nauk, SSSR, Ser. Khim., 2738 (1970);Chem. Abstr. 74,99366 (1971). 51. Berry, A. D., and Fox, W. B., J . Org. Chem. 43, 365 (1978). 52. Berry, A. D., and Fox, W. B., J . Fluorine Chem. 6,175 (1975). 53. Terjeson, R. J., and Gard, G. L., J . Fluorine Chem. 35, 653 (1987). 54. Mir, Q. C., DeBuhr, R., Huang, C., White, H. F., and Gard, G. L., J.Fluorine Chem. 16,373 (1980). 55. Roberts, H. L., U.S. Patent Nos. 3,063,972and 3,063,922(1962);Canadian Patent No. 630,704(1961);Dale, J. W., US. Patent No. 3,126,366(1964);Sherratt, S., British Patent No. 929,990(1963). 56. Banks, R. E., Barlow, M. G., Haszeldine, R. N., and Morton, W. D., J . Chem. SOC. Perkin I , 1266 (1974). 57. Hoover, F. W., and Coffman, D. D., J . Org. Chem. 29,3567 (1964). 58. Canich, J . M., Ludvig, M. M., Paudler, W. W., Gard, G. L., and Shreeve, J . M., Inorg. Chem. 24, 3668 (1985). 59. Wang, Q. C., White, H. F., and Gard, G. L., J . Fluorine Chem. 13,455 (1979). 60. Potter, B., and Seppelt, K., Znorg. Chem. 21,3147 (1982). 61. Cohen, B., and MacDiarmid, A. G., Angew. Chem. 75,207 (1963). 62. Dobbie, R. C., J . Chem. SOC.A, 1555 (1968). 63. Winter, R., and Gard, G. L., J . Fluorine Chem. 52, 73 (1991). 64. Kitazume, T., and Shreeve, J . M., J . A m . Chem. SOC.100,492 (1978). 65. Dudley, F.B., Cady, G. H., and Eggers, D. F., Jr., J . A m . Chem. SOC.78,1553 (1956). 66. Ruff, J. K., and Lustig, M., Znorg. Chem. 3, 1422 (1964). 67. Ruff, J . K.,Inorg. Synth. 11, 131 (1968). 68. Kennedy, R. C., and Cady, G. H., J . Fluorine Chem. 3,41 (1973/1974). 69. Dudley, F. B., J . Am. Chem. SOC.85, 3407 (1963). 70. Merrill, C. I., and Cady, G. H., J . A m . Chem. SOC.83, 298 (1961). 71. Hale, W. H., Jr., and Williamson, S. M., Znorg. Chem. 4,1314 (1965). 72. Shreeve, J. M., Duncan, L. C., and Cady, G. H., Znorg. Chem. 4,1516 (1965). 73. Ruff, J. K., Znorg. Chem. 4, 1788 (1965). 74. Pass, G., and Roberts, H. L., Inorg. Chem. 2, 1016 (1963). 75. Williamson, S.M., and Cady, G. H., Inorg. Chem. 1, 673 (1962). 76. Czerepinski, R., and Cady, G. H., J . A m . Chem. SOC.90,3954 (1968). 77. Lustig, M., and Shreeve, J. M., Adv. Fluorine Chem. 7 , 175 (1973). 78. Place, R. D., and Williamson, S. M., J . Am. Chem. SOC.90, 2550 (1968). 79. Shreeve, J. M., Adv. Znorg. Chem. Radiochem. 26, 119 (1983). 80. Prezel, E. J . , Christen, D., Oberhammer, H., Mallela, S. P., and Shreeve, J. M., J . Mol. Struct. 248,415 (1991).

164

VERMA, KIRCHMEIER, AND SHREEVE

81. Gould, D. E., Anderson, L. R., Young, D. E., and Fox, W. B., J . A m . Chem. Soc.

91, 1310 (1969). 82. Schack, C. J., Wilson, R. D., Muirhead, J . S., and Cohz, S. N., J . A m . Chem. SOC.

91, 2907 (1969). Schack, C. J., and Maya, W., J . A m . Chem. SOC.91,2902 (1969). Schack, C. J., Wilson, R. D., and Cohz, S. N., J . A m . Chem. SOC.91, 2907 (1969). Smardzewski, R. R., and Fox, W. B., J . Phys. 79, 219 (1975). Young, D. E., Anderson, L. R., Gould, D. E., and Fox, W. B., Tetrahedron Lett., 723 (1969). 87. Seppelt, K., Angew. Chem. Znt. Ed. 16, 44 (1976); 2.Anorg. Allgem. Chem. 428, 35 (1977). 88. Kloeter, G., and Seppelt, K., J . A m . Chem. SOC.101, 367 (1979). 89. Seppelt, K., Angew. Chem. Int. Ed. 16, 322 (1977). 90. Walker, N., and DesMarteau, D. D., J . A m . Chem. SOC.79, 13 (1975). 91. Hopkins, M. J., Walker, N., and DesMarteau, D. D., J . Org. Chem. 41, 1467 (1976). 92. Anderson, L. R., Young, D. E., Gould, D. E., Jurik-Hogan, R., Nuechterlein, D., and Fox, W. B., J . Org. Chem. 35,3730 (1970). 93. Schack, C. J., Wilson, R. D., and Criste, K. O., J . Fluorine Chem. 45, 283 (1989). 94. Durrell, W. S., Stump, E. C., Jr., Westmoreland, G., and Padgett, C. D., J . Polymer Sci. A 3, 4065 (1965). 95. Krumm, B., Kirchmeier, R. L., and Shreeve, J . M., private communication. 96. Gombler, W., J . Fluorine Chem. 9, 233 (1977). 97. Tyczkowski, E. A., Chern. Abstr. 82, 124632 (1975). 98. Abe, T., and Shreeve, J. M., J . Fluorine Chem. 3, 187 (197311974). 99. Barlow, M. G., Dean, R. R., and Lee, J., Trans. Furuduy Soc. 65, 321 (1969). 100. Marsden, C. L., Christen, D., and Oberhammer, H., J . Mol. Struct. 131,299 (1965). 101. Baba, H., Kodaira, K., Nagase, S., and Abe, T., Bull. Chern. SOC.(Jpn.)50, 2809 (1977). 102. Baba, H., Kodaira, K., Nagase, S., and Abe, T., Bull. Chern. SOC.(Jpn.)51, 1891 (1978). 103. Abe, T., Nagase, S., and Baba, H., Bull. Chem. SOC.(Jpn.)46, 3845 (1973). 104. Abe, T., Nagase, S., and Baba, H., Bull. Chem. SOC.(Jpn.)43, 1812 (1970). 105. Huang, H., Roesky, H. W., and Lagow, R. J., Znorg. Chem. 30, 789 (1991). 106. Shimp, L. A., and Lagow, R. J., Inorg. Chem. 16,2974 (1977). 107. Waterfeld, A., and Mews, R., J . Fluorine Chem. 23,325 (1983). 108. Clifford, A. F., El-Shamy, H. K., Emelkus, H. J., and Haszeldine, R. N., J . Chem. SOC.,2372 (1953). 109. Gupta, K. D., Mews, R., Waterfeld, A., Shreeve, J. M., and Oberhammer, H., Znorg. Chem. 25,275 (1986). 110. Kleemann, G., and Seppelt, K., Angew. Chem. 90,547 (1978). 111. Kleeman, G., and Seppelt, K., Chem. Ber. 112, 1140 (1979). 112. Witucki, E. F., Energy Res. Abstr. 3(31471) (1978);Chem. Abstr. 89,199878 (1978). 113. Emeleus, H. J., and Tattershall, B. W., J . Znorg. Nucl. Chem. 28, 1823 (1966). 114. Silvey, G. A., and Cady, G . H., J . A m . Chem. SOC.72, 3624 (1950). 115. Damerius, R., Leopold, D., Schulze, W., and Seppelt, K., 2.Anorg. Allg. Chem. 578, 110 (1989). 116. Seppelt, K., in “Synthetic Fluorine Chemistry” (G. A. Olah, R. D. Chambers, and G. K. Surya Prakash, Eds.). Wiley, New York, 1992. 117. Berry, A. D., and Fox, W. B., J . Fluorine Chem. 7,449 (1976). 118. De Marco, R. A., and Fox, W. B., J . Fluorine Chem. 12, 137 (1978). 83. 84. 85. 86.

PENTAFLUOROSULFANYL COMPOUNDS

165

119. Gard, G. L., and Woolf, C., J . Fluorine Chem. 1, 487 (1971/1972). 120. Vorob’ev, M. D., Filatov, A . S., and Englin, M. A., Zh. Org. Khim. 10, 407 (1974);

Chem. Abstr. 80, 120152 (1974). 121. Kirchmeier, R. L., and Shreeve, J. M., Inorg. Chem. 12, 2886 (1973). 122. Woolf, C., and Gard, G. L., U.S. Patent No. 3,448,121 (1969); Chem. Abstr. 72,

12547 (1970);Gard, G. L., Woolf, C., and Shaw, R. M., U S . Patent No. 3,519,725 (1970);Chem. Abstr. 73, 11443 (1970). 123. Marcellis, A. W., and Eibeck, R. E., J . Fluorine Chem. 5, 71 (1975). 124. Witucki, E. F., J . Fluorine Chem. 20, 803 (1982). 125. Noftle, R. E., and Fox, W. B., J . Fluorine Chem. 9, 219 (1977). 126. Hutchinson, J., J . Fluorine Chern. 3,429 (1973/1974). 127. Oates, G., and Winfield, J. M., J . Fluorine Chem. 4, 235 (1974). 128. Winter, R., and Gard, G. L., J . Fluorine Chem. 50, 141 (1990). 129. Debuhr, R., Howbert, J., Canich, J. M., White, H. F., and Gard, G. L., J . Fluorine Chem. 20,515 (1982). 130. Kadyrov, A. A., and Rokhlin, E. M., Usp. Khim. 57, 1488 (1988). 131. Wozniak, L., and Chojnowski, J., Tetrahedron 45, 2465 (1989). 132. Wessolowski, H., Roschenthaler, G.-V., Winter, R., and Gard, G. L., Z . Naturforsch. B 46, 126 (1991). 133. Wessolowski, H., Roschenthaler, G.-V., Winter, R., and Gard, G. L., 2.Naturforsch. B 46, 123 (1991). 134. Hoover, F. W., and Coffmann, D. D., J . Org. Chem. 29,3567 (1964). 135. Canich, J., Ludvig, M. M., Paudler, W. W., Gard, G. L., and Shreeve, J. M., Inorg. Chem. 24, 3668 (1985);Terjeson, R. J., Canich, J., and Gard, G. L., Inorg. Synth., 27, 329 (1990). 136. Csaszar, A. G., Hedberg, K., Terjeson, R. J., and Gard, G. L., Znorg. Chem. 26, 955 (1987). 137. Kovacina, T. A., De Marco, R. A., and Snow, A. W., J.Fluorine Chem. 21,261 (1982). 138. Berry, A. D., De Marco, R. A., and Fox, W. B., J . Am. Chem. Soc. 101,737 (1979). 139. Wang, Q . C., White, H. F., and Gard, G. L., J . Fluorine Chem. 13, 455 (1979). 140. Kovacina, T. A., De Marco, R. A., and Snow, A. W., U.S.Patent No. 4,535,011 (1985); Chem. Abstr. 104,90558f (1985). 141. Allied Chemical Corp., British Patent No. 1,167,228 (1969); Chem. Abstr. 72, 16735 (1970). 142. Banks, R. E., and Haszeldine, R. N., British Patent No. 1,145,263 (1969); Chem. Abstr. 70, 97807 (1969). 143. Yonekura, M., Jpn. Kokai 74, 123 (1974); 986 (1974); Chem. Abstr. 82, 175255 (1975). 144. Gilbert, E. E., and Gard, G. L., U S . Patent No. 3,475,453 (1969);Chem. Abstr. 72, 20877 (1970). 145. Allied Chemical Corp., British Patent No. 1,167,122 (1969); Chem. Abstr. 72, 21366 (1970). 146. Duncan, L. C., Rhyne, T. C., Clifford, A. F., Shaddix, R. E., and Thompson, J. W., J . Inorg. Nucl. Chem. (Suppl.), 33 (1976). 147. Clifford, A. F., and Shanzer, A., J . Fluorine Chem. 7 , 6 5 (1976). 148. Haas, A., Chem. Ber. 98, 111 (1965). 149. Thrasher, J. S., Howell, J. L., and Clifford, A. F., Inorg. Chem. 21, 1616 (1982). 150. Thrasher, J. S., Howell, J. L., and Clifford, A. F., Chern. Ber. 117, 1707 (1984). 151. Thrasher, J. S., Clark, M., Nielsen, J. B., Alvarado, C., and Anderson, M. T., Inorg. Synth. 29, 38 (1992).

166

VERMA. KIRCHMEIER, AND SHREEVE

152. Kuhle, E., Anders, B., Klauke, E., Tarnow, H., and Gumach, G., Angew. Chem. 81, 18 (1969);Angew. Chem. Int. Ed. Engl. 8 , 2 0 (1969). 153. Krumm, B., Kirchmeier, R. L., and Shreeve, J . M., private communication. 254. Oberhammer, H., and Seppelt, K.,J . Mol. Struct. 101, 325 (1983). 155. Thrasher, J . S., and Clifford, A. F., J . Fluorine Chem. 23, 593 (1983). 156. Roesky, H. W., Angew. Chem. Int. Ed. Engl. 8, 136 (1969). 157. Bott, S. G., Clark, M., Thrasher, J. S., and Atwood, J. L., J . Crystallogr. Spectrosc. Res. 17, 187 (1987). 258. Sitzmann, M. F., and Gilligan, W. H., US. Patent No. 4,831,186 (1989);Chem. Ahstr. 111, 60565h (19891. 259. Sitzmann, M. E., J . Fluorine Chem. 57, 195 (1991).

160. Sitzmann, M. E., Gilligan, W. H., Ornelass, D. N., and Thrasher, J . S., J . Energ. Muter. 8, 352 (1990); Chem. Abstr. 114, 210037r (1990). 161. Appel, R., and Montenarh, M., Chem. Ber. 110, 2368 (1977). 262. Roesky, H. W., and Sidiropoulos, G., Chem. Ber. 110, 3730 (1977). 163. Patai, S., “The Chemistry ofcyanates and Their Thio Derivatives,”Part 112.Wiley, New York, 1977. 164. King, C., J . Org. Chem. 25,352 (1960). 165. Roesky, H. W., and Hoff, A., Chem. Ber. 101, 162 (1968). 266. Klauss, K., Friedrich, H.-J., and Jensen, H., Liebigs. Ann. Chem., 561 (1974). 167. Biener, H., Liehigs. Ann. Chem. 686, 102 (1965). 2676. Clark, M., Kellen-Yuen, C. J., Robinson, K. D., Zhang, H., Yang, Y . , Midappat. K. V . , Fuller, J. W., Atwood, J . L., and Thrasher, J. S., Eur. J . Solid State Inorg. Chem. 29, 809 (1992). 168. Clifford, A. F., and Howell, J., J . Fluorine Chem. 10, 431 (1977). 169. Thrasher, J. S., Howell, J., Clark, M., and Clifford, A. F., J . A m . Chem. SOC.108, 3526 (1986). 170. Schmuck, A., and Seppelt, K., Angew. Chem. Znt. Ed. Engl. 26, 134 (1987). 272. Hagemann, H., Art, D., and Ugi, J., Angew. Chem. Int. Ed. Engl. 8, 606 (1969). 172. Seppelt, K., and Oberhammer, H., Znorg. Chem. 24, 1227 (1985). 173. Zylka, P., Mack, H.-G., Schmuck, A., Seppelt, K., and Oberhammer, H., Inorg. Chem. 30, 59 (1991). 174. Attawya, J. A., Groth, R. H., and Bigelow, L. A,, J . A m . Chern. SOC.81,3599 (1959). 175. Smith, W. C., Tullock, C. W., Smith, R. D., and Engelhardt, V. A,, J . A m . Chem. SOC.82, 551 (1960). 176. Griffiths, J. E., and Sturman, D. F., Spectrochim. Actu A 25, 1355 (1969). 177. Vorob’ev, M. D., Filatov, A. S., andEnglin, M. A.,Zh. Obshch. Khim. 43,1653 (1973). 178. Losking, O., and Willner, H., Angew. Chem. Znt. Ed. Engl. 28, 1255 (1989). 179. Jacobs, J., McGrady, G. S., Willner, Christen, D., Oberhammer, H., and Zylka, P., J . Mol. Struct. 245, 275 (1991). 180. Thrasher, J. S., and Mudappat, K. V.,Angew.Chem. Znt. Ed. Engl. 28,1256 (1989). 181. Lentz, D., J . Fluorine Chem. 24,523 (1984). 182. Clifford, A. F., and Duncan, L. C . , Znorg. Chem. 5, 692 (1966). 183. Clifford, A. F., and Zeilenga, G. R., Inorg. Chem. 8, 1789 (1969). 184. Clifford, A. F., and Thompson, J. W., Inorg. Chem. 5, 1424 (1966). 185. Hofer, R., and Glemser, O., 2. Naturforsch. B 30, 458 (1975). 286. Waterfeld, A., and Mews, R., unpublished results. 187. Clifford, A. F., and Goel, G., Inorg. Chern. 8, 2004 (1969). 188. Hofer, R., and Glemser, O., Z. Anorg. Allg. Chem. 416, 263 (1975).

PENTAFLUOROSULFANYL COMPOUNDS

167

189. Thrasher, J . S., Hosmane, N. S., Maurer, D. E., and Clifford, A. F., Inorg. Chem.

21,2506 (1982). 190. Thrasher, J. S., Howell, J., and Clifford, A. F., J . Fluorine Chem. 24,431 (1984). 191. Patai, S. (Ed.), “The Chemistry of Carbon-Nitrogen Double Bonds.” Interscience,

New York, 1970. Roesky, H. W., and Giere, H. H., Chem. Ber. 102,3707(1969). Roesky, H. W., Angew. Chem. Int. Ed. Engl. 8,136 (1969). Clifford, A. F., and Zeilenga, G. R., Inorg. Chem. 8, 979 (1969). Waterfeld, A., and Mews, R., Angew. Chem. 93, 1075 (1981);Angew. Chem. Znt. Ed. Engl. 20, 1017 (1981). 196. Waterfeld, A., and Mews, R., J. Chem. Soc., Chem. Comm., 839 (1982). 197. Thrasher, J. S., and Seppelt, K., 2.Anorg. Allg. Chem. 507, 7 (1983). 198. Kricke, P., Stahl, I., Mews, R., and Glemser, O., Chem. Ber. 114,3467 (1981). 199. Kloter, G., Lentz, W., Seppelt, K., and Sunderrneyer, W., Angew. Chem. 89, 754 (1977); Angew. Chem. Int. Ed. Engl. 16,707 (1977). 200. Klotter, G.,and Seppelt, K., J. Amer. Chem. SOC.101,347 (1979). 201. Leidinger, W.,and Sundermeyer, W., Chem. Ber. 115, 2992 (1982). 202. Stump, E. C., Jr., Padgett, C. D., and Brey, W. S., J.Fluorine Chem. 2,648 (1963). 203. Cady, G.H., Eggers, D. F., Jr., and Tittle, B., Proc. Chem. SOC.,65 (1963). 204. Haase, J., Oberhammer, H., Zeil, W., Glemser, O., and Mews, R., 2.Naturforsch. A 26, 1333 (1971). 205. Lustig, M., and Ruff, J . K., Inorg. Chem. 4, 1444 (1965). 206. Roesky, H. W., Angew. Chem. 80,626 (1968). 207. Roberts, H.L., Quart. Reus. 15,55 (1961). 208. Dobbie, R. C., J . Chem. SOC.A , 1555 (1966). 209. Moldavskii, D.D., and Temchenko, V. G., Zh. Org. Khim. 6, 185 (1970). 210. Thrasher, J. S., Nielsen, J . , Bott, S. G., McClure, D. J., Morris, S. A., and Atwood, J . L., Inorg. Chem. 27,570 (1988). 211. Sekiya, A,, and DesMarteau, D. D., J. Fluorine Chem. 17,463 (1981). 212. Waterfeld, A., and Mews, R., Chem. Ber. 24, 1674 (1985). 213. Singh, S.,and DesMarteau, D. D., Inorg. Chem. 25,4596 (1986). 214. Waterfeld, A., Oberhammer, H., and Mews, R., Angew. Chem. Int. Ed. Engl. 21, 355 (1982);Angew. Chem. (Suppl.), 834 (1982). 215. Foropoulos, J., Jr., and DesMarteau, D. D., Inorg. Chern. 23, 3720 (1984). 216. Ang, H.G., and Syn, Y.C., Adu. Inorg. Chem. Radiochem. 16,l (1974). 21 7. Nielsen, J . B., and Thrasher, J. S., J. Fluorine Chem. 48,407 (1990). 218. Thrasher, J . S., and Seppelt, K., Inorg. Chem. 24,4171 (1985). 219. Peterman, K. E., and Shreeve, J. M., Inorg. Chem. 14, 1223 (1975). 220. Sarwar, G., Kirchrneier, R. L., and Shreeve, J. M., Inorg. Chem. 28,2187 (1989). 221. De Marco, R. A., and Shreeve, J. M., J.Fluorine Chem. 1,269 (1971/1972). 222. De Marco, R. A., and Shreeve, J. M., J. Chem. SOC., Chem. Commun., 788 (1971). 223. Moldavskii, D. D., and Temchenko, V. J.,Zh. Obshch. Khim. 39, 1393 (1969). 224. Swindell, R. F., Zaborowski, L. M., and Shreeve, J. M.,Znorg. Chem. 10,1635 (1971). 225. Choudhry, M. R., Harrell, J . W., Jr., Nielsen, J. B., and Thrasher, J. S., J. Chem. Phys. 89, 5353 (1988). 226. Young, J . A., Durrell, W. S., and Dresdner, R. D., J.Am. Chem. SOC.82,4553(1960). 227. Young, J. A., Tsoukalas, S. N., and Dresdner, R. D., J.Am. Chem. SOC.80,3604 (1959). 228. Thrasher, J . S., and Neilsen, J., J . Am. Chem. SOC.108, 1108 (1986). 192. 193. 194. 195.

168

VERMA, KIRCHMEIER, A N D SHREEVE

229. Waterfeld, A,, and Mews, R., Angew. Chem. Znt. Ed. Engl. 21, 354 (1982). 230. Haase, J., Oberhammer, H., Zeil, W., and Glemser, O., 2. Nuturforsch. A 26, 1333 (1971). 231. Thrasher, J. S., and Clifford, A. F., J . Fluorine Chem. 19,411 (1982). 232. Roesky, H. W . , Angew. Chem., 81 (1969) 119; Angew. Chem. Znt. Ed. Engl. 8, 136 (1969). 233. Bonnet, R., in “The Chemistry of Carbon-Nitrogen Double Bond” ( S . Patai, Ed.), Chap. 13, pp. 598-662. Interscience Publishers, New York, 1970. 234. Kiihle, E., Anders, B., Klauke, E., Tarnow, H., and Zurnach, G., Angew. Chem. 81, 18 (1969);Angew. Chem. Znt. Ed. Engl. 8, 20 (1969). 235. Roesky, H. W., and Tutkunkardes, S., 2. Anorg. Allg. Chem. 378, 177 (1970). 236. Roesky, H. W., Giere, H. H., and Babb, D. P., Znorg. Chem. 9, 1076 (1970). 237. Roesky, H. W., and Giere, H. H., 2. Anorg. Allg. Chem. 378, 177 (1970). 238. Patel, N. R., Kirchrneier, R. L., and Shreeve, J. M., Znorg. Chem. 32, 4802 (1993). 239. Zheng, Y. Y., Patel, N. R., Kirchmeier, R. L., and Shreeve, J . M., Znorg. Chem. 31, 488 (1992). 240. John, E. O., Mack, H. G., Oberhammer, H., Kirchmier, R. L., and Shreeve, J. M., Znorg. Chem. 32, 287 (1993). 241. Sekiya, A., and DesMarteau, D. D., Znorg. Chem. 19, 1330 (1980). 242. Falardeau, E. R., and DesMarteau, D. D., J . A m . Chem. SOC.98, 3529 (1976). 243. Cohen, B., Hopper, T. R., and Peacock, R. D., J . Chem. Soc., Chem. Commun., 32 (1965). 244. Thrasher, J. S., Howell, J. L., Maurer, D. E., and Clifford, A. F., J . Fluorine Chem. 24, 205 (1984). 245. Waterfeld, A., Stahl, I., and Mews, R., unpublished results; Ref. 64 in Glemser, O., and Mews. R., Angew. Chem. Znt. Ed. Engl. 19, 883 (1980). 246. Thrasher, J. S., Innaccone, G . A., Hosmane, N. S., Maurer, D. E., and Clifford, A. F., J . Fluorine Chem. 18, 537 (1981). 247 Abe, T., and Shreeve, J. M., Inorg. Chem. 19, 3063 (1980). 248. Clifford, A. F., and Goel, R. G., Znorg. Chem. 8, 2004 (1969). 249. Lustig, M., Znorg. Chem. 5, 1317 (1966). 250. Roesky, H. W., and Mews, R., Angew. Chem. Znt. Ed. Engl. 7, 217 (1968). 251. Glemser, O., Halasz, S., and Biermann, M., Znorg. Nucl. Chem. Lett. 4,591 (1968). 252. Stahl, I., Mews, R., and Glernser, O., Chern. Ber. 113, 2430 (1980). 253. Heilrnann, W., and Mews, R., J . Fluorine Chem. 52, 377 (1991). 254. Haszeldine, R. N., and Kid, J. M., J . Chem. Soc., 4228 (1954). 255. Gramstad, T., and Haszeldine, R. N., J . Chem. SOC.,2640 (1957). 256. Temple, S., J . Org. Chem. 33, 344 (1968). 257. England, D. C . , Dietrich, M. A,, and Lindsey, R. V., J. A m . Chem. SOC.82, 6181 (1960). 258. Howells, R. D., and McCown, J. D., Chem. Reus. 77, 69 (1977). 259. Mohtasharn, J., and Gard, G. L., Coord. Chem. Revs. 112, 47 (1992). 260. Canich, J. M., Ludvig, M. L., Gard, G. L., and Shreeve, J . M., Znorg. Chem. 23, 4403 (1984); J. Mohtasharn, R. J. Terjeson, and G. L. Gard, Znorg. Synth. 29, 33 (1992). 261. Gard, G. L., Waterfeld, A., Mews, R., Mohtasharn, J.,and Winter, R., Znorg. Chem. 29, 4588 (1990). 262. Terjeson, R. J., Mohtasham, J., and Gard, G. L., Znorg. Chem. 27, 2916 (1988). 263. Winter, R., and Gard, G . L., Znorg. Chem. 27, 4329 (1988).

PENTAFLUOROSULFANYL COMPOUNDS

169

264. Terjeson, R. J., Mohtasham, J., Sheets, R. M., and Gard, G. L., J . Fluorine Chem. 38,3 (1988). 265. Khalilolahi, J., Mohtasham, J., Lerchen, M. E., Sheets, R. M., and Gard, G. L., Inorg. Chem. 26,2307 (1987). 266. Huang, T. J., Dong, Z. X., and Shreeve, J. M., Inorg. Chem. 26, 2304 (1987). 267. Pressprich, M. R., Willet, R. D., Terjeson, R. J., Winter, R., and Gard, G. L., Inorg. Chem. 29,3058 (1990). 268. Winter, R., Peyton, D. H., and Gard, G. L., Inorg. Chem. 28,3766 (1989). 269. Krugerke, T.,Buschmann, J., Kleemann, G., Luger, P., and Seppelt, K., Angew. Chem. Int. Ed. Engl. 26, 799 (1987). 270. Willenbring, R. H., Mohtasham, J., Winter, R., and Gard, G. L., Can. J . Chem. 67, 2037 (1989).

This Page Intentionally Left Blank

ADVANCES IN INORGANIC CHEMISTRY, VOL.

41

THE HUNTING OF THE GALLIUM HYDRIDES ANTHONY J. DOWNS and COLIN R. PULHAMt *inorganic Chemistry Laboratory, University of Oxford, South Parks Road, Oxford, OX1 3QR, United Kingdom; and tDepartment of Chemistry, University of Edinburgh, West Mains Road, Edinburgh, EH9 3JJ, United Kingdom

They sought it with thimbles, they sought it with care; They pursued it with forks and hope; They threatened its life with a railway-share; They charmed it with smiles and soap. Lewis Caroll, “The Hunting of the Snark”

I. Introduction 11. History and Chemical Background A. The Search for Free Gallane B. Complexes of Gallane 111. Conduct of the Hunt: Practical Methods of Attack

IV.

V.

VI.

VII.

A. Handling B. Physical Methods of Detection and Analysis Toward Gallane: Preparations for the Hunt A. Dimethylgallium tetrahydroborate B. Hydridogallium bis(tetrahydrob0rate) C. Dimethylgallane D. Monochlorogallane Gallane at Last! A. Preliminaries B. Synthesis and Characterization C. Chemical Properties D. Summary Hydrogen-Rich Derivatives of Gallane A. Introduction B. Gallaborane, GaBH, C. 2-Galla-ut-uchno-tetraborane(l0), 2-GaB3HIo Hydrides of the Other Group 13 Metals: Preliminaries and Prospects A. Introduction B. Aluminum Hydrides C. Indium and Thallium Hydrides References

171 Copyright 0 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.

172

DOWNS AND PULHAM

I. Introduction

Boron dwarfs the other members of Group 13 in the richness of its hydride chemistry (1-9). At a recent count (9) about 100 binary borane species are now known, typically as discrete molecules remarkable for their stoicheiometries and structures, which have done much to shape our understanding of chemical bonding at large. By contrast, there is known only one binary hydride of aluminium stable under normal conditions, and that is a polymeric solid, the a-form being isostructural with AlF, and featuring six-coordinate aluminium atoms (10-13 1. There is no doubt that the vapors of a Group 13 metal, M, react with hydrogen, under appropriate conditions of excitation, to give transient hydride molecules like MH [M = A1 (141,Ga (14-16), In (14,15, 171, or T1 (14,1511 or MH3 (M = Al, Ga, or In) (18).The MH molecules have appreciable bond dissociation energies, ranging from 294 for M = A1 to 190 kJ mol-’ for M = T1, but the heats of atomization of the elements M and H2are such that the monohydrides are highly unstable with respect to the elements at normal temperatures. The transients GaH and GaH, apart, the uncoordinated binary hydride of gallium, [GaH,],, , has been for many years something like the fabulous Snark of Lewis Carroll’s fertile imagination (191, with its existence befogged by both claimants and counterclaimants. Given the history of the compound, it is not surprising that one authoritative view delivered in 1988 (20) was that “[GaH,], is questionable.” In this account we shall be concerned with establishing the true credentials of gallane and its derivatives. After a brief outline of earlier searches for gallane, including the characterization of coordinatively saturated gallanes like GaH,- and Me,N-GaH, , we shall describe the sort of procedures that it has been necessary to devise in order to hunt not only gallane itself, but also compounds like gallaborane, [GaBH,], , and the tetraborane(l0) derivative, 2-GaB3Hlo,which are little or no less reactive and only slightly more robust with regard to thermal decomposition. Of the instruments used to pursue the Snark, only care and hope have played a major part in our pursuit of the gallium hydrides; more prosaically we dwell on the physical methods that have enabled us to know the “warranted genuine” compounds. How the base-free gallium hydrides have ultimately been tracked down is then revealed, with due note of the pivotal role of monochlorogallane, [H2GaC1],, itself the recent fruit of pure serendipity. Such physical properties of the compounds as have been ascertained so far are recorded. Although these are still early days, our account will aim also

GALLIUM HYDRIDES

173

to give some idea of the potential for new chemistry that the compounds hold. In this prospect alone there is much to relish. Whether there is any moral to the tale is for the reader to judge. Only two decades ago the hydrides of the heavier Group 13 metals might have appeared as no more than scientific curiosities, although the preparation of ether-free aluminium hydride was classified under US. Air Force contracts in the early energy-hungry 1960s. Such a perception has changed out of all recognition with the technological rise of 111-V (13-15) compounds (10) and the need to form thin films of these or of the metals, e.g., for the growth of GaAlAs and the “metallization” of semiconductor devices. Chemical vapor deposition (CVD) implicating organometallic precursors in reactions such as

is of primary importance in this connection, although the success of such methods demands close control of the conditions and, even then, cannot always be guaranteed t o deliver deposits with the desired purity or physical properties. Volatility and thermal instability, allied to the relative cleanness of the decomposition reactions, confer on hydrides significant advantages over organometallic sources of the Group 13 metals, notably in the formation of carbon-free deposits. For example, trimethylamine alane, Me3N.A1H3,has been found to offer considerable attractions as a precursor to pure aluminium films (21-26).Compounds as volatile and thermally frail as the elusive gallane, [GaH,],, are therefore of more than passing interest, not only as a means of vapor transport of the metal at low temperatures, but also as a source of solid metal or metal-bearing films. In addition, compounds like gallane may be either active intermediates or parents to such intermediates in the thermolysis reactions associated with useful CVD processes and our current knowledge of which owes more to speculation than to wellsubstantiated fact.

11. History and Chemical Background

A. THE SEARCH FOR FREEGALLANE Earlier searches starting with elemental gallium having been to no avail, the first substantive claim to the synthesis of gallium hydride

174

DOWNS AND PULHAM

was staked out just over 50 years ago. It depended on more than one route, viz. (27-30)

- l,iGaH4, E t 2 0

Me3Ga + H2

electric

discharge

1

- [GaH,],, n

GaCI3 1

-2 [Me2GaH12

EtJN

Et3N.GaMe3+ Ga,H,.

(3)

With the benefit of hindsight, it seems odd that the use of ether as a reaction medium should yield not a complex like Et20.GaH3 but the base-free hydride [GaH,], . That a deficiency of triethylamine should favor ligand redistribution in dimethylgallane with the formation of free gallane and the complex Et,N.GaMe, is perhaps less surprising if we suppose that the propensity to associate makes gallane a weaker Lewis acid than Me,Ga. Altogether more surprising were the reports that free gallane is stable at temperatures up to 130°C. Doubts about the authenticity of some of these claims were raised about 20 years later by the failure of attempts to reproduce the experiments (31,321.On the other hand, analytical and spectroscopic evidence was presented (32,33)to support the low-temperature displacement reaction

-

Me3N,GaH3(c)+ BF3(g)

258K

1 - [GaH31,,(1)+ Me3N.BF3(c), n

(4)

which for more than two decades was widely accepted as a route to uncoordinated gallane (8,321. The compound was described as a viscous oil unstable at room temperature. The vapor exhibited infrared absorptions centered near 1980 and 700 cm-', attributable t o Ga-H, stretching and bending modes, respectively (H, , terminal hydrogen atom). The physical properties clearly pointed to an oligomeric or polymeric product and so it was curious that the infrared spectrum betrayed no sign of features attributable to motions of Ga-Hb-Ga units (Hb, bridging hydrogen atom). Yet, on the evidence of mass-spectrometric and matrix-isolation studies (34, 351, the vapor species derived from the interaction of Me,N.GaH, with BF, include B,H, and BH,F3-n species in proportions implying that halide-hydride exchange, and not displacement, is the predominant reaction pathway. A product answering to the description of the material alleged to be uncoordinated gallane has been prepared, but the signs are that it is a mixture of fluorogallanes of the type [H,GaF3-,], ,probably incorporating coordinated Me,N ( x = 1 or 2 and n > 2) (35-37).

175

GALLIUM HYDRIDES

All these vicissitudes must, we feel, call in question the preparation of anything approaching pure gallane in the period before 1989. Monosubstituted derivatives of gallane, [H,GaX],, have been no less elusive. The displacement reaction Me,N.GaH,Cl(c)

-

-1 [H2GaCll,(1) + Me3N.BF3(c)

+ BF,(g)

(5)

was reported to afford a thermally unstable oil, believed to be monochlorogallane, [H,GaClI, (38).Little headway was made, however, with the characterization of the product, and, in the light of the doubts cast on the analogous preparation of gallane itself and of more recent studies (to be described) (36, 37, 39, 401, this claim too must be treated with caution. By contrast, the same period had witnessed the convincing authentication of several disubstituted derivatives of gallane. These included compounds of the type [HGaX212,where X = C1 or Br, prepared by metathesis (41, 42): 2MeaSiH + GazXs

253 K

[HGaXz12+ 2Me3SiX.

(6)

The chloride forms white crystals that melt at 29°C with decomposition and, when heated to 150”C, decompose quantitatively into gallium “dichloride,” Ga[GaCl,], and hydrogen. Several diorganogallanes were also described. Among these were [Et,GaHI, (43-45) and [Bu’,GaHI, (46)prepared, for example, by reactions such as (7) and (8):

-

Et3Ga + NaGaH, 1 - [Bu’,GaCl],

2

+ LiH

70°C

EtpO. 40-45°C

1 [Et2GaHln+ NaGaH3Et n

1 - [Bu’,GaHI,

(7)

+ LICI.

Only recently, however, with the isolation and more detailed interrogation of the compounds [Me,GaH], (47) and HGa(BH4)2(48-50), has it been possible to discover with any sureness the properties-including the structures-of such compounds. Insofar as these properties have a bearing on the hunting of other gallium hydrides, it will be better that we elaborate on them in due course (in Section IV).

B. COMPLEXES OF GALLANE The confusion besetting the status of gallane itself has certainly not extended to coordinated derivatives, usually of the type LaGaH, ,where

t

/ I

-

c

0

c

GALLIUM HYDRIDES

177

L is a basic species like Me3N or H- (12, 32). Prepared by methods along the lines indicated in Scheme 1, many of these have been known for some years, in several cases as compounds long-lived a t ambient temperatures. Numerous neutral 1: 1 complexes have been prepared with stabilities varying in the sequences Me2NH > Me3N > C5H5N> Et3N > PhNMe, s=Ph3N; Me,N = Me3P > Me2PH; and R3N > RzO or R,S. More or less tetrahedral coordination of the gallium center is the norm, as illustrated in Fig. 1 for the gaseous molecules Me3N.GaH3 (511 and [Me2NGaH212(52),whose structures have been determined by electron diffraction. Measuring about 150 pm, the Ga-H, bonds are somewhat shorter here than in the monohydride GaH [re= 166.21 pm ClS)], but comparable with Ge'"-H and As"'-H bonds. The Ga-H, stretching modes of L.GaH3 moieties have wave numbers that vary with the nature of L (and with the medium) in the range 1720-1880 cm-' (32,5345). Partial replacement of the hydrogen by more electronegative substituents like chlorine shifts v(Ga-H,) to higher energy, as exemplified by the mean wave numbers (in cm-') in the following series: Me3N.GaH3, 1836; Me3N.GaH,Cl, 1902; and Me,N-GaHCl,, 1959. Ill.

Conduct of the Hunt: Practical Methods of Attack

A. HANDLING The gallium hydrides known prior to 1989 are, without exception, unusually susceptible to attack by air or moisture. None of them is thermally robust and some evidently decompose a t subambient temperatures. Of necessity, therefore, the compounds are nearly always handled in uacuo. However, because they tend to dissolve in, or react with, vacuum greases, it is usually a minimum requirement that they be handled in apparatus incorporating greaseless valves and unions with vacuum seals made in a material like Teflon. Yet even these measures are inadequate sometimes to guard against decomposition or reaction with impurities adsorbed on the surfaces of the apparatus. To succeed with the synthesis and manipulation of such hydrides has demanded scrupulous attention to practical techniques, including the development of special procedures (37,56).There follows a brief outline indicating some of the practical and strategic considerations. 1. Vapor Transfer and Sampling In the pursuit of gallane and related hypersensitive compounds, we have found that operations must be carried out at pressures mm

a

b

as deduced from the electron-diffraction patterns FIG.1. Molecular structures of the gallane derivatives (a)Me3N.GaH, and (b) [Me2NGaH2I2 of the vapors (reproduced with permission from Refs. 51 and 52).

GALLIUM HYDRIDES

179

Hg in all -glass apparatus individually constructed and incorporating appropriately sited break-seals and constrictions (37, 56). Figure 2a illustrates a typical assembly. Such apparatus possesses two crucial advantages. In the first place, cooling of all parts can be effected by blowing a stream of cold gas over the exposed surfaces. Second, the apparatus can be rigorously preconditioned by heating under continuous pumping to eliminate volatile impurities adsorbed on the inner walls. Distillation trains are kept short, each trap normally being equipped (a)with constrictions to permit isolation of a sample by sealing under vacuum and (b) with a break-seal permitting access to the contents to be regained under equally rigorous conditions. Even so, the properties of some gallium hydrides impose severe limitations on what

i1 Break-seal ampoule containing gallane

flCold nitrogen gas

containing

FIG.2. Pyrex glass apparatus used (a) for the synthesis and sampling of a base-free gallane and (b) for the admission of the gallane vapor to the chamber of the electrondiffraction apparatus. In (a) A is a sample of [H,GaCI],; B,, B,, and B3 are greaseless valves; C is freshly prepared LiGaH,, LiBH4,or [Bu",NI[B,H,]; D,,Dz,and D3are U-tube traps for fractionation of the volatile components of the reaction mixture; and E is an NMR tube (reproduced with permission from Ref. 56; copyright 1991, American Chemical Society).

180

DOWNS AND PULHAM

can be realistically achieved by way of physical and chemical characterization. Thus, the proclivity to decompose means that vapor pressures must be kept low and this tends to thwart a variety of potentially informative studies involving, for example, the use of Raman or NMR spectroscopy to investigate the vapor. The problems are less acute with more sensitive techniques, like modern FTIR spectroscopy offering sampling times in the order of 1-2 min, and the lifetime of the vapor sample can be extended by going to lower pressures with a compensating increase in path length (as in a multiple-reflection cell) (37, 56). There is also scope, as yet unexploited, for introducing the molecules into a supersonic jet and investigating their microwave, rovibrational spectra, or both, taking advantage of the enhanced signal strength and spectroscopic resolution made possible by the internal cooling induced under these conditions (57-59). Without rigorous sampling techniques, however, there is the risk of confusion caused by reactions of the hydride with adventitious impurities; such is the case, for example, with conventional mass spectrometric analysis. Accordingly, for electron-diffraction measurements it has been necessary to construct a special allglass inlet assembly (37, 56). As illustrated in Fig. 2b, this provides for the direct injection of the gallium hydride vapor from a storage ampoule into the chamber of the diffraction apparatus via a glass channel that can be suitably passivated and then cooled to whatever temperature may be needed to forestall thermal decomposition. But problems, like sorrows, “come not single spies.” The strongly reducing vapor of the gallium hydride is liable to react with the emulsion of the photographic plates used to record the electron-diffraction patterns; the resulting fogging effects can be alleviated, but not eliminated, by various measures (37, 56).

2. Trapping Experiments Such is the thermal instability of the gallium hydrides that meaningful studies of the condensed phases are largely confined to samples at low temperatures. In these circumstances, the tracking and identification of the compounds have called for trapping. This may be achieved physically by quenching the vapor on a cold surface either alone or with an excess of a suitable diluent [as in matrix isolation (SO)];questions of identity and likely structure are then addressed by reference to the infrared or Raman spectrum of the deposit, the attribution of the features being checked by examining the effects of isotopic enrichment of the sample. Alternatively the experimenter may resort to chemical trapping of the hydride by treating it with a compound likely to undergo a fast and quantitative reaction, yielding a known product. Trime-

GALLIUM HYDRIDES

181

thylamine is such a compound and the discovery that the addition reactions (9)and (10) take place cleanly at 178 K has provided irresistible proof of the identities of the relevant gallanes: 1

2 [H,GaC112 + mNMe3 1

- IGaH,],

+ mNMe,

178 K

178 K

H2ClGa(NMe3), (36,39,40)

H3Ga(NMe3), (37,56,61).

(9) (10)

3. Chemical Analysis

To determine the composition of a compound that is more or less short-lived at normal temperatures, it is necessary to devise a method of chemical analysis that can be carried out on site, that is, with the aid of a glass vacuum line. The prime desideratum is to find a suitable reaction of the compound that can be engineered to proceed quantitatively in an evacuated, preconditioned glass ampoule to deliver stable products that are amenable to direct assay, for example, by weighing, tensimetric measurements, or elemental analysis. A good illustration is provided by hydridogallium bidtetrahydroborate) (see Section IV.A), which decomposes at or just above room temperature in accordance with Eq. (11)(48-50):

In this case a Toepler pump can be used to remove the volatile products derived from a known mass of the gallane, and an efficient trap cooled to 77 K to separate the condensable B,H, from the noncondensable H2; the two fractions are then assayed tensimetrically. The mass balance is completed simply by weighing the residue of elemental gallium. Similar measures can be adopted to determine the stoicheiometry of a reaction engaging a gallane with another compound, e.g., NH, ,NMe, , or HC1.

B. PHYSICAL METHODSOF DETECTIONAND ANALYSIS It will be evident from the preceding account that not all physical methods lend themselves equally well to the detection and specification of the gallium hydrides. There are, for example, obvious difficulties in trying to grow single crystals of a low-melting, thermally unstable compound, and, although the case of hypofluorous acid (62)shows what can be achieved ultimately, X-ray methods do not obviously commend

182

DOWNS AND PULHAM

themselves at the first sighting of the molecular quarry. In any case, there is ample circumstantial evidence that aggregation is a primary motif in gallium hydride chemistry and that the form of the compound may well vary from one phase to another. Three methods have in the event formed the spearhead of the present hunt, namely, (a) vibrational spectroscopy; (b) NMR spectroscopy; and (c) electron diffraction of the vapor. Some of the more robust molecules, like HGa(BH,), (48-50) and H2GaB3H,(63), can be identified by their mass spectra, and UV photoelectron spectroscopy has also been turned to account in studies of gaseous tetrahydroborate molecules like ANBH,), (64 and HGa(BH,), (501, notably for the light it sheds on the mode of coordination of the BH, ligand. Compared with the three primary methods, however, such sources of supplementary information are barely in the running. 1. Vibrational Spectroscopy

Identification of a gallium hydride sample has been accomplished in the first place more often than not by reference to its infrared spectrum, supported sometimes by its Raman spectrum. To test and elaborate upon the inferences drawn from such a spectrum, we can appeal to various particulars; most informative of these are (i) signs of rotational fine structure associated with the infrared absorptions of the vapor, (ii) the response to isotopic changes, and (iii) analogies with the vibrational properties of related, authenticated molecules. As with the characterization of metal carbonyls, it is the stretching vibrations that by their energies and intensities offer the most telling commentary on the molecular identity. Where v(C-0) modes differentiate terminal from bridging carbonyl functions, so v(M-H) modes differentiate terminal from bridging M-H functions in hydride derivatives of a Group 13 metal, M. Studies carried out in Oxford supplement earlier measurements in establishing, for example, that the stretching modes of terminal Ga-H bonds give rise to strong infrared absorptions in the range 1720-2050 cm-' (Refs. 32,37, 38, 41, 42,48-50, 54-56, 63, and 65, for example). This contrasts with bridging Ga-H-Ga units for which v(Ga-H) occurs with variable intensity in the range 900-1720 cm-' (37,47,56,66,67). A similar distinction can be drawn between terminal Al-H and bridging A1-H-A1 moieties (12, 47, 68, 69), just as the number, energies, and intensity patterns of v(B-H) modes are familiar signatures for the BH, ligand in different types of coordination (70, 71 ). For an M-H-M bridge there are two stretching vibrations depicted

183

GALLIUM HYDRIDES

4 Ga. . .

. ..Ga

Ga . .

.....Ga

(lb)

(la)

schematically in l a and lb, which may be described as antisymmetric bas) and symmetric (us), respectively. The wave numbers of these two vibrations are linked via the simple relationship (12) to the M-H-M bond angle 8: uas/us= tan(8/2).

(12)

The relation is subject to certain approximations, namely (a) that the mass of M is effectively infinite compared with that of hydrogen; (b) that the angle-bending force constant, k,, is much smaller than the bond-stretching force constant, k,, and that neither u, nor u,, experiences significant mixing with any other motion of the molecule at large; and (c) that the stretch-stretch interaction constant, k,, is negligible (72). As 8 approaches go", so, according to Eq. (12), the ratio u,,/u, approaches unity; i.e., the separation between u,, and u, tends to zero, and, as 0 approaches 180",the ratio and frequency separation increase. There is a concomitant change in the relative intensities of the two infrared absorptions corresponding to these modes, such that the ratio Ia,/I, runs from near unity to infinity as 0 ranges from 90 to 180". Such considerations have been invaluable in the first sighting of hydrogenbridged molecules like dimethylgallane, [MezGaH], (471, and gallane itself, [GaH,], (56). At the same time, a relatively simple molecule like Ga2H6(37,561 or GaBH6(37,651 may be expected to betray its presence in the gas phase by infrared absorptions modulated by rotational structure. Scrutiny of this structure, preferably at high resolution, opens up the prospect of determining the symmetry properties of the associated vibrational transitions and of securing at least a rough estimate of one or more of the rotational constants. Thus, GazH6with a diborane-like structure (2) featuring two massive gallium atoms on a common symmetry axis H

184

DOWNS AND PULHAM

has a momenta1 ellipsoid approximating to that of a highly prolate symmetric top with ZA G ZB = I,. Such a molecule may be expected to show infrared absorptions of two kinds according to whether the dipole change associated with the vibrational transition is parallel or perpendicular t o the Ga-Ga axis. A parallel-type band should be dominated by more or less structured P- and R-branches flanking a single, very narrow Q-branch, whereas a perpendicular-type band should be dominated not by the P- and R-branches, which form an unresolved or only partially resolved background, but by a regular series of Q-subbranches with a spacing approximating to 2(A’ (where the rotational constants A ’ and B ’ refer to the u = 1 vibrational state and B ’ is the average of the two constants B’ and (2’). Under modest resolution a parallel-type feature is expected to show a P-R branch separation, AuPR,to which we may appeal, through another approximate relation, viz. Eq. (131, for an estimate of the mean rotational constant B:

a’)

Because the magnitude ofB is governed mainly by the Ga...Ga distance, it is thus possible to gauge this distance to a useful approximation. A similar approach was used to arrive at the first estimate of the Xe-F bond length in the linear molecule XeFz (73).How the results of such studies in practice point irresistibly to the fugitive species GazH6 is more aptly treated later (Section V.B.2).

2 . NMR Spectroscopy The normal edge that NMR measurements enjoy in so many matters of detection and characterization has been somewhat blunted where gallium hydrides are involved. The studies carried out to date have been concerned exclusively with hydrocarbon solutions of the compounds, sometimes at room temperature, more often at low temperatures (down to 190 K). One immediate problem is whether the compound has the same form under these conditions as in the vapor at low pressure or in the solid phase. The signs are that aggregation becomes a serious issue in solution where a given gallium hydride often exists in more than one oligomeric form. Further problems stem from the quadrupolar character of the two naturally occurring gallium isotopes, 69Ga (60.108%, Z = I) and 71Ga(39.892%, Z = $1 ( l o ) ,which causes marked broadening of the ‘H resonances originating in hydrogen atoms bound t o gallium. The ‘H signals do sharpen when the solution is cooled, but

GALLIUM HYDRIDES

185

the spectra hold relatively little structural information, beyond the finding that terminal hydrogen atoms have chemical shifts in the range 6 4.0-5.5, whereas bridging hydrogen atoms resonate at lower frequency (6 1-3.1) (32, 36, 37). The 71Ganucleus (which is superior in its receptivity and width factor to the more abundant 69Ga)has featured in numerous NMR experiments (74,751, but the line widths associated with gallium atoms in less than symmetrical environments make it relatively unattractive as a marker for a gallium hydride. Rather more instructive are the NMR spectra of mixed gallium boron hydrides like GaBH, and GaB,H,, , with which “B measurements help to build up a fuller picture. Yet further complications are in store. First, there is the susceptibility of any hydrogen-bridged framework to undergo facile exchange. Fluxionality is a familiar phenomenon in metal tetrahydroborates, most investigations implying magnetic equivalence of the four protons of each BH4 ligand irrespective of its coordination mode and of the temperature (71,76).Compounds containing a metal coordinated to an octahydrotriborate group, B,H,, show a wider repertoire of properties, ranging from the essentially rigid behavior of Mn(B,H&CO), to the rapid exchange of boron and hydrogen atoms in Be(B,H,)(C,H,) and M+[B,H,l- salts (63, 77-78). A rigid system involving bidentate ligation of the B,H, group, with the adoption of a B4Hl,-like structure, gives a distinctive “B NMR spectrum featuring two resonances with intensities in the ratio 2 : 1 and that exhibit characteristic chemical shifts. On the other hand, coupling patterns frequently belie the simplicity of their appearance, being invariably subject to second-order effects. Another potential source of confusion comes from relaxation phemomena. This problem is exemplified by the effects of cooling a metal tetrahydroborate sample, resulting typically in the collapse of the multiplet structures of the ‘H and “B resonances. Such a change reflects the decrease in the spin-lattice relaxation time T,brought about by the increased viscosity of the sample at low temperatures; the accelerated interconversion between nuclear spin states can lead ultimately to decoupling of the ‘H and “B nuclei, a phenomenon referred to as “thermal” or “correlation-time” decoupling (711. 3. Electron Diffraction Electron diffraction needs no introduction as a relatively simple and direct tool for determining the interatomic distances within a molecule in the vapor state, in which it is free from the constraints and perturbations of the condensed phases (79,801. In an ideal situation, the method is capable of defining quite precisely the positions of hydrogen atoms

186

DOWNS AND PULHAM

because the electron scattering, unlike X-ray scattering, originates from close to the nuclei and the intensity is determined by the product of scattering factors for the components of a given atom pair. There is scope to determine the structures of quite complicated molecules, always provided that the inherent shortcomings of the method are recognized. One problem arises from the low resolution of electron-diffraction measurements. The experimental results, after various preliminary manipulations, can be presented as a plot of scattering intensity against scattering angle, which contains one-dimensional information only. Fourier transformation of the complex wave form gives, as in Fig. 3, g radial distribution curve that contains one peak for each distance in the molecule. If two or more of these distances are similar, the corresponding peaks will overlap and it may be impossible to resolve them. In these circumstances the geometric and vibrational parameters used to specify the molecular model tend to act in concert or to be subject to correlation, and, unless one or more of the parameters can be constrained on the basis of additional, independent information, there must be a degree of uncertainty about the structure analysis. The problem is clearly exacerbated for a relatively weakly scattering atom pair (like Ga-HI, which has a separation more or less coincident with that of a more strongly scattering pair (like Ga-N), to the detriment of the accuracy with which the separation of the first pair can be determined. Sometimes, too, the scattering pattern may fail to distinguish clearly between different model structures for a particular compound. This has proved to be the case, for example, with gaseous Be(BH,),, the nature of which has still to be resolved (79, 80). The fortunes of Be(BH4I2illustrate inter alza the potential mischief that impurities may cause, particularly if the impurity molecules carry interatomic distances similar t o those in the sample molecule. Understandably, therefore, it is imperative that pure samples be used and that steps be taken to prevent decomposition or contamination of the sample en route to the diffraction chamber. Another problem arises from the fact that each distance determined by electron diffraction is averaged over the vibrations of the molecule. The effect is most acute for a molecule undergoing one or more vibrations of large amplitude, as with the bending vibration of linear C1-Hg-C1 or the out-of-plane puckering vibration of the central fourmembered ring of Cl,Ga(p-Cl),GaCl, ; certain nonbonded distances then appear on average to be appreciably shorter than expected. Unless a correction is made for the so-called “shrinkage”effect (ideally through calculations requiring a detailed knowledge of the vibrational proper-

187

GALLIUM HYDRIDES

I , Scattering

0 .

I

'

I

40

3;io

intensity

slnm-1

Fourier transformation

Ga-H, C-N

I

--. L

L

a

Radial

l o Ga. . H ,

c

distribution curve

r Ipm

N FIG.3. The molecularscattering intensity pattern for the Me3N.GaH3molecule related by Fourier transformationto the corresponding radial distribution curve. The difference between the observed and calculated radial distribution curves is also shown (adapted from Ref. 51 ).

188

DOWNS AND PULHAM

ties of the molecule), the analysis may give one to believe that a linear molecule is bent or that a planar one is nonplanar. Shrinkage need not be a problem with a molecule whose vibrational properties are well characterized or when there is independent evidence from some other source affording a realistic estimate of the magnitude of the effect. Otherwise, though, there is no escaping the uncertainty implicit in the vibrational-averaging of interatomic distances. IV. Toward Gallane: Preparations for the Hunt

A. DIMETHYLGALLIUM TETRAHYDROBORATE One of the first compounds containing a Ga-H bond to receive detailed attention was the mixed gallium boron hydride Me,GaBH,. First reported in 1943 as the product of the interaction of Me3Ga with B2H6 (811, the compound is more satisfactorily prepared by the reaction of Me,GaCl with freshly recrystallized lithium tetrahydroborate in the absence of a solvent at 258 K (see Scheme 2)(82). The preferred method is significant in pointing the way to what has generally proved to be the most productive and dependable strategy for the synthesis of hydrides of the heavier Group 13 elements. Avoidance of a solvent is dictated by the need to minimize the risks of contamination of the free hydride. Nevertheless, a relatively efficient reaction can be engineered, with yields depending chiefly on the precise state and purity of the solid reagents. A similar approach has been adopted previously for making volatile tetrahydroborates of other metals, e.g., AUBH,), (83) and Zr(BH,), (84 1. Dimethylgallium tetrahydroborate melts at ca. 274 K and has a vapor pressure of 13-14 mm Hg at 273 K. It decomposes at ambient temperatures; at a pressure of 10 mm Hg the vapor has a halflife in the order of 3 hr, decomposing to give elemental gallium and hydrogen together with a mixture of methylated boranes. The vapor is composed of the diborane-like molecules Me2Ga(pH),BH2,in which fourfold coordination of the gallium is completed by bidentate ligation of the BH4 group (82,851. The resemblance to diborane is underlined by the response to nitrogen bases. Thus, ammonia induces heterolytic cleavage of the Me,Ga(p-H),BH2 skeleton with the formation of the salt-like product [Me2Ga(NH,),1+BH4-. By contrast, trimethylamine forms a molecular adduct Me,N.GaMe,(BH,) at low temperatures and this appears to undergo homolytic cleavage at temperatures above 228 K to give Me,N-BH, with gallium metal and dihydrogen, as well as traces of an unstable intermediate containing Ga-H bonds.

189

GALLIUM HYDRIDES Ga + H2

+ methylboranes

Me2GaBH4

Me2GaCl + +2H6

+ H2

Me3N.!3H3

+ Ga + H2

SCHEME 2. Formation and some reactions of dimethylgallium tetrahydroborate, Me2GaBH4(82).

B. HYDRIDOGALLIUM ~~~(TETRAHYDROBORATE) The reaction with Me2GaC1suggests that lithium tetrahydroborate should also undergo metathesis with gallium(II1) chloride, as it does with aluminium(II1) chloride (83),to give the tris(tetrahydrob0rate) derivative of the Group 13 metal. In fact, it was found in 1976 that the powdered solids react at ca. 228 K to afford not Ga(BH& but the hydridogallium compound HGa(BH,), (48);the same product is formed when dichlorogallane, [HGaCl2I2(41,42 1, is substituted for the trichloride (Scheme 3). Hydridogallium bis(tetrahydrob0rate) melts at about

190

DOWNS AND PULHAM Me3N-BH3 + Me3N.GaH3 Ga + B2H6 + -$I2 3

Me2GaBH4 + MeHGaBH4

Excess Me3N

+a+[GaC14]-

+ B2H6 + $H2

t

293 K

[HGa(NH3) 3 ] 2+[BH43-2

SCHEME 3. Preparation and some reactions of hydridogallium bis(tetrahydroborate1, iIGa(BH& (48-50).

200 K t o give a colorless, relatively mobile liquid that has a vapor pressure of ca. 10 mm Hg at 228 K. It decomposes to gallium metal, dihydrogen, and B,H6 at room temperature [see Eq. (ll)];at a pressure of ca. 10 mm Hg the vapor has a half-life on the order of 10 min. The compound is a rare example of an M3H9species involving one or more Group 13 elements, M, and for which theoretical studies predict a cyclic structure (3) with the three M atoms linked in pairs through single hydrogen bridges (86).On the evidence of the density, mass and vibrational spectra, and electron-diffraction pattern, however, the vapor consists of molecules in the form not of 3 but of HGa[(p-H),BH212,with a single terminal Ga-H bond and two bidentate BH, groups making

GALLIUM HYDRIDES

191

up a five-coordinated gallium center (see Fig. 4 and Table I) (48-50). On the other hand, the physical properties of the compound imply that the monomer is subject to loose aggregation in the condensed phases; in particular, the ‘H and IIB NMR spectra of C6D5CD, solutions at temperatures in the range 190-270 K suggest that the monomer and an oligomer [HGa(BH,),l, (where n = 2 in all probability) coexist in equilibrium. Noteworthy chemical features revealed in Scheme 3 include the behaviors of bases as varied as CO, NMe,, and NH,, which bring about homolytic or heterolytic fission of the Ga(p-H),BH, fragments. The reaction with CO provides a link with the thermally less

FIG.4. Structureof the gaseous molecule HGa(BH& as determinedby electron diffraction. H atoms are numbered 1, 2, 2’, 3, 3’, 4, 4’, 5, and 5’ (48-50).

192

DOWNS AND PULHAM

stable gallaborane, H2GaBH4, whose detailed characterization postdates that of HGa(BH,), by some 14 years (see Section V1.A). Of Ga(BH4)3,however, there is still no vestige, even a t low temperatures. These findings appear to emphasize the comparative reluctance of gallium, compared with aluminium, to rise to coordination numbers greater than four (10).

C. DIMETHYLGALLANE After earlier sightings of what was probably the impure compound, dimethylgallane, [Me2GaH], , was prepared relatively efficiently in 1984-1986 by way of the exchange reaction between trimethylgallium and freshly prepared lithium or sodium tetrahydrogallate (Scheme 4) (47). Careful fractionation in U ~ C U Ogives a viscous liquid (m.p., ca. 273 K; v.p. at 293 K, ca. 1mm Hg) that decomposes in a matter of days at room temperature. The compound resembles physically the product reported some 40 years previously (27-30) as having been isolated from the discharge reaction between Me,Ga and H2 [Eq. (311.It resembles too the corresponding aluminium compound, [Me2A1Hl,, which exists in the vapor and solution phases as a mixture of oligomers with n = 2 and, probably, 4 (35, 69, 87-89). On the evidence of its mass and infrared spectra and electron-diffraction pattern (471, the vapor of dimethylgallane at low pressures and near-ambient temperatures consists predominantly of the dimeric molecule (4) (see Fig. 3 and Table I). This represented, therefore, the first gallium hydride to be shown positively to contain a Ga(p-H),Ga bridging unit in a structure closely resembling that of the corresponding alane, [Me2A1H], (87-89). At 261.0 pm, the Ga-SGa separation is comparable with the shorter Gaa-Ga distances (247-307 pm) displayed by elemental gallium in its different guises (101, and the Ga-H bond, measuring 170.8 pm, is notable for being about 20 pm longer than the terminal Ga-H bonds of molecules like Me,N.GaH, (51)and HGa(BH,), (49, 50). The infrared spectrum

(4)

of the vapor includes, in addition to the bands associated with internal motions of the Me,Ga groups, two conspicuous features at 1290 and 1185 cm-’, which shift to 971 and 893 cm-l, respectively, when the hydrogen bound to gallium is replaced by deuterium. These must repre-

GALLIUM HYDRIDES

193

2

Me 2GaBH4

SCHEME4. Preparation and some reactions of dimethylgallane, [Me2GaH], (reproduced with permission from Ref. 47).

sent to a fair approximation the antisymmetric (la;v,,, 1290/971cm-'1 and symmetric (lb; v,, 1185/893 cm-') stretching vibrations of the central Ga(p-H/D),Ga moiety of 4. The Ga-H-Ga bond angle, 8, is estimated on the basis of Eq.(12) to be ca. 95"; the electron-diffraction measurements imply a value of 99.6" 2 1.4". Condensation of the dimethylgallane at low temperatures gives a solid with a significantly different pattern of infrared absorptions attributable to v,, and v,; vas now gives rise to an intense, broad band centered near 1700/1250 cm-' and v, to a much weaker feature near 965/850 cm-'. The obvious inference is that the degree of aggregation has increased on condensation with the formation, possibly, of a trimer akin to 3 or, more likely, the

194

DOWNS AND PULHAM

(5)

tetramer (5) akin to [Me2A1H14(89).In either case the Ga-H-Ga bridge angle has opened out to at least 120".As we shall see shortly, dimethylgallane foreshadows in its structural and spectroscopic properties the most distinctive features of gallane itself, [GaH,], ,

D. MONOCHLOROGALLANE The next step involves not so much a logical progression as a stroke of luck. We have noted already that dichlorogallane, [HGaCl,], , is produced by metathesis between gallium(II1) chloride and trimethylsilane, in accordance with Eq. (6) (41,421. In an experiment designed to prepare dichlorogallane by this route, an excess of trimethylsilane was inadvertently taken. Exchange between the neat reagents took place, as expected, at 250-258 K, but fractional condensation of the volatile products in uucuo gave not a white solid, which dichlorogallane is known to be (41,421, but a viscous, colorless liquid freezing a t ca.183 K and with a vapor pressure at ambient temperatures on the order of 0.5 mm Hg (36, 39, 40). The compound was found to decompose over a period of days at room temperature with the quantitative formation of equimolar proportions of H, and an involatile white solid having the composition GaC1. Hence exchange with an excess of the silane has given rise to monochlorogallane, which, on the evidence of its infrared and Raman spectra, exists as the dimer [H2GaC11,in both the condensed and vapor phases: Ga,CI, + 4Me3SiH

-

[H,GaCI],

IA

+ 4Me3SiC1.

2Hz + 2GaCI

(14)

The electron-diffraction pattern of the vapor can be satisfactorily interpreted, like the vibrational properties, on the basis of a molecule H,Ga(p-Cl),GaH, with D,, symmetry, as illustrated in Fig. 3. Here

GALLIUM HYDRIDES

195

the Ga-H distance is back to the 150-pm mark that typifies terminal Ga-H bonds and the central Ga(p-Cl),Ga unit is comparable in its dimensions to those in [Me,GaC112 (47) and Ga2C16(90). Quantum mechanical methods confirm that chlorine-bridging takes precedence over hydrogen-bridging and predict dimensions and vibrational properties in keeping with the experimental findings (91, 92). More recently the monomeric molecules HGaC1, (93) and H2GaC1 (94)have both been prepared by quite a different approach. This involves the photolytically induced addition of HC1 or H2 to molecular GaCl isolated in a n argon matrix: GaCl + HX

-

HXGaCl (X = C1 or H).

(15)

Identification rests firmly on the infrared spectra of the natural and deuteriated molecules, the results being borne out by ub initio calculations. Whether this method would lend itself to the synthesis of monochlorogallane on the larger scale remains to be seen. With the unearthing of monochlorogallane good fortune has delivered into the stalkers’ hands the most important agent yet in the pursuit of hydrogen-rich gallanes. Scheme 5 illustrates some of its more important reactions (36,37,39,40).Its decomposition is intriguing in offering a low-temperature route to what appears to be metastable gallium(1) chloride (10).Independent studies have shown that GaCl, maintained in a mixed toluene/diethyl ether solvent a t subambient temperatures, can be used as a source of other gallium(1) compounds, e.g., C,H,Ga (95, 961, and the decomposition of a monosubstituted gallane opens up further synthetic possibilities. Predictably bases cleave the Ga(pCl),Ga skeleton of monochlorogallane. For example, trimethylamine forms the 2 : 1adduct (Me,N),GaH2C1at low temperatures; this dissociates at room temperature to the 1: 1 adduct Me,N.GaH,Cl (38) and free NMe, . Ammonia, by contrast, brings about heterolytic cleavage with the generation of the salt-like product [H,Ga(NH,),]+Cl-. The ability of Ga-H bonds to add t o the double bonds of alkenes has also been shown to extend from dichlorogallane (41,42)to monochlorogallane, ethene reacting in stages to give first cis- and trans-[EtHGaCl], and then [Et2GaC11,. Most compelling of all, though, is the role of monochlorogallane as a precursor to other derivatives of the type [H,GaX], , typically through the interaction with a salt of the X- anion under solvent-free conditions. From this point the hunt for gallane itself is truly entered.

196

DOWNS AND PULHAM

Et

NH,.

[Me3N]2.GaH2CI

19) K Room

temperacure

Me3N.EaH2CI

+

Me3N

2 [H2Ga ( NH3 )2] +CI

SCHEME 5. Preparation and reactions of [H2GaC112(reproduced with permission from Ref. 10). V. Gallane at Last!

A. PRELIMINARIES Earlier attempts to prepare gallane in the Oxford laboratory (35,361 compassed a variety of potential routes, viz., (a) displacement reactions involving an adduct of GaH,, e.g., Me3N.GaH, or MGaH, (M = Li or Na), and an acid, e.g., BF, or HCl; (b) gas-phase pyrolysis or matrix photolysis of an adduct of GaH,; and (c) the interaction of a tetrahydrogallate MGaH, with a gallium compound, e.g., Ga,Cl,. The only one of these to give any encouragement was the interaction of gallium(II1) chloride with a tetrahydrogallate, the solid mixture yielding under

197

GALLIUM HYDRIDES

solvent-free conditions at ambient temperatures small amounts (in the order of 1-2 mg or less) of a volatile, thermally unstable product, in addition to substantial quantities of elemental gallium and hydrogen. The condensate formed by quenching the vapor of the product on a CsI window held at 77 K was typically characterized by the infrared spectrum reproduced in Fig. 5, with three main absorptions a t 1978 (s), 1705 (s,br),and 550 cm-' (s,vbr) (97).The spectrum differs significantly from the one reported by Greenwood and Wallbridge (33)for the product they identified as free gallane for it includes a prominent, broad absorption near 1700 cm-', as well as the much sharper feature near 1980 cm- The vibrational properties of known gallium hydrides, as outlined in earlier sections, prompt the bands at 1978 and 1705 cm-' most plausibly to be identified with v(Ga-H,) and v(Ga-Hb-Ga) fundamentals, respectively, in an aggregate with a comparatively wide Ga-Hb-Ga bridge angle [cf. [Me,GaHI, (47).Irrespective of the conditions of the experiment, however, it proved impossible to isolate a product that was entirely free from chloride. Thus, chemical analysis showed the proportion Ga:C1 to be typically 5:1, and the 'H NMR spectrum of C6D,CD3 solutions at low temperatures confirmed that

'.

2000

1800

1600

1400

1200

1000

800

600

400

20

2000

1800

1600

1400

1200

1000

800

600

400

20

;/cm-' FIG.5. The IR spectrum of the annealed solid film formed by condensing the volatile products of the reaction between solid gallium(II1)chloride and (i) NaGaHl or (ii)NaGaDl on a CsI window held at 77 K (reproducedwith permission from Ref. 56; copyright 1991, American Chemical Society).

198

DOWNS AND PULHAM

the product contained more than one gallium hydride derivative. It appears, therefore, that a hydride-rich product including species like [GaH,], is formed through the interaction of neat gallium(II1) chloride and a tetrahydrogallate, but only in very low yields (<1%) and with little prospect for the isolation of pure gallane. Altogether superior t o gallium(II1) chloride as a precursor to gallane, however, is the newly discovered monochlorogallane, [H,GaCl], . Not only has hydride/chloride exchange already gone two-thirds of the way to completion, but monochlorogallane is also a liquid even at quite low temperatures and therefore susceptible to more efficient mixing with a hydride ion source like LiGaH, . B. SYNTHESIS AND CHARACTERIZATION 1. Synthesis

Monochlorogallane has indeed been the turning point in the hunt for gallane and its derivatives. For it reacts in uucuo with lithium tetrahydrogallate at 243-250 K to give not only substantial quantities of elemental gallium and hydrogen but also gallane in yields up to 50%, based on Eq. (15) (37, 56, 61):

The identity of the volatile, thermally perishable product has been established unequivocally by chemical analysis, by its vibrational and 'H NMR spectra, and by chemical trapping with trimethylamine to give the known molecular adduct (Me3N),GaH, as the sole product at 178 K. The reaction affording gallane is carried out between freshly prepared reagents under solvent-free conditions. The choice of hydride ion source is dictated by its activity and by the need, at least in the first place, to avoid the formation of mixed hydride derivatives. Thus, lithium tetrahydroaluminate is less efficient in this role than the tetrahydrogallate, and lithium tetrahydroborate gives high yields of gallaborane, [GaBH,], (see Section V1.A) (37, 65).The volatile products of the reaction are removed under continuous pumping and essentially pure gallane can be isolated by fractional condensation. Here it is imperative that operations be carried out at pressures less than mm Hg in preconditioned all-glass apparatus of the sort described previously (see Section III.A.l). In addition, the thermal frailty of gallane requires that all apparatus to which it has access must be maintained at temperatures less than 263 K; without such cooling the fate

GALLIUM HYDRIDES

199

of the gallane is signaled by the formation of gray deposits of elemental gallium on the walls of the apparatus. Efforts to improve the yield of gallane by using a solvent to give better control over the metathesis reaction have so far met with only limited success. The choice of medium is obviously restricted by the need to minimize both the basic properties and the susceptibility to reduction of the would-be solvent. Thus, the mildly basic properties characterizing a n ether, say, may enhance its ability to dissolve the reagents, but only at the expense of producing a sample of the gallane that can be freed from the solvent, if at all, only with difficulty. On the other hand, such a solution may be perfectly suitable if the gallane is needed merely as a synthetic intermediate or for the growth of galliumcontaining films. Just such a n approach has been used with signal success, for example, to open up the synthetic chemistry of AlCl (98-102) and GaCl (95, 96). Preliminary studies of reaction media show that n-octane and methylcyclohexane offer no advantage in the yields of gallane they afford; some improvement may be achieved by the use of toluene, but fractional condensation gives only small quantities of pure gallane, the bulk of the product remaining with the solventrich fraction. 2. Physical Properties and Chemical Analysis Gallane condenses at low temperatures as a white solid that melts at ca. 223 K to form a colorless, viscous liquid. The rate of vaporization of the solid a t 210 K is consistent with a vapor pressure on the order of 1mm Hg. Samples of the material in the condensed phase (liquid or solution) decompose to the elements at temperatures in excess of 243 K. At a pressure of 10 mm Hg the vapor has a half-life of about 2 min at ambient temperatures. Decomposition to the elements in accordance with Eq. (161, 1 - [GaH,],

-

Gab) +

3

H,(g),

(16)

furnishes a method of chemical analysis for the gallane, albeit one which is rendered less than ideal for small amounts of the material by competition from side reactions with traces of adsorbed moisture or other hydroxylic impurities. A superior method of quantitative assay exploits the reaction with an excess of anhydrous hydrogen chloride at much lower temperatures (178 K): -[GaH3], 1

+ 3HC1-

iGazCl8 1 + 3Hz.

(17)

100

000

1600

1200

800

41

80

60 40

2c

0.54

I

0.5

0.44

2000

1200

1600

800

400

;/cm-’ 2

‘0 2050

1950

2000

100

l! 0

100

A

;90

90

m

.=

L

E

E

I-

s 80 2 0

60

2050

2000

’

i ~ /mc

1950

l! 0

GALLIUM HYDRIDES

201

Qualitative and quantitative analysis for chlorine confirms, moreover, that the compound can be made free from contamination by chlorogallanes.

3. Vibrational Spectra A film of the annealed solid compound at 77 K displays an infrared spectrum resembling that of the condensate formed by the vapors derived from the reaction of an excess of metal tetrahydrogallate with gallium(II1) chloride (see Fig. 4) (35-37,56,61). Hence it appears that small amounts of impure gallane are indeed generated by chloride-hydride exchange starting from gallium(II1) chloride. The three main absorptions at 1978, 1705, and 550 cm-' are observed to shift to 1422, 1200, and 400 cm-', respectively, for the perdeuteriated compound. The corresponding Raman spectrum of the solid also includes just three significant features, namely, at 1979 (s), 700 (w),and 520 cm-' (w) (971, which shift to 1412,500, and 375 cm-', respectively, upon deuteriation. Very different infrared spectra are exhibited by the vapor or by solid matrices formed by codepositing the vapor with an excess of a suitable inert gas at ca. 20 K (see Fig. 6). Here the pattern and energies of the absorptions, and particularly the rotational structure of individual absorptions of the vapor, testify to the presence of a relatively simple molecule with a momenta1 ellipsoid in which at least one ofthe principal moments of inertia is unusually small for a gallium derivative. Three aspects of the spectra turn out to be critical to the unambiguous identification of the gaseous molecule. (i) The infrared spectrum of gallane vapor includes four distinct absorptions attributable to u(Ga-H) [or u(Ga-D)] fundamentals. Of these two occur at high energy-near 1980 (1430) cm-l-in the region diagnostic of the stretching motions of terminal Ga-H bonds in a neutral gallium hydride (see Section III.B.l). The other two occur at substantially lower energy, namely, 1180-1300 (840-950) cm-'. So closely does gallane match dimethylgallane (see Section 4.3) in this region of the spectrum that it is hard to resist a similar interpretation, the two absorptions being most plausibly identified with the stretching

FIG.6. (a)The IR spectrum of gallane vapor (i) trapped in a solid nitrogen matrix at ca. 20 K and (ii) at a pressure of ca. 5 mm Hg and temperature near 270 K (contained in a cell fitted with CsI windows and having a path length of 10 cm). (b)Part of the IR spectrum of gallane vapor at a pressure of ca. 0.05 mm Hg contained in a multiplereflection cell set to a path length of 6.5 m and maintained at ca. 290 K (reproduced with permission from Ref. 56; copyright 1991, American Chemical Society).

202

DOWNS AND PULHAM

vibrations of one or more Ga-H-Ga bridges. In that case, Eq. (12) leads to an estimate of ca. 93” for the Ga-H-Ga bond angle, 19, in gallane, thereby adducing strong circumstantial evidence that the absorber is in fact Ga2H6with a diborane-like structure (2). (ii) Most of the bands in the infrared spectrum of gallane vapor carry unmistakable signs of rotational structure; no better example is provided than that of the two absorptions near 1980 cm-’, the appearance of which, as measured at moderately high resolution with the aid of a multiple reflection cell, is illustrated in Fig. 5. There are, it appears, two types of band. One type, exemplified by the band at 1976 cm-’, displays all the features characteristic of a parallel band of a highly prolate symmetric top molecule (see Section III.B.l). The other type has the attributes of aperpendicular band of such a molecule (as witness the band at 1993 cm-’ in Fig. 5). Under modest resolution, the parallel-type features assume envelopes with an average P-R branch separation of 10.3 ? 0.5 cm-l. According to Eq. (13), this implies a mean rotational constant B of about 0.066 cm-’; with reasonable allowance for the small contribution made to the B value by the hydrogen atoms, a molecule with the structure of 2 is then estimated to have a Ga-eGa distance on the order of 260 pm. A result so tantalizingly close to the corresponding distance in the dimethylgallane dimer, Me,Ga(p-H),GaMe, [261 pm as measured by electron diffraction (4711 may not be wholly beyond the reach of blind chance, but the odds must be stacked heavily in favor of H,Ga(pH),GaH, as the principal component of gallane vapor. Detailed elucidation of the rotational structure associated with the observed vibrational transitions must await the services of an infrared spectrometer having a resolving power compatible with the B value of the molecule and so with the capacity to resolve the closely spaced rotational lines forming the P- and R-branches of the parallel-type bands. This will inevitably be challenging work, but some idea of the potential rewards may be gained from a preliminary analysis of the Q-branch pattern so conspicuous in the perpendicular-type band at 1993 cm-’. Hence comes an estimate of 1.5095 ? 0.0010 cm-’ for A”-B”, the difference between the appropriate rotational constants of the absorber molecule in its vibrational ground state (37). Even this morsel of information carries significant weight, as will be evident shortly in relation to what can be learned about the dimensions of the molecule by experiment. (iii) In the pattern and wave numbers of the vibrational transitions observed in infrared absorption, there is no change of consequence when the vapor species are trapped in a solid inert matrix at low

GALLIUM HYDRIDES

203

temperatures. What is remarkable, however, is that annealing of the matrix (at temperatures no higher than 35 K for an argon matrix) causes the bands associated with the vapor species to decay with the simultaneous appearance and growth of bands akin to those of solid gallane (q.u.1. Facile aggregation of GazH6molecules appears therefore to be the rule in the condensed phases, giving, at least in the first instance, a discrete oligomer; the infrared spectrum argues that this must involve a change in the mode of hydrogen bridging while retaining terminal Ga-H bonds [cf. a-AlH, (1311.One possible formulation of the oligomer is the tetramer [GaH,], with a cyclic configuration such as 5. Such a structure is certainly consistent with the energies and relative intensities of the infrared absorptions associated with the stretching vibrations of bridging Ga-H functions, and which imply that, as with dimethylgallane, the Ga-H-Ga angle, 8, has now opened out to at least 120". The physical properties of solid gallane-its volatility and solubility in solvents like toluene-favor the belief that it too consists of discrete oligomers like [GaH,], , rather than an extended one-dimensional polymer along the lines of 6.

Although the full interpretation of the vapor spectrum of normal and perdeuteriated gallane is far from straightforard, the main features can be successfully analyzed in terms of the vibrational properties to be expected of the molecules Ga@, and Ga,D,, with the structure (2) belonging, like B@6, to the D,,point group. The relevant assignments, given in Table 11, rest on six principal criteria: (a) the selection rules expected to govern the activity of vibrational transitions in infrared absorption; (b)the effect of deuteriation on the energy of a given transition (with due appeals to product-rule calculations, etc.); (c) the rotational fine structure modulating individual vibrational bands; (d) analogies with the vibrational properties of related molecules, notably [H,GaClI, (36,37,39, 40, 1031,[MezGaH], (471,H2Ga(p-H),BH2(37), Me,N.GaH, ( 5 4 ) , and B,H, (104);(e) the harmonic frequencies and infrared intensities calculated for the molecules on the basis of ab initio MO techniques (86,105-109); and (f) the results of normal coordinate

204

DOWNS AND PULHAM

TABLE I DIMENSIONS O F S O M E GASEOUSGALLIUM HYDRIDES A S DETERMINED BY ELECTRON DIFFRACTION" rtGa-H,) lpml

r(Ga-Hbl lpml

151.9 (3.51 158.6 (0.81

171.0 I3 8) 182.6 (0.81

X

=

X

=

Ga, 258.0 (0.21 B, 217.9 10.2)

Me2Gaip-HJ2BH2

-

179.1 (3.01

X

=

B, 216.3 10.8)

H2Galp-CI12GaH2

155.9 11.91

-

X

=

Ca, 324.1 (071

Me2Galp-H l,GaMe,

-

170.8 11.4)

X

=

Ga, 261.0 10.51

H2GaB3HH

144.2 (1.1)

176.0 (2.8)

X

=

B, 231.2 (0.1)

H3Ga,NMeR

149.7 11.51

-

X

=

N, 212.4 10.71

lH2GaNMezl~

148.7 13.61

-

X

=

N, 202.7 (0.41

Molecule H,Gaip-H),GaH, H2Ga(p-H12BH,

" Estimated standard

r(Ga...X) lpm)

Other parameters (Ga-Hb-Ga 97.9 (3.21" rlB-Hbl 133.4 (0.81, rlB-H,I 123.4 (0.81 pm; (Hb-Ca-Hb 75 3 (1.2r, (Hb-B-H, 113 4 (2.7)' rlGa-C) 194.4 (0.41. rlB-Hbl 121.7 (1.91, r(B-H,l 119 2 11.9) pm; (Hb-Ga-Hb 68.4 (4.61" r(Ga-Clb) 234.9 10.31 pm; (ClbGa-Clb 92.8 10.8)" rtGa-CI 195.4 (0.41 pm; (GaHh-Ga 99.6 11.41" rlBil,3)-Bl4)l185.2 (1.31, rlB-HbJ 126.4 (0.7) pm. dihedral angle 114.4 10.61" rlN-C) 148.2 (0.5)pm; (GaN-C 109.9 (0.5)" rtN-C) 146.3 I 1 31 pm: (GaN-Ga 90.6 (0.8)"

Refer

37. 37,

8i

39, 4;

37,

deviations are given in parentheses.

analysis calculations. Most of the fundamentals active in infrared absorption have thus been accounted for; only the b,, deformation mode uI4 (dominated by the GaH, wagging motion) has not been pinned down as positively as the others, and the b2uring-puckering mode, expected to occur near 200 cm-'(105-109), has so far escaped detection. 4 . Electron Diffraction of the Vapor

With cooling of the specially constructed inlet system described in Section 1II.A to temperatures between 253 and 258 K, it has been feasible to record an electron diffraction pattern for gallane vapor (37, 56). For reasons that will by now be self-evident, the results are not of the highest quality. Nevertheless, the pattern is wholly consistent with the inferences drawn from the infrared spectrum of the vapor, and its analysis gives access to the first experimentally determined dimensions of the Ga2H6molecule. Figure 7 depicts the experimental radial-distributon curve, which, like that of B2H6(110),is distinguished by just two prominent peaks corresponding t o scattering from M-H and M...M atom pairs (M = B or Ga). That atom pairs separated by more than 300 pm do not contribute appreciably to the scattering endorses Ga2H6as the predominant vapor species under the prevailing

51

5:

205

GALLIUM HYDRIDES

TABLE I1 OBSERVED AND CALCULATED VIBRATIONAL PROPERTIES OF THE Ga2H6AND Ga2D6MOLECULES(56,105,109)" Ga&k Number and approximate description of modeb

Symmetry cIassb

Observed Harmonic lanharmonicl vibrational vibrational frequency (cm-9 frequency 1cm-l) ~ a l c . ~ , ~ in IR absorptione ~

WK

ui up it3

u4

a" b2,

u5 ub:

b2u

us

u,

wg

vio

bl,

vll

bl"

vi3

ul2 u14

b3, b3u

uI5 u16

iq7 uLs

utGa-H,) utGa-Hb) StH,-Ga-H,) uICa.-Gal Twist utGa-Hb) H,-Ga-H, wag uiGa-H,l p(H,-Ga-H,I Ring pucker utGa-H,) ptH,-Ga-H,) ulGa-Hb) H,-Ga-H, wag Twist vIGa-H,) u(Ga-Hbl SIH,-Ga-H,l

~~

2099 1577 782 241 473 1356 406 2101 (353) 837 (178) 225 (71 2096 493 1294 I2911 682 1165) 808 2093 1120) 1412 (1229) 722 (607)

Harmonic vibrational frequency (cm-') calc.e

0bserved (anharmonic) vibrational frequency ( c m - ' ~ in IR absorptionP

~~

ia ia ia ia ia ia ia 1993 m 760 w

no ia ia 1202 s

ca. 650 m ia 1976 m 1273 s 671 vs

1488 1116 558 237 335 960 297 1502 598 159 1498 359 924 487 572 1482 1004 518

~~

ia ia ia ia ia ia ia 1439 m 555 w

no ia ia 860 s 439 mw

ia 1416 m 923 s 484 vs

" Abbreviations: s, strong; m. medium; w, weak; v, very; la. inactive in infrared absorption; no, not observed. Numbering and classification of the modes conform to those given for B,H, and B,D6 in Ref. 104. Results basedon configuration interaction methods with single anddouble excitations(C1SD)with double-zeta plus polarization basis sets isee Refs. 105 and 109).Computed frequencies are typically some 5% higher than we values determined experimentally, which are in turn larger than observed, anharmonic frequencies, by up to ca. 60 cm-' for vibrations like uiGa-HI with large amplitudes. The frequencies have been calculated for the molecules 69Ga2H6and 69Ga,D6. "Calculated infrared intensities given in parentheses; values in km mol-i. ' Frequencies are those for Ga,H, and Ga2Db:molecules with gallium of natural isotopic composition.

conditions (pressure, ca. 1 mm Hg; temperature, ca. 256 K); heavier oligomers like [GaH,], (q.u.) would surely include Ga...Ga atom pairs with separations exceeding 300 pm. Within the limits of experimental uncertainty, a GazH6 molecule having the structure of 2 is found satisfactorily to account for the measured scattering. The two independent distances r(Ga... Ga) and r(Ga-HImeanare well determined, but correlation between the parameters specifying the bridging and terminal Ga-H distances obviates an exact differentiation between these distances. The H,-Ga-H, angle, which is specified by the dimensions of the GaGaH, triangle, is correspondingly ill-determined. Irrespective of the precise basis set adopted, all the ab initio calculations reported so

206

DOWNS AND PULHAM

I00

200

300

400

500

600

-

-

rPm

FIG.7. Observed and difference radial distribution curves, P ( r ) l rversus r, for digallane vapor; before Fourier transformation the data were multiplied by s.exp[(-0.000 020 sz)/(ZG,- fcJ2] (reproducedwith permission from Ref. 56; copyright 1991, American Chemical Society).

far (86,105-109) seem to converge on a value near 130”for this angle; indeed, according to other theoretical calculations involving the molecules H,Ga(p-H),BH, (111,112)and H,Ga(p-Cl),GaH, (91,92), such a value seems to be endemic to the H,Ga(p-X), moiety (X = H or C1). According to the results of the diffraction study, the mean B-value, B ”, is ca. 0.0676 ? 0.001 cm-’ (pleasingly close to the first rough estimate elicited from the P-R branch separations of the parallel-type infrared bands). The rotational analysis of the infrared-active fundamental va having already delivered a value of A”-B”, it follows that A“ = 1.5771 0.002 cm-’. Because A ” depends only on the positions of the light hydrogen atoms with respect to the Ga...Ga axis, it provides an independent means of appraising the H,-Ga-H, angle. A value near 138” is thus deduced. Uncertainty necessarily persists while we lack a realistic audit of the effects of vibrational averaging (implicit in the electron-diffraction input to this calculation). Such an audit, together with more precise information about the rotational constants of natural

*

207

GALLIUM HYDRIDES

and isotopically enriched Ga,H, ,will be needed before a more complete and exact picture of the molecular structure can be secured. Suffice it to say that the best estimates at present available give dimensions in keeping not only with those of related molecules, e.g., Me,Ga(pH),GaMe, (471,H2Ga(pu-C1),GaH2 (37,39,40),andMe3N.GaH3(511, but also with the predictions of recent theoretical calculations (105-109). 5 . ' H NMR Spectrum

The 'H NMR spectrum of gallane dissolved in [2H,]toluene at 208 K consists of two singlets at 6 4.41 and 1.11, with relative intensities in the ratio 2 : 1 (see Fig. 8);both show the broad line widths typical of protons directly bound to gallium atoms (32).The gallane is relatively volatile under these conditions and so it is tempting to infer the presence of Ga2H6molecules having the structure of 2. Such an interpretation appears questionable, however, in the light of the ready aggregation of Ga2H6revealed by the matrix experiments. In fact, the spectrum does not differentiate between Ga2H6and any other oligomer, like

-30°C

* 1

6.0

5.0

4.0

3.0

1

2.0

1.0

Chemical Shift bHIppm

FIG. 8. The 'H NMR spectrum of a C6D5CD3solution of gallane recorded with the sample held successively at -65, -30, and 0°C. The resonance near 8H 2.0 is due to residual C7D7Hin the solvent (reproducedwith permission from Ref. 56; copyright 1991, American Chemical Society).

208

DOWNS AND PULHAM

[GaH,], , containing terminal and bridging hydrogen atoms in the proportion 2 : 1. Warming the sample to 243 K results in broadening and coalescence of the two resonances, indicating rapid exchange between bridging and terminal proton sites. The single broad resonance that the sample now displays occurs at 6 3.7, a chemical shift appreciably different from the weighted mean (6 3.31) of the shifts characterizing the separate resonances observed with the sample at lower temperatures. In all probability, therefore, more than one molecule is implicated in the exchange process. Further warming to 273 K results in rapid decomposition of the sample. Studies of the decomposition have failed as yet to pick up any scent of gallium hydride intermediates formed en route to the ultimate products, namely, elemental gallium and hydrogen. It is possible, however, that polygallane intermediates like Ga,Hlo enter into the decomposition [compare, for example, the thermolysis of diborane (1-8)l and that they will prove t o be amenable to physical or chemical trapping.

C. CHEMICAL PROPERTIES Some of the chemical properties of gallane have now been charted, with the results summarized in Scheme 6. The reactions appear mostly to parallel those of diborane (1-8). Thus, symmetrical cleavage of the Ga,H, moiety occurs, at least formally, with trimethylamine or phosphine to give the corresponding molecular adduct L,GaH, [L= Me,N, n = 1 or 2; L = PH,, n = 1 (37, 56, 113)l. The phosphine derivative H,P-GaH3 is a frail vessel, even by the standards of gallane chemistry. That it is formed, however, on cocondensation of gallane with an excess of nitrogen doped with phosphine at ca. 20 K is signaled by the infrared spectrum of the resulting matrix (56, 113);natural and deuteriated versions of the molecule H3P-GaH3have thus been characterized in some detail. By contrast, ammonia causes unsymmetrical cleavage of the Ga-H-Ga bridges of gallane at 178 K, with the formation of a saltlike product most aptly formulated as [H,Ga(NHJ,] [GaH,l-. Stepwise insertion into the Ga-H bonds is the path taken by the reaction with ethene, affording mainly diethylgallane, [Et,GaHl,. +

D. SUMMARY The experiments described in the preceding sections leave little room for doubt that the base-free trihydride of gallium has at last been run to earth. It can be prepared in amounts up to 500 mg. While decomposing to the elements at ambient temperatures, gallane vaporizes at low

v

N

c

-1-

+

I

v

-I

0 0

I

0

A

c

0

I

c

0 c

N

0 W

Y

w

210

DOWNS AND PULHAM

pressures as the diborane-like molecule H,Ga(p-H),GaH,, the main dimensions of which have been deduced from the electron-diffraction pattern of the vapor. Despite this progress, there is much still to be done. For example, the nature of the compound in the condensed phases has still to be unraveled. Growing a single crystal will be a tall order indeed, but neutron-diffraction studies on crystalline powder samples of the wholly or partially deuteriated gallane at 4.5 K are already in train (114). Preliminary analysis of the results suggests a bodycentered tetragonal unit cell with a = 1262.8 and c = 499.6 pm; more detailed analysis may expose the molecular form of the solid to give a definitive explanation of such properties as its volatility and the apparent retention of terminal as well as bridging Ga-H bonds. Extending our chemical knowledge is never going to be easy with a compound so predisposed to decomposition. On the evidence of the reaction with phosphine, however, dissociation of [GaH,], aggregates appears to be no less facile than aggregation, and there is clearly scope for further cocondensation experiments, for example to afford a glimpse of the arsine adduct H3As.GaH3, the stable existence of which receives support both from the recent matrix isolation of the complex H,As.GaMe, (115)and from the diagnosis of ab initio calculations (116). Such a molecule is of more than esoteric interest, as testified by very recent spectroscopic studies of the adsorption of hydrogen atoms on the (1x 6) reconstruction of GaAs(100) (117,118); unmistakable evidence has thus been found for the formation of gallium hydride species with both terminal and bridging Ga-H functions. Another facet inviting exploration is the possibility of forming open or closed cages in gallium species analogous to B10H14and Bl2HlZ2-,with the potential for kinetic and thermodynamic stabilization of the Ga-H bonds that may thus be achieved. Gallium clusters are already known in intermetallic compounds (10). Typically they are not discrete but interlinked, as with the Gazl unit found in the phase Rb0.6Na6,25Ga20.02 (119);consisting of a pair of confacially linked Gal, icosahedra, 7 is joined by its vertices to other such units as well as to single Gal, icosahedra. Added encour-

GALLIUM HYDRIDES

211

agement to the search for polygallane derivatives comes too from the icosahedral anion [AlI2Buil2l2-, which has just seen the light of day as its potassium salt (120).

VI. Hydrogen-Rich Derivatives of Gallane

A. INTRODUCTION Monochlorogallane has been of crucial importance in the entree it has given to gallane. It is also the obvious precursor to other monosubstituted gallanes of the type [H2GaX],,typically through its interaction with a salt of the X- anion in U ~ C U Oand under solvent-free conditions (see Scheme 5). Hence it has been possible to achieve 80-90% conversion of the monochlorogallane to the simple mixed hydride GaBH, (37, 65) or the tetraborane(l0) derivative 2-GaB,Hlo (37, 63). As with the synthesis of gallane, operations need to be carried out in preconditioned, all-glass apparatus along the lines illustrated in Fig. 2a, and the vapors of the volatile product need to be kept at low pressure and exposed to wall temperatures not exceeding 263 or 283 K for GaBH, or GaB,Hl,, respectively. The new compounds have been authenticated by chemical analysis, by their spectroscopic properties, and by electron diffraction of the vapors. In addition, a reconnaissance of their chemistries has been mounted.

B. GALLABORANE, GaBH, (37, 65) In its synthesis and early characterization, gallaborane has followed closely the precedents set by gallane. Preparation involves the reaction between monochlorogallane and freshly recrystallized lithium tetrahydroborate at 250 K. Fractional condensation in uucuo gives a product even more volatile and marginally more robust to thermal decomposition than gallane. The solid condensate registers a vapor pressure on the order of 1 mm Hg at 195 K and melts near 228 K. In the condensed phases gallaborane decomposes at temperatures exceeding 238 K; at a pressure near 100 mm Hg, the vapor enjoys a half-life on the order of 2 min at room temperature. Decomposition results in the quantitative formation of elemental gallium, diborane, and dihydrogen in the stoicheiometric proportions required by Eq. (18) and so provides a means of chemical assay:

212

DOWNS AND PULHAM

The infrared spectra of natural and perdeuteriated samples of gallaborane imply the presence of a molecular unit common to the vapor and matrix-isolated states. The results advance strong circumstantial evidence that the molecule is H,Ga(p-H),BH2 (8) with a bidentate

(8)

BH, group and possessing Czusymmetry. This conclusion is strongly endorsed by the well-resolved rotational structure displayed by most of the infrared absorptions of the vapor. For example, as illustrated in Fig. 9, the absorption near 2550 cm-', associated with a B-H, stretching fundamental, clearly shows the P- and R-branches characteristic of a perpendicular-type transition of a rotor approximating to a symmetric top. Strongly in evidence are the individual Q-subbranches corresponding to the transitions betwen different K-levels; analysis of these gives a value of 1.7675 ? 0.0010 cm-' for the difference between the rotational constants A"-B" (cf. Section V.B.3). As in the case of digallane, such a large value can be reconciled only with a molecule having a very small moment of inertia about one principal axis, here the Ga-..Baxis; it therefore implies that only hydrogen atoms lie off this axis. By contrast, the second infrared band originating in a v(B-H,) fundamental, located near 2480 cm-' (see Fig. €9,displays all the features characteristic of a parallel -type transition, the partially resolved rotational structure serving notice of a value of B" on the order of 0.25 cm-'. Once again, therefore, we must be dealing with a distinctly prolate symmetric top. In its information relating to the vibrational and rotational properties and dimensions of the GaBH, molecule, the spectrum will undoubtedly be a happy hunting ground of the future, once the appropriate conditions of resolution can be met. Unlike digallane, the dipolar GaBH, molecule also invites high-resolution microwave measurements (79). By taking account of the rotational structure, the effects of deuteriation, the selection rules expected to govern the activity of vibrational transitions in infrared absorption, and the vibrational properties of related molecules, e.g., [H,GaClI, (39, 40, 103), Ga2H, (37,561,B2Hs (104), and MezGa(p-H),BH2(82), it has been possible to identify most of the vibrational fundamentals of the H,Ga(p-H),BH, molecule. Strong support for the proposed assignment comes from the

213

GALLIUM HYDRIDES

a 261 5 100 I

2600

2575

2550

2525

251 0 I100

b 2475

I

50

30 2510

2500

2450 2 15 100

50

2475

30 2450 24 5

btcm-1 FIG. 9. Portions of the IR spectrum of gallaborane vapor at a pressure of 0.15 mm Hg and maintained a t ca. 290 K, showing the two v(B-H,) fundamentals; the vapor is contained in a multiple-reflection cell set to a path length of 3.9 m (37,65).

214

DOWNS AND PULHAM

harmonic frequencies and infrared intensities computed on the basis of ab initio procedures (111,112). Electron-diffraction measurements on the vapor at low pressures (ca. 1mm Hg) have yet to be completed at more than one camera distance, for technical reasons connected with the reactivity and thermal instability of the compound. According to the preliminary measurements made at a single camera distance, however, the scattering pattern meets all the primary requirements of a simple GaBH, molecule with the structure of 8. Neither the infrared spectrum nor the electrondiffraction pattern gives any grounds for believing that any oligomer [GaBH,],, with n = 2, 3, etc., makes up an appreciable fraction of the vapor. Table I includes the best estimates attainable in present circumstances for the dimensions of the GaBH, molecule. In that the results come close not only to the corresponding distances and angles in molecules like Ga2H6(56),Me,GaBH, (85),and HGa(BH,), (48-501, but also to the prognoses of detailed quantum mechanical studies (111, 112),they are unlikely to be too wide of the mark. There is reassuring substance here to bring to the increasingly lively theoretical speculation excited by this and related discoveries (86, 91, 92,105-109, 111, 112). The Ga-H bond in the central Ga(p-H),B bridge is attenuated by some 10 pm compared with its counterpart in the symmetrical Ga(p.-H),Gaunit of Ga2H6;this reflects presumably the greater charge separation between the gallium and hydrogen atoms in the unsymmetrical Ga-H-B bridges, as well as the steric demands imposed by BH, as a bidentate ligand with a peculiarly short "bite." Less clear is the state of gallaborane in the condensed phases, although aggregation appears to be a recurring theme of both the solid and solutions. At temperatures in the range 153-238 K, solutions of the gallane in toluene or methylcyclohexane display (37, 65) the 'H and "B magnetic resonances characteristic of a "rigid" terminal GaH, and a fluxional BH, unit (71 in a molecule having the empirical formulation H2GaBH,. In addition, though, the spectra reveal the coexistence of a second component in dynamic equilibrium with the first. This has 'H and "B resonances close to those believed t o originate in monomeric H,GaBH, ,but the 'H{"B} spectrum discloses, intriguingly, that, unlike the monomer, it is characterized not by a singlet but by a doublet BH, resonance with a separation of 8.5 Hz due to spin-spin coupling to a second nucleus. We have still to find a wholly convincing explanation of these findings, but the properties of the second component (including its response to changes of concentration) point to a loosely bound aggregate, possibly a dimer [H,GaBH,], having a cyclic structure similar to 5 (89)with alternating GaH, and BH, groups linked via single hydrogen

215

GALLIUM HYDRIDES

bridges. Bidentate coordination of the BH, group is retained in solid gallaborane, on the evidence of its infrared and Raman spectra, but as yet there are few other clues to the nature of the solid. Chemical studies are still only in their infancy, but Scheme 7 summarizes the results of a preliminary survey. Scrutiny of the products of thermal decomposition, engineered under various conditions, has failed so far to spot any clear-cut signs of new gallaborane intermediates, e.g., Ga2B2HIo. Homolytic cleavage of the Ga(p-HI2Bbridge is brought about by the bases Me,N and Me,P, which give rise to an equimolar mixture of the adducts LmGaH, and LaBH, (L = Me,N or Me,P). At lower temperatures gallaborane and Me,N appear to unite in a molecu-

Me3P*GaH3 Y Me3P'BH3 b

\ I2[ H 2 GaC1I2 p w .

LiBH4

H2GaBHQ

250 K

/ Me3p

178

K

1 \I NH3 195 K

Me3N *GaH2(BH4)

Excess Me3N

[H2Ga(NH3) ,]+[BH4]-

Me 3N.GaH3 + Me3N.BH3

SCHEME 7. Synthesis and some reactions of gallaborane (37, 65).

216

DOWNS AND PULHAM

lar adduct, Me3N.GaH2(BH4),in which bidentate ligation of the BH4 group is preserved, thereby implying a structure such as 9 with five-

(9)

fold coordination of the gallium center; the infrared spectrum of this product hints at an obvious kinship to another unstable complex, Me3N.GaMe2(BH4)(121),as well as the more stable aluminium analogues Me3N.A1R2(BH4)[R = H (122) or Me (121, 123)l. Ammonia differs from the larger bases Me3N and Me3P in promoting heterolytic cleavage of the GaBH, molecule with the formation of a relatively robust product most aptly formulated as a cationic gallane derivative, [H2Ga(NH3 121 [BHJ . +

C. 2-GALLA-ARACHNO-TETRABORANE( lo), 2-GaB3H,, (37, 63) Metathesis with tetra-n-butylammonium octahydrotriborate, [Bu",N]+[B3Hel-,at temperatures near 243 K converts monochlorogallane into gallatetraborane, GaB3Hlo,in yields not far from those implied by

243-268

&[H2GaC1I2 + [BU"~N]~[B~H~I-

K

GaB3Hlot [Bun4NI'Cl-.

(19)

The product melts at ca. 178 K to give a colorless mobile liquid, the vapor pressure of which reaches 1 mm Hg at temperatures near 210 K. Its thermal stability shows a distinct advance on those of gallane and gallaborane. Although samples of the material in the condensed phase decompose rapidly at temperatures above 283 K, the half-life of the vapor at room temperature and a pressure of ca. 100 mm Hg extends to about 30 min. Handling is also made easier in that the vapor will even survive the passage through a greaseless (Teflon)valve provided that steps are taken to precondition the surfaces of preliminary exposure to a portion of the vapor, which is subsequently pumped to waste. Decomposition of the vapor at ambient temperatures proceeds in accordance with

-

GaB3Hlo(g)

GaBp(s)+ ?B,H&g) f G M g ) ,

(20)

GALLIUM HYDRIDES

217

and the resulting quantitative assay serves to fix the chemical composition of the gallane. The involatile solid residue answering to the composition GaB, may be an authentic example of a rare breed, namely, a gallium boride. The course of decomposition is complicated, however, by the existence of more than one channel. As the local concentration of the gallane increases, for example, with the move to the condensed phases, bimolecular processes compete with the unimolecular one, which is presumed to hold the key to the gas-phase reaction (20).Thus, decomposition of gallatetraborane in toluene solution gives metallic gallium, dihydrogen, and tetraborane(l0) as the principal products, as in Eq. (211, although these are accompanied by small amounts of unidentified boranes generated by secondary reactions: 4GaB3H,,

-

4Ga

+ 3B4HI0+ 5H2.

(21)

Gallatetraborane has been characterized (a) by its mass spectrum, (b) by the vibrational spectra of the vapor (sampled directly at room temperature or trapped in a solid nitrogen matrix at ca. 20 K) and of the annealed solid at 77 K, and (c) by the 'H and "B NMR spectra of [ 2H,]toluene solutions of the compound. That the vibrational spectra are essentially invariant with phase and temperature points to the presence of a molecular unit common to the vapor and solid states. Comparison with the spectra of compounds known to contain terminal GaH2 groups (37,39, 40,52, 56, 63) and bidentate octahydrotriborate groups, (p-H),B3H6[e.g., Me,GaB,H, (124,125) and B4H10(12611leaves little doubt that the new compound is a derivative of tetraborane(lO), with gallium replacing boron a t the 2 position (see Fig. 10).Unequivocal evidence of identity comes in any case from the NMR spectra. At temperatures between 193 and 283 K solutions of the hydride exhibit two llB resonances with relative intensities of 2 : 1(see Fig. 11): the triplet resonance at S("B) -12.9 is due to the unique apical boron atom [B(4) in Fig. lo], whereas the resonance with the unusual triplet splitting pattern centered a t 6("B) -44.0 is due to the "hinge" boron atoms [B(1) and B(3) in Fig. 101. The details of the 'H NMR spectrum have been decoded by 'lB decoupling, as illustrated in Fig. 11. The broad AB-type signal a t tiH 4.02, 4.56 must be associated with the exo- and endo-protons of the GaH, unit, the corresponding doublet a t aH 2.09, 2.78 with the apical BH, unit; the resonance at 6H 1.16 is due to the terminal hydrogens of the hinge boron atoms and those a t tiH-0.80 and - 1.49 are due to the four bridging hydrogens. The NMR properties echo those of the dimethylgallium derivative Me2GaB3H8(124) and related compounds featuring bidentate ligation of the B3H8- anion to a metal center (77, 78). Such systems run the whole gamut of fluxional-

218

DOWNS AND PULHAM

n

W Ht

k. "b'

A

7

WHt'

*

FIG.10. The structure of the molecule 2-GaB3H,oas deduced from the electron-diffraction pattern of the vapor (reproduced with permission from Ref. 63).

ity from the essentially static behaviors of H2GaB3H8and (OC),MnB,H, (78) to the rapid exchange of boron and hydrogen atoms that distinguishes (C,H,)BeB,H, (78) and M+[B3H81-salts (127),even at very low temperatures. The electron-diffraction pattern of the vapor at ca. 273 K, like the spectroscopic properties, has all the features to be expected of the GaB3Hlomolecule with the tetraborane(l0)-like structure illustrated in Fig, 10. Such a structure has yielded satisfactorily to refinement, despite some problems of correlation, giving the dimensions detailed in Table I. The measured geometrical parameters hold few surprises, being concordant not only with those of the related species B,HIo (128), Me2MB3H8(M = A1 or Ga) (125),(Ph3P),CuB3H8(129),and ( T ~ - C , H ~ ) ~ NbB3H8 (1301, but also with some theoretical forecasts (131). The known reactivity of tetraborane(l0) (1-8) holds out the promise of a varied and potentially fertile chemistry for its 2-gallium-substituted derivative. That promise remains largely untested at present, although gallatetraborane has been shown to react with an excess of ammonia at 195 K, yielding a white solid long-lived at room temperature and that, on the evidence of its infrared spectrum, is another cationic gallium hydride derivative, viz., [H,Ga(NH3)21+[B3H81-. It will be interesting t o see whether homolytic cleavage of the GaB, skeleton

219

GALLIUM HYDRIDES

is also possible and, if so, whether gallium is eliminated as a complex of the type L-GaH, or whether it is retained in a trinuclear GaBz cluster, also as a complex L.GaB,H7 (e.g., L = a phosphine). Another noteworthy reaction of tetraborane(l0) is that with CO, leading to the

a

I

I

-5

-10

-15

1

I

1

-20 -25

-30

-35 -40 -45

-50 -5

Chemical Shift BB/ppm

FIG.11. (a) I'B NMR spectrum of a C6D5CD3solution of 2-GaB3Hloat 233 K (measured at 96.25 MHz). (b) IH NMR spectrum of the same solution at 233 K (measured at 300 MHz), showing the effects of broad-band and selective IlB decoupling. The sharp resonance in the 'H spectrum near SH 2.0 is due to the solvent (reproduced with permission from Ref. 63).

220

DOWNS AND PULHAM

A

510

410

2.0 1.0 0.0 Chemical Shlft d ~ / p p m

3.0

-1.0

I

-2.

FIG. 1 lb-Continued.

elimination of H2 and the formation of B,H&O, and so the action of CO on the gallium derivative would surely repay investigation. It remains also to be seen whether controlled, rapid pyrolysis or copyrolysis offers a route to other gallium-substituted boranes.

GALLIUM HYDRIDES

221

VII. Hydrides of the Other Group 13 Metals: Preliminaries and Prospects

A. INTRODUCTION By degrees, a picture of the gallium hydrides is starting to take shape. With the identification and characterization of the thermally fragile but discrete molecules Ga&, , GaBH, , and GaB3Hlo,it appears that gallium, in its hydride chemistry, comes closer than aluminum to shadowing the unique behavior of boron (1-8, 10). This can be correlated with the valence electrons, which are bound somewhat more tightly in the Ga than in the A1 atom, as a result of the imperfect screening of the increased nuclear charge by the additional 3d’O shell. The tendency for gallium to remain four-coordinate in situations in which aluminum favors sixfold coordination is another sign of the reversion to boron-like properties on the part of gallium. In the light of what we now know, it is reasonable to expect that the chemistry of gallium hydrides is capable of considerable expansion, if not to match that of the boron hydrides, at least to surpass that of aluminum hydrides. It is appropriate, nevertheless, that we should conclude this chapter with a stocktaking of the other Group 13 metals, aluminum, indium, and thallium, to see what, if any, scope there may be for an enlargement of their hydride chemistry.

B. ALUMINUMHYDRIDES (10) Theoretical analysis of the equilibrium molecular structures, vibrational properties, and binding energies of Group 13 hydrides gives no hint of intrinsic instability in the molecules AlH, , Al,H6, and A13H9 (86,105-109). Hence they tend to underline the seemingly anomalous position of aluminum, vis-a-vis boron and gallium, in that aluminum gives an involatile polymeric hydride [AlH,], (10-131, but only fleeting signs of discrete molecules like A12H6.With hydride as a ligand it seems that aluminum is seldom to be found in four-coordinated environments, and the tendency to polymerize, more than any thermal instability, is the principal barrier to the isolation of tractable aluminum hydrides, although there are exceptions, e.g., diorganoalanes of the type [RzAIHl,, alane adducts like (Me3NI2A1H3,the dimethylamido derivative [Me2NA1H213 (681,and the mixed hydride AUBH,),. Another practical problem is that there are fewer synthetic options for the formation of A1-H bonds. There is, for example, no prospect of being able to prepare mono- or dichloroalane by reactions analogous to (141,once it is

222

DOWNS AND PULHAM

appreciated that the reagent MeaSiHis itself made by chloride-hydride exchange typically initiated by the A1-H bonds of LiAlH, (132). Instead one must look to metathesis reactions involving alkali-metal hydrides or compounds already containing A1-H bonds, e.g., LiAlH, or [R2A1Hl,. One possibility is to turn to direct synthesis from the elements. The molecule A1H has certainly been detected spectroscopically as a shortlived species on appropriate activation of gaseous A1/H2mixtures (14). AlH, and Al2H, have also been detected by mass spectrometry when aluminum is evaporated into a hydrogen atmosphere (18,133-135). Indeed, evaporation of aluminum from a hot tungsten filament into a hydrogen atmosphere (e.g., at a pressure of 2-3 mm Hg) yields at 77 K a heterogeneous solid condensate including elemental aluminum and a hydride with the empirical composition AlH, (136).The hydride appears t o be formed mainly not in the gas phase but on condensation of the vapors, and the tungsten filament is probably responsible for activating the reaction (by generating H atoms, by the ultraviolet radiation it emits, or both). Circumstantial evidence based on the overall composition of the deposit implies that x > 1.Warming the deposit to ambient temperatures results in decomposition of the AlH, with the regeneration of H2. As with gallane (56),however, this decomposition can be forestalled by chemical trapping with an excess of trimethylamine (136);warming then affords the known, relatively stable alane adduct (Me3NI2A1H3,identified by its infrared and 'H and 27AlNMR spectra (137).Although AIH, may thus be indicated as the hydride precursor, the inhomogeneous nature of the condensate means that disproportionation, as in a reaction such as (22), cannot be ruled out as a pathway to (Me3NI2A1H,: 3A1H2 + 4Me3N

-

2(Me3NI2A1H3+ Al.

(22)

Matrix isolation is a well-tried stratagem (60)for exploring more closely the reactions of metal atoms and other unsaturated species at low temperatures. For example, as noted earlier, it has witnessed the formation of the gallane molecules GaH,Cl,-, [n = 1 (93) or 2 (9411by their infrared spectra. Likewise the binary hydride molecules GaH, Ga2H2,and GaH, have been identified in cryogenic reactions engaging H2and Ga atoms or Gaz molecules (138).Such studies have also shown that matrix-isolated A1 atoms are inserted into H2 upon selective photoexcitation, with the formation of the species AIHz and hence, by thermal or photolytic dissociation, A1H (139).Under the action of broad-band irradiation and with a noble-gas matrix relatively rich in H2, it has now been established that planar, monomeric AlH3 is the

223

GALLIUM HYDRIDES

major product (136).Thus, irradiation of such a matrix for 10-40 min leads, as illustrated in Fig. 12, to the appearance and growth of three new infrared bands. With H, in an argon matrix these occur at 1882.7, 783.5, and 697.6 cm-'; with D2 each of these bands shifts to substantially lower energy, viz., 1882.7 + 1376.5, 783.5 + 568.4, and 697.6 + 513.9 cm-', thereby evincing large H(D) displacements for the relevant vibrational mode. Experiments with mixtures of H2 and D, show no less than 15 infrared absorptions attributable to different A1HnD3-, isotopomers. The identification of the product is strongly sanctioned by the results of MP2 ab initio calculations (see Table 111)(136).Hence it is possible to account for the doublet structure of the absorption near 1880 cm-' and the three absorptions occurring in the region 1350-1380 cm-' in the spectrum of a photolysed matrix containing both H, and D,; even more telling are the prediction and observation of just 10 distinct vibrational fundamentals in the region 500-800 cm-'. The

600

500

600

500

L I0

?

------Ll 0

0 /cm-l

FIG.12. IR spectra of argon matrices containing A1 atoms and H,or H2+ D,, showing the effects of broad-band photolysis: (a) A1 with H2 after photolysis for 20 min; (b) A1 with H2 + D2 in 1 : 1 proportions after photolysis for 40 min. The species A, B, C, and D correspond to AIH, , AlH,D, AIHD,, and AID,, respectively. X corresponds to a hydride and X' to the related deuteride impurity (reproduced with permission from Ref. 136).

224

DOWNS AND PULHAM

TABLE I11 VIBRATIONAL FREQUENCIES( V in cm-') FOR AIHB(A),AIHzD (B),AIHDz tc), AND AID3 (D) (a) OBSERVED FOR THE MOLECULES ISOLATED I N Ar MATRICES, (b) CALCULATED BY THE MP2 AB INITIO METHOD,AND (C) CALCULATED BY NORMAL COORDINATE ANALYSIS (n.c.a.) (136) Calculated

Species AlH3, A

AlHzD, B

AlHDz, C

AID,, D

Experiment, Ar matrix 1882.7 n.o.h 783.5 697.6 1882.7 1881.2 1367.3 780.6 651.8 645.6 1881.2 1377.9 1355.2 716.4 586.7 570.8 1376.5 n.o.h 568.4 513.9

Assignment

MP2 ab initio method (intensity)"

n.c.a.

2024.2 (268) 2021.0 (0) 830.7 (249) 737.6 (394) 2024.1 (271) 2022.1 (93) 1455.6 (99) 830.4 (2461 687.0 (170) 680.2 (335) 2023.0 (186) 1468.4 (155) 1442.6 (49) 761.6 (206) 617.4 (276) 601.9 (130) 1468.4 (155) 1429.7 (0) 601.2 (128) 547.4 (217)

1881.4 1880.6 782.5 697.8' 1881.3 1880.9 1366.8 782.2 649.3 643.4' 1881.0 1378.0 1355.6 717.4 584.0' 570.1 1378.0 1344.4 569.5 517.8'

' Intensities in km mol-', as obtained from the frequency calculation. For direct comparison with observed intensities (as in Fig. 12, for example), these values need to be scaled according to the probabilities of formation of the different AlH,D3_, isotopomers in matrices containing both Hz and Dz. *n.o., not observed. Fundamental is silent in IR absorption for an AIH, or AlD:, molecule with D3hsymmetry. Harmonic frequency for the out-of-plane deformation; anharmonic frequencies have been estimated for the other fundamentals (136).

spectra give us to believe that A1H is an intermediate, and the finding that a n equimolar mixture of H, and D, gives AlH,, AlH,D, AlHD, , and AlD, in highly nonstatistical proportions approaching 1: 1: 1: 1 favors the mechanism,

225

GALLIUM HYDRIDES

A1 + H,

-

A1

hu,

-

H+AI-H

I

AlH,

(23)

in which the ultimate step parallels the oxidative addition of H2 to GaCl (94).To date there are no signs of the dimer A12H,. Whether molecular alane has the thermal stability t o survive at temperatures appreciably greater than those of a solid noble-gas matrix and, if so, whether this approach is viable as a method of synthesis on the larger scale are other issues still to be settled. Altogether easier of access, more stable, and more tractable are complexes in which AlH3 is coordinated by one or more base molecules, e.g., (Me3N),AlH3(n = 1 or 2). These are of interest as hydride sources for hydroalumination of unsaturated substrates (10, 1401, as precursors to the hydrides of other metals, and, as already noted, in the chemical vapor deposition of aluminum metal in thin-film technology (21-26). The 1 : 1 trimethylamine adduct, Me3N-A1H3, is monomeric in the vapor but, in common with PhCH2Me2N.A1H3 and

dH,CH=CHCH2CH2fiMe.A1H3, takes the form of a dimer in the solid state (141).The dimer, as exemplified by [PhCH2Me2N.A1H312 in Fig. 13, consists of an A12H6unit with highly unsymmetrical Al-H...Al bridges. It can thus be viewed as a derivative of the elusive dialane, A12H6.Polydentate tertiary amine complexes of alane may be polymeric (1421, or they may incorporate cationic as well as anionic aluminum hydride moieties (143). Another recent development of aluminum hydride chemistry has been the preparation of a variety of aluminohydride complexes of transition metals (144).Here too a notable feature is the double hydrogen-bridged moiety, M(P-H)~A~, involved in the bonding of pairs of A1 atoms or in the attachment of A1 to the transition metal center. It remains to be seen whether the heavier Group 13 metals enter into the formation of similar compounds. C. INDIUM AND THALLIUM HYDRIDES (10) Such is the weakness of In-H and T1-H bonds that the chemistry of indium and thallium hydrides is unlikely to rival that of the corresponding aluminum and gallium compounds. This is certainly the inference to be drawn from the thermal instability of the known compounds LiInH, and LiTlH, , both of which decompose rapidly at 273 K (similar

226

DOWNS AND PULHAM

C

C

FIG.13. Projection of [PhCH2Me2N.AlH3l2in the crystalline solid with 20% thermal ellipsoids for non-hydrogen atoms and arbitrary radii for hydrogen atoms where shown (adapted with permission from Ref. 141 ; copyright 1991, American Chemical Society).

decomposition of LiAlH, and LiGaH, occurs at ca. 400 and 320 K, respectively). Despite mass spectrometric evidence of InH, in the gas phase (181,earlier claims to the synthesis of the binary hydrides [InH31, and [T1H3], in the condensed phases have yet to be substantiated. Until recently the only reasonably well-characterized compounds with In-H bonds were a few thermally unstable alkylindates of the types M[R21nH2]and M[RInH,] (M = alkali metal; R = Me, Et, or Me3SiCH2)(43). With the aid of the large tris(trimethylsily1)methyl (Tsi)group, however, it has been possible to synthesize an alkylhydroindate that is sufficiently robust to permit characterization in the solid

GALLIUM HYDRIDES

227

state and in solution (145).The crystal structure of the solid complex [Li(thf )21[(Tsi)zIn,H,](thf, tetrahydrofuran) implies the presence of a complex anion [(Tsi)H,In(p-H)InHz(Tsi)l-, the In atoms of which are linked to the Li+ ion via two of the H atoms. With its single hydrogen bridge the anion has obvious affinities with [Me,Al(p-H)AlMe,]-, which is known in the form of its sodium salt (146). Earlier records of indium or thallium tetrahydroborates cite the preparation of an unstable adduct In(BH,),.thf by the reaction of trimethylindium with diborane in tetrahydrofuran solution. By exploiting solvent-free conditions, we have found that trimethylindium can be made to react with gaseous diborane at ambient temperatures to give a white crystalline solid, MeJnBH, ,which has been characterized by elemental analysis and by its vibrational and NMR spectra (147). The compound is long-lived at room temperature and can be vaporized without decomposition at ca. 350 K. The infrared spectrum of the vapor isolated in a solid nitrogen matrix at ca. 20 K indicates that the predominant species are molecules of the type MezIn(p-H)2BHz(101, incorporating

(10)

a bidentate BH, group, and therefore analogous to the corresponding gallium compound (82,85).At the same time the vibrational properties hint a t a formulation that comes closer to an ion-pair Me,In+BH,- than to a molecule with a significant In-H bonding interaction. Studies along these lines are continuing as part of a wider search for tractable indium hydrides. Whatever the odds may be, there is a n appealing congruity about this development, for was not Me,GaBH, (see Section 1V.A) the first step in the hunting of the gallium hydrides? ACKNOWLEDGMENTS It is not just a moral duty but a pleasure to acknowledge the many contributions of our colleagues over the years. Their names are cited in the appropriate references, but we owe a particular debt to Drs. P. D. P. Thomas, M. T. Barlow, C. J. Dain, P. L. Baxter, and M. J. Goode (who was responsible for carrying out the first experiments leading to the isolation and identification of monochlorogallane and gallane) and also to Dr. H. E. Robertson and Professors I. M. Mills, D. W. H. Rankin, and H. Schnockel for the expert technical assistance and advice that they have given so generously. To the SERC and the Royal Society we are indebted for financial support.

228

DOWNS AND PULHAM

REFERENCES 1. “Gmelin Handbook of Inorganic Chemistry,” 8th edn., “Boron Compounds,” Syst. No. 13, Parts 14,18, and 20,1977-1979; 2nd Supplement, Vol. 1,1983; 3rd Supplement, Vol. 1, 1987; Springer-Verlag, Berlin and Heidelberg. 2 . “Supplement to Mellor’s Comprehensive Treatise on Inorganic and Theoretical Chemistry,” Vol. V, “Boron.” Part BI: “Boron-Hydrogen Compounds.” Longman, London, 1981. 3. Greenwood, N. N., and Thomas, B. S., “The Chemistry of Boron.” Pergamon Press, Oxford, 1975. 4 . Muetterties, E. L. (Ed.), “Boron Hydride Chemistry.” Academic Press, New York, 1975. 5. Liebman, J . F., Greenberg, A., and Williams, R. E. (Eds.),“Advances in Boron and the Boranes.” VCH, Weinheim, 1988. 6. Housecroft, C. E., “Boranes and Metalloboranes.” Ellis Horwood, Chichester, 1990. 7. Olah, G. A., Wade, K., and Williams, R. E. (Eds.), “Electron Deficient Boron and Carbon Clusters.” Wiley, New York, 1991. 8. Greenwood, N. N., and Earnshaw, A., “Chemistry of the Elements.” Pergamon, Oxford, 1984. 9. Greenwood, N. N., Chem. SOC.Rev. 21,49 (1992). 10. Downs, A. J . (Ed.), “Chemistry of Aluminium, Gallium, Indium and Thallium.” Blackie, Glasgow, 1993. 11. Wade, K., and Banister, A. J., “The Chemistry of Aluminium, Gallium, Indium and Thallium.” Pergamon, Oxford, 1975. 12. Wiberg, E., and Amberger, E., “Hydrides of the Elements of Main Groups I-IV.” Elsevier, Amsterdam, 1971. 13. Turley, J. W., and Rinn, H. W., Inorg. Chem. 8, 18 (1969). 14. Huber, K. P., and Herzberg, G., “Molecular Spectra and Molecular Structure. IV. Constants of Diatomic Molecules.” van Nostrand Reinhold, New York, 1979. 15. Urban, R.-D., Birk, H., Polomsky, P., and Jones, H., J . Chem. Phys. 94,2523 (1991). 16. Garton, W. R. S., Proc. Phys. SOC.M A , 509 (1951); Ginter, M. L., and Innes, K. K., J. Mol. Spectrosc. 7,64 (1961); Urban, R.-D., Magg, U., and Jones, H., Chem. Phys. Lett. 154, 135 (1989). 17. Bahnmaier, A. H., Urban, R.-D., and Jones, H., Chem. Phys. Lett. 155,269 (1989). 18. Breisacher, P., and Siegel, B., J. Am. Chem. SOC. 87, 4255 (1965). 19. Carroll, L., “The Hunting of the Snark.” Macmillan, London, 1876. 20. Cotton, F. A., and Wilkinson, G., “Advanced Inorganic Chemistry,” 5th ed. WileyInterscience, New York, 1988. 21. Gladfelter, W. L., Boyd, D. C., and Jensen, K. F., Chem. Muter. 1, 339 (1989). 22. Baum, T. H., Larson, C. E., and Jackson, R. L., Appl. Phys. Lett. 55, 1264 (1989). 23. Beach, D. B., Blum, S. E., and LeGoues, F. K., J . Vuc.Sci. Technol. A 7,3117 (1989). 24. Abernathy, C. R., Jordan, A. S., Pearton, S. J., Hobson, W. S., Bohling, D. A., and Muhr, G. T., Appl. Phys. Lett. 56,2654 (1990). 25. Wee, A. T. S., Murrell, A. J., Singh, N. K., O’Hare, D., and Foord, J. S., J . Chem. SOC.,Chem. Commun., 11 (1990). 26. Elms, F. M., Lamb, R. N., Pigram, P. J., Gardiner, M. G., Wood, B. J., and Raston, C. L., J . Chem. SOC.,Chem. Commun. 1423 (1992). 27. Wiberg, E., and Johannsen, T., Naturwissenschuften 29,320 (1941). 28. Wiberg, E., and Johannsen, T., Chemie 55, 38 (1942).

GALLIUM HYDRIDES

229

29. Wiberg, E., Johannsen, T., and Stecher, O., 2. Anorg. Allg. Chem. 251, 114 (1943). 30. Wiberg, E., and Schmidt, M., 2. Naturforsch. B 7, 577 (1952). 31. Shriver, D. F., Parry, R. W., Greenwood, N. N., Storr, A., and Wallbridge, M. G. H., Inorg. Chem. 2, 867 (1963). 32. Greenwood, N. N., in “New Pathways in Inorganic Chemistry” (E. A. V. Ebsworth, A. G. Maddock, and A. G. Sharpe, Eds.), p. 37. Cambridge Univ. Press, Cambridge, 1968. 33. Greenwood, N. N., and Wallbridge, M. G. H., J . Chem. SOC.3912 (1963). 34. Shirk, A. E., and Shirk, J . S., Inorg. Chem. 22, 72 (1983). 35. Baxter, P. L, D. Phil. thesis, University of Oxford, 1984. 36. Goode, M. J., D. Phil. thesis, University of Oxford, 1987. 37. Pulham, C. R., Part I1 thesis, University of Oxford, 1987; D. Phil. thesis, University of Oxford, 1991; unpublished results. 38. Greenwood, N. N., and Storr, A., J . Chem. SOC.3426 (1965). 39. Goode, M. J.,Downs, A. J., Pulham, C. R., Rankin, D. W. H., and Robertson, H. E., J . Chem. SOC.,Chem. Commun., 768 (1988). 40. Goode, M. J., Brain,P. T.,Downs, A. J.,Pulham, C. R., Rankin,D. W. H.,Robertson, H . E., and Souter, P. F., unpublished results. 41. Schmidbaur, H., Findeiss, W., and Gast, E., Angew. Chem., Int. Ed. Engl. 4, 152 (19651. 42. Schmidbaur, H., and Klein, H.-F., Chem. Ber. 100, 1129 (1967). 43. Tuck, D. G., in “Comprehensive Organometallic Chemistry” (G. Wilkinson, F. G. A. Stone, and E. W. Abel, Eds.), Vol. 1, p. 683. Pergamon, Oxford, 1982. 44. Eisch, J. J., J . A m . Chem. SOC.84, 3830 (1962). 45. Gavrilenko, V. V., Kolesov, V. S., and Zakharkin, L. I., J . Gen. Chem. USSR (Engl. Transl.)44, 1867 (1974). 46. Markova, V. V., Kormer, V. A., and Petrov, A. A,, J . Gen. Chem. USSR (Engl. Transl.)37, 1662 (1967). 47. Baxter, P. L., Downs, A. J., Goode, M. J., Rankin, D. W. H., and Robertson, H. E., J . Chem. SOC.,Chem. Commun.,805 (1986); J . Chem. SOC.,Dalton Trans.,2873 (1990). 48. Downs, A. J., and Thomas, P. D. P., J. Chem. SOC.,Chem. Commun., 825 (1976). 49. Barlow, M. T., Dain, C. J., Downs, A. J., Laurenson, G. S., and Rankin, D. W. H., J . Chem. SOC.,Dalton Trans., 597 (1982). 50. Downs, A. J., Harman, L. A., Pulham, C. R., Rankin, D. W. H., Robertson, H. E., Souter, P. F., and Thomas, P. D. P., unpublished results. 51. Baxter, P. L., Downs, A. J., and Rankin, D. W. H., J . Chem. SOC.,Dalton Trans., 1755 (1984). 52. Baxter, P. L., Downs, A. J., Rankin, D. W. H., and Robertson, H. E., J . Chem. SOC., Dalton Trans., 807 (1985). 53. Shirk, A. E., and Shriver, D. F., J . A m . Chem. SOC.95,5904 (1973). 54. Durig, J. R., Chatterjee, K. K., Li, Y. S., Jalilian, M., Zozulin, A. J., and Odom, J. D., J . Chem. Phys. 73, 2 1 (1980). 55. Odom, J. D., Chatterjee, K. K., and Durig, J. R., J . Phys. Chem. 84, 1843 (1980). 56. Pulham, C. R., Downs, A. J., Goode, M. J., Rankin, D. W. H., and Robertson, H. E., J . Am. Chem. SOC.113, 5149 (1991). 57. Miller, T. A., Science 223, 545 (1984). 58. Foster, S. C., Kennedy, R. A., and Miller, T. A., in “Frontiers of Laser Spectroscopy of Gases” (A. C. P. Alves, J. M. Brown, and J . M. Hollas, Eds.), NATO AS1 Series C, Vol. 234, p. 421. Kluwer Academic Publishers, Dordrecht, 1987.

230

DOWNS AND PULHAM

59. Smalley, R. E., Wharton, L., and Levy, D. H., Acc. Chem. Res. 10, 139 (1977). 60. Almond, M. J., and Downs, A. J., “Advances in Spectroscopy,” Vol. 17. Wiley, Chichester, 1989. 61. Downs, A. J., Goode, M. J., and Pulham, C. R., J . A m . Chem. SOC.111,1936 (1989). 62. Poll, W., Pawelke, G., Mootz, D., and Appelman, E. H., Angew. Chem., Int. Ed. Engl. 27, 392 (1988). 63. Pulham, C. R., Downs, A. J., Rankin, D. W. H., and Robertson, H. E., J . Chem. SOC.,Chem. Commun., 1520 (1990); J . Chem. SOC.,Dalton Trans., 1509 (1992). 64. Downs, A. J., Egdell, R. G., Orchard, A. F., andThomas, P. D. P., J . Chem. Soc., Dalton Trans., 1755 (1978). 65. Pulham, C. R., Brain, P. T., Downs, A. J., Rankin, D. W. H., and Robertson, H. E., J . Chem. Soc., Chem. Commun., 177 (1990). 66. Grady, A. S., Mapplebeck, A. L., Russell, D. K., and Taylorson, M. G., J . Chem. Soc., Chem. Commun., 929 (1990). 67. Grady, A. S., Markwell, R. D., and Russell, D. K., J . Chem. Soc., Chem. Commun., 14 (1991). 68. Downs, A. J., Duckworth, D., Machell, J. C., and Pulham, C. R., Polyhedron 11, 1295 (1992). 69. See, for example, Grady, A. S., Puntambekar, S. G . ,and Russell, D. K., Spectrochim. Acta A 47, 47 (1991). 70. Marks, T. J., Kennelly, W. J., Kolb, J. R., and Shimp, L. A,, Inorg. Chem. 11, 2540 (1972). 71. Marks, T. J . , and Kolb, J. R., Chem. Rev. 77, 263 (1977). 72. Howard, M. W., Jayasooriya, U. A., Kettle, S. F. A,, Powell, D. B., and Sheppard, N., J . Chem. SOC.,Chem. Commun., 18 (1979). 73. Smith, D. F., in “Noble-Gas Compounds” (H. H. Hyman, Ed.), p. 295. Univ. of Chicago Press, Chicago, 1963. 74. Hinton, J . F., and Briggs, R. W., in “NMR and the Periodic Table” (R. K. Harris and B. E. Mann, Eds.), p. 279. Academic Press, London, 1978. 75. Akitt, J. W., in “Multinuclear NMR” (J. Mason, Ed.), p. 259. Plenum, New York/ London, 1987. 76. There are exceptions involving coordination of the tetrahydroborate ligand in certain transition-metal complexes. See, for example, Wink, D. J., and Cooper, N. J., J . Chem. SOC.,Dalton Trans., 1257 (1984). 77. Kennedy, J. D., Prog. Znorg. Chem. 32, 519 (1984). 78. Gaines, D. F., and Hildebrandt, S. J., Inorg. Chem. 17, 794 (1978); Gaines, D. F., Walsh, J. L., Morris, J. H., and Hillenbrand, D. F., Inorg. Chem. 17, 1516 (1978). 79. Ebsworth, E. A. V., Rankin, D. W. H., and Cradock, S., “Structural Methods in Inorganic Chemistry,” 2nd ed. Blackwell Scientific Publications, Oxford, 1991. 80. Rankin, D. W. H., Chem. Br. 18,426 (1982). 81. Schlesinger, H. I., Brown, H. C., and Schaeffer, G. W., J . A m . Chem. SOC.65, 1786 (1943). 82. Downs, A. J., and Thomas, P. D. P., J . Chem. SOC.,Dalton Trans., 809 (1978). 83. Schlesinger, H. I., Brown, H. C., and Hyde, E. K., J . A m . Chem. SOC.75,209 (1953). 84. Reid, W. E., Jr., Bish, J. M., and Brenner, A., J . Electrochem. SOC.104, 21 (1957). 85. Barlow, M. T., Downs, A. J., Thomas, P. D. P., and Rankin, D. W. H., J . Chem. Soc., Dalton Trans., 1793 (1979). 86. Duke, B. J., Liang, C., and Schaefer, H. F., 111, J . A m . Chem. SOC.113,2884 (1991). 87. Wartik, T., and Schlesinger, H. I., J . A m . Chem. SOC. 75,835 (1953).

GALLIUM HYDRIDES

23 1

88. Almenningen, A,, Anderson, G. A., Forgaard, F. R., and Haaland, A., Actu Chem. Scand. 26,2315 (1972). 89. Brain, P. T., Downs, A. J., Jones, L. A., Pulham, C. R., Rankin, D. W. H., and Robertson, H. E., unpublished results; Jones, L. A., D. Phil. thesis, University of Oxford, 1993.Compare also the structure of [Me2AIF14:Gundersen, G., Haugen, T., and Haaland, A., J . Orgunomet. Chem. 54,77 (1973). 90. Petrov, V. M., Giricheva, N. I., Girichev, G. V.,Titov, V. A,, and Chusova, T. P., Zh. Strukt. Khirn. 32,56 (1991). 91. Lammertsma, K., and Leszczynski, J.,J . Chern. Soc., Chem. Commun., 1005 (1989). 92. Duke, B. J., Hamilton, T. P., and Schaefer, H. F., 111,Inorg. Chem. 30,4225(1991). 93. Koppe, R., Tacke, M., and Schnockel, H., Z . Anorg. Allg. Chem. 605,35 (1991). 94. Koppe, R., and Schnockel, H., J . Chem. Soc., Dalton Trans., 3393 (1992). 95. Tacke, M., Kreienkamp, H., Plaggenborg, L., and Schnockel, H., Z. Anorg. Allg. Chem. 604,35 (1991). 96. Loos, D., Schnockel, H., Gauss, J.,and Schneider, U., Angew. Chem.,Znt. Ed. Engl.

31, 1362 (1992). 97. s, strong; w, weak; br, broad; v, very. 98. Tacke, M., and Schnockel, H., Znorg. Chern. 28,2895 (1989). 99. Schnockel, H., Leimkuhler, M., Lotz, R., and Mattes, R., Angew. Chem., Znt. E d .

Engl. 25,921 (1986). 100. Ahlrichs, R., Haser, M., Schnockel, H., and Tacke, M., Chem. Phys. Lett. 154,

104 (1989). 101. Dohmeier, C., Mattes, R., and Schnockel, H., J . Chem. Soc., Chern. Commun.,

358 (1990). 102. Dohmeier, C., Robl, C., Tacke, M., and Schnockel, H., Angew. Chern.,Znt. Ed. Engl.

30,564 (1991). 103. Pulham, C. R., Greene, T. M., Mills, I. M., and Downs, A. J., unpublished results. 104. Duncan, J. L., McKean, D. C., Torto, I., and Nivellini, G. D., J . Mol. Spectrosc. 85,

16 (1981). 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116.

Liang, C., Davy, R. D., and Schaefer, H. F., 111, Chem. Phys. Lett. 159,393 (1989). Lammertsma, K.,and Leszczynski, J., J . Phys. Chem. 94,2806 (1990). Duke, B. J.,J . Mol. Struct. (THEOCHEM) 208, 197 (1990). Trinquier, G., and Malrieu, J.-P.,J . A m . Chem. SOC.113,8634 (1991). Shen, M.,and Schaefer, H. F., 111, J . Chem. Phys. 96,2868 (1992). Bartell, L. S., and Carroll, B. L., J . Chem. Phys. 42, 1135 (1965). Barone, V.,Minichino, C., Lelj, F., and Russo, N., J . Comput. Chem. 9,518(1988). van der Woerd, M. J., Lammertsma, K., Duke, B. J., and Schaefer, H. F., 111, J. Chem. Phys. 95, 1160 (1991). Pulham, C. R., and Downs, A. J., unpublished results. Pulham, C. R., Knight, K., and Downs, A. J., unpublished results. Piocos, E. A,, and Auk, B. S., J . Am. Chem. SOC.111,8978 (1989). Dobbs, K. D., Trachtman, M., Bock, C. W., and Cowley, A. H., J . Phys. Chem. 94,

5210 (1990). Gee, P. E., and Hicks, R. F., J . Vuc. Sci. Technol. A 10,892 (1992). Qi, H., Gee, P. E., and Hicks, R. F., Phys. Rev. Lett., 72,250 (1994). Charbonnel, M., and Belin, C., J . Solid State Chern. 67,210 (1987). Hiller, W., Klinkhammer, K.-W., Uhl, W., and Wagner, J., Angew. Chem., Int. Ed. Engl. 30, 179 (1991). 121. Thomas, P. D. P., D. Phil. thesis, University of Oxford, 1977.

117. 118. 119. 120.

232 122. 123. 124. 125.

DOWNS AND PULHAM

Ruff, J. K., Inorg. Chem. 2, 515 (1963). Oddy, P. R., and Wallbridge, M. G. H., J . Chem. SOC.,Dalton Trans., 869 (1976). Borlin, J., and Gaines, D. F., J . Am. Chem. SOC.94, 1367 (1972). Dain, C. J., Downs, A. J., and Rankin, D. W. H., J . Chem. Soc., Dalton Trans., 2465 (1981).

126. Dahl, A. J., and Taylor, R. C., Inorg. Chem. 10,2508 (1971). 127. Norman, A. D., and Schaeffer, R., J . Phys. Chem. 70, 1662 (1966); Marynick, D., and Onak, T., J . Chem. SOC.A 1160 (1970); Beall, H., Bushweller, C. H., Dewkett, W. J., and Grace, M., J . Am. Chem. SOC.92,3484 (1970). 128. Dain, C. J . , Downs, A. J . , Laurenson, G. S., and Rankin, D. W. H., J . Chem. Soc., Dalton Trans., 472 (1981). 129. Lippard, S. J., and Melmed, K. M., Inorg. Chem. 8, 2755 (1969). 130. Grebenik, P. D., Leach, J . B., Pounds, J. M., Green, M. L. H., and Mountford, P., J . Organornet. Chem. 382, C 1 (1990). 131. Duke, B. J., and Schaefer, H. F., 111, J . Chem. SOC.,Chem. Commun., 123 (1991). 132. Finholt, A. E., Bond, A. C., Jr., Wilzbach, K. E., and Schlesinger, H. I., J . Am. Chem. SOC.69, 2692 (1947). 133. Breisacher, P., and Siegel, B., J . Am. Chem. SOC.86, 5053 (1964). 134. Siegel, B., J . Am. Chem. SOC.82, 1535 (1960). 135. Hara, M., Domen, K., Onishi, T., and Nozoye, H., J . Phys. Chem. 95, 6 (1991). 136. Kurth, F. A,, Eberlein, R. A., Schnockel, H., Downs, A. J., and Pulham, C. R., J . Chem. SOC.,Chem. Commun., 1302 (1993). 137. Fraser, G. W., Greenwood, N. N., and Straughan, B. P., J . Chem. SOC.3742 (1963); Heitsch, C. W., and Kniseley, R. N., Spectrochim.Acta 19,1385 (1963);Mastryukov, V. S., Golubinskii, A. V., and Vilkov, L. V., J . Struct. Chem. (Engl. Transl.) 20, 788 (1979); Whitaker, T. J., Part I1 thesis, University of Oxford, 1990; Moore, C. B., Part I1 thesis, University of Oxford, 1991. 138. Xiao, Z. L., Hauge, R. H., and Margrave, J . L., Inorg. Chem. 32, 642 (1993). 139. Parnis, J. M., and Ozin, G. A., J . Phys. Chem. 93, 1215, 1220 (1989). 140. Cloke, F . G. N., Dalby, C. I., Henderson, M. J., Hitchcock, P. B., Kennard, C. H. L., Lamb, R. N., and Raston, C. L., J . Chem. SOC.,Chem. Commun., 1394 (1990). 141. Atwood, J . L., Bennett, F. R., Elms, F. M., Jones, C., Raston, C. L., and Robinson, K. D., J . Am. Chem. SOC.113,8183 (1991). 142. Atwood, J . L., Bennett, F. R., Jones, C., Koutsantonis, G. A,, Raston, C. L., and Robinson, K. D., J . Chem. SOC.,Chem. Commun., 541 (1992). 143. Atwood, J . L., Robinson, K. D., Jones, C., and Raston, C. L., J . Chem. SOC.,Chem. Commun., 1697 (1991). 144. Bulychev, B. M., Polyhedron 9,387 (1990). 145. Avent, A. G., Eaborn, C., Hitchcock, P. B., Smith, J. D., and Sullivan, A. C., J . Chem. SOC.,Chem. Commun., 988 (1986). 146. Atwood, J. L., Hrncir, D. C., Rogers, R. D., and Howard, J. A. K., J . Am. Chem. SOC.103, 6787 (1981). 147. Pulham, C . R., and Downs, A. J . , unpublished results.

ADVANCES IN INORGANIC CHEMISTRY, VOL. 41

THE STRUCTURES OF THE GROUP 15 ELEMENT(II1) HALIDES AND HALOGENOANIONS GEORGE A. FISHER and NICHOLAS C. NORMAN Department of Chemistry, The University of Newcastle upon Tyne, Newcastle upon Tyne. NE1 7RU, United Kingdom

I. Introduction

This chapter deals with the solid-state structures of the element(II1) halides and halogenoanions of arsenic, antimony, and bismuth with brief mention of some compounds of phosphorus where relevant. A 233 Copyright 0 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.

234

FISHER AND NORMAN

number of related topics will not be considered here; these include lower oxidation state halides, or subhalides, element(V1 compounds, and mixed oxidation state species. It does not attempt to be fully comprehensive but instead highlights important structural classes and references a number of representative examples. Coordination complexes involving element halides or halogenoanions and neutral twoelectron donor ligands L (or multidentate Ln), although a large and rapidly expanding field, and a number of cationic species such as [Sb2F,I2+and [Sb6Fl3l5+(1)will also not be considered. This area has not been reviewed extensively before although Davidovich and Buslaev (2) have considered the coordination chemistry of bismuth in a general way, in which the structures of halides and halogenoanions are featured. Sawyer and Gillespie ( 3 ) have commented previously upon the stereochemistry of antimony(II1) halides (mostly fluorides). The element trihalides will be considered first, followed by the many classes of anionic structures that are known. Finally, a section on bonding and on the various ideas that have been used to rationalize some of the observed trends is included. II. Element Trihalides, EX,

The structures of the element trihalides EX, are covered in a number of textbooks on structural inorganic chemistry (4,5 ) , and these will not be discussed in great detail here. It is, however, worth mentioning some of the salient structural features. In most cases, a molecular trigonal pyramidal EX, unit consistent with VSEPR theory predictions is readily apparent in the solid-state structure, although there are usually a number of fairly short intermolecular contacts or secondary bonds present. A general description of the structures as molecularly covalent but as having a tendency toward macromolecular or polymeric networks is therefore reasonable. Only in the case of the fluorides is an ionic model appropriate. In SbF, ( 6 )the antimony atom is bonded to three fluorines (av. Sb-F, 1.92 A) in a trigonal pyramidal arrangement with three other fluorines at greater distances (av. 2.61 A) approximately trans to the primary Sb-F bonds, such that the overall coordination geometry around the antimony center is that of a distorted octahedron. In bismuth trifluoride, a primary BiF, trigonal pyramid is much less evident and the bismuth atom has a larger coordination number with eight nearest neighbors at distances ranging from 2.233 to 2.503 A in a bicapped

GROUP 15 ELEMENT(II1) HALIDES AND HALOGENOANIONS

235

trigonal prismatic geometry; a ninth and more distant fluorine (3.105 A) is present in the remaining capping site (7). The solid-state structures of SbC1, (8) and P-SbBr, (9) are isomorphous with BiF, , although in both cases the primary SbX, pyramid is clearly discernable. Thus in SbC1, there are three short Sb-C1 bonds (av., 2.359 A), with five longer bonds ranging from 3.457 to 3.736 A, whereas in P-SbBr, the three short Sb-Br bonds average 2.49 A, with the longer distances at 3.6 A and greater. The structure of BiCI, (10) is very similar (although not strictly isomorphous) with three short Bi-C1 bonds (av., 2.500 A) and five longer distances of 3.216 to 3.450 A, as shown in Fig. 1. The compound a-SbBr, (11) is isomorphous with AsBr, (121,and in both cases the EX, pyramid is also evident, although, unlike the structures mentioned above, there are only three short intermolecular contacts. For a-SbBr, the primary and secondary Sb-Br bonds average 2.50 and 3.75 A, respectively, whereas in AsBr, the corresponding As-Br distances are 2.36 and 3.77 A. Bismuth tribromide also exists as two modifications (13).In a-BiBr, the bismuth atom is bonded to three bromines (av., 2.663 A) with a trigonal pyramidal coordination geometry, with three longer secondary bonds (av., 3.316A) approximately trans to the primary bonds, as shown in Fig. 2; one phase of SbI, (14) is isomorphous with corresponding average distances of 2.765 and 3.675 A. In P-BiBr, a trigonal pyra-

FIG.1. Part of the structure of BiC13, showing the three short and five long Bi-Cl bonds about the bismuth center as filled and open bonds, respectively.

236

FISHER AND NORMAN

FIG.2. Part of the structure of u-BiBr3showing the three short and three long Bi-Br bonds as filled and open bonds, respectively.

midal BiBr, unit is not evident and the bismuth resides in a fairly regular octahedron of bromines for which the average Bi-Br distance is 2.81 A, intermediate between the primary and secondary bond distances in a-BiBr, . Octahedral coordination is also encountered in the structures of a second phase of SbI, (15),which is isomorphous with A d 3 (16)and BiI, (151, and these three structures provide an interesting comparison in terms of the relative lengths of the primary and secondary bonds. Thus the structures can be viewed as comprising a hexagonally close-packed array of iodines in which 4 of the octahedral holes (strictly 3 of the octahedral holes in alternate layers) are occupied by As, Sb, or Bi. In AsI, there are three short As-I bonds (2.591 A) and three longer bonds (3.467 A), whereas in BiI, all six Bi-I distances are equal at 3.1 A; in Sb13 the situation is intermediate, with short and long Sb-I bonds of 2.868 and 3.32 A, respectively. Clearly there are a number of interesting trends that are apparent here and that require explanation. Table I lists the approximate differences between the primary and the secondary E-X bond lengths (A) for the element trihalides EX3; common superscripts reflect isomorphous (or nearly so) structures for which a comparison is particularly appropriate. Before proceeding, however, the terms primary bond and secondary bond should be defined more thoroughly. This matter, and the concept

237

GROUP 15 ELEMENT(II1) HALIDES AND HALOGENOANIONS

TABLE I APPROXIMATE DIFFERENCES(TO THE NEAREST0.05 A) BETWEEN THE LENGTHSOF THE PRIMARY AND SECONDARY E-X BONDS(A) FOR THE ELEMENT TRIHALIDES EX,"

(A)

E-F

A

Sb-F Bi-F

0.70

O

-*

E-CI

A (A)

E-Br

Sb-CI Bi-CI

1.25* 0.85*

As-Br Sb-Br Bi-Br

A

(A)

1.40q 1.257 (a),1.10* ( p ) 0.65# (a),0 (PI

E-I AS-I Sb-I Bi-I

A

(A)

0.906 0.451, 0.91# OB

Common superscripts indicate isomorphous or nearly isomorphous structures

of secondary bonding in general, has been addressed by Alcock (171, wherein a primary bond is taken as a normal covalent bond with a bond length within usual ranges, whereas a secondary bond is a significantly longer interaction, although considerably less than the sum of the van der Waals radii for the two elements concerned. Moreover, it is usually observed that secondary bonds lie approximately trans to primary bonds and that there is a strong correlation between the trans-related primary and secondary bond distances such that, as the secondary distance gets shorter, the primary distance gets longer; when the two distances become equal (or nearly so) the distinction between primary and secondary is lost; i.e., A = 0. This latter feature, or trans effect as it is sometimes called, has been commented upon in detail by Alcock (17, 18) and Sheldrick (19) and is consistent with the idea that the acceptor orbitals on the EX, unit, which are responsible for the presence of secondary bonding or, more generally, element center Lewis acidity, are the E-X cr*-orbitals(as opposed to a more conventional explanation based on the use of vacant d orbitals), a matter that has been discussed in some detail and which will not be reiterated here (17,201. Similar trans effect correlations have also been observed for related selenium and tellurium halide complexes by Knop and coworkers and by Krebs and Ahlers (19);a number of structural similarities between group 16 halides and halogenoanions are evident and these are mentioned where appropriate in the following section. One thing that is immediately clear from Table I is that for a given halide there is a strong tendency for the difference between the primary and the secondary bond lengths to decrease on moving from As to Sb to Bi, this being particularly well illustrated in the isomorphous series of EI, structures. Furthermore, for a given group 15 element, there is tendency for A to decrease on moving to heavier halides as is clear from the structures of BiCl, , BiBr, , and BiI, , although the fluorides are an obvious exception here. This last point is most likely a result of a large E-F electro-

238

FISHER AND NORMAN

negativity difference (two to three times greater than that between other bond pairs in this table using modified Pauling electronegativities), and the structures are probably best considered as ionic, for which equal or nearly equal interionic distances are not unexpected; covalent bonding and hence a discussion of primary and secondary bonds are therefore not really appropriate here. For all the other cases in which the electronegativity differences are much smaller, covalent bonding is appreciable and this is reflected in the observed bond localization in many of the structures, i.e., the appearance of a definite EX, pyramid. The remaining trend, therefore, is that, in situations in which covalent bonding is important, i.e., excluding the fluorides, the secondary bonding interactions become more pronounced on moving to both heavier group 15 elements and to heavier halides and therefore that A approaches zero. In BiI, and p-BiBr, all Bi-I and Bi-Br distances are equal and the limit of A = 0 is reached. This is, in fact, quite a general observation in that secondary bonding interactions tend to increase (as defined by a decreasing A) as the electronegativity of the elements decrease and, importantly, as the element size increases (17, 28,201. In BiI,, therefore, it is an equivalence of primary and secondary bonding interactions that results in a regular octahedral coordination of six iodines around the bismuth center; the structure is not this way because it is ionic, as is sometimes stated, because the electronegativity difference between bismuth and iodine is far too small to result in ionic bonding. A complementary explanation for BiI, and P-BiBr, is that, as the element electronegativities become similar and fall within the range typical of metalloids, metalloid-type structures are encountered, i.e., polymeric or macromolecular solid-state structures with equal bond distances and a delocalized but covalent bonding scheme. Another feature, and one that will be met again in many of the structures of the anions to follow, is that, in P-BiBr, and BiI, ,the lone pair of electrons associated with the bismuth(II1) center is stereochemically inactive. A number of these points will be returned to in Section IV. 111. Element(lll) Halogenoanions

The element halogenoanions are a structurally diverse group of compounds and will be considered in the following sections according to common structural types. There is no unique way in which to do this, but, throughout, the relationships between the various forms will be highlighted.

GROUP 15 ELEMENT(II1) HALIDES AND HALOGENOANIONS

239

A. [EX41 The simplest species results from formal addition of one halide anion t o one EX, molecule to give the [EX,]- anion. Such species usually dimerize or polymerize, but a genuine example of a discrete mononuclear species is found in the structure of [Et,N][PCl,] (1) (21). The structure is shown in Fig. 3 and the coordination geometry around the phosphorus center may be adequately described as an equatorially vacant, trigonal bipyramid. Such a structure is predicted by VSEPR rules for an AX,E (E = lone pair here) species, but it is worth commenting on the inequivalence of the axial P-C1 bond lengths. The equatorial P-C1 distances average 2.046 A, whereas the distances to the axial chlorines are 2.118 and 2.850 A. This is fully consistent with the c*bonding model outlined above and will be considered again in Section IV. Thus, the structure of 1 may be viewed as an adduct of PC1, with C1- in which the chloride is approaching trans to C1(2),resulting in a slight lengthening of this bond. A second structure which should be considered here is that of [Pr4nNl[PBr,] (2) (221, shown in Fig. 4. The essential unit is similar to that in 1, although the axial P-Br bonds are much more similar in length, but there is clearly a degree of dimerization present as evidenced by the P-Br(1') distance of 3.460 A, which, although much longer than the corresponding primary distance [P-Br(l), 2.527 A], is certainly smaller than the sum of the van der Waals radii for P and Br (3.85 A). Note also that the Br(l'I-P-Br(4) angle of 167.9" places Br(1') almost trans to Br(l), and the geometry around the phosphorus center can

CK21

FIG.3. The structure of [PClJ in 1.

240

FISHER AND NORMAN

Brll'l

_ -

). V

FIG.4. The structure of [PBrJ in 2.

therefore be considered as a five-coordinate square-based pyramid. The fact that the bromide shows the beginnings of five-coordination a s opposed to the four-coordinate chloride complex is consistent with the increasing importance of secondary bonding for the heavier halides as mentioned in the previous section. B. [EzXslZThe structure of 2 is clearly on the borderline between being a monomer and being a dimer, and in this section compounds for which a dimeric formulation is unambiguous will be considered. Clearly there exists, in principle a t least, a continuum from a weakly bound to a more strongly bound to a symmetrically bound dimer, extremes of which are shown in the diagrams A and B, and any distinction that is made will necessarily be somewhat arbitrary.

A

B

With the above caveat in mind, some examples of complexes for which the dimeric formula [E2Xa12- is clearly appropriate are [PhMeNHzlz[AszC181 (3) (231, [Ph,Plz[AszBr81(41 (241, [Pr4"Nl2[As2Br8l (51 (241, [Pr4"Nlz[AszI81(6) (251, [Ph4Plz[SbzIal(7) (261, [ButNH,l,

GROUP 15 ELEMENTW) HALIDES AND HALOGENOANIONS

24 1

FIG.5. The structure of [As,Br,J- in 5.

[Sb,Cl,] (8)(271,and [BiC12(18-crown-6)][BizC18] (9) (28);a view of 5 is shown in Fig. 5. A couple of points deserve further comment. In all structures the element center is five-coordinate and the overall structure can be described as edge-shared, bis-square-based pyramidal with the apical halides anti rather than syn; moreover, the halide bridges are fairly symmetric and a representation as in A is appropriate. Furthermore, the structures of 5 and 6 are isomorphous with the phosphorus bromide complex 2, which provides a further illustration of the trend toward increasing secondary bonding for arsenic and antimony as opposed to phosphorus. This statement about secondary bonding can be restated as a trend toward increasing Lewis acidity on progressing to the heavier elements in group 15, and it is therefore interesting to note that the structures of 4 and 7 are not isomorphous and that in 7 there is a close contact between a phenyl ring of the [Ph,P]+ cation and the vacant site at the antimony center, which is absent in the structure of 4. A further example of anti [E,X,12- is [Pr,"Nlz[Sb2Cl81(10)(29) (not isomorphous with 2,5, or 6), in which there are two crystallographically independent anions, one of which has fairly asymmetric Sb-C1-Sb bridges (2.716 and 3.111 A as opposed to 2.794 and 2.961 A), but of more interest is the structure of [Bu4"Nl2[SbzCls1(11) (29),which has a syn arrangement of apical chlorides as shown in C;the Sb-C1-Sb bridges are quite symmetric in this case.

C

242

FISHER AND NORMAN

C. [{EX,Jnln Probably the most commonly encountered form of the [EX,]- unit is as a one-dimensional polymer, which may be represented as [{EX,}, 1"(although in the formulas to follow, [EX,] is simpler), in which the coordination about the group 15 element has increased to six and the geometry is approximately octahedral. Examples include [C,H,NHI[AsBr,I (12) (301,[C5H5NHI[As141 (13) (31), [C5H5NHI[SbC1,1(14) (321, [Mg(MeCN)61[SbC1412 (15) (331, [Et2NH21[BiC1,1(16) (341, [Fe(qC,H,),][BiCl,] (17) (351, [2-picoliniurn][BiBr4](18) (36),and [2-picolinium][BiI,] (19) (36).The structure of a section of the polymer in 17 is shown in Fig. 6 and, more generally, in D, from which it is apparent that adjacent E,(p-X), planes are perpendicular and that the terminal or nonbridging halides are always cis.

D

There is a considerable variation in the degree of bridge asymmetry present in these structures and it is often found that alternate bridging units have quite different degrees of asymmetry. For example, in 12, the relevant distances are 2.688 and 3.130 A for one As,(p-Br), unit and 2.690 and 3.129 A for the adjacent unit k e . , both are the same within experimental error). This is clearly a polymeric structure and the average difference between the primary and the secondary bond lengths is 0.440 A, a value that is fairly typical of most structures. In

FIG.6. Part of the polymeric structure of [{BiCI,},l"- in 17.

GROUP 15 ELEMENT(II1) HALIDES AND HALOGENOANIONS

243

15, however, the difference in asymmetry between alternate bridging units is considerable, the relevant pairs of distances being 2.8221 3.006 A (difference, 0.184) and 2.43313.580A (difference, 1.147).In this latter case, a description of the structure as a loosely bound polymer of dimers is more appropriate. Before leaving this section, mention should be made of a number of other structures that are also polymeric but that have rather different overall arrangements of the monomeric units. In Na[SbF,] (20) (37) a basic [SbF41-unit is clearly discernable, with Sb-F distances from 1.94 to 2.07 A and two secondary interactions a t 2.68 and 2.86 A, such that the overall structure can be described as two linked one-dimensional polymeric chains, as shown in E.

E

In K[SbF,l (21)(38) the [SbF41- unit is less distinct, there being three short Sb-F distances together with two longer and one longer still. Some mixed chloro-fluoro compounds are also known. These include K[SbClF,I (22) (39) and Cs[SbClF,] (23)(40),the former having a structure in which the antimony center is bonded to three fluorines (av., 1.95 A) with three much longer contacts to three chlorine atoms (av., 3.12 A) and with an arrangement around the antimony center as shown in F, which can be described as a CBudistorted octahedron. The structure is perhaps best thought of as comprising SbF, molecules together with chloride and potassium ions.

I

I

Cl

F

244

FISHER AND NORMAN

Finally, in the structure of [NH,][BiF,] (24) (411,for which an ionic description is appropriate, the bismuth center is coordinated by nine fluorines in a fashion similar t o that found in the structure of BiF, .

D. [EqX1614Anions of the formula [E4Xl6I4-can be thought of either as tetramers of [EX,]- or, in some cases, as dimers of [E2X,12-.Three types of structure have been characterized, the first of which corresponds to the description as a dimer of [E2Xa12-units. Examples include [HLI, [Sb4Br16] (L = 2-amino-1,3,4-thiadiazole)(25) (421, [Mg(MeCN),I, [Bi,ClI6] (26) (431, and [Fe(~-C5H,)21,[Bi,Br161 (27) ( 4 4 ) .A view of 27 is shown in Fig. 7 and a line diagram is shown in G . Any perspective from G has been omitted but one [E2XaI2-is highlighted to indicate how the structure can be viewed as a dimer of syn [E2XaI2-units. An alternative description, which will be useful in a later discussion of higher nuclearity clusters, is an [E2Xl0I4-edge-shared bioctahedral structure, itself discussed later, to which a neutral E,X, unit has been added, as is illustrated in H, or to which two face capping EX, units have been added, as shown in I. The overall structure is analogous to one form of TeI, . A second structural type is found in [Et,N],[Sb,C116] (28) (291,which is shown in Fig. 8. This structure can also be thought of in terms of a dimerization of syn [E2Xa12-units but in which the E-E vectors are

FIG.7. The structure of [Bi4Br,6]4-in 27.

GROUP 15 ELEMENT(II1) HALIDES AND HALOGENOANIONS

245

FIG.8. The structure of [Sb,C1,,14- in 28.

perpendicular rather than parallel, as found in the previous type, shown in Fig. 7 and G-I. In one case there are four p3-X groups as opposed to two p2- and two p3-X groups in the other. In the cubic form, the anion is isostructural with the neutral TeC1, and TeBr, structures. The third type for which a tetramer of [EX4]- units is an appropriate description is found in K4[Sb4F,,l (29) (451, shown in Fig. 9. This is

G

.$!

.,+. ..

.' .' -.. *.

I

:':

. I .

I

I I

!

H

I

246

FISHER AND NORMAN

F

FIG.9. The structure of [Sb4F,6]4-in 29.

a much more open structure in which the antimony atoms are fivecoordinate with a square-based pyramidal coordination geometry.

Turning now to a system that results from a formal addition of two halide anions to a EX, molecule, i.e., [EX512-,examples in which this unit occurs as a discrete species will be addressed first before turning to dimeric and polymeric forms. In fact, monomeric or essentially monomeric examples of [EX,I2- are not particularly common, two examples being [NH,],[SbCl,] (30)( 4 6 )and K2[SbC1,1 (31) (47),in which a squarebased pyramidal [EX,]’- unit is present, as shown in J. In both cases, the axial Sb-C1 distance is considerably shorter than the basal Sb-Cl lengths, although this is unsurprising because the axial bond may be viewed as a two-center, two-electron bond, whereas the basal plane is composed of two perpendicular, three-center, four-electron Cl-Sb-C1 units. These units may be symmetrical or unsymmetrical as mentioned before for primary and secondary bonds. In 30 all basal Sb-Cl lengths are about 2.62 A, whereas in 31 one unit has nearly equal lengths of 2.622 and 2.625 A and the other is somewhat asymmetric, with distances of 2.385 and 2.799 A. The closest interionic Sb-Cl contacts are 3.710 and 3.881 A.

J

GROUP 15 ELEMENT(II1) HALIDES AND HALOGENOANIONS

247

F. [E2X1ol4A more commonly encountered form of [EX,]’- is a dimer, i.e., [E2XIOl4-,which was mentioned briefly in Section 1II.D. Examples include K,[Bi,Br,,I (32)(481, [NH,14[Bi2Br,ol(33)(481,[Sr(H20)812[Bi2Br,,] (34)(491, and [HL],[Bi,Br,,] (L = 2,5-diamino-1,3,4-thiadiazole) (35)(50);a view of 35 is shown in Fig. 10. In all cases, the structure can be described as edge-shared bioctahedral and none of the interbond angles deviate from idealized values by more than a few degrees. The Bi-Br distances to the bridging bromines, in the examples given, are longer than those to the terminal bromines, as expected, but in all cases, the Bi-Br-Bi bridging units are quite symmetrical. G. [{EX5)n

A few examples of higher nuclearity aggregates composed of [EX,]’are known. In the structure of [2,2’-bipyridiniuml[SbCl,l(36)(51), a tetrameric arrangement is found, as shown in K.Each [SbC1,I2- unit is linked to two others through approximately linear chlorine bridges that are somewhat asymmetric, the relevant distances being 2.8041 3.218and 2.83613.054A. A polymeric form shown in L is found in [4,4’-bipyridinium][SbC151 (37)(511 and also in [piperidinium],[BiBr,l (38)(52).In 37 the [EX,]’units are discernable in that the linear chlorine bridge is quite asymmetric (2.588vs 3.182A), whereas in 38 the analogous Bi-Br distances

FIG. 10. The structure of [Bi2Brlo14-in 36.

248

FISHER AND NORMAN

I I

r

------K

are more nearly symmetrical (3.02vs 3.13 A), such that a more regular octahedral coordination is found around the bismuth centers. In the mixed chloride/fluoride structure [NH,],[SbCl,F,] (39) (531, a polymeric structure related to L is present, as shown in M, the main difference being that the bridging chlorides (all fluorides are terminal) are bent rather than linear, as in L.

L

M

H. [EX,],Formal addition of three halide ions to EX, affords the mononuclear octahedral species [EX,I3-. Examples include Rb,[BiBr,] (40) (541, [Et,NH,],[BiBr,] (41) (55), [Me2NH213[BiBr61 (42) (56), Na,[BiBr,I-

GROUP 15 ELEMENT(111) HALIDES AND HALOGENOANIONS

249

A

FIG.11. The structure of [BiBr6I3- in 40.

[Bi,Br,,].18H20 (43) (571, Cs2Na[BiC1,] (44) (58), [Me2NHzI4[BiC1,]C1 (45) (591, Rb,[BiI,l[13]I~2H20(46) (601, and a range of alkali metal [BiF,]” salts (611; a view of 40 is shown in Fig. 11. ions are close to regular In general the structures of the octahedral, with only small deviations from idealized angles and small differences in bond lengths. Such deviations as do occur tend to be of a C,,-type distortion in which three mutually cis bonds are slightly longer than the bonds of their trans partners. The nature of these distortions will be commented upon in Section IV, when bonding is considered in a little more detail.

1. [E2Xg13 [E2Xgl3-can exist either as discrete dinuclear units, addressed here, or as polymeric forms, which are considered in the following section. The structure of the dinuclear form can be described as a confacial bioctahedron and examples include [C5H5NHl3[As2Clg1 (47) (62),[C5H5NH13[As2Brgl(48) (631, [piperidinium14[As,BrglBr(49) (64), [EtzNH213[Bi21gl(50) (651,[Ph4P131Bi2Brgl (51) (66),[Me4N13[SbzBrgl.Br2 (52) (671, [Me4N],[SbzBrg] (53) (681, and the mixed halide complex [Me4N13[Sb2Cl,(p-Br)3](54) (68);a view of 48 is shown in Fig. 12. In most cases the bridging halides are quite symmetric (the exception being 49, in which there is some degree of asymmetry) and the coordination around the group 15element is close to that of a regular octahedron. The deviations that do occur are such that the terminal E-X bonds

250

FISHER AND NORMAN

FIG.12. The structure of [AszBr913-in 48.

are shorter than the bridging E-X bonds (as expected) and also that the interbond angles at E between the terminal X groups are usually slightly greater than go", whereas those between the bridging X group are less than 90". This latter distortion in angles could be interpreted as the beginnings of localization of the lone pairs trans to the E-E vector, but an alternative explanation is that an overall antibonding interaction between the E centers leads to a lengthening of the E-E distance with a concomitant change in the relevant bond angles. The lengths of E-E distances and the E-X-E angle for these and a series of related transition metal compounds in terms of any E-E bonding present have been addressed by Cotton and Ucko (69). An alternative way of thinking about this structural type (as opposed to a confacial bioctahedron) is as an [EXs13- octahedron in which one face of the octahedron is capped by a neutral EX3 unit. This is a somewhat contrived point of view but, in later sections, this capping principle will be quite useful in rationalizing a number of structural types. One other example of a different dinuclear structure is found in [Co(NH3)&Sb2F91 (54) (70).This is a more open structure, as has been observed before for some fluoride complexes, and may be described as two square-based pyramids that share a vertex, as shown in N.

N

GROUP 15 ELEMENT(II1) HALIDES AND HALOGENOANIONS

251

With the structure of 54 in mind, the related structure of the [Sb,F,lanion should be considered, two forms of which have been observed. In Cs[Sb,F,] (55) (71) this anion is present as a discrete dimer, with a structure as shown in 0. A reasonable description is of two equatorially vacant, trigonal bipyramids that share an axial vertex. The terminal Sb-F bonds average about 1.95 A, whereas the bridging Sb-F distance is 2.24 A (a C2 axis passes through the p-F atom); the Sb-F-Sb angle is 125.3'. In the related compound K[Sb2F71(56) (72) discrete anions are not present and a one-dimensional polymeric structure is found, as shown in P. As is clear from the diagram, there are two types of antimony center, one of which is five coordinate and the other four coordinate.

0

P

J. [IE&g}n 13n-

As well as discrete dinuclear units of the form [E2X913-,as discussed in the previous section, a number of polymeric forms are also known. The compound [C5H5HNl3[Sb2Cl9] (57) (68) has a double chain structure, as shown in Fig. 13, in which each antimony atom is octahedrally coordinated to six chlorines, three of which bridge to three different antimony centers. The structure of [Me3NHl3[Sb2Cl91 (58) (73) is similar, as are the structures of P-Cs3[Sb2C1,1(59) (74,751 and P-Cs3[Bi2Cl91 (60) (75). In the last two, the main difference is in the packing arrangement of the double chains. Figure 13 and a discussion in Ref. 68 stress the covalent nature of the interactions between the antimony and the chlorine atoms, although the structures of 59 and 60 have been described more as an ionic type of structure involving close-packed chlorines, with the antimony atoms occupying 3 of the octahedral holes. A comparison with the description of Bi13 is interesting in this respect.

252

FISHER AND NORMAN

d FIG. 13. Part of the polymeric structure of IC,H,HNI:,ISb2C1,1 in 57

Similar polymeric structures are also found in a-Cs,[Sb,Cl,] (61) (761, a-Cs,[Bi,Cl,I (62) (751, and Cs3[As,C1,1 (63) (77), which are all

isomorphous, together with Cs3[Bi2BrS1(64) (78) and the isomorphous pair Cs,[Sb,I,] (65) (79) and Cs,[Bi,I,] (66) (79). A different polymeric form is found in the structure of Rb2[SbC1,F,] (67) (53).In this case a polymeric chain of edge-shared bioctahedral units linked through linear chloride bridges, as shown in Q , is present, the repeat unit being [SbzCl3F6I3-,together with isolated [SbC1,I3octahedra and Rb' cations. A more descriptive formula is Rb6[Sb,C13F,I[SbC&I.

Q

K. [E3X1213There are a number of compounds that contain discrete anions of the formula [E3Xl2l3-and these are found in two distinct forms. One

GROUP 15 ELEMENT(II1) HALIDES AND HALOGENOANIONS

253

FIG. 14. The structure of [Bi,I,,13- in 69.

type is what may be described as linear and examples include [K(15crown-5),1,[Sb31,,1 (68) (801, [Bu,"Nl3[Bi,Il21 (69) (81), and [Bi(dmpu),][Bi,I,,] (dmpu, dimethylpropylene urea) (70) (82);a view of 69 is shown in Fig. 14. In all cases, the element E has octahedral coordination and the halides are either terminal or doubly bridging. A useful description in terms of the capping principle mentioned earlier is a n [Ex,],- octahedron in which two opposite faces are capped by EX, units, resulting in a threefold rotational symmetry for the anion. A second type of [E,X,,l3- structure is found in the complexes [Et3NH13[A~3Br121 (71) (83, 84) and [Me3NHl3[As3Il21(72) (84); a view of 72 is shown in Fig. 15. In this instance, the structure can be viewed as being derived from an [EX613-octahedron in which two faces sharing a common vertex have been capped by EX, units. This results in termi-

u

u I(21

FIG.15. The structure of [As3Il2I3-in 72.

I1121

254

FISHER AND NORMAN

nal and p2-X groups but also in a rather unusual triply bridging halide with a T-shaped geometry.

L. [E,X,,13The capping principle can be taken a little further and used to understand a number of structures with the general formula [E,X,,I3-, three isomeric forms of which exist. The structure of [H{OP(NMe2)3}213[Sb51181 (73) (85)is shown in Fig. 16 and can be viewed as based on a central [SbI,13- octahedron in which four faces are capped by SbI, units, such that the overall structure has D2,,symmetry. Removal of an adjacent pair of EX, units that share a common vertex (but not an edge) results in the structure type shown in Fig. 15, whereas removal of an opposite pair of EX, units results in the structure type shown in Fig. 14. A second isomeric form is found in [Me,N],[Sb,I,,l (74) (861,as shown in Fig. 17. This may also be viewed as an [SbI,13- octahedron associated with four SbI, units, but, in this case, it is four edges sharing a common vertex that are bridged, such that the overall structure is that of a square-based pyramid with C,,,symmetry. Moreover, in the four SbI, units, only two of the iodines on each antimony are terminal; the others bridge to adjacent units. A third type of structure is found in [Ph,P]3[Bi,Il,l (75) (87), in which the whole arrangement is linear with symmetry. In this case, the structure can be built up not by capping four faces or bridging four edges of an octahedron, but by a progressive capping of the face opposite

FIG. 16. The structure of [Sb5II8l3-in 73.

GROUP 15 ELEMENT(III1 HALIDES AND HALOGENOANIONS

255

the starting octahedron as each new BiI, unit is added. The structure is represented in R.

R

M. [EJ,,15The anion IE,Xl,15- is found in two compounds, [MeNH31,[Bi,Brll] (76) (88) and the isomorphous [MeNH,15[Bi,C11,1 (77) (89);a view of 76 is shown in Fig. 18. Both bismuth centers are octahedrally coordi-

nated and linked by a linear halide bridge.

N.

CE,X1816-

An anion of this formula is found in [C5H5NHl6[Bi,Cl1,](78) (901, the structure of which is shown in Fig. 19. Each bismuth center is

256

FISHER AND NORMAN

1

,-, Brl241

v Brl21)

u

Br112)

FIG. 18. The structure of [Bi2Br,,15-in 76.

octahedrally coordinated and structural similarities are evident in comparison with the [E2X,,I4- and [E2X,,I5-structures, the latter in terms of the linear halide bridges that are present. A relationship between this unit and the polymeric structure found in 67 is also apparent.

An anion with this formula is found in [Cu(MeCN),12[Sb31,,l(79) (86) and is shown in Fig. 20. Each antimony center is octahedrally coordinated, although this is somewhat distorted as a result of the fourand six-membered rings within the structure.

FIG. 19. The structure of [Bi4C1,8]6-in 78.

GROUP 15 ELEMENTUII) HALIDES AND HALOGENOANIONS

257

FIG.20. The structure of [Sb,I11I2- in 79.

Hexanuclear anions of the formula [E6Xz2I4-are formally dimers of the unit encountered in 79, although the structural similarities are not great. Three examples, which comprise two distinct but related structural forms, are known. One of these forms is found in the compounds [Fe(77-C5H5)214[Sb61221 (80)(86) and [Et4P]4[Bi6122] (81)(911, and a view of 80 is shown in Fig. 21. This structure can be viewed as based on the tetranuclear species [E4XI6l4-,as described in Section III.D, in which two opposite X, faces have been capped by EX, units (analogous to I) or, alternatively, in which the structure has been extended by addition of a further E2Xsunit (analogous t o H). Because the [E4Xl6I4structure can itself be viewed as being derived from the [E2Xlo14-type, described in Section III.F, also by the addition of two EX, units (or an

FIG.21. The structure of [Sb&I4- in 80.

258

FISHER AND NORMAN

E,X, unit), a class of tetraanion may therefore be defined with the general formula [E2nX6n+4]4-, the structures [E,X,,,14-, [E4X1,l4-, and [E6XZ2l4forming the first three members of the series, for which n = 1, 2, and 3, respectively. The other structural form is found in the compound [Fe(l,lO-phen),],[SbGI,,].2MeCN(82) (921, which is shown in Fig. 22. This can also be derived from the basic structure of the tetranuclear [E,X1,l4- cluster but, whereas in 80 and 81 it is faces in the plane of the E, unit that are capped, in 82, it is faces above and below, such that the E, core is no longer planar; Sb(3) and its symmetry related partner in Figs. 21 and 22 are the capping groups. An alternative description is a double cube that highlights a relationship between this cluster and the cubic structure of 28 in which two adjacent EX, units have been added to one face of the cube.

Q. [E&814Considering the general formula [E,, X6,+,l4-, described in the previous section, there are five structures known with the formula [E8X28]4-, which extend the series to n = 4. Three structural forms are known. In [Ph4P14[Sb81281 (83) (931 and [ H ( d m p ~ ) , I ~ [ S b ~C,H,Cl I ~ ~ l ~(84) x (85), the eight antimony atoms are essentially coplanar and the structure can clearly be thought of as based on the [E,X,,l4- cluster shown in Fig, 21, to which another two EX, units (or an E,X, unit) have been

FIG.22. The structure of [SbGIz214in 82.

GROUP 15 ELEMENT(II1) HALIDES AND HALOGENOANIONS

259

FIG.23. The structure of ISb8IZ8l4-in 83.

added in much the same way as [E,X,,14~can be built up from [E4Xl6I4and two EX,; a view of 83 is shown in Fig. 23. A second structural form is found in [Me4N14[Sb81,,l (85) (94) and [Me3Sl,[Sb,I,,l (86) (94).In these cases, the structure is more three dimensional and can be thought of as built up from a [E4Xl6I4-cluster, as described in Section III.D, in which two faces are capped by EX, units, Sb(4), and two edges are bridged by EX, units, Sb(3), none of which are adjacent to each other. A view of 85 is shown in Fig. 24.

FIG. 24. The structure of [SbsIzs14-in 85.

260

FISHER AND NORMAN

P

611121

FIG. 25. The structure of [As81zs14-in 87.

A third type of structure is found in [Et3NH14[As81z81 (87) (31), which is shown in Fig. 25. This may also be viewed as based on an [E4X,,I4core in which opposite sides are bonded to E2& units in a way slightly different from that found in the first type of structure mentioned in this section, i.e., 83 as shown in Fig. 23. As(3) and Ad41 are the atoms of the E2& units.

R. [E&3,16 The largest discrete anion is the [Bi,C13,]6- anion, found in the structure of [Et4N]6[Bi,C13,](88) (951,which is shown in Fig. 26 and which has similarities to a polymeric structure described in the next section. The central unit is clearly structurally similar to the [E4X16I4- unit, described in Section 1II.D (Fig. 71, and the structure of 88 can be envisaged as built up from this core by the addition of two [EzX71-anions to opposite ends.

Two structures of antimony fluorides with this empirical formula were mentioned in Section 111.1because they were related to structures

GROUP 15 ELEMENTW) HALIDES AND HALOGENOANIONS

261

FIG.26. The structure of [Bi,CI,o]6- in 88.

discussed at that point. In this section, another example, which is found in the compound [Me3N(CHzPh)l[SbzI,l(89) (941, is considered. The anion is polymeric, is shown in Fig. 27, and is based on the [E4XI6l4structure (Section 1II.D) in such a way that the polymer can be constructed from these units by the sharing of opposite edges.

T* [{EJ1&n 1"Three types of polymeric structure with this formula are known. A view of the anion in [Ph4P][Sb,Il,] (90) (93) is shown in Fig. 28 and may be thought of as a polymer of [E4X1614-units in which opposite E atoms are shared. The structural similarities to the [Bi,C130]6- anion shown in Fig. 26 are also obvious.

FIG.27. Part of the polymeric structure of the [{Sbz17}n]"anion in 89.

262

FISHER AND NORMAN

FIG.28. Part of the polymeric structure of the [{Sb3110}n In- anion in 90.

In [C(NMe2)31[Sb31,0] (91) (96)a different structure is found (Fig. 29). The repeat unit in this structure is derived from the [E,X,,I4- anion shown in Fig. 21, in which opposite edges are shared in a way similar though not identical to that in which the structure of 89, shown in Fig. 27, was derived from the [E4XI6l4-unit. A third structural type is observed in the structure of the fluoroantimonate complex Na[Sb3Flo1(92) (971,which is shown in Fig. 30. As is often found for the fluoride complexes, the structure is more open and the antimony atoms are five coordinate with a square-based pyramidal geometry. The essential unit is a triangular [Sb3F9]fragment, with three additional fluorines, one per antimony, forming bridges to adja-

GROUP 15 ELEMENT(II1) HALIDES AND HALOGENOANIONS

FIG. 30. Part of the polymeric structure of the [{Sb3FI,,},l"- anion in 92.

FIG. 31. Part of the polymeric structure of the [{Sb4FI3},l"- anion in 93.

263

264

FISHER AND NORMAN

cent triangles, such that the resulting polymeric structure is a twodimensional infinite sheet.

u. [{E,X,,)"

1"

The complexes Q[Sb,Fl31 (Q = K, Rb, Cs, T1, NH,) form an isomorphous series (98,99)in which the [E4Fl31-unit is present as a tetramer of SbF3 units, with a central fluoride anion in the center of the Sb, square and weaker Sb-F-Sb bridging interactions between the tetramers, giving rise to an infinite two-dimensional polymeric sheet. The structure of the potassium salt (93) is shown in Fig. 31.

IV. General Comments

One feature that is clearly apparent from Section I11 is the great variety of halogenoanion structures that are known. In this regard it is interesting to note a parallel to the structural chemistry of the copper(1) halogenoanions reviewed recently by Jagner and Helgesson (100).These authors comment on the wide range of structures and the possible factors that affect the particular type that is formed. Of the many possible factors that might influence the solid-state structure of a given anion, undoubtedly the most important is the nature of the cation, such as the size, shape, and localization of charge. These features are not easy to quantify and it is therefore difficult if not impossible t o predict or rationalize many of the observed structures, but a number of general points do emerge. For example, Jagner and Helgesson comment on the fact that the smaller copper(1)halogenoanions are generally associated with smaller cations, i.e., that the anion size and the cation size (or collective size of the cations if more than one is needed to balance the charge of the anion) tend to be similar. The importance of ion size and shape has been addressed in general terms by Mingos and Rohl (1011, and a useful discussion of the importance of relative ion sizes with respect to lattice energies has been given by Alcock (18). Nevertheless, the difficulties associated with any rationalizations are clear if a number of structures mentioned in Section I11 are compared. For example, compounds 15 and 26 have the same [Mg(MeCN),12+ cation but, in the former case, the [SbCl,I- anion is polymeric, whereas, in the latter case, the [BiCl,]- anion is a tetramer. A similar situation is observed for 17, with a polymeric [BiCl,I- anion, and 27, with a tetrameric [BiBr,I- anion, the cation in both these cases being [Fe(q-

GROUP 15 ELEMENTW) HALIDES AND HALOGENOANIONS

265

C5HJ2It. Moreover, in 80, four [Fe(+2,H5),1+ cations are associated with the [Sb,I2,l4- anion rather than a tetrameric [Bi4BrI6l4-anion, as found in 27. There are many other pairs of structures that differ in having the same number and type of cation but different anion structures, e.g., 47/57 and 41/50, which further illustrate this point. Factors such as the solvent and concentration of the reactants are also probably important in determining the type of structure that crystallizes; as is found in the copper(1) chemistry, the species present in solution are probably simple mononuclear or dinuclear units. It should be borne in mind that many of the structures formed may be determined as much by the kinetics of crystallization as by thermodynamic factors. Turning now to the structures themselves, it is worth reiterating the point, made in Section.11,that, as with the halides themselves, the structures of the anions reveal strongly correlated trans bond lengths and that the primary and secondary bond lengths are generally more nearly equal the heavier the elements involved. This feature is also related to the degree of bridge asymmetry in E2(p-XI2units, which tends to be greatest for chlorides and least for iodides, although in polymeric structures there may be considerable variation in the asymmetry of adjacent bridging units. With regard to coordination geometries, it is clear that, despite the wide variety of structural types, in the vast majority of cases the group 15 element is six coordinate and that the coordination geometry is close to octahedral. This raises the problem, alluded to earlier, of the stereochemical activity or significance of the lone pair of electrons that is present in all element(II1) complexes. In the four- and five-coordinate structures, there is little problem in understanding the coordination geometry; the structures are as predicted by VSEPR theory and are also consistent with the u*-orbital model for Lewis acidity outlined in Section 11. Thus in 1, for example, the equatorially vacant trigonal bipyramidal structure is anticipated from VSEPR but the anion can also be understood as a complex of PC1, and C1- in which the C1- interacts with one P-Cl u*-orbital, which leads to a lengthening of the P-C1 bond trans to the coordinated chloride. In 2 the axial P-Br bonds are more nearly the same length as the distinction between primary and secondary becomes less distinct, but both are longer than the bonds to the terminal bromines. Also, the beginnings of five coordination, which leads to a square-based pyramidal geometry as a second cT*-orbital is employed, are apparent. The way in which four- and five-coordinate structures can be built up from a trigonal pyramid is illustrated by considering one and two X- groups, respectively, in S , where the lobes represent the E-X u*-orbitals.

266

FISHER AND NORMAN XI I

S

In the vast majority of cases in which six coordination is observed, the bonding can be viewed as arising from the interaction of all three u*-orbitals with a halide anion, i.e., all three X- in S . Because the three orbitals are all trans to the primary E-X bonds, such a situation leads naturally to octahedral coordination. Moreover, in cases in which the primary and secondary bonds are the same length, i.e., where A = 0 and a three-center, four-electron bonding model is appropriate, a regular octahedron is the result. Such a structure is clearly at odds with simple VSEPR theory, which is predicated on the lone pair(s) occupying specific stereochemical sites, but stereochemical inactivity of the lone pair tends t o be the rule rather than the exception in sixcoordinate, seven-electron pair systems; Ng and Zuckerman (102)have reviewed this topic for p-block compounds in general. The problem of stereochemical activity of lone pairs is, in fact, more subtle than either of the simplified models outlined above, and this point can be illustrated with reference to some specific examples. Thus, in an illuminating paper, Wheeler and Kumar (103)comment on the fact that in compound 46, the octahedron is regular (undistorted), whereas, in [H3N(CH2),NH313CBiC1612-2H20 (94) (1041,the [Bic16l3- octahedron shows a marked trigonal CSudistortion in which three angles are slightly less than 90" and three are slightly greater. Furthermore, the three mutually cis Bi-Cl bonds in 94 that have interbond angles greater than 90" are significantly longer than the other three (2.644vs 2.792 A), both of these factors being consistent with the beginnings of lone pair localization in this face of the octahedron. In 94 and in the isomorphousbromide complex, the beginnings of stereochemical activity of the lone pair, which is clearly inactive in 46, is evident. An extended Huckel molecular orbital analysis on both of these complexes revealed that this structural difference between the chloride and the iodide can be traced to a second-order Jahn-Teller distortion, which is predicted to occur more readily for the chloride. Specifically,

GROUP 15 ELEMENT(II1) HALIDES AND HALOGENOANIONS

267

the calculations indicate that the relevant orbital mixing (which will not be discussed in detail here) is expected to be greater in an octahedral chloride, as opposed to an iodide, because the relevant orbitals are closer in energy, with the result that the chlorides should show a greater tendency to be distorted, as is indeed observed. Furthermore, this distortion is expected to take place along a C3,coordinate, through changes either in bond angles or in bond lengths, precisely as seen in 94.' This distortion is associated with the onset of the stereochemical activity of the lone pair, which becomes localized on one face of the octahedron. Moreover, a C,, distortion is predicted to be of lower energy than any alternative modes, such as localization of the lone pair along one edge of the octahedron with a concomitant enlargement of one angle: a C2, distortion. It should be borne in mind that the difference in energy between distorted and undistorted structures is likely to be small and of the same order as crystal packing forces; for example, the [BiC1,I3- anion in 44 is a regular octahedron and it should be anticipated that changes in the cation will have significant effects. Nevertheless, the trend is clear in that more regular coordination geometries should be expected for the iodides, in terms of both bond lengths and bond angles, which is broadly in line with what is observed. Similar arguments can be advanced to account for why bismuth halides and halogenoanions have more regular geometries than arsenic or antimony systems. In this regard, Wheeler and Kumar carried out calculations on the solid-state structures of Bi13 and Sb13. As will be recalled from Section 11, these two solids (and Ad,) are isomorphous and differ only in the degree to which the group 15 element center is displaced toward three mutually cis iodides. The degree of distortion increases from Bi to As and this is also traced to a second-order Jahn-Teller distortion, which becomes more pronounced for the lighter element. In this case, the distortion is manifest not only as an asymmetry in the E-I distances but also as a decrease in the interbond angles between the longer E-I bonds (and a corresponding increase in the angles between the shorter bonds), the opposite of what is observed in molecular species. This results from the constraints imposed on the

'

A distinction should be made between two variations of this CSudistortion. In one case, an increase in the angles of three mutually cis E-X bonds may occur without any significant changes in bond lengths. Alternatively, three mutually cis E-X bonds may lengthen without a significant change in any bond angles. Combinations of both are, of course, possible and these types of change account for most of the observed distortions from regular octahedral geometry.

268

FISHER AND NORMAN

iodide positons by the solid lattice and is an unusual example of a lone pair becoming localized in a region of decreasing bond angles. In conclusion, these studies provide a theoretical foundation for rationalizing many of the trends in coordination geometry regularity (bond lengths and angles), which have been commented upon throughout this chapter. More simplified descriptions of second-order Jahn-Teller distortions in some of these systems can be found in Refs. 105 and 106.

ACKNOWLEDGMENTS We thank A. G. Orpen, S. Pohl, and W. S. Sheldrick for their comments and for sharing results prior to publication.

REFERENCES 1. Gillespie, R. J., Slim, D. R., and Vekris, J, E., J . Chem. Soc., Dalton Trans., 971

(1977);Edwards, A. J., and Slim, D. R., J . Chem. SOC.,Chem. Commun., 178 (1974). 2. Davidovich, R. L., and Buslaev, Y . ,Koord. Chem. 14, 1011 (1988). 3. Sawyer, J. F., and Gillespie, R. J., Prog. Znorg. Chem. 34, 65 (1986). 4. Wells, A. F., “Structural Inorganic Chemistry,” 5th ed. Oxford Univ. Press, Ox-

ford, 1984. 5 . Muller, U., “Inorganic Structural Chemistry,” Wiley, Chichester, 1992. 6. Edwards, A. J., J . Chem. SOC.A , 2751 (1971). 7. Gries, O., and Martinez-Ripoll, M., Z . Anorg. Allg. Chern. 436, 105 (1977);Cheetham, A. K., and Norman, N.,Acta. Chem. Scand. A 28, 55 (1974). 8. Lipka, A., Acta Crystallogr. B 35, 3020 (1979). 9. Cushen, D. W., and Hulme, R., J . Chem. Soc., 2218 (1962). 10. Nyburg, S. C., Ozin, G. A., and Szymanski,J . T.,Actu Crystallogr. B 27,2298 (1971). 11. Cushen, D. W., and Hulme, R., J . Chem. Soc., 4162 (1964). 12. Trotter, J., Z. Kristdogr. 122, 230 (1965). 13. von Benda, H., Z. Kristullogr. 151, 271 (1980). 14. Pohl, S., and Saak, W., Z. Kristallogr. 169, 177 (1984). 15. Trotter, J., and Zobel, T., 2.Kristallogr. 123, 67 (1966). 16. Enjalbert, R., and Galy, J., Actu Crystallogr. B . 36, 914 (1980). 17. Alcock, N. W., Adv. Znorg. Chem. Radiochem. 15, 1 (1972). 18. Alcock, N. W., “Bonding and Structure: Structural Principles in Inorganic and Organic Chemistry.” Ellis Horwood, Chichester, 1990. 19. Sheldrick, W. S., and Martin, C., Z . Naturforsch. B 46, 639 (1991); James, M. A,, Knop, O., and Cameron, T. S., Can. J . Chem. 70,1795 (1992);Krebs, B., and Ahlers, F.-P., Ado. Inorg. Chem. 35,235 (1990). 20. Clegg, W., Compton, N. A,, Errington, R. J., Fisher, G. A,, Hockless, D. C. R., Norman, N. C., Williams, N. A. L., Stratford, S. E., Nichols, S. J., Jarrett, P. S., and Orpen, A. G., J . Chern. Soc., Dalton Trans., 193 (1992);Clegg, W., Errington,

GROUP 15 ELEMENTW) HALIDES AND HALOGENOANIONS

269

R. J., Fisher, G. A., Hockless, D. C. R., Norman, N. C., Orpen, A. G., and Stratford, S. E., J . Chem. Soc., Dalton Trans., 1967 (1992);Clegg, W., Errington, R. J., Fisher, G. A., Flynn, R. J., and Norman, N. C., J. Chem. Soc., Dalton Trans., 637 (1993), and references therein. 21. Dillon, K. B., Platt, A. W. G., Scmidpeter, A., Zwaschka, F., and Sheldrick, W. S., Z. Anorg. Allg. Chem. 488,7 (1982). 22. Sheldrick, W. S., Scmidpeter, A., Zwaschka, F., Dillon, K. B., Platt, W. G., and Waddington, T. C., J. Chem. Soc., Dalton Trans., 413 (1981). 23. Kaub, J., and Sheldrick, W. S., 2.Naturforsch. B 39, 1252 (1984). 24. Sheldrick, W. S., and Horn, C., Z. Nuturforsch. B 44,405 (1989). 25. Sheldrick, W. S., and Kiefer, J., Z. Naturforsch. B 47, 1079 (1992). 26. Pohl, S.,Saak, W., and Haase, D., Angew. Chem., Znt. Ed. Engl. 26, 467 (1987). 27. Belz, J., Weber, R., Roloff, A., and Ross, B., 2.Kristullogr. 202, 281 (1992). 28. Alcock, N. W., Ravindran, M., and Willey, G. R., J. Chem. SOC., Chem. Commun., 1063 (1989). 29. Ensinger, U., Schwarz, W., and Schmidt, A., 2.Naturforsch. B 37, 1584 (1982). 30. Kaub, J., and Sheldrick, W. S., Z. Naturforsch. B 39, 1257 (1984). 31. Sheldrick, W. S., Hausler, H.J., and Kaub, J., Z. Nuturforsch. B 43,789 (1988). 32. Porter, S.K., and Jacobson, R. A., J . Chem. SOC.A , 1356 (1970). 33. Drew, M. G. B., Claire, P. P. K., and Willey, G. R., J. Chem. Soc., Dalton Trans., 215 (1988). 34. Blazic, B., and Lazarini, F., Actu Crystullogr. C 41, 1619 (1985). 35. Mammano, N. J., Zalkin, A., Landers, A., and Rheingold, A. L., Znorg. Chem. 16, 297 (1977). 36. Robertson, B. K., McPherson, W. G., and Meyers, E. A., J. Phys. Chem. 71,3531 (1967). 37. Habibi, N.,Bonnet, B., and Ducourant, B., J . Fluorine Chem. 12,237 (1978). 38. Habibi, N., Ducourant, B., Bonnet, B., and Fourcade, R., J . Fluorine Chem. 12, 63 (1978). 39. Ducourant, B., Fourcade, R., Philippot, E., and Mascherpa, G., Rev. Chim. Miner. 12,485 (1975). 40. Ducourant, B., Fourcade, R., and Mascherpa, G., J.Fluorine Chem. 11,149(1978); Reu. Chim. Miner. 20,314 (1983). 4 1 . Aurivillus, B., and Lindblom, C.-I., Acta Chem. Scand. 18,1554 (1964). 42. Antolini, L., Benedetti, A., Fabretti, A. C., and Giusti, A., J. Chem. Soc., Dalton Trans., 2501 (1988). 43. Willey, G . R., Collins, H., and Drew, M. G. B., J. Chem. SOC.,Dalton Trans., 961 (1991). 44. Rheingold, A. L., Uhler, A. D., and Landers, A. G., Znorg. Chem. 22, 3255 (1983). 45. Bystrom, A,, Backlund, S., and Wilhelmi, K.-A., Ark. Kemi. 4, 175 (1952). 46. Edstrand, M., Inge, M., and Ingri, N., Acta Chem. Scand. 9,122 (1955). 47. Wismer, R. K., and Jacobson, R. A,, Inorg. Chem. 13, 1678 (1974). 48. Lazarini, F., Actu Crystallogr. B 33, 1954 (1977). 49. Lazarini, F., and Leban, I., Acta Crystullogr. B 36,2745 (1980). 50. Benedetti, A., Fabretti, A. C., and Malavasi, W., J . Crystullogr. Spectrosc. Res. 22, 145 (1992). 51. Lipka, A., Z. Naturforsch. B 38, 1615 (1983). 52. McPherson, W.G., and Meyers, E. A., J . Phys. Chem. 72,532 (1968). 53. Udovenko, A. A., Volkova, L. M., and Davidovich, R. L., Sou. J. Coord. Chem. (Engl. Transl.)4, 234 (1978).

270

FISHER AND NORMAN

Lazarini, F., Acta Crystallogr. B 34, 2288 (1978). Lazarini, F., Acta Crystallogr. C 41, 1617 (1985). McPherson, W. G., and Meyers, E. A., J . Phys. Chem. 72, 3117 (1968). Lazarini, F., Acta Crystallogr. B 36, 2748 (1980). Morss, L. R., and Robinson, W. R., Acta Crystallogr. B 28, 653 (1972). Herdtweck E., and Kreusel, U., Acta Crystallogr. C 49, 318 (1993). 60. Lazarini, F., Acta Crystallogr. B 33, 1957 (1977). 61. Coussin, A., Vedrine, A., and Cousseins, J.-C., C.R.Acad. Sci.Paris 274,864 (1972). 62. Kaub, J., and Sheldrick, W. S., 2.Naturforsch. B 39, 1252 (1984). 63. Kaub, J., and Sheldrick, W. S., 2. Naturforsch. B 39, 1257 (1984). 64. Sheldrick, W. S., and Horn, C., 2. Naturforsch. B 44, 405 (1989). 65. Lazarini, F., Acta Crystallogr. C 43, 875 (1987). 66. Lazarini, F., Acta Crystallogr. B 33, 2686 (1977). 67. Hubbard, C. R., and Jacobson, R. A., Inorg. Chem. 11, 2247 (1972). 68. Hall, M., Nunn, M., Begley, M. J . , and Sowerby, B. D., J . Chem. SOC.,Dalton Trans., 1231 (1986). 69. Cotton, F. A., and Ucko, D. A., Inorg. Chim. Acta 6, 161 (1972). 70. Schroeder, D. R., and Jacobson, R. A., Inorg. Chen. 12, 515 (1973). 71. Ryan, R. R., Mastin, S. H., and Larson, A. C., Inorg. Chem. 10, 2793 (1971). 72. Mastin, S. H., and Ryan, R. R., Inorg. Chem. 10, 1757 (1971). 73. Kruger, F. J., Zettler, F., and Schmidt, A,, 2. Anorg. Allg. Chem. 449, 135 (1978). 74. Kihara, K., and Sudo, T., Acta Crystallogr. B 30, 1088 (1974). 75. Meyer, G., and Schonemund, A., 2. Anorg. Allg. Chem. 468, 185 (1980). 76. Kihara, K., and Sudo, T., Z . Kristallagr. 134, 142 (1971). 77. Howard, J. L., and Goldstein, L., J . Chem. Phys. 3, 117 (1935). 78. Lazarini, F., Acta Crystallogr. B 33, 2961 (1977). 79. Chalbot, B., and Parthe, E., Acta Crystallogr. B 34, 645 (1978). 80. Borgsen, B., Weller, F., and Dehnicke, K., 2. Anorg. Allg. Chem. 596, 55 (1991). 81. Gieser, U.,Wade, E., Wang, H. H., and Williams, J. M., Acta Crystallogr. C 46, 1547 (1990). 82. Carmalt, C. J., Farrugia, L. J., and Norman, N. C., submitted for publication. 83. Sheldrick, W. S., and Hausler, H . J . , Angew. Chem., Znt. Ed. Engl. 26,1172 (1987). 84. Sheldrick, W. S., and Kiefer, J.,2. Naturforsch. B 47, 1079 (1992). 85. Carmalt, C. J., Farrugia, L. J., and Norman, N. C., Polyhedron 12, 2081 (1993). 86. Pohl, S., Lotz, R., Saak, W., and Haase, D., Angew. Chem., Int. Ed. Engl. 28, 344 (1989). 87. Pohl, S., personal communication. 88. Matuszewski, J., Jakubas, R., Sobczyk, L., and Glowiak, T., Acta Crystallogr. C 46, 1385 (1990). 89. Lefebvre, J., Carpenter, P., and Jakubas, R., Acta Cryslallogr. B 47, 228 (1991). 90. Aurivillius, B., and Stalhandske, C . , Acta Chem. Scand. A 32, 715 (1978). 91. Clegg, W., Errington, R. J., Fisher, G. A., Green, M. E., Hockless, D. C. R., and Norman, N. C., Chem. Ber. 124,2457 (1991). 92. Pohl, S., Haase, D., Lotz, R., and Saak, W., 2. Naturforsch. B 43, 1033 (1988). 93. Pohl, S., Saak, W., and Haase, D., 2. Naturforsch. B 42, 1493 (1987). 94. Pohl, S., Lotz, R., Haase, D., and Saak, W., 2.Naturforsch. B 43, 1144 (1988). 95. Zaleski, J., Glowiak, T., Jakubas, R., and Sobczyk, L., J . Phys. Chem. Solids 50, 1265 (1989). 96. Pohl, S., Saak, W., Mayer, P., and Schmidpeter, A,, Angew. Chem., Int. Ed. Engl. 25, 825 (1986). 54. 55. 56. 57. 58. 59.

GROUP 15 ELEMENTU) HALIDES AND HALOGENOANIONS

271

97. Fourcade, R., Mascherpa, G., and Philippot, E., Acta Crystallogr.B 31,2322(1975). 98. Bystrom, A., and Wilhelmi, K.-A., Ark. Kemi. 3, 17 (1951). 99. Ducourant, B.,Fourcade, R., Philippot, E., and Mascherpa, G., Rev. Chim. Miner. 12,553 (1975). 100. Jagner, S., and Helgesson, G., Adu. Inorg. Chem. 37, 1 (1991). 101. Mingos, D.M. P., and Rohl, A. L., J . Chem. SOC.,Dalton Trans., 3419 (1991). 102. Ng, S.W., and Zuckerman, J. J., Adv. Inorg. Chem. Radiochem. 29,297 (1985). 103. Wheeler, R. A., and Kumar, P. N. V. P., J . A m . Chem. SOC.114,4776 (1992). 104. du Bois, A,, and Abriel, W., 2. Naturforsch. B 43, 1003 (1988);du Bois, A., and Abriel, W., 2. Kristallogr. 182,36 (1988). 105. Gimarc, B. M., “Molecular Structure and Bonding.” Academic Press, New York, 1979. 106. Albright, T. A., Burdett, J. K., and Whangbo, M. H., “Orbital Interactions in Chemistry.” Wiley-Interscience, New York, 1985.

NOTEADDEDIN

PROOF

Since the submission of this chapter some additional halogenoanion structural types have been characterized. A third structure type for the [E6Xz2]4-anion (Section 1II.P) has been found in the isomorphous pair of compounds [Et,N(CHzPh)l4[E6Izz1(E = Sb, Bi) (95)(107). This structure can also be derived from the [E4X,,I4- structure (Section 1II.D) and two EX3 units, but in which the EX3 units cap X3 faces in a manner different from that encountered in the structures shown in Fig. 21 (80) and in Fig. 22 (82).The relationship between all three structural types is discussed in (107). A fourth structural type of the [E&,l4- anion is observed in the structure of (CH3(CH,)2COS(CH2)zNMe314 [Sb,Iz8](96)(108),which can be derived from the structure of the double cubic [E6Xzz14anion in 82 in which two opposite X2 edges have been bridged by SbI, units. Finally, a polymeric iodoantimonate anion is found in the structure of [2-(4-nitrophenyl)allylltrimethylammonium][Sb31,,] (97)(109).The empirical formula of the anion is the same as those discussed in Section III.T, although the structure is quite different, comprising a polymer in which some of the six antimony atoms in the repeat unit, [Sb&~l~-,are disordered between the seven available octahedral interstices of the close-packed iodine atom array. 107. Pohl, S., Peters, M., Haase, D., and Saak, W., 2. Naturforsch. B , in press. 108. Carmalt, C.J., Farrugia, L.J., and Norman, N. C., 2. Anorg. Allg. Chem., in press.

109. Carmalt, C. J., Farrugia, L. J., and Norman, N. C., Polyhedron, in press.

This Page Intentionally Left Blank

ADVANCES IN INORGANIC CHEMISTRY, VOL.

41

INTERVALENCE CHARGE TRANSFER AND ELECTRON EXCHANGE STUDIES OF DINUCLEAR RUTHENIUM COMPLEXES ROBERT J. CRUTCHLEY The Ottawa-Carleton Chemistry Institute, Carleton University, 1125 Colonel By Dr., Ottawa, Ontario, K1S 586, Canada

I. Introduction 11. Mixed-Valence Complexes A. Theory of Mixed-Valence Systems B. Class I11 Mixed-Valence Complexes C. Class I1 Mixed-Valence Complexes 111. Electron Exchange A. Theory B. Studies of Exchange Coupled Dinuclear Ruthenium Complexes IV. Future Studies V. Glossary of Abbreviations and Ligand Structures References

I. Introduction

Dinuclear ruthenium complexes form the largest group by far of any mixed-valence system and are the exclusive subject of this chapter. Ruthenium is the transition metal of choice to study electron transfer or exchange because it is relatively inexpensive and forms stable Ru(II1) and Ru(I1) coordination complexes. In addition, the synthetic coordination chemistry of ruthenium is well developed (1). The intellectual push to study mixed-valence complexes was provided by the publication in 1967 of two review articles, by Allen and Hush (2) and Robin and Day (3)’on the physical properties of mixed-valence systems. These were followed by Hush’s publication ( 4 ) of his theoretical model of intervalence transitions, which provided a link between the properties of mixed-valence complexes in solution and the Marcus theory of intermolecular electron transfer (5,6). The review by Robin and Day classified mixed-valence complexes into three types: class I, 213 Copyright 0 1994 by Academic h e a s , h e . All rights of reproduction in any form reserved.

274

ROBERT J. CRUTCHLEY

completely valence trapped (no coupling between metal ions); class 11, valence trapped (weak coupling between metal ions); and class 111, delocalized valency (strong coupling between metal ions). These ideas, complete with comparisons of mixed-valence complex properties with Marcus electron transfer theory, formed the basis of Creutz’s review of mixed-valence complexes in 1983 (7). Since the publication of Creutz’s review, many new complexes have been synthesized and important developments have occurred. This chapter surveys the literature from 1982 to August of 1993, and it is meant to be an extension of the Creutz review and a guide to the experimentalist in the design of novel dinuclear ruthenium complexes for the study of superexchange phenomena. The reader should be familiar with the above reviews, and it would be useful to have a basic understanding of excited-state (8)and thermal (9)electron transfer. A quantum mechanical approach to these subjects can be found in a recent review by Newton (10). This chapter is broken up into two main sections: mixed-valence complexes and electron exchange. The section on mixed-valence complexes discusses the various theoretical approaches with a strong emphasis on the Hush model, a summary of new class I11 complexes, and a summary of class I1 complexes. The section on electron exchange begins with an introduction to magnetic exchange theory, in which a charge-transfer model for superexchange is discussed, and ends with a discussion of dinuclear ruthenium complexes that undergo antiferromagnetic spin coupling. The section on future studies summarizes important developments and research that is needed to increase understanding of superexchange phenomena. The practical applications of this research to molecular devices and materials is also pointed out. Finally a glossary of abbreviations and organic ligand structures is included at the end of the review. It.

Mixed-Valence Complexes

A. THETHEORY OF MIXED-VALENCE SYSTEMS 1. The Hush Model

The conceptual framework that is used to unGa-stand thermal and photoinduced electron transfer is illustrated by a classical potential energy-configuration diagram. The simpliest case for electron transfer in which the electron is coupled between donor and acceptor by a single oscillator having the same frequency in both the initial and the final states is illustrated in Figs. l a and lb.

DINUCLEAR RUTHENIUM COMPLEXES

275

a b FIG.1. Potential energy-configuration diagram of initial and final states for (a)symmetric mixed valence complex and (b) asymmetric mixed valence system.

For Figs. l a (symmetric case) and l b (asymmetric case), the lefthand curve is that for the initial state and the right-hand curve, that for the final state. For the symmetric case there is no energy difference between the initial and the final states (Eo= 0) but for the asymmetric case E , > 0. The vertical transition between initial and final states is the energy required to photoexcite an electron between states. This is the intervalence transition (IT) or, in the case of the dinuclear ruthenium complexes of this chapter, metal-to-metal charge transfer (MMCT). A thermal electron transfer route arises from the vibronic coupling of initial and final states and has the activation energy Eth. When donor and acceptor interact, initial and final states mix and this causes a distortion in the potential energy curve (Fig. 2). The thermal activation energy Ethbecomes smaller with increasing magnitude of the resonance exchange integral Had.The Marcus theory ofbimolecular electron transfer rates (5, 6 ) used the above conceptual framework to

FIG. 2. Distortion of potential energy curves resulting from mixing of initial and final states.

276

ROBERT J. CRUTCHLEY

understand the dependence ofEthon the nature of weakly coupled donor and acceptor complexes. It was Hush ( 4 ) who applied this conceptual framework to the problem of intervalence transitions. For the weak coupling case (class I1 complexes), in which the distortion seen in Fig. 2 is small, Hush derived the following relationships of intervalence band properties: For the symmetric system (Fig. la), Eop = W h ,

Avli2

(2310~,,,)'/~.

For the asymmetric system (Fig. lb),

The band width at half peak height AVl/2 assumes a Gaussian band shape and is in cm-'. AFl,z is defined as the value of AF at which

but that it is most commonly evaluated as AF, for which Z/Zmm = B. For weak coupling cases, Hush showed that the intensity of an intervalence transition was related to the extent of coupling between donor and acceptor. The derivation (11 begins by considering the theoretical expression for oscillator strength, f

=

1.085

X

lO"GF,,,D,

(6)

where G refers t o the degeneracy of the states concerned, V,, is approximated to be the energy of the intervalence band in cm-' at maximum extinction coefficient E,,, , and D is the dipole strength for an electricdipole-allowed transition. For a one-electron system, D is related to the charge-transfer transition dipole moment M, by

where $, is the ground-state wave function, Qe is the excited-state wave function, and er is the transition dipole operator. The square of the electric charge, e2, is usually incorporated into the prefactor of Eq. (6). Hush considered isolated donor &, and acceptor 4, wave functions to

DINUCLEAR RUTHENIUM COMPLEXES

277

be weakly coupled and, following Murrell (121,the resulting groundand excited-state wave functions can be written

These wave functions are orthogonal and so

where Sadis the overlap integral between donor and acceptor wave functions. For the weak coupling case, Sad = 0 and the mixing coefficients (Yad = - a d a = a. Then by perturbation theory

where Hadis the resonance exchange integral in cm-', which is a measure of electronic coupling, and Ed - E, is the difference in energy in cm-' between donor and acceptor wave functions, which can be approximated by VmaX. Substituting Eqs. (8) and (9) into Eq. (7) yields after simplification (131,

where R is the transition dipole length (the distance through which charge is carried) in centimeters. Equations (11) and (12) can be used to solve for dipole strength in Eq. (7) and the result substituted into Eq. (6)yields f

=

1.085 x 10"GHa~R2/V,,.

(13)

The oscillator strength of a Gaussian band can be determined experimentally (1I 1 by using the expression f

=

4.6 x 10-9~,axV1/2.

(14)

Substituting Eq. (14)for f and rearranging gives Had

=

2.06 X

R

.jj )1/2* (ijmax*&max 112

(15)

278

ROBERT J. CRUTCHLEY

The Hush model allows the experimentor to estimate the amount of metal-metal coupling (Hadand a ) provided an intervalence band is experimentally observable and R can be estimated. In the case of strong coupling (class I11 complexes), for which it is no longer appropriate to use perturbation theory, the interaction of states is so great that bonding and antibonding potential energy curves result (the potential energy diagram for a symmetric mixed valence complex is shown in Fig. 3). The magnitude of Hadhas become so great that Eth = 0. The odd electron can now move freely between metal centers (delocalized)and the energy of the intervalence transition is a measure of electronic coupling between donor and acceptor wave functions,

Because the ground state is delocalized between metal ions, it is not strictly appropriate to describe the intervalence transition band of a class I11 complex as a metal-to-metal charge transfer transition. Nevertheless, the MMCT designation will be used for the sake of simplicity. The dependence of intervalence transition energy on the activation barrier to thermal electron transfer in Eqs. (1) and (3)allows the solvent dependence of intervalence transitions to be understood in terms of relationships developed for the Marcus Theory of electron transfer (5, 6). For a weakly coupled symmetric system, Eo, can be written ( 4 ) approximately as

FIG.3. Potential energy diagram for a symmetric class I11 mixed valence complex.

DINUCLEAR RUTHENIUM COMPLEXES

279

where AE ' accounts for any additional energy associated with excitation to either a spin-orbit or a ligand-field excited state (14-16),xi is the inner sphere reorganization parameter, which can be written,

and xo is the outer sphere reorganization parameter, which can be expressed in the form,

where d2,d3 and f2, f3 are the metal-ligand bond lengths and the force constants for the metals in oxidation states I1 and 111, respectively, r is the molecular trapping site radius, d is the distance between metal centers, and Do, and D,are the optical and static dielectric constants of the sovlents. From Eq. (191, Eopshould be linearly proportional to the solvent term (l/Dop- l/Ds). This relationship has been demonstrated for weakly coupled mixed valence complexes (see Section (II.C.2.c) and this gives qualitative support to the Hush model. The recent application of electroabsorption (Stark effect)spectroscopy to mixed valence complexes has shown that the actual charge transfer distance [R in Eq. (1511 is considerably smaller than the metal-tometal separation (17,18).This was not unexpected because it has been previously suggested (7) that the donor or acceptor wave function can extend itself onto the LUMO or HOMO of the bridging ligand, respectively. A consequence of this is that values of Hadderived using Eq. (15)are significantly underestimated in the Creutz review (7) and Table VI of this chapter, which used metal-metal separation for R. In addition, the larger molecular trapping radius r means that a key boundary condition in the derivation of Eq. (19) (i.e., d > 2r) may no longer be satisfied (19).A more appropriate approach to the problem of solvent reorganizational energy under these conditions may be the ellipsoidal cavity treatment (20,211, where

280

ROBERT J. CRUTCHLEY

and where

and

A0

-

[(A2/A2- B2)]"2.

(23)

In Eqs. (20)-(231,R is the interfocal length,Di, is the dielectric constant within the cavity, 5 k = d / R , P , and Q, are Legendre polynomials of the first and second kinds, A is the length of the semimajor axis, and B is the length of the two semiminor axes of the ellipsoid. 2 . The Comproportionation Equilibrium

The comproportionation constant K, (7)is determined by the equilibrium for class I and I1 complexes, (11-11) + (111-111)

+2(111-11),

(24)

and for class I11 complexes

K, is usually deduced from the difference in potential between metalcentered reduction couples. For class I and I1 complexes, K, is usually less than lo3 and is determined by contributions from four terms,

AG,

=

AG,

+ AGe + AG, + AGi,

(26)

where AGc is the free energy of comproportionation, AGr is the free energy of resonance exchange, AGe is the free energy that takes into account the repulsion of like charged metal centers, AGs is an entropic factor (bRT In f) that reflects a statistical distribution of Eq. (241, and AGi is an inductive or synergistic factor due t o stabilization of M(I1) and M(II1) or vice versa. For a class I complex, only AG, makes a significant contribution to AG,. For a class I1 complex, Sutton and Taube (22) have shown that all terms can contribute, with AG, being relatively minor and for symmetric systems given by

AG,

=

Had2/E,,.

(27)

DINUCLEAR RUTHENIUM COMPLEXES

28 1

It is possible to estimate K , when it deviates only slightly from the entropic factor of 4 by an amount related to the Coloumb repulsion energy of the metal charges (23). Thus,

K,

=

4 exp(l/sr,,kT)

(28)

where E is the dielectric constant of the material between the metal atoms separated by the distance rmm. For class I11 complexes, K, is very large. This reflects the strong bonding interaction between donor and acceptor wave functions, which results in delocalization of the odd electron. The greatest contribution to AGc is derived from AGr, where AHr = Had. 3. Other Theoretical Approaches

The Hush model is the preferred method of analysis of mixed valence complexes for the experimentalist because of its readily understandable derivation, its overlap with the Marcus theory of electron transfer, and the facility of its application. However, it is applicable only to weakly coupled class I1 complexes. A quantitative theory that is applicable to all mixed valence complexes is obviously desirable. Piepho, Krausz, and Schatz (24)have developed a vibronic coupling model (PKS model) that is applicable to both class I1 and I11 mixed-valence complexes and allows explicit vibronic eigenvalues and eigenvectors to be determined. The PKS model is a two-site, one-dimensional model that uses a valence-bond type of electronic basis and three parameters ( E electron coupling, A vibronic coupling, and F,the wave number of the totally symmetric metal-ligand stretching vibration, which is usually taken as 500 cm-'). The E and A parameters are adjusted so that the position and width of the predicted spectrum match those of the observed spectrum. Buhks (25) and Wong and Schatz (26)have extended the PKS model to incorporate the motion of solvent as well as the internal motion. The PKS model has been criticized (27-29) for the assumption of a single frequency and its unsatisfactory description of the magnetic circular dichroism (MCD) found for the intervalence band of the Creutz-Taube ion (30).In a series of papers (31),Ondrechen and coworkers developed a more realistic three-site model for delocalized (class 111) bridged mixed-valence complexes. This model incorporates many features of the PKS model but differs in that it explicitly includes the bridge, it uses a molecular orbital basis of one-electron functions, and the choice of important vibrational modes is not the same. The near-IR band line shape of the Creutz-Taube ion has been reasonably

282

ROBERT J. CRUTCHLEY

analyzed by using the three-site model, which by interation fits only one Hamiltonian parameter (31f). However, the model is appropriate only for delocalized systems. Piepho has responded to the criticisms of the PKS model by developing an improved version, the MO vibronic coupling model for mixedvalence complexes (32).Multicenter vibrations are now considered and a molecular orbital basis set (as with the three-site model) is used. This model was used t o calculate band shape and g values for the Cretuz-Taube ion (33).The MO vibronic coupling model is admittedly more empirical than the three-site model but it has the advantage in being applicable to all mixed-valence complexes. In a recent study, Curtis and coworkers (34) developed an analysis of electronic coupling in mixed-valence complexes that combines electrochemical results with an extension of Mulliken’s theory of donoracceptor interactions. The approach was applied to class I1 and near class I11 complexes. It was found that the degree of electronic coupling in these systems is at least three times that which would be predicted based solely on spectroscopic measurements (Eq. 15of the Hush model). However, the electronic coupling determined for complexes that were very close to, if not of, class I11 character was significantly smaller than that predicted by Eq. (16).

B. CLASSI11 MIXED-VALENCE COMPLEXES 1. Class ZZZ Criteria

Table I compiles the electrochemical and metal-metal absorption band spectral data of novel mixed-valence complexes (35-46) that are good candidates for valence delocalized systems. The decision as t o where to draw the line between class I1 and I11mixed-valence complexes is a difficult one to make. The Creutz-Taube ion has undergone an extremely rigorous physical characterization by just about every method known (see below), and it is only recently that the preponderance of evidence strongly favors a class I11description. The same degree of examination should be applied to the complexes of Table I to fully justify their class I11 assignment. In preparing Table I two criteria were used. First, the comproportionation constant for symmetric mixed-valence complexes should be large. It is important to recognize that for complexes in which the metal-metal distance is small significant contributions to K , arise from factors other than those associated with metal-metal coupling. Second, the breakdown of Hush theory ( 4 )for the spectral properties of MMCT transitions should be apparent. In practice this means that the pre-

283

DINUCLEAR RUTHENIUM COMPLEXES

TABLE I RuTHENIuM(III/II)

Complex

COUPLES,COMPROPORTIONATION CONSTANTS, AND MMCT BANDPROPERTIES OF CLASS 111 COMPLEXES El/, iV) 0.69O 1.58 1.52a 2.02 1.67' 2.22 0.38O

0.89 -0.09' t0.46 -0.06' t0.51 -0.130" t0.160 0.9of 1.29 0.50'

0.34 1.40" 1.70 0.325" 0.575 0.220' 0.575 0.117 0.536 0.242c 0.582 0.128 0.563 0.347c 0.669 0.351' 0.672 0.260' 0.628 0.336' 0.699 0.0W 0.440

Versus SCE. *Not observed. Versus ferroceniumlferrocene couple Measured at 240 K. Not reported. f Versus NHE.

AE

Amar

"112

EmUr

(mV)

K,

(nm)

(em-')

1M-l crn-l)

890

lOI5

1450

=2000

500

3 x 108

b

CH3CN

37

550

2 x 109

b

CH&N

37

510

4 x

CICH2CHzCI

38

550

3 x 109

CH,CN

39

570

6 x lo9

CH,CN

39

290

8

HZO

40

390

4 x 106

CH3CN

41

160

600

1820

CH3CN

42

300

lo5

1520

CH3CN

43

250

16000

1600

929

331

C6H5CI

44

355

lo6

1703

1970

3040

CH3CN

45, 46

419

lo7

1624

1980

4300

CH,CN

46

340

6 x lo5

1685

1850

5940

CH,CN

45, 46

435

lo7

1595

2420

3050

CH3CN

46

305

lo5

1690

1940

9100

CH3CN

46

x

lo8

lo4

1410

2600

i8ood

1500

500

7800 15000~

1240

2820

4300

27000

Solvent Acidic H,O

Ref. 35, 36

321

3 x

lo5

1692

1940

4360

CH3CN

46

368

2 x lo6

1648

1770

6000

CH3CN

46

363

lo6

1651

1790

4570

CH3CN

46

425

lo7

1597

1760

4160

CH3CN

46

284

ROBERT J . CRUTCHLEY

dicted band width of the MMCT band by Hush theory is significantly greater than that seen experimentally, the extinction coefficient E,,, is large (usually >lo3 M-' cm-' ), and Vmax is solvent independent.

2 . Studies of Class III Mixed-Valence Complexes a. Metal-Metal Coupling. It might be expected that complexes possessing a large comproportionation constant and metal-metal coupling will have a correspondingly large MMCT extinction coefficient. However, such a correlation should be made with caution. For example, the complex [{(NH3)4R~}2(p-bpt~)]5+ (35,36) has a large K, but only a small MMCT E,,, (500 M-' cm-'1. In contrast, the complex [(tterpyRu),(p-tpbp)I3+(42)has a relatively small K, (600) but a large cmax(27,000 M-' cm-'1. In a comparison (36) between [{(NH3),Ru),(p-bptz)15+and the Creutz-Taube ion [MMCT,,,A = 1596 nm and E = 7620 M-' cm-' in acetonitrile (46)1,the authors pointed out that, although the metal-metal distances are the same, the d.rr orbitals in [{(NH3),Ru},(p-bptz)15+ are directed toward the middle of the a-diimine chelate system and are not linearily oriented toward each other as in the Creutz-Taube ion. Exactly how this explains the difference in MMCT extinction coefficient or more quantitatively the oscillator strength was left to speculation. The above examples would seem t o be excellent candidates for analysis by the MO vibronic coupling (33) and three-site (31) models that were discussed previously. Brant and Stephenson (47)have suggested that X-ray photoelectron spectroscopy (PEW can be used to distinquish between strongly and weakly coupled metals in delocalized class I11 complexes. The PES spectra of Creutz-Taube type complexes show two Ru 3d512 photoionization states of equal intensity and split by 3.4-3.6 eV. This was initially thought to be evidence for class I1 properties. However Hush (48) suggested that this behavior was not inconsistent with class I11 properties if the delocalized valence orbital is highly polarizable so that under the influence of the localized metal core hole the unpaired electron associated with this orbital becomes trapped and two final states are realized. If the metal-metal coupling in class I11 complexes grows strong, the polarizability of the delocalized valence orbital is diminished until finally the metal core level spectrum must approach a single final state. This condition was met in the mixed-valence complex [{[As(p-tol)3],C1Ru}2(p-Cl~31, for which in the PES spectrum only one Ru 3d5,z photoionization state was observed (47).

b. Nature ofthe Bridging Ligand. Whether a class I1 or class I11 mixed-valence complex is formed obviously depends a great deal on

DINUCLEAR RUTHENIUM COMPLEXES

285

the nature of the bridging ligand. To be effective,the molecular orbitals of the bridging ligand must be symmetry and energy matched and interact simultaneously with the donor and acceptor orbitals of the metal ions. Having framed the requirements in this way, it is easy t o see why in class I11 complexes the bridging ligands all involve extended 7 systems. An examination of class I11 complexes in Table I and in the Creutz review (7) shows that there are two main types of bridging systems. The nitrogen donor atom heterocyclic bridging ligands are 7~ acceptors. This means that the Ru(I1)7-donor wave function can extend onto the 7~ LUMO of the bridging ligand and the dominant mechanism for metal-metal coupling is called electron transfer superexchange. Conjugated anion bridging ligands are 7~ donors. This allows the Ru(II1) .rr-acceptor wave function to extend itself onto the 7 HOMO of the bridging ligand and the dominant mechanism for metal-metal coupling is then called hole-transfer superexchange. The above descriptions are rather simplistic. Theoretical studies (49-51 ) show that metal-metal coupling is mediated by the full set of bonding and antibonding orbitals of the bridging ligand. The superexchange pathways described above are those which are expected to make the greatest contribution to metal-metal coupling. If Ru orbitals closely match the energy of the HOMO or LUMO of the bridging ligand, covalency is approached and the distinction between formal oxidation states becomes blurred, resulting in the extreme case of a radical bridging ligand. The situation is illustrated below:

For the complex [{(bpy),R~},(p-adc)]~+ (38), researchers invoked a contribution of resonance form I1 to the ground state in order to explain ESR and electronic absorption spectroscopy results. Alternatively, if Ru orbitals closely match the energy of the LUMO of the bridging ligand, the following resonance forms may result:

Elaborations of these basic concepts can be found in studies of the comproportionation constant of mixed-valence complexes (37, 52,53). It is a synthetic challenge for the inorganic coordination chemist to construct class I11 systems whose ground states have a significant

286

ROBERT J. CRUTCHLEY

contribution from resonance forms I1 or IV. These systems will have properties that will be directly applicable to the construction of molecular materials and devices. c . Nature of the Ruthenium Ion. The energy of Ru(I1) donor and Ru(II1) acceptor orbitals and the ability of these orbitals to interact with the bridging ligand can be significantly affected by the nature of the ligands which make up the coordination sphere. It is well known that the potential of the Ru(III/II) couple becomes increasing anodic with increasing coordination of T acceptor ligands [the [ R U ( N H ~ ) ~ ] ~ + ” couple = 0.051 V (54),whereas the [ R ~ ( b p y ) ~ ] couple ~ + ’ ’ = 1.272 V vs NHE in aqueous solution (55)l.It follows from the previous discussion that symmetric mixed-valence complexes are more likely to have delocalized valency if the (NH,),Ru(II) moiety is bonded to a T acceptor bridging ligand and if a (diimine),Ru(III) moiety is bonded to a T donor bridging ligand. In an impressive body of work, Curtis and coworkers ( 4 5 , 4 6 )have synthesized derivatives of the Creutz-Taube ion to probe the influence of the coordination sphere on the magnitude of metal-metal coupling. Class I11 properties could be transformed to class I1 properties, simply by changing an ammine ligand to a strong T-acceptor ligand. Both the increase in MMCT band AVl,z (as shown in Fig. 4) and the decrease in extinction coefficient were consistent with a transformation to class I1 properties. In addition, the symmetric complexes [LRu(NH,),(p-pyz)Ru(NH3),L15+ (46) have smaller MMCT ATliz compared with their asymmetric analogues. This suggests that asymmetry helps trap out the odd electron and promotes conversion to class I1 properties. Similar effects are seen due to outer-sphere perturbations (discussed below). Examination of these Creutz-Taube derivatives using Piepho’s MO vibronic coupling model (32)would be a good test of the model and, if successful, could reveal some important details concerning the dependence of metal-metal coupling on the nature of the metal atom as perturbed by its coordination sphere. Modification of metal-metal coupling is not just restricted to the inner coordination sphere. Hupp and coworkers (56)have shown that addition of the crown ethers, dibenzo-36-crown-12 (DB-36-C-12)or dibenzo-30-crown-10 (DB-30-C-10), to a nitromethane solution of the Creutz-Taube ion or trans-[((NH,),R~(py)}~(ppyz)]~+ can result in dramatic changes to the MMCT band because of the affinity of the crown ether oxygens for ruthenium-ammine hydrogens. Figure 5 shows the changes observed for the MMCT of truns-[{(NH,),Ru(py)}~~p-pyz)15+. The maximum effects occur with the 1: 1stoichiometry of crown ether +

287

DINUCLEAR RUTHENIUM COMPLEXES 0

2,6-Me2pyz

2,2'-bpy

0.100 0.150 ti[E t (Rub)] (volts VS. SCE)

0.050

5

FIG. 4. A;, 2 , the band width a t half-maximum for the MMCT band vs G[E,,,(Rub)], the potential shift in the second redox wave relative to the Creutz-Taube second wave for complexes t r a n ~ - [ L ( N H ~ ) ~ R u ( ~ - p y z ) R u ( Nwhere H ~ ) ~L] ~=+ 3,5-Me2py, , py, 3-Clpy, 3-Fpy, 2,6-Me2pyz, and bpy. [Reproduced with permission from Ref. (45).Copyright (1985) American Chemical Society.]

1 .o

0.8

8

-e

$51

0.6 0.4

0.2 0.0

1200

1600

Wavelength, nm FIG.5. MMCT absorption spectra for 8.6 m M of t r a n ~ - [ { ( N H ~ ) ~ R u ( p y ) ) ~ ( ~in- p y z ) l ~ ~ nitromethane-d, with (a) no added crown, (b) 1 eq. of DB-36-C-12, and (c) 16 eq. of DB-36-C-12 (56).[Reproduced with permission from Ref. (45). Copyright (1985)American Chemical Society.]

288

ROBERT J. CRUTCHLEY

to complex. It was suggested that the initially delocalized class I11 complex becomes valence-localized class I1 with the formation of the asymmetric 1: 1 adduct. Further addition of crown ether forms the symmetric 2 : 1 adduct and a return to class I11 properties. Roberts and Hupp (57)have seen a similar affect although not as dramatic by adding dimethyl sulfoxide to an acetonitrile solution of trans[{(NH,),Ru(py)},(p-pyz)15+.The changes in the intervalence spectrum of the complex were suggested to arise from a valence-trapped ground state due to the selective solvation of ammines bonded to Ru(II1). Importantly, the Creutz-Taube ion’s intervalence transition was largely insensitive to DMSO concentration and this may reflect the magnitude of metal-metal coupling in this complex. d . Physical Characterization of the Creutz-Taube Ion, [(NH3)$u(p-pyz)Ru(NH,),15+. The controversy concerning the classification of the electronic structure of the Creutz-Taube ion appears to have reached a consensus view in favor of the delocalized class I11description. A number of studies (17, 16, 58-66) have appeared since the Creutz review (7) and these have been summarized in Table 11. The near-IR, IR, and MCD spectra of the Creutz-Taube ion are of particular interest because of theoretical predictions of low-energy tunneling transitions near 2000 and 3200 cm-’ (30,67).FTIR spectroelectrochemistry of the TABLE I1 THE CLASSIFICATION OF THE ELECTRONIC STRUCTURE OF [{(NHg)5Ru}z(~~-pyz)]’ ’ Experiment Far IR

Conditions

6 K Docosane mull

Time resolution

Year

10-13

1981

X ray IR and MCD Solvent dependence Mossbauer Near and middle IR Electroabsorption

Single crystal 3 K 100 K,295 K 1.4 K In polymer support Room temperature Liquid He temperature D,O solution 50% Glycerol/H,O Liquid Nz

58

59

10-9

1984

10-18 10-13

1985 1986

10-13 10-7

1987 1989

Delocalized Delocalized

10-13

10-19

Ref.

Delocalized Delocalization is possible Delocalization is possible Delocalized Delocalized

Mossbauer, X ray, EPR EPR

Conclusion

Delocalized 1989 1989- 1991 Delocalized

60 61 62, 63

64 65

66 17,18

DINUCLEAR RUTHENIUM COMPLEXES

289

complex in D,O (66) showed a broad band at ca. 2000 cm-' ( E = 300 M-'cm-' and Ai71,2 = 1400 cm-') together with a suggestion of a band between 4500 and 5000 cm-'. The latter band is partially obscured by the intense 6320-cm-' absorption but appears to have a band width and an intensity similar to that of the 200-cm-' absorption. A near-IR MCD spectrum of the Creutz-Taube ion in a PVA foil (63) showed a weak C term in the region of the main band at 6400 cm-'. The MCD intensity was consistent with an approximately 1% or greater (XY) polarization of the absorption intensity with 2-polarized (metal-metal axis) intensity dominant. The band at 2000 cm-' was also MCD active (621, in agreement with electronic coupling models (30, 31e).

C. CLASSI1 MIXED-VALENCE COMPLEXES 1 . General Description

There have been numerous class I1 mixed-valence complexes synthesized since the Creutz review ( 7 ) .Tables I11 and IV are compilations of symmetric deca- or octaamminediruthenium (68-80) and tetrakisdiiminediruthenium complexes (37,41,81-90),respectively. Table V is a compilation of asymmetric mixed valence complexes (45,46, 77, 83, 86,87,91-102). Table VI summarizes the Hush theory treatments of mixed-valence complexes that have appeared in the literature. A class I1 valence trapped complex is generally recognized by its small comproportionation constant (K, is usually <1000), the weak intensity of its MMCT absorption band, a value of ATliz that is greater than the values given by either Eq. (2) or (4) (see Table VI), and MMCT energy that is sensitive to the nature of the solvent. The experimental band width for a class I1 is usually greater than the theoretical bandwidth because of anharmonicity and other effects, such as a significant spin-orbit term for ruthenium systems (1000 cm-'1 (103). When the experimental Ai& is less than or equal to the theoretical AV,,z, class I11 behavior is approached and electroabsorption spectroscopy (1 7,18,104)would seem to be the best method to unimbiguously establish the degree of delocalization. 2 . Studies of Class 11 Mixed-Valence Complexes a. Electroabsorption (Stark Effect) Spectroscopy. The recent application of electroabsorption spectroscopy to the MMCT bands of mixedvalence complexes is a n important development ( 1 7,18,104).An external electric field will perturb the electronic/vibronic absorption envelope of an MMCT band in a manner, which is directly relatable to

290

ROBERT J. CRUTCHLEY

TABLE I11 COMPROPORTIONATION CONSTANTS AND MMCT BANDPROPERTIES OF SYMMETRIC CLASSI1 COMPLEXES, I{(NH3),Ru},(p-L)lm *

1040 1110 -1000 a 0 0

40

-1400 =1200 -10 =lo08 =4 808 =4 690 1.2 x lo5 1510 55 975 340 1100 155 1220 75 1250 = 10 990 2400 1550 343 1400 33.2 1058 11.2 752 23.4 862 7.0 781 c 8.5 10 -770 10 -770 c 9 10 752 4 800 P 10 C 28 r 21 13 ~1220 640 5.4 x lo4 1690 7730 1140 a

Broad and ill defined from 850 to 1500 nm. Not corrected for comproportionation constant. No MMCT band was observed.

1430 4900 4700 1250 =4800 -1220 =160h ~ 1 6 ~ ~0.4" =500 =40 48 9 2 1 6400 2.3 2 0.7 7000 430 5040 150 2820 6200 6000 2530 2300 5950 1600 3700 6200 6000 41 98 5400 5860 8 5600 507 5900 135 20 140 12

-575 40 1525 920

NOzPh NOLPh NO2Ph CHJN CH3CN CHJN D20 D2O 02MDCI 02MDCl 02MDCl D2O CH3CN 01MDC1 01MDCI 0.1 M DCl

DMF

1 M DzS04 1 M DLSO, 1 M D2S04 1 M D2S0, 1 M DlS04 1 M DJSO, 1 M DZSO, ~ 6 0 0 0 1 M D,SO, =6000 1 M DLSOl 1 M DLSO, ~ 6 0 0 0 1 M D,SO, =6000 1 M D,SO, D2O D2O D2O ~ 4 3 0 0 DzO Acetone 3310 CH,CN 4010 CHSCN

68 68 68 69 69 69 70 70 71 72 72 73 74 75, 7 75 75 76 76 77 77 77 77 77 77 77 77 77 77 77 78 78 78 78, i 80 46 46

291

DINUCLEAR RUTHENIUM COMPLEXES

TABLE IV COMPROPORTIONATION CONSTANTS AND MMCT BANDPROPERTIES OF SYMMETRIC CLASS11 COMPLEXES. [{(DIIMINE)~Ru}~(~L)]"* Complex" 105 105 105 8.2 x 104 9 x 104

748 5.6 x 105 2000 1600 1100 510 3400 22 49 49 49 107

O1

1800 1850 1850 1062 1950 1011 2000 1370 1700 2000

2400 2200 2400 1300 3400 1700 4650 <100 2000

3100 2520

1000 42 650 458 735 589 1618 729 709

4340 16900 5000 4500 4700 4700 4008 6036 4934

3300 4200 4170 2300

83

b

b b

b b

b

1639 910 1320 1270 1330 1300 1580 1295 1150

81 81 81 82-84

85 85 86 84 37 41 73 87 87 88 89 90 90 90 90 43 43 43

In acetonitrile except where noted. A MMCT band was not observed. In D20.

changes in the polarizability (Aha) and the permanent electric dipole moment (Ap) resulting from the MMCT transition. This will allow unequivocal answers to questions regarding the extent of electronic delocalization and may also reveal and permit characterization of MMCT transitions that overlap with more intense MLCT or LMCT bands. The change in absorption, AAb),which results from the application of an external electric field, Fext,across a sample of nonoriented and

292

ROBERT J . CRUTCHLEY

TABLE V COMPROPORTIONATION CONSTANTSAND MMCT BANDPROPERTIES CLASS I1 COMPLEXES

OF

ASYMMETRIC

C om p I e x <350 350 3.4 x 10s 3.6 x 10' 3.5 x 103

1.2 x 107 1.0 x lop 1.0 x 102 6.1 x 10' 7.1 x loS 7.1 x 103 5.9 x 102 5.4 x 10:' 910 510 330 5.1 x 10; 4.7 x 10' 4.6 x lo7 4.7 x 107 1.9 x 108 3.4 x 108 4.4 x 104 6370 1.3 x lo6 3.5 x 106 1.2 x 107 1.4 x 107 2.2 x 107 1.5 x 10' 4.9 x 107 4660 6360 5240 4660 4660 4480 3550 944 2.7 x 107 1.3 x 109 2400 1100 6 x loL9 3 x 10'8 1017 1.5 x 10'

945 940 735

450 41 1 36

820

150

994 1855 1030 1445 1140 960 920 935 900 1035 1180 1160 1176 1266 1190 1308 1232 1304 1313 1310 1490 1508 1346 1153 1441 1541 1020 847 908 892 860 873 860 850 840 1026 1028 1050 1042 1035 1033 960 994 720 761 1640

1700 3200 2100 2160 730 575

1300 1240 713 1340 918 518 10A2 490 868 400 826 3400 2420 2430 1650 2920 2790 1275 23 1010 1145 560 480 1000 940 660 890 1580 1780 1480 1080 940 450 660 295 900 350

I,

3700 4820

5470 5430 5650 5480 5120 5440 4950 5910 5360 5650 5940 3180 3350 3880 4040 3160 3370 4980 5100 5300 4900 5170 5280 5320 5280 5420 5740 4840 4850 5070 5080 5000 6050 5970 5060 4620 5500

b

694 675 781 640 1094 b

3500 2840 3800 3250 7000

~

a

5445 5464 4565 4845

Where geometric isomerism is possible, all complexes are trans unless otherwise noted Not observed.

4500 4520 4300 5200

CH,CN CH3CN CH$N CH3CN CH-,CN

92 92 92

CH&N

83

CH,CN CHCI, CHCI, CHCI CHCI, CHJCN CH,CN CH,CN CH3CN CH,CN CHJN CHJN CH3CN CH,CN CH,CN CH,CN CH,CN CH,CN CH3CN CH ,CN CHICN CH$N CH,CN 1 M D2S0, CHJN CH3CN CHJCN CHJCN CH,CN CH,CN CHJCN CH3CN CH,CN CHJCN

86 9.9

CH3CN CH3CN CH$N CHJN CH3CN HdO H2O CH,CN CH3CN CHICN

H2O HZO HZO Acetone HZO HZO

91

8,'1

9.3

9.3

93 45, 9 94 94 4.5. 9 94 94 45. 9 94 94

94 94 4.5. 9 4.5. Y 45. 9 .15. Y 41;

46

__

41;. i /,

46 4t1 4h 46 4t1 Jh

46 4h

46

46 46 46

46 46 46

95 95 96

8; 8; 9; 9x. I

9; 1O( 101 10:

TABLE VI HUSHMODELCALCULATED PARAMETERS OF CLASS 11 MIXED-VALENCECOMPLEXES

5464 4565 4845 5445 3300 4200 4170 2300 3100 2520 3700 4340 16900 4900 4700 =4800

6400 7000 5000 4500 4700 4700 3500 5040 6200 6000

5950 5470 5430 5650 5480 5120 5440 4950 5910 5360 5650 5940 3700 5060 4620 5170 5100 6000 5400 5860 5600 5900 -6000 -6000 -6000 -6000 4300 4008 6036 4934 Calculated from data provided

4959 5605 5307 4994 3580 3530 3530 3440 4100' 3685" 4300 1780 5040 4710 4560 =4810

5300 5800 4190 4270 4160 4220 3400 4867 4580 4350 4300

7

I 7 7.1 6.184 6.184 6.184 5.5 12-15 12-15 6.9 8

4.60 15.8 18.1 20.6

14.4 17.6 6.9 6.9 6.9 6.9 7 10.1

453 138' 276O 465 700 740 770 725 60-80 240-300 456" 4700" 400 340 260 =240 476-780 -266

467a 379" 478' 4349 279 179'

6.9 6.9 6.9 6.9 6.9 6.9 6.9

7790 710" 550' 7450 570"

6.9

420' 560a 390' 575"

6.9 6.9 6.9

460' 6OOn

3900

4060 4670 5540 5180 5440 5480 5480 5370 5540 4350 3820 4220 4480

185 0.10 0.51 1.90 16 19

20 20 0.07-0.12° 1.67-2.6O0 2.7 5.9" 1.3 1.20 0.8'

~0.6

92 92 92 91 81 81 81 82 85 85 86 88 89 68 68 68

70 70 72 72 3.P 2.3O 4.1"

3.2"

90 90 90 90

0.72 0.3'

73 75 75, 76 75 75

6.5O 7.0° 4.1 7.6 5.2 3.0" 62 2.7 5.30 2.6' 5.7'

94 94 94 94 94 94 94 94 94 94 94

76 9 9 9.3 8.2 6.0 5.9 10.2 11.8 12.1 8.1 14.0 12.0 10.2 13 11 15.5 20

300

510 525 93 143 236 50 314 169 104 148 48

62 <200

380 240 180

95 95

77 77 77 77 77 77 77 77 77 77 77 78, 79 43 43 43

294

ROBERT J. CRUTCHLEY

immobilized molecules, can be described (18) by a linear combination of zeroth, first, and second derivatives of the absorption band:

where A , = Dl3 + (3 cos2 x - 1)E/30; B, = 5F + (3 cos2 x - 1)G; C, = 5H + (3 cos2 x - 1)Z;D = S ' ; E = [3S2 - 223'1; F = (i)Tr(Aa) + R'.Ap; F + G = (#)m.Aa.m+ Rz.Ap;H = lAp12;H + Z = 3 ( m ~ A p ) ~ . x specifies the experimental angle between the external electric field and the light polarization at frequency u, and h is Planck's constant. The scalars S1 and S2, and the vectors R' and R2 are functions of the transition moment polarizability and hyperpolarizability tensors. m is a unit vector oriented along the transition dipole moment and Flntis the internal electric field at the molecule, which depends on the externally applied field Fextsuch that

where f is a tensor that corrects for the local electric field and depends on the dielectric constant of the medium. Figures 6 and 7 show absorption and electroabsorption spectra of [{(NH3)5R~}2 (p-pyz)15' and [{(NH3)5Ru}Z(p-4,4'-bpy)15 , respectively. The change in AA as a function of x is uniform for the bands, which indicates that the molecular properties that give rise to AA are identically oriented with respect to the transition dipole moment. The electroabsorption spectra in the near-IR region (MMCT bands) give the greatest differences between complexes when analyzed with Eq. (31) and these are shown in Fig. 8. For the Creutz-Taube ion (Fig. 8A1, the spectrum does not satisfactorily reduce to a sum of derivatives but nevertheless shows that AA(v) line shape to be modeled primarily by a negative zeroth derivative (A,) term, especially at energies below 6500 cm-'. The fit in this case yields a value for IApI = 0.7 ? 0.1 D, which when compared with the maximum permanent electric dipole moment (lAplmax= 32.7 D, assuming a metal-to-metal distance) is strong evidence for a delocalized ground state. Contrast this result with the analysis of the electroabsorption spectrum of [{(NH3),Ru),(p-4,4'bpy)15+shown in Fig. 8B. In Fig. 8B, the linear combination of derivatives closely models the AA (v)line shape of the near-IR band. The second-derivative component +

295

DINUCLEAR RUTHENIUM COMPLEXES

I

- 7 . 7

1.0-

a

Oa5i0.0

1

I

,

.

,

A \------J I

I

.

.

.

p54.7' -

-30

nI\

,

'A

I I

--

. .

. '

.

I

1

.

I 1

' I

','

\I

. .

I

10000

I

.

.

I

20000

15000

ENERGY (c rn-1)

FIG.6 . (A)Absorption and (B) electroabsorption spectra of [ { ( N H B ) 5 R ~ } 2 ( ~ - payt~ ) ] 5 ' 90" (-) and x = 54.7" (---) are shown. Fifty percent glycerol/H,O, 77 K, FeKt = 4 x 10' Vicm. [Reproduced with permission from Ref. (18).Copyright (1991)American Chemical Society.]

x

=

1500 . n

'."r

'

WAVELENGTH (nm) 1000 500 1

8

,

'

I

-

A

A

a

10000

15000

.

.

.

.

20000

ENERGY (c rn-1)

FIG.7. (A)Absorption and (B)electroabsorption spectra of [{(NHB)5Ru}z(~-4,4'-bpy)]5+ a t x = 90"(-)and x = 54.7"(---I are shown. Fifty percent glycerol/H20,77 K,F,,, = 4 x lo5 Vicm. The near-IR absorption has been magnified 5 x . (Reproduced with permission from Ref. (18).Copyright [1991]American Chemical Society.)

296

ROBERT J. CRUTCHLEY

6000

7000

ENERGY (cm-1)

10

In

0 F

X

d

0

6000

8000

ENERGY

10000

(ern-1

to electroabsorption data (0) for I{(NH,)jRu},(~-pyz)l~ a t x = 90" in the near-IR region of the spectrum. Zeroth (D),first (+), and second ( A )components for [{(NH3)5R~}2of the fits are also shown. (B) Fit (-i to electroabsorption data (0) (p-4,4'-bpy)I5+at x = 90" in the near-IR region of the spectrum. Zeroth (D),first (+), and second (Ai components of the fits are also shown. [Reproduced with permission from Ref. (18).Copyright (1991) American Chemical Society.]

FIG.8. ( A )Fit

(-)

dominates the spectrum and the fit yields a value for IApI = 28.5 f 1.5 D, indicating that the states involved are highly dipolar and the unpaired electron is essentially localized. The value of I Ap I is substantially smaller than [ApImax= 54.3 D. There are a number of possible explanations of this difference (111, but it seems likely that the effective separation Re between metal ions is reduced by delocalization of the donor wave function onto the bridging ligand. The consequences of a smaller separation between metal ions has already been discussed in Section II.A.l, where the application of an ellipsoid model for outersphere reorganization energy was introduced. Hush theory calculations of Hadand the mixing coefficient (Y [in the Creutz review (7) and Table VI of this chapter], which use R = metal-metal distance, are largely underestimated. For example, Sutton, Sutton, and Taube (105) have reported Hadand a values for [{(NH,),R~)~(p-4,4'-bpy)]~+ of 460 cm-' and 0.057 [R = 11.3 A in Eq. (15)]. Hupp, Dong, Blackbourn, and Lu (106) have estimated for the

DINUCLEAR RUTHENIUM COMPLEXES

297

above complex Re = 5.1 & 0.7 A, which yields Had= 1020 ? 130 cm-’ and a = 0.126. Clearly, electroabsorption studies of all class I1 complexes should be done to get more reasonable estimates off?,, and a.

b. Raman and Electronic Absorption Spectroscopy. The intervalence-enhanced Raman scattering from [(CN),Ru(p-CN)Ru(NH,),] allowed a mode-by-mode assessment of the Franck-Condon barrier to intramolecular electron transfer (107). The results agreed well with X-ray crystallographic measurements. An attempt to apply this technique to the symmetric mixed-valence complexes [((bpy),RuC1},(4,4’bpy)I3+and [{(NH,),R~}~(4,4’-bpy)]~+ was unsuccessful due to the overlap of MLCT and MMCT bands (108).The symmetric (CN),Ru analogue proved to be thermally unstable. The resonance Raman spectrum of trans-[{(bpy),R~(CN)},(p-CN)]~+ under near-resonance conditions with the MMCT band showed enhancement of the bridging cyanide stretching as expected for this type of electronic transition (109).Analysis of the IR spectrum supports the valence-localized model in contrast to a previous study (110). The broadness of the intervalence electronic absorption band and the lack of structure has generally lent themselves to a rather simplistic singly degenerate state-to-state description. Multiple MMCT transitions have been observed for osmium dimers ( I l l ) , for which the spin orbit coupling term (A) is large (for Os, A = 3000 cm-’ and, for Ru, A = 1000 cm-’1. In addition, the lowering the symmetry of the ruthenium coordination sphere will split the degeneracy of dn orbitals and result in multiple MMCT transitions (112).These factors may explain the multiple MMCT transitions seen for ruthenium diimine complexes, which incorporate the 2,2-bibenzimidazolate (BiBzIm) bridging ligand (82, 83). ~

c. Solvent Effects. Studies into the solvent dependence of MMCT bands have proven to be especially fruitful in the years since the Creutz review (7). According to Marcus-Hush theory, the dependence of E,, on the outersphere reorganization energy is given by Eq. (19).Thus, for a given class I1 mixed-valence complex in which xi and AE’ in Eq. (17) can be considered solvent independent, shifts in E,, with the nature of the solvent should show a linear correlation with l/Dop-l/Ds.Figure 9 shows the linear correlation between E,, and l/Dop-l/Ds for the complex [{(NH,),R~},(p.-4,4’-bpy)]~+ (112).Similar correlations for other mixedvalence complexes can be found in the literature (74,91,92,113). The data point for D,O is more than three standard deviations above the best-fit line for the remaining solvents. This nonconformity had been ascribed to hydrophobic solvent effects or to solvent-induced nonlocal

298

ROBERT J. CRUTCHLEY

1

95

,

7.51 0.4

I

0.5

FIG.9. MMCT band energies vs l/DOp- l / D s .Key to the solvents: BN, benzonitrile; NB, nitrobenzene; DMSO, dimethyl sulfoxide; DMF, dimethylformamide; DMA, dimethylacetamide; PC, propylene carbonate; FA, formamide; NM, nitromethane. [Reproduced with permission from Ref. (112). Copyright (1987) American Chemical Society.]

dielectric properties (114,115). However, a plot of MMCT band E,, for the complex [((NH,),R~}~(p-4,4’-bpy)]~+ (at infinite dilution) vs l/Dop-l/Ds yielded a n E,, data point for D 2 0within two standard deviations of the best-fit line (106). Other solvent studies have shown a dependence of the MMCT band energy on complex concentration, added electrolyte (116-1 181, pressure (119,1201, and even the nature of the oxidant used to generate the mixed-valence complex (121). In asymmetric complexes of the type [(bpy),RuCl(p-pyz)Ru(NH,),LI4+, studies (94)revealed that there is a solvent donor-number (DNI-dependent contribution to the Frank-Condon barrier of approximately 0.006 eV/DN, which completely overwhelms the dielectriccontinuum-theory-derived (l/D,,,-1/Ds) solvent dependence typically observed in symmetrical dimers. In this case, variations in MMCT E,, with solvent give linear correlations when plotted against solvent dependent the difference in potential between the two ruthenium(III/II) couples, as shown in Fig. 10. The microscopic origin of this solvent effect was described by Curtis, Sullivan, and Meyer (122) in their study of solvatochromism in the charge transfer transitions of mononuclear Ru(I1) and Ru(II1) ammine complexes. The dependence

299

DINUCLEAR RUTHENIUM COMPLEXES

0.00

0.20

0.40

0.G 0

0.80

A E,,2(volts) FIG. 10. E , (or E,,J vs AE1,* as solvent is varied for the complexes [(bpy),RuCl(p-pyz)Ru(NH3).,LI4+, where L = NH3 (V)and L = pyridine (0). [Reproducedwith permission from Ref. (94).Copyright (1986) American Chemical Society.]

of charge transfer E,, on DN suggested a hydrogen-bonding type of interaction involving the solvent’s nonbonding electrons and the N-H bond. The strength of this hydrogen-bonding interaction will depend on the acidity of ammine hydrogen, which is greater for an ammine bound to a Ru(II1) ion. The net effect of this interaction is the transfer of electron density from the partially deprotonated ammine to the ruthenium ion. Thus, for the asymmetric mixed-valence complexes in Fig. 10, the Ru(II1)oxidation state is stabilized by strong donor solvents and hEliBincreases with increasing DN. An additional consequence is that the solvent stabilization of the Ru(II1) coordination sphere must decrease delocalization of the odd electron and the magnitude of metal-metal coupling. Preferential solvation of the Ru(II1) coordination sphere in the asymmetric mixed-valence complex [(bpy),RuCl(p-py~)Ru(NH,)~Ll~+ by DMSO in acetonitrile was found to make a solvent trapping barrier, which must be considered in situations involv-

300

ROBERT J. CRUTCHLEY

ing electron transfer processes of preferentially solvated solutes (123, 124).The proper choice of ligands in the design of the complex trans[(py)(NH,),Ru(p-4cp)Ru(bpy),C114+permitted the study of solventinduced or crown-ether-induced redox isomerization (125).Preferential solvation also occurs for symmetric ruthenium ammine mixed-valence complexes (118,126),because the solvent's interaction with the Ru(II1) ammine coordination sphere is greater than that for Ru(I1). Crown ethers were shown to have the same effect (127).Hupp (128)examined the effect of solvent on the extent of electronic coupling involving complexes of the type [(bpy)2RuC1(p-pyz)R~(NH3)4L14+. Electronic coupling increased with decreasing solvent-perturbed Eo [Eq.(311.In an unexpected result for a 7~ acceptor bridging ligand, these solvent effects were rationalized (in part) by a hole-transfer pathway for metalmetal coupling. The thermochromism of MMCT bands has been the subject of recent investigation (129-131 1. The MMCT band of asymmetric class I1 complexes was shown to shift to higher energy with decreasing temperature, whereas for symmetric class I1 complexes only a negligible temperature dependence was observed. This effect was shown to arise mainly from the temperature dependence of the redox asymmetry term E , in Eq. (3). Importantly, the MMCT bands of symmetric and asymmetric class I11 complexes show only a slight temperature dependence (113). It was speculated that this was due to a weak temperature dependence of Had. d. Distance between Ruthenium Ions. From electron transfer theory (10, 132), the distance dependence of the electron exchange integral

Had obeys the relationship

where p is a measure of the electron transmitting properties of the bridging medium and d is the separation between donor and acceptor wave functions. Thus, for a series of bridging ligands in which the value of p is constant, an exponential dependence of Hadon the distance of separation is expected. Figure 11 shows the distance dependence of Hadfor a series of mixed-valence complexes (77) of the type [{(NH,),R u } ~ ( ~ - L ) ]A ~ ' .similar relationship was found for a series of mixedvalence complexes of the general type [(NH,),Ru-py-(CH=CH),-pyRu(NH3)J5+,where py is pyridine (68,133). Reimers and Hush reanalyzed the raw experimental data for these complexes (23) and deconvoluted the electronic spectra into contributions from T + r*,metal-

DINUCLEAR RUTHENIUM COMPLEXES

301

1.01

-1

N u m b e r of C o n j u g a t e d A t o m s

FIG.11. Plot of log VAB(or Had)vs number of conjugated atoms in the bridging ligand for complexes of the type [{(NH3),Ru),(p-L)I5+. [Reproduced with permission from Ref. (77). Copyright ( 1983)American Chemical Society.]

to-ligand, ligand-to-metal, and intervalence transitions. Metal-metal separation was increased from 8 to 20 A in mixed-valence complexes of the type [{(NH3)5Ru-py-(Ph),-py-R~(NH3)5]5+, where Ph is phenyl (69).Only weak coupling was observed and it was concluded that aromatic spacers alone are not sufficient to promote strong electronic coupling. Beratan and Hopfield (134)developed a semiempirical theory to predict the dependence of Had on distance and linker geometry for mixed-valence dithiaspirocyclobutane molecules (72). Siddarth and Marcus (135)applied an extended Huckel method to evaluate Hadfor the mixed-valence dithiaspirocyclobutane molecules and obtained somewhat better agreement with experiment. For a series of similar class I1 mixed-valence complexes that differ only in the nature of the bridging ligand, MMCT band energy will increase with increasing distance between the metal ions. This dependence has its origin in the magnitude of xo as described by Eq. (19) and has been the subject of study (136,137).The relationship between E,, and the barrier to thermal electron transfer [Eq.(111 was nicely demonstrated by a correlation between electron-transfer rate constants and intervalence-transfer energies (138)and is shown in Fig. 12.

e. Nature of the Bridging Ligand. As shown in Fig. 11,Richardson and Taube (77) demonstrated the dependence of Had on the distance of metal-metal separation. The observed trends were predicted theoretically by combining a molecular orbital description of the bridging li-

302

ROBERT J . CRUTCHLEY

0

E,, IkJ/moll

FIG.12. Plot of activation free energy for electron transfer in [(H20)(NH,),R~"-LC O ) ( N H ~ )vs ~ ]intervalence ~' band energy in [(NHB)Ru"-L-Ru"'lNH, )il'- for six bridging ligands. [Reproduced with permission from Ref. (138).Copyright (1987) American Chemical Society.]

gand with a semiempirical measure of metal-ligand charge-transfer interactions. In addition to the distance between metal ions, the dependence of metal-metal coupling on bridging ligand size, orientation of substituents, and saturation was also demonstrated theoretically and experimentally. Hale and Ratner (139) reexamined these systems and split up the MMCT transition moments into contributions from holetransfer and electron-transfer superexchange mechanisms. Simple modifications of the bridging ligand can have dramatic results. For example, the bridging ligand 4,4'-azopyridine (azpy) is a far poorer medium for metal-metal coupling (70) compared with 4,4'dithiodipyridine (dtdp)(401,and this difference is deserving of a theoretical analysis. Joachim (140) has explored theoretically the effect of donor and acceptor substitution of the bridging ligand 1,4-dicyanobenzene on through-bond electronic coupling. Recently, there has been surge of interest in anionic 7~ aromatic bridging ligands (76, 78, 79, 81-85, 88, 91, 92, 95, 141). The dominant pathway in these systems is expected to be hold-transfer superexchange. In saturated systems (121,134) without the possibility of direct metal-metal orbital overlap, the superexchange pathway is most likely via the cr-symmetry HOMO. These systems provide an advantage in that metal-metal coupling can

DINUCLEAR RUTHENIUM COMPLEXES

303

be investigated for both Ru(II)-Ru(III) and Ru(III)-Ru(III) complexes. This will be discussed in Section 1II.B. f. Transient Absorption Studies. For the mixed-valence complexes [(dpte)2C1Ru(p-L)RuCl(bpy)213t [dpte, PhSCH,CH,SPh; L, 4,4'-bpy, trans- 12-bis(4-pyridyl)ethylene,and 1,2-bis(4-pyridyl)ethane],photol-

ysis into the T * + d.rr[R~"Cl(dpte)~I chromophore resulted in population of the bpy-based "LCT excited state [-Ru"'Cl(bpy7)bpy]* (142, 143). Thus, ligand-to-ligand electron transfer, [(dpte),C1Rur1'(p-L71R ~ ~ ~ ' C l ( b p y )*,+ ] ~ [(dpte),CIRulll(p-L)RurllCl(bpy' ' )bpy)I3+*, occurred and is competitive with excited-state nonradiative decay by ligand to metal electron transfer. Picosecond infrared studies (144-146) of the dynamics of [(NO5R u " C N R U " ' ( N H ~ ) following ~I~~ MMCT photolysis permitted observation of the formation and decay of the MMCT excited state and the evaluation of vibronic coupling and energy-transfer dynamics. The experimental results were in agreement with recent electron-transfer theories that have been used to predict excited vibrational populations resulting from back electron transfer in the Marcus "inverted region" (146).

g . Molecular Electronics and Molecular Materials. The purposeful construction of a single supermolecule for signal processing is a practical application of mixed-valence complexes and a major synthetic challenge. The control of signal transfer (an electron) from one metal site to another obviously depends on control of metal-metal coupling. Theoretical studies have focussed on the effect of photoinduced conformational changes of the bridging ligand on the interaction between metals (147).The effect on metal-metal coupling must be extreme to be useful because a change in conformation can open new coupling pathways, which although not as efficient serve to diminish the effect (148).Chemical approaches such as protonation of the bridging ligand (70,85,149) or outer-sphere coordination of crown ethers (57) are too slow to be of much use. A photoinduced conformational change that causes a n outersphere perturbation might be a possible approach. Sano and Taube (80) prepared [ 1,5-dithiacyclooctane 1-oxide)bis(pentaammineruthenium)15+,which upon oxidation underwent linkage isomerism. The process was reversible and is a n example of molecular hysteresis with potential application to high-density memory storage. Novel molecular materials may be prepared by considering a mixedvalence complex as a molecular building block. Polymerization and

3 04

ROBERT J. CRUTCHLEY

retention of electron delocalization between blocks then lead t o a longchain molecular wire. Researchers (150) have found varying degrees of conductivity in bisaxially coordinated macrocyclic ruthenium complexes [(Mac)RuL],. For example, when Mac is phthalocyanine and when L is pyrazi ne or 1,2,4,5-tetraazine, the conductivity was found S cm-' (151 and lo-' S cm-' (1521,respectively. to be Other researchers ave investigated intramolecular energy and electron transfer in supramolecular species based on ruthenium(I1) polypyridine systems (1531. These supramolecular species could perform the same function as light harvesting structures found in the photosynthetic apparatus. The potential development of optoelectronic devices based on nonlinear polarization of the MMCT transitions of mixed-valence complexes has been the subject of an investigation (100).The hyperpolarizabilities measured for the complexes [(CN),Ru(pu-CN)Ru(NH,),3- and [(775-CSH,)Ru(PPh3),(p-CN)Ru(NH3)513t are among the largest known for a solution species. The tunability of MMCT energy is an added advantage for the construction of nonlinear optical devices.

Ill. Electron Exchange

A. THEORY 1 . Magnetic Electron Exchange If we consider a system composed of metal A with Spin S, and metal B with spin Sb,which are bridged by an organic molecule L,

then the Hamiltonian for the intramolecular magnetic interaction between spins can be written

where the exchange coupling constant J is positive if the spins are parallel and negative if they are paired. In the literature, Eq. (35)can be found with -J , 2J,or J instead of - 2J and so the proper conversion should be made when comparing data. If two metal ions with S = 4 are antiferromagnetically coupled, a

DINUCLEAR RUTHENIUM COMPLEXES

305

singlet ground state and an excited triplet state that are separated in energy by - 2J are created: 35

t -2 J

'S

Ferromagnetic coupling results in the formation of a triplet ground state and excited singlet state: 'S

Whether a singlet or triplet ground state is formed depends on the relative magnitude of ferromagnetic ( J F ) and antiferromagnetic (JAF) terms,

where JF represents the overall spin pairing energy and J A F is the energy released as a result of the coupling of spins (a weak bonding interaction). The value of JFis usually thought to be small relative to J A F , although ab initio calculations of dinuclear Cu(I1) complexes with short metal-metal separations have shown J F to make a significant contribution to the total exchange constant 2J (154). 2 . The Charge Transfer Model for Antiferromagnetic Exchange

The magnitude O f J A F has been interpreted by using parameters that reflect the extent of electron exchange stabilization (155-1591, the

306

ROBERT J . CRUTCHLEY

extent of overlap between coupled metal orbitals (160),or both. The approach presented in this chapter is developed from the same conceptual basis as that for Marcus electron-transfer theory and the Hush model for intervalence transitions. The electron-exchangeinteraction can be described (155)as resulting from the mixing of the charge transfer state F with the initial state I,

t

MA-L-MB

I

1

Eb,

MA-L-M,

ti

(37)

F

This mixing can occur only if the spins in state I are antiparallel and result in the stabilization energy J A F . The potential energy-configuration diagram representing initial and final states in which EA is not large is shown in Fig. 13. EAPis the charge transfer energy between initial and final states at the same nuclear configuration. H a b is the resonance energy (or tunneling matrix element Tab)between the two states. When Eb is large, poor mixing of states and consequently very weak antiferromagnetic exchange result. The analogy between Fig. 13 and the potential energy-configuration diagram (Fig. lb) for asymmetric mixed-valence complexes should be obvious. However, an important difference is the addition of a spin pairing energy term (161), estimated to be approximately 6000 cm-' if Ru(1V) has spin = 0, which will shift E,& to higher energies. Expressions for the antiferromagnetic exchange term JAF have appeared in the literature (155-159). Bertrand (159) gives a complete derivation with the final result,

This result is an approximation but nevertheless is a valuable guide to the design of exchange coupled systems. More rigorous ub initio methods that take into account second-ordereffects are available ( 1 5 4 ) . The predicted MMCT transition of an antiferromagnetically exchanged coupled system has not been observed experimentally. This may be because bridging ligands that give rise to strong antiferromagnetic coupling will also give rise to an intense ligand-to-metal chargetransfer transition in the visible region (see Section III.A.31, which may overlap and obscure the MMCT transition. If so, electroabsorption spectroscopy may be of some help in deconvoluting LMCT and MMCT

DINUCLEAR RUTHENIUM COMPLEXES

307

FIG.13. The potential energy-configuration diagram representing initial I and final F states.

bands. An antiferromagnetically exchange coupled system will have an MMCT transition band that will be treatable by the Hush model or the MO vibronic coupling model for intervalence transitions. A theoretical study of a delocalized exchange coupled system that uses the vibronic coupling model has appeared in the literature (162).It does, however, make the assumption that the ground state is not a singlet. 3. Superexchange

In the discussion so far the nature of the bridging ligand has not been treated and yet it is of paramount importance in determining the magnitude of Haband hence JAF. For the systems of this chapter, the metal ions are too far apart to interact directly as shown in Eq. (371, and, for metal-metal coupling to occur, bridging ligand orbitals must play an important role. It was Kramers in 1934 (163)who proposed that ligand-to-metal charge-transfer LMCT excited states were involved,

t

tl

MA-L-M,

1

q,,

t

+ t+

MA-L-M,

(39)

In the LMCT excited state, the unpaired electrons can couple but can do so only if their spins are opposite. The mixing of the LMCT excited state with the ground state therefore stabilizes a ground state in which the electrons are antiparallel. This ligand mediated electron exchange mechanism is called superexchange and is analogous to the holetransfer superexchange mechanism discussed in Section II.B.2.b. The optimization of superexchange in Eq. (39) requires that the m a g netic orbitals (orbitals containing one electron) overlap simultaneously with a single ligand molecular orbital, the superexchange pathway.

308

ROBERT J . CRUTCHLEY

To overlap effectively,magnetic orbitals must have the same symmetry and be as close in energy as possible to the ligand molecular orbital. The interaction will result in an LMCT transition whose energy is a measure of the gap between magnetic orbitals and the superexchange pathway (78) and whose oscillator strength is a measure of the electronic interaction (164-166). 4. Experimental Evaluation of the Electron Exchange Constant J

The magnetic properties of a compound are evaluated by measuring its magnetic susceptibility, x, as a function of temperature (167). For an antiferromagnetically coupled dimer complex, the molar magnetic susceptibility is determined by the thermal population distribution of singlet ground and triplet excited states and is usually given by the modified Bleaney-Bowers expression (168),

where g is the powder average g value, C/(T - 8) is a Curie-Weiss term that corrects for any paramagnetic impurity at low temperatures, and xp is a correction for any background paramagnetism. The treatment of the temperature dependence of magnetic susceptibility is almost completely dominated by spin, but for Ru(III), orbital angular momentum may make a significant contribution. Drillon and Georges (169) have developed procedures for including the orbital contribution for tZgnconfigurations, but analytical expressions are not available. A good estimate of the electron exchange constant J for Ru(II1) dimers can be had by assuming isotopically coupled, S = f spins and by using Eq. (40).

B. STUDIES OF EXCHANGE COUPLEDDINUCLEAR RUTHENIUMCOMPLEXES There have been only a handful of studies on exchange coupled ruthenium dimers and yet the information available on metal-metal coupling is potentially as valuable as that obtained from mixed-valence complexes. The reason for this lack of activity has been the greater familiarity of researchers with the chemistry of Ru(I1) bound to .rr-acid ligands. The synthetic pathways to complexes of this type are well explored (7) and possess the tremendous advantages of stability and ease of handling. Ru(I1) complexes that incorporate anionic .rr donor

DINUCLEAR RUTHENIUM COMPLEXES

309

ligands are unstable with respect to ligand substitution and this lack of complex integrity is an extreme practical disadvantage. On the other hand, Ru(II1) complexes that incorporate n-donor ligands are quite stable, and so the synthesis of stable Ru(II1) dinuclear complexes that have n-donor bridging ligands needs only the development of synthetic pathways to become easily accessible for study. The first attempt to study electron exchange in dinuclear Ru(II1) complexes produced a negative result. Complexes of the type [{(bpy)Ru(C1)},(p-L)]6+, where L is the n accepting ligands pyrazine, 4,4'-bipyridine and 1,2-bis(4-pyridyl)ethane,showed little if any evidence for magnetic interactions down to liquid-helium temperatures (169). This led the researchers to suggest that the primary orbital mechanism for metal-metal coupling in the mixed-valence analogues is by dnRu(I1)n "(L) mixing (electron transfer superexchange) because oxidation leads to orbital contraction of the d n orbitals and to their stabilization relative to the LUMO of the bridging ligand. Electron exchange is extremely weak in these complexes because of the large difference in energy between the Ru(II1) d n orbitals and either the n symmetry LUMO or HOMO of the bridging ligand. These same arguments could explain the weak antiferromagnetic coupling (25= -6 cm-') that was seen for the fully oxidized Creutz-Taube ion [{(NH,),Ru},(p-pyz)lG+(59, 1701, and together these results serve to emphasize the unsuitability of n acceptor bridging ligands as a medium for electron exchange in Ru(III)-Ru(III) dinuclear complexes. The first real indication of antiferromagnetic superexchange in ruthenium dimers that possessed polyatomic bridging ligands was found for the complex [{(NH3)5R~}2(p-t-BuMN)15t (761, where

The magnetic susceptibility of this complex (0.98 pB) is severely depressed relative to those of mononuclear low-spin RuUII) complexes, which range from 1.9 to 2.07 p B (1711. Unfortunately, the temperature dependence of the magnetic susceptibility of this complex and other derivatives was not tested for and so a measure of antiferromagnetic J is not available. The lone electron pair of the bridging ligand's central carbon can delocalize into the n systems of both nitrile groups. This provides a superexchange pathway of 7~ symmetry that spans the entire molecule and can interact simultaneously with both metal ions. It is

310

ROBERT J. CRUTCHLEY

not unexpected that the mixed-valence complex also shows evidence of strong metal-metal coupling and is a delocalized class I11 system. The possibility that long-range antiferromagnetic coupling might occur between magnetic orbitals bridged by an easily polarizable, extended T HOMO system led researchers to prepare (78) the dinuclear Ru(II1) complexes [{(NH,),Ru}~(~-L)]~+, where L is Dicyd2-, Me,Dicyd2-, C12Dicyd2-,and C1,Dicyd2-. The temperature dependence of magnetic susceptibility showed the complexes to be antiferromagnetically coupled systems, as is illustrated in Fig. 14. The best fits of the magnetic data gave values of -J for the complexes, which varied from a low of -61.9 to 2400 cm-’. Antiferromagnetic coupling of this magnitude at an estimated through-space separation between metal ions of 13.2 was unprecedented for any metal ion dimer system (78, 173) and was shown to be due to a continuous and energetically favorable superexchange pathway involving the magnetic d~ orbitals of the Ru(II1) ions and the 7~ HOMO of the bridging Dicyd2- ligand. The details of this study deserve further mention. 0.31

0.27

-

.

0.24

=I

-E, 0.20

N

w

0 7 X

.$ .- 0.17 0 c

a 0,

0.13 m 0.10

0.06 Temperature (K)

FIG.14. Experimental (0) and modeled (-) by using Eq. (40) temperature dependen[Reproduced cies of the molar magnetic susceptibility of [{(NH,)Rujz(~-Dicyd)][OTsl,. with permission from Ref. (78). Copyright (1992) American Chemical Society.]

DINUCLEAR RUTHENIUM COMPLEXES

311

Extended Huckel calculations were performed using the crystal structure data (174 1 of the free dianion ligands. The frontier orbitals of the Dicyd2- are illustrated in Fig. 15 and show that both HOMO and SHOMO are of 7~ symmetry and span the entire molecule. A crystal structure determination of the complex [{(NH,)RU}~(~-D~C~~)][OTS

7F2

FIG.15. Frontier orbitals of Dicyd2- from EHMO calculations. Minor orbital corrections are omitted for clarity. [Reproduced with permission from Ref. (78). Copyright (1992) American Chemical Society.]

312

ROBERT J . CRUTCHLEY

(where OTs is tosylate anion) was also done and showed that planarity of the bridging ligand was retained upon coordination. However, two conformations of the dinuclear complex were observed. For conformer A, both Ru(II1)-cyanamide bonds are essentially linear and, for conformer B, both Ru(II1)-cyanamide bonds are bent. Because of the different Ru-NCN bond angles, both HOMO and SHOMO are potential superexchange pathways. The energy gap between magnetic orbitals and the HOMO of the Dicyd2- derivatives was estimated from the energy of the LMCT band associated with the Ru(II1)-cyanamide chromophore for the complexes in aqueous solution. After a correction to compensate for the change in inner and outer coordination sphere configurations resulting from a transition into an excited vibrational level of the excited LMCT state, the energy gaps between magnetic orbitals and the superexchange pathway were estimated a t 0.48, 0.63, 0.74, and 0.94 eV for [{(NH3)5Ru}2(p-L)14+, where L is Me2Dicyd2-,Dicyd2-, C1,Dicyd2-, and C14Dicyd2-, respectively. Qualitatively, the trend in the value of -J increased with decreasing energy gap as expected. The complex with the smallest energy gap, [{(NH,)5Ru}2(p-Me2Dicyd)14 +, is diamagnetic at room temperature and only a lower limit of -J 2 400 cm- could be estimated. The effect of Dicyd2- ligand planarity on the magnitude of superexchange was illustrated by the study (175) of the complex [{(NH&R ~ } ~ ( p - M e ~ D i c y dTemperature-dependent )]~+. magnetic susceptibility data gave, after modeling with Eq. (401, -J = 110 cm-'. For the Me,Dicyd2- ligand, steric repulsion forces the cyanamide groups into a n anticonfiguration out of the phenyl ring plane. An EHMO calculation of the ligand showed a HOMO with electron density mostly localized on the phenyl ring. This would result in poor overlap with the magnetic orbitals and a reduction in superexchange compared with the planar Dicyd2- derivatives. For the complex [((NH,),R~}~(p-Me,Dicyd)l~+, the energy gap between magnetic orbitals and the superexchange pathway can be estimated a t 0.76 eV. This value is very close to that of [{(NH,),Ru},(p-C1,Dicyd)14+,for which -J = 95.9 cm-', and suggests that the energy gap between magnetic orbitals and Dicyd2- orbitals may be the most important factor in determining the magnitude of -J . The complex [{(NH3)5Ru}2(p-Dicyd)]4+ showed a n interesting dependence of - J o n the counteranion (78). When the counterion was C104-, the complex was nearly diamagnetic at room temperature, with a n estimated -J 9 400 cm-'. When the counterion was OTs-, -J = 100 cm- l. Counterion effects are usually ascribed to intermolecular exchange interactions. However, in this case, IAJI = 300 cm-' is far too large to be explained by relatively weak intermolecular interactions

DINUCLEAR RUTHENIUM COMPLEXES

313

between octahedral Ru(II1) ions (168, 171). The solvent-dependent metal-metal coupling that was seen for mixed-valence complexes (discussed in Section II.C.2.c) may have an application here. It is suggested that hydrogen bonding interactions in the crystal lattice between counterions and solvent molecules with ammines bonded to Ru(II1) can help decouple the Ru(II1) ion from the bridging ligand (165, 166). Solventdependent antiferromagnetic exchange was indicated by the observation that [((NH,),R~},(p-Me~Dicyd)]~+ forms diamagnetic solutions in nitromethane, whereas those of dimethyl sulfoxide are paramagnetic (176). The solution 'H NMR spectra show that complex integrity is maintained in both solvents. Attempts to see an MMCT band for the Ru(II1)-Ru(II1) dinuclear complexes were unsuccessful (78, 79). The intense low-energy LMCT bands are likely to obscur these transitions. The values of Hab for the complexes [((NH,),RU},(~.-L)~~+ were estimated from Eq. (38) and compared with Hadcalculated for the mixed-valence complexes using Eq. (15). Ha, was significantly larger than Hadand it was suggested that solid-state geometry was optimal for metal-metal coupling. Metal-metal coupling in the mixed-valence complexes is also attenuated by strong solvent donor interactions with the ammine ligands (176) and this may partially account for the difference between Hab and Had.This is not to say that Haband Hadshould be equivalent because they are derived from different state descriptions. IV. Future Studies

Electroabsorption spectroscopy and its application to MMCT transitions (17, 18, 104) are important developments that will test current theories and increase our understanding of the factors that control superexchange phenomena. It is to be hoped that the use of electroabsorption spectroscopy becomes commonplace. Selective perturbation of metal-metal coupling by either innersphere or outer-sphere manipulations will continue. Aside from increasing our understanding of metal-metal superexchange coupling, these studies will probe the nature of solvent-solute interactions and perhaps be applicable to the design of molecular devices. The synthesis of novel nonlinear optical materials that incorporate mixed-valence metal systems will be the subject of much interest (100). There are very few examples of exchange coupled dinuclear ruthenium complexes and a clear opportunity exists for further study. All dinuclear complexes that possess anionic bridging ligands (76,81-85,

3 14

ROBERT J. CRUTCHLEY

88, 91, 92, 95, 141) and can form stable Ru(III)-Ru(III) complexes

should be evaluated for antiferromagnetic exchange. More results are needed to elucidate the dependence of electron exchange on the relative symmetry and energy of the superexchange pathway. The distance dependence of electron exchange poses interesting questions. A fully delocalized system should have very little distance dependence. The synthesis of [{(NH3),Ru}2(p-Dicyd)14+complexes and the observation of strong antiferromagnetic coupling between metal ions separated by 13 8, are a consequence of the very special properties of the bridging Dicyd2- ligand. The Dicyd2- molecule is easily oxidized, is highly polarizable, and possesses a T HOMO that spans the entire molecule. Novel bridging ligands with similar properties should be synthesized. Ondrechen et al. (177, 178) have explored the design of conductive mixed valency oligomers. The polarizability of the bridging ligand had a direct impact on both spectroscopic and conductive properties. What is now needed is the practical application of superexchange phenomena to the synthesis of novel solid-state materials. V. Glossary of Abbreviations and Ligand Structures

DINUCLEAR RUTHENIUM COMPLEXES

bpbirnH2

315

%--p " 00

dpirnbH2

120

316

ROBERT J. CRUTCHLEY 3-dpph

1Ldcb

N

a:: qCN

13dcb

CN 14dcb dts

s3cs dtsd

\ I

sm

dpmn

” ;

NC

dtst

CN

13dca

..

I

1

o o c ~ ~ ~ 2 c- o o N

I

CN

Me2dicyd2-NcN*

1 2 NCN

NC

317

DINUCLEAR RUTHENIUM COMPLEXES

C12dicyd2-

NcN*

12-

CI

26dcn NCN NC

CI

C14dicyd2dto

ci s

0s

NCN

PYZC -

+kNC1’CI

CI

@foo ‘

dpqa$0 & 0 0 N

phen

furn isn

NC/cN

318

ROBERT J. CRUTCHLEY MeZphen

PY

Q Me4 phen

PPdW

fN

C/N

w

DINUCLEAR RUTHENIUM COMPLEXES

319

ACKNOWLEDGMENTS The financial support of NSERC (Canada) and Carleton University is gratefully acknowledged. I would also like to thank NSERC (Canada) for the award of a University Research Fellowship and Johnson-Matthey P.L.C. for the loan of ruthenium trichloride hydrate in support of my research program.

REFERENCES 1. Seddon, E. A,, and Seddon, K. R., “The Chemistry of Ruthenium,” 2nd ed. Elsevier, Amsterdam, 1984. 2 . Allen, G. C., and Hush, N. S., Prog. Inorg. Chem. 8, 357 (1967). 3. Robin, M. B.,and Day, P., Adu. Znorg. Chem. Radiochem. 10,247 (1967). 4 . (a) Hush, N. S., Prog. Inorg. Chem. 8,391 (1967);(b) Hush, N. S., Trans. Faraduy SOC.57, 557 (1961);(c) Hush, N.S., Electrochim. Acta 13, 1005 (1968). 5. Marcus, R.A., J . Chem. Phys. 24,966 (1956). 6. Marcus, R. A., J . Chem. Phys. 26, 867 (1957);872 (1957). 7. Creutz, C.,Prog. Inorg. Chem. 30, 1 (1983). 8. Meyer, T. J . , Prog. Inorg. Chem. 30,389 (1983). 9. Sutin, N., Prog. Inorg. Chem. 30, 441 (1983). 10. Newton, M. D.,Chem. Rev. 91,767 (1991).

320

ROBERT J. CRUTCHLEY

11. Mataga, N., and Kubota, T., “Molecular Interactions and Electronic Spectra.”

Dekker, New York, 1970. Murrell, J . N., Quart. Reu. (London) 15,191 (1961). Chen, P., Curry, M., and Meyer, T. J.,Inorg. Chem. 28,2271 (1989). Creutz, C., Inorg. Chem. 17,3723 (1978). Hupp, J. T., and Meyer, T. J., Inorg. Chem. 26, 2332 (1986). Kober, E. M., Goldsby, K. A., Narayana, D. N. S., and Meyer, T. J.,J . A m . Chem. SOC.105,4303 (1985). 17. Oh, D. H., and Boxer, S. G., J . A m . Chem. SOC.111, 1130 (1989). 18. Oh, D.H., Sano, M., and Boxer, S. G., J . A m . Chem. Sac. 113,6880(1991). 19. Marcus, R. A., J . Chem. Phys. 43,679 (1965). 20. Cannon, R. D., Chem. Phys. Lett. 49,299 (1977). 21. Brunschwig, B. S., Ehrenson, S., and Sutin, N., J . Phys. Chem. 90,3657 (1986). 22. Sutton, J . E., and Taube, H., Inorg. Chem. 20, 3125 (1981). 23. Reimers, J . R., and Hush, N. S., Inorg. Chem. 29,3686 (1990);4510 (1990). 24. (a) Piepho, S. B., Krausz, E. R., and Schatz, P. N., J . A m . Chem. SOC.100,2996 (1978);(b) Wong, K. Y., and Schatz, P. N., Prog. Inorg. Chem. 28,369 (1981). 25. Buhks, E., Ph.D. Thesis, Tel Aviv University, Tel Aviv, 1980. 26. Wong, K. Y., and Schatz, P. N., in “Mechanistic Aspects o f Inorganic Reactions” (Rorabacher, D. B., and Endicott, J . F., Eds.), Chap. 12,p. 281.American Chemical Society, Washington, D.C., 1982. 27. (a)Hush, N. S., in “Mechanistic Aspects ofInorganic Reactions” (Rorabacher, D. B., and Endicott, J. F., Eds.), Chap. 13,p. 301.American Chemical Society, Washington, D.C., 1982;(b) p. 330.American Chemical Society, Washington, D.C., 1982. 28. Hush, N. S., in “Mixed-Valence Compounds” (Brown, D. B., Ed.), p. 151.D. Reidel, Dordrecht, Holland, 1980. 29. KO,J., and Ondrechen, M. J., J . A m . Chem. Sac. 107,6161 (1985). 30. Dubicki, L., Ferguson, J., and Krausz, E. R., J . A m . Chem. SOC.107, 179 (1985). 31. (a)Root, L. J.,and Ondrechen, M. J., Chem.Phys. Lett. 93,421(1982); (b) Ondrechen, M. J., KO,J., and Root, L., J . Phys. Chem. 88,5919 (1984); (c) KO,J., and Ondrechen, M. J., Chem. Phys. Lett. 112,507(1984);(d) KO,J., Zhang, L.-T., and Ondrechen, M. J., J . Am. Chem. SOC.108,1712(1986);(e) Zhang, L.-T., KO,J., and Ondrechen, M. J., J . Am. Chem. SOC.109,1666(1987);(f) Ondrechen, M. J., KO,J., and Zhang, L.-T., J . Am. Chem. SOC.109,1672(1987);(g) Zhang, L.-T., KO,J., and Ondrechen, M. J., J . Phys. Chem. 93,3030 (1989). 32. (a)Piepho, S.,J . Am. Chem. Chem. SOC.110,6319(1988);(b)Piepho, S.B., in “MixedValency Systems: Applications in Chemistry, Physics and Biology” (Prassides, K., Ed.), p. 329.Kluwer Academic, Dordrecht, 1991;(c)Schatz, P. N . , in “Mixed-Valency Systems: Applications in Chemistry, Physics and Biology” (Prassides, K., Ed.), p. 7. (d) Riley, M. J., and Krausz, E . R., in “Mixed-Valency Systems: Applications in Chemistry, Physics and Biology” (Prassides, K., Ed.), p. 341. 33. Piepho, S., J . Am. Chern. Chem. SOC. 112,4197 (1990). 34. Salaymeh, F., Berhane, S., Yusof, R., de la Rosa, R., Fung, E. Y., Matamoros, R., Lau, K. W., Zheng, Q., Kober, E. M., and Curtis, J. C., Inorg. Chem. 32,3895(1993). 35. Johnson, J . E. B., de Groff, C., and Ruminski, R. R.,lnorg. Chirn. Actu 187,73(1991). 36. Poppe, J., Moscherosch, M., and Kaim, W., Inorg. Chem. 32,2640 (1993). 37. Ernst, S.,Kasack, V., and Kaim, W., Inorg. Chem. 27, 1146 (1988). 38. Kaim, W., Kasack, V., Binder, H., Roth, E., and Jordanov, J.,Angew. Chem. Znt. Ed. Engl. 27, 1174 (1988). 12. 13. 14. 15. 16.

DINUCLEAR RUTHENIUM COMPLEXES

32 I

39. Bardwell, D., Jeffery, J. C., JouliB, L., and Ward, M. D., J . Chem. SOC.Dalton Trans., 2255 (1993). 40. de Sousa Moreira, I., and Franco, D. W., J . Chem. SOC.Chem. Commun., 450 (1992). 41. Ruminski, R. R., Cockroft, T., and Shoup, M., Inorg. Chem. 27,4026 (1988). 42. Beley, M., Collin, J.-P., Louis, R., Metz, B., and Sauvage, J.-P., J . A m . Chem. SOC.

113,8521 (1991). 43. Collin, J.-P., Laine, R., Launay, J.-P., Sauvage, J.-P., and Sour, A., J . Chem. SOC. Chem. Commun., 434 (1993). 44. Palaniappan, V., Yadav, S. K. S., and Agarwala, U. C., Polyhedron. 4,1457 (1985). 45. de la Rosa, R., Chang, P. J., Salaymeh, F., and Curtis, J . C., Znorg. Chem. 24,

4229 (1985).

F.,Berhane, S., Yusof, R., de la Rosa, R., Fung, E. Y., Matamoros, R., Lau, K. W., Zheng, Q., Kober, E. M., and Curtis, J. C., Inorg. Chem. 32,3895(1993). 47. Brant, P., and Stephenson, T. A,, Inorg. Chem. 26,22 (1987). 48. Hush, N.S., Chem. Phys. 10,361 (1975). 49. Larsson, S.,J . A m . Chem. SOC. 103,4034 (1981). 50. Kosloff, R., and Ratner, M. A., Israel J . Chem. 30, 45 (1990). 51. Broo, A., and Larsson, S., Chem. Phys. 161,363 (1992). 52. Kaim, W., and Kasack, V.,Inorg. Chem. 29,4696 (1990). 53. Cooper, J. B., MacQueen, D. B., Petersen, J. D., and Wertz, D. W., Inorg. Chem. 29, 3701 (1990). 54. Matsubara, T., and Ford, P. C., Inorg. Chem. 15,1107 (1976). 55. Szentrimay, R., Yeh, P., and Kuwana, T., in “Electrochemical Studies of Biological Systems” (Sawyer, D. T., Ed.). American Chemical Society, Washington, D.C., 1977. 56. Dong, Y., Hupp, J. T., and Yoon, D. I., J . A m . Chem. SOC.116,4379 (1993). 57. Roberts, J. A., and Hupp, J. T., Inorg. Chem. 31, 160 (1992). 58. Krausz, E.,Burton, C., and Broomhead, J., Inorg. Chem. 20, 434 (1981). 59. Furholz, U.,Burgi, H.-B., Wagner, F. E., Stebler, A., Ammeter, J. H., Krausz, E., Clark, R. J. H., Stead, M. J., and Ludi, A., J . A m . Chem. SOC.106, 121 (1984). 60. Stebler, A.,Ammeter, J. H., Furholz, U., and Ludi, A., Inorg. Chem. 23,2764(1984). 61. Furholz, U.,Stefan, J., Burgi, H.-B., and Ludi, A., Inorg. Chem. 24, 943 (1985). 62. Krausz, E.R., and Mau, A. W. H., Inorg. Chem. 25, 1484 (1986). 63. Krausz, E.,and Ludi, A., Znorg. Chem. 24,939 (1985). 64. Creutz, C., and Chou, M. H., Znorg. Chem. 26, 2995 (1987). 65. Joss, S.,Hasselbach, K. M., Burgi, H.-B., Wordel, R., Wagner, F. E., and Ludi, A., Inorg. Chem. 28, 1815 (1989). 66. Best, S. P., Clark, R. J. H., McQueen, R. C. S., and Joss, S., J . Am. Chem. SOC. 111, 548 (1989). 67. Piepho, S. B., Krausz, E. R., and Schatz, P. N., Chem. Phys. Lett. 55,539 (1978). 68. Woitellier, S.,Launay, J. P., and Spangler, C. W., Inorg. Chem. 28, 758 (1989). 69. Kim, Y., and Lieber, C. M., Znorg. Chem. 28,3990 (1989). 70. Launay, J.-P., Tourrel-Paagis, M., Libskier, J.-F., Marvaud, V., and Joachim, C., Inorg. Chem. 30, 1033 (1991). 71. Stein, C. A., and Taube, H., J . A m . Chem. SOC.103,693 (1981). 72. Stein, C. A., Lewis, N. A., and Steitz, G., J . Am. Chem. Soc. 104, 2596 (1982). 73. Sedney, D., and Ludi, A., Znorg. Chim. Acta 47, 153 (1981). 74. Amer, S. I., Dasgupta, T. P., and Henry, P. M., Inorg. Chem. 22, 1970 (1983). 75. Sutton, J. E., Krentzien, H., and Taube, H., Inorg. Chem. 21, 2842 (1982). 46. Salaymeh,

322

ROBERT J. CRUTCHLEY

76. Krentzien, H., and Taube, H., Znorg. Chem. 21,4001 (1982). 77. Richardson, D. E., and Taube, H., J . A m . Chem. Soc. 105,40 (1983). 78. Aquino, M. A. S., Lee, F. L., Gabe, E. J., Bensimon, C., Greedan, J. E., and Crutchley, R. J., J . Am. Chem. SOC.114,5130 (1992). 79. Aquino, M. A. S., Bostock, A. E., and Crutchley, R. J.,Znorg. Chem. 29,3641(1990). 80. Sano, M., and Taube, H., J . A m . Chem. SOC.113,2327 (1991). 81. Hage, R., Haasnoot, J. G., Reedijk, J., Wang, R., and Vos, J. G., Znorg. Chem. 30, 3263 (1991). 82. Haga, M., Matsumura-Inoue, T., and Yamabe, S., Znorg. Chem. 26,4148 (1987). 83. Haga, M., and Bond, A. M., Znorg. Chem. 30,475 (1991). 84. Rillema, D.P., Sahai, R., Mathews, P., Edwards, A. K., Shaver, R. J., and Morgan, L., Znorg. Chem. 29, 167 (1990). 85. Haga, M., Ano, T., Kano, K., and Yamabe, S., Znorg. Chem. 30,3843 (1991). 86. Goldsby, K. A., and Meyer, T. J., Znorg. Chem. 23,3002 (1984). 87. Wallace, A. W., Murphy, W. R., andpetersen,J. D.,Znorg.Chim.Acta 166,47(1989). 88. Ohno, T., Nozaki, K., and Haga, M., Znorg. Chem. 31,4256 (1992). 89. Geno, M. J. K., and Dawson, J. H., Znorg. Chem. 23, 1182 (1984). 90. Cloninger, K. K., and Callahan, R. W., Znorg. Chern. 20, 1611 (1981). 91. Palaniappan, V., and Agarwala, U. C., Znorg. Chem. 27,3568 (1988). 92. Palaniappan, V., Sathaiah, S., Bist, H. D., and Agarwala, U. C., J . A m . Chem. SOC. 110,6403 (1988). 93. Das, B. K., and Chakravarty, A. R., Znorg. Chem. 30,4978 (1991). 94. Chang, J. P., Fung, E. Y., and Curtis, J. C., Znorg. Chern. 25,4233 (1986). 95. Chou, M. H., Creutz, C., and Sutin, N., Inorg. Chem. 31, 2318 (1992). 96. Fagalde, F., and Katz, N. E., J . Chem. Soc. Dalton Trans., 571 (1993). 97. (a) Bignozzi, C. A,, Paradisi, C., Roffia, S., and Scandola, F., Znorg. Chem. 27,408 (1988);(b) Roffia, S.,Paradisi, C., and Bignozzi, C. A,, J . Electroanal. Chem. 200, 105 (1986);(c) Bignozzi, C. A., Roffia, S., and Scandola, F. J., J . A m . Chem. SOC. 107, 1644 (1985). 98. Siddiqui, S., Henderson, W. W., and Shepherd, R. E., Znorg. Chem. 26,3101(1987). 99. (a) Vogler, A., and Kisslinger, J.,J . Am. Chem. SOC.104,2311(1982); (b) Ludi, A., Chirnia 26,647 (1972). 100. Laidlaw, W. M., Denning, R. G., Verbiest, T., Chauchard, E., and Persoons, A., Nature 363,58 (1993). 101. Ismail, K. Z., Tunuli, M. S., and Weber, S. G . , Znorg. Chem. 26, 1555 (1987). 102. Henderson, W. W., and Shepherd, R. E., Inorg. Chem. 24,2398 (1985). 103. Curtis, J. C., and Meyer, T. J . , Znorg. Chem. 21, 1562 (1982). 104. Reimers, J.R., and Hush, N. S., in “Mixed-Valency Systems: Applications in Chemistry, Physics and Biology” (Prassides, K., Ed.), p. 29. Kluwer Academic, Dordrecht, 1991. 105. Sutton, J. E., Sutton, P. M., and Taube, H., Znorg. Chem. 18,1017 (1979). 106. Hupp, J.T., Dong, Y., Blackbourn, R. L., and Lu, H., J . Phys. Chem. 97,3278 (1993). 107. Doorn, S.K., and Hupp, J. T., J . A m . Chem. SOC.111, 1142 (1989). 108. Doorn, S.K., Hupp, J. T., Porterfield, D. R., Campion, A., and Chase, D. B., J . A m . Chem. SOC.112,4999 (1990). 109. Bignozzi, C. A., Argazzi, R., Schoonover, J. R., Dyer, R. B., and Scandola, F . , Znorg. Chem. 31,5260 (1992). 110. Cooper, J . B., Vess, T. M., Kalsbeck, W. A., and Wertz, D. W., Znorg. Chem. 30, 2286 (1991).

DINUCLEAR RUTHENIUM COMPLEXES

323

111. Kober, E. M., Goldsby, K. A,, Narayana, D. N. S., and Meyer, T. J., J . Am. Chem. SOC.105,4303 (1985). 112. Hupp, J. T., and Meyer, T. J., Inorg. Chem. 26,2332 (1987). 113. Brunschwig, B. S., Ehrenson, S., and Sutin, N., J . Phys. Chem. 90,3657 (1986). 114. Kornyshev, A. A,, and Ulstrup, J., Chem. Phys. Lett. 126,74 (1986). 115. Ulstrup, J.,J . Phys. Chem. 91,5153 (1987). 116. (a) Hammack, W. S., Drickamer, H. G., Lowery, M. D., and Hendrickson, D. N., Chem. Phys. Lett. 132,231(1986);(b) Lowery, M. D., Hammack, W. S., Drickamer, H. G . , and Hendrickson, D. N., J . Am. Chem. SOC.109,8019(1987);(c) Hammack, W. S., Drickamer, H. G., Lowery, M. D., and Hendrickson, D. N., Inorg. Chem. 27,

1307 (1988). 117. (a) Lewis, N. A,, and Obeng, Y. S., J . Am. Chem. SOC.110,2306 (1988);(b) Lewis, N. A., Obeng, Y . S., and Purcell, W. L., Inorg. Chem. 28, 3796 (1989). 118. (a) Blackbourn, R. L., and Hupp, J. T., Chen. Phys. Lett. 150, 399 (1988); (b) Blackbourn, R. L., and Hupp, J. T., J . Phys. Chern. 94,1790 (1990). 119. Lewis, N. A., Obeng, Y. S., Taveras, D. V., and van Eldik, R., J . Am. Chem. SOC. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133.

134. 135. 136. 137. 138. 139. 140. 141. 142.

111,924(1989). Burgess, J., Spectrochim. Actu A 45, 159 (1989). Lewis, N. A., and Obeng, Y. S., J . Am. Chem. SOC.111,7624 (1989). Curtis, J. C., Sullivan, B. P., and Meyer, T. J., Znorg. Chem. 22, 224 (1983). Ennix, K. S., McMahon, P. T., de la Rosa, R., and Curtis, J. C., Inorg. Chem. 26, 2660 (1987). Curtis, J. C., Blackbourn, R. L., Ennix, K. S., Hu, S., Roberts, J. A., and Hupp, J. T., Inorg. Chem. 28,3791 (1989). Curtis, J. C., Roberts, J. A,, Blackbourn, R. L., Dong, Y., Massum, M., Johnson, C. S., and Hupp, J. T., Inorg. Chem. 30, 3856 (1991). Hupp, J. T., and Weydert, J., Inorg. Chem. 26,2657 (1987). Todd, M. D., Dong, Y., and Hupp, J. T., Inorg. Chem. 30,4685(1991). Hupp, J. T., J . Am. Chem. SOC.112, 1563 (1990). Hupp, J. T., Kober, E. M., Neyhart, G. A., and Meyer, T. J., in “Mixed-Valency Systems: Applications in Chemistry, Physics and Biology” (Prassides, K., Ed.), p. 51. Kluwer Academic, Dordrecht, 1991. Dong, Y., and Hupp, J. T., Znorg. Chem. 31,3322 (1992). Hupp, J. T., and Dong, Y., J . Am. Chem. SOC.115,6428 (1993). (a) Marcus, R. A., and Sutin, N., Biochim. Biophys. Actu 811,265 (1985);(b) Hush, N. S., Coord. Chem. Reu. 64, 135 (1985). (a) Joachim, C., Launay, J. P., and Woitellier, S.,Chem. Phys. 147, 131 (1990); (b) Launay, J. P., in “Mixed-Valency Systems: Applications in Chemistry, Physics and Biology” (Prassides, K., Ed.), p. 321.Kluwer Academic, Dordrecht, 1991. Beratan, D. N., and Hopfield, J. J., J . Am. Chem. Soc. 106,1584 (1984). Siddarth, P., and Marcus, R. A., J . Phys. Chem. 94,2985 (1990). Isied, S.S., Vassilian, A., and Wishart, J. F., J . Am. Chem. SOC. 110,635 (1988). Katz, N. E., Creutz, C., and Sutin, N., Inorg. Chem. 27, 1687 (1988). Gesolowitz, D. A,, Inorg. Chem. 26,4135 (1987). Hale, P. D., and Ratner, M. A., Znt. J . Quunt. Chem. 18, 195 (1984). Joachim, C., Chem. Phys. Lett. 185,569(1991). Hage, R.,Haasnoot, J. G., Reedijk, J., and Vos, J. G., Chemtructs 4,75 (1992). Curtis, J. C . , Bernstein, J. S., Schmehl, R. H., and Meyer, T. J., Chem. Phys. Lett. 81,48 (1981).

324

ROBERT J. CRUTCHLEY

143. Curtis, J . C., Bernstein, J . S., and Meyer, T. J., Inorg. Chem. 24, 385 (1985). 144. Doon, S. K., Stoutland, P. O., Dyer, R. B., and Woodruff, W. H., J . Am. Chem. SOC.

114, 3133 (1992). 145. Klinger, D. A. V., Tominaga, K., Walker, G. C., and Barbara, P. F., J. Am. Chem.

SOC. 114, 8323 (1992). 146. Doon, S. K., Dyer, R. B., Stoutland, P. O., and Woodruff, W. H., J . Am. Chem. SOC.

115, 6398 (1993). Joachim, C., and Launay, J. P., Chem. Phys. 109,93 (1986). Woitellier, S., Launay, J. P., and Joachim, C., Chem. Phys. 131,481 (1989). Marvaud, V., and Launay, J.-P., Inorg. Chem. 32, 1376 (1993). Hanack, M., Deger, S., and Lange, A., Coord. Chern. Reu. 83, 135 (1988). Kobel, W., and Hanack, M., Inorg. Chem. 25, 103 (1986). Keppeler, U., Deger, S., Lange, A., and Hanack, M., Angew. Chem. 99,349 (1987). For example see (a) Petersen, J . D., in “Supramolecular Photochemistry” (Balzani, V., Ed.), p. 135. Reidel, Dordrecht, 1987; (b) Campagna, S., Denti, G., Serroni, S., Ciano, M., Juris, A,, and Balzani, V., Inorg. Chem. 31,2982 (1992); (c) Arana, C. R., and Abruna, H. D.,lnorg. Chem. 32,194 (1993);(d)Serroni, S., Denti, G., Campagna, S., Juris, A,, Ciano, M., and Balzani, V., Angew. Chem. Znt. Ed. Engl. 31, 1493 (1992); and references cited therein. 154. Daudey, J. P., De Loth, P., and Malrieu, J. P., in “Magneto-Structural Correlations in Exchange Coupled Systems” (Willet, R. D., Gatteschi, D., and Kahn, O., Eds.), p. 87. Reidel, Dordrecht, 1985. 155. Anderson, P. W., Phys. Rev. 115, 2 (1959). 156. Hopfield, J. J., in “Tunneling in Biological Systems” (Chance, B., De Vault, D. C., Frauenfelder, H., Marcus, R. A., Schrieffer, J. R., and Sutin, N., Eds.). Academic Press, New York, 1979. 157. Okamura, M. Y., Fredkin, D. R., Isaacson, R. A., and Feher, G., in “Tunneling in Biological Systems” (Chance, B., De Vault, D. C., Frauenfelder, H., Marcus, R. A,, Schrieffer, J . R., and Sutin, N., Eds.), p. 729. Academic Press, New York, 1979. 158. Hay, P. J., Thibeault, J.C., and Hoffmann, R., J . Am. Chem. SOC.97,4884 (1975). 159. Bertrand, P., Chem. Phys. Lett. 113, 104 (1985). 160. (a) Kahn, O., and Briat, B., J . Chem. SOC., Faraday II, 268 (1976); 1441 (1976); (b) Kahn, O., in “Magneto-Structural Correlations in Exchange Coupled Systems” (Willet, R. D., Gatteschi, D., and Kahn, O., Eds.), p. 37. Reidel, Dordrecht, 1985. 161. Lever, A. B. P., “Inorganic Electronic Spectroscopy,” 2nd ed., p. 220. Elsevier, Amsterdam, 1984. 162. Hale, P. D., Ratner, M. A,, and Hofacker, G. L., Chem. Phys. Lett. 119,264 (1985). 163. Kramers, H. A., Physica 1, 182 (1934). 164. Crutchley,R. J.,McCaw, K., Lee, F. L., andGabe, E. J., Inorg. Chem.29,2576(1990). 165. Saleh, A. A., and Crutchley, R. J., Inorg. Chem. 29, 2132 (1990). 166. Crutchley, R. J., Saleh, A. A., McCaw, K., and Aquino, M. A. S., Mol. Cryst. Liq. Cryst. 194, 93 (1991). 167. For a good introduction to magnetism, see Drago, R. S., “Physical Methods for Chemists,” 2nd. ed. Chap. 11. Saunders, Orlando, 1992. 168. Carlin, R. L., “Magnetochemistry.” Springer-Verlag, BerlidHeidelberg, 1986. 169. Drillon, M., and Georges, R., Phys. Rev. B 24, 1278 (1981). 170. Johnson, E. C., Callahan, R. W., Eckberg, R. P., Hatfield, W. E., and Meyer, T. J., Inorg. Chem. 18,618 (1979). 171. Bunker, B. C., Drago, R. S., Hendrickson, D. N., Richman, R. M., and Kessell, S. L., J . Am. Chem. SOC.100, 3805 (1978).

147. 148. 149. 150. 151. 152. 153.

DINUCLEAR RUTHENIUM COMPLEXES

325

172. Figgis, B. N., Prog. Inorg. Chem. 6, 173 (1964). 173. Hendrickson, D. N., in “Magneto-Structural Correlations in Exchange Coupled Systems” (Willet, R. D., Gatteschi, D., and Kahn, O., Eds.), p. 523. Reidel, Dordrecht, 1985. 174. Aquino, M. A. S., Lee, F. L., Gabe, E. J., Bensimon, C., and Crutchley, R. J., Actu Crystullogr. C 49, 1543 (1993). 175. Aquino, M. A. S., Lee, F. L., Gabe, E. J.,Bensimon, C., Greedan, J. E., and Crutchley, R. J., Inorg. Chem. 30, 3234 (1991). 176. Naklicki, M. L., and Crutchley, R. J., submitted for publication. 177. Ondrechen, M. J., and Wu, X. M., in “Mixed-Valency Systems: Applications in Chemistry, Physics and Biology” (Prassides, K., Ed.), p. 335. Kluwer Academic, Dordrecht, 1991. 178. Ondrechen, M. J., Gozashti, S., Zhang, L.-T., and Zhou, F., in “Electron Transfer in Biology and the Solid State” (Johnson, M. K., King, R. B., Kurtz, D. M., Jr., Kutal, C., Norton, M. L., and Scott, R. A., Eds.), Chap. 12. American Chemical Society, Washington, D.C., 1990.

This Page Intentionally Left Blank

ADVANCES IN INORGANIC CHEMISTRY, VOL.

41

RECENT SYNTHETIC, STRUCTURAL, SPECTROSCOPIC, AND THEORETICAL STUDIES ON MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES J. CLADE, F. FRICK, AND M. JANSEN lnstitut of Inorganic Chemistry, University of Bonn, 53121 Bonn, Germany

I. Introduction 11. Molecular Phosphorus Oxides A. Syntheses B. Crystal and Molecular Structures C. Spectroscopy D. Theoretical Studies 111. Molecular Phosphorus Oxide Sulfides A. Syntheses B. Crystal and Molecular Structures C. Spectroscopy D. Theoretical Studies IV. Comparative Considerations V. Concluding Remarks References

I. Introduction

The phosphorus oxides and oxide sulfides considered in this chapter have one topological feature in common: they all consist of four tetrahedral (PX,) or q-tetrahedral [EPX, (X = chalcogen, E = lone pair)] groups that are connected by three of their four vertices to form cagelike molecules. Phosphorus thus is surrounded by three bridging and one terminal chalcogen atoms, the latter position being occupied by a lone pair in the case of trivalent phosphorus. If all tetrahedral groups are generare identical, molecules of high symmetry (point group Td) ated, and the bridging chalcogen atoms are arranged as an ideal octahe327 Copyright B 1994 by Academic F’ress, Inc. All rights of reproduction in any form reserved.

328

CLADE,FRICK,A N D JANSEN

dron that is interpenetrated by a regular tetrahedron of phosphorus atoms. This type of P4& skeleton frequently is referred to as adamantane like, (CH)4(CH2)6. On adding, removing, or substituting one or more of the terminally bound chalcogen atoms, the symmetry is lowered and the geometry of the P4xG cage is distorted considerably. Because of the wide variety of possible modifications, this family of molecules is particularly well suited for comparative structural, spectroscopic,and theoretical studies aiming at a better understanding of the propagation of perturbations within molecules. This aspect will be emphasized throughout the chapter. As a typical example for the compounds under consideration, P408 is depicted in Fig. 1. The legend gives the nomenclature that will be used in order to describe subunits of the molecules. Especially the phosphorus oxides are classics of inorganic chemistry and have gained considerable relevance for basic research and industrial chemistry as well. Single representatives of the title compounds have been mentioned already in the early times of quantitative chemistry, e.g., P4010and P,Os in 1816 (1). At the turn of the last century, some fundamental work on their synthesis and constitution had been

FIG.1. Molecular structure of P408;schematic illustration of the different P-0 linkP(III)-Ob; ) (c3) PW-0,; and ( 0 )P(V)-O,. To specify the ages. (El) P(III)-08; (I different atoms forming the phosphorus oxide molecules, the following nomenclature will be used: P(II1) denotes a formally trivalent phosphorus atom; P(V), a formally pentavalent phosphorus atom. 0, denotes a bridging oxygen atom between two P(II1) atoms; Ob,a bridging atom between a P(II1)and a P(V)atom; and 0,, a bridging oxygen atom between two P(V) atoms. Oddenotes a terminally bonded oxygen atom.

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

329

TABLE I KNOWNPHOSPHORUS OXIDESAND OXIDESULFIDES, GENERALIZED FORMULAP,O,S, n + m 10

n

10 9 8

9

8

7

6

-

p4°10

X

7 6 5 4 3 2 1

p407

P;O&

p4°3s6

X X

X X X X X

p4°6s

X X X X X

p4°6

X X X X X

accomplished (2-5). From that time to the mid-l960s, almost no activities took place in this field. Then, with increasing applications of X-ray crystallography to inorganic solids, the first reliable crystal structure determinations became available (6, 7). These results, addressed with structural and preparative issue, have been reviewed thoroughly by M. Meisel (8).Since then considerable progress has been made: The knowledge of the structures is far more complete, the synthetic approaches have been improved considerably, and extensive comparative spectroscopic investigations have been performed. These results deserve to be reviewed and discussed within a common context. The compounds that will be included are collected in Table I; the respective literature will be covered until the middle of 1993. Omitted are all investigations including matrix isolation techniques and the binary phosphorus sulfides. II. Molecular Phosphorus Oxides

A. SYNTHESIS 1. General Remarks The preferred route of preparing P406and P4010is the oxidation of elemental phosphorus with dry air, oxygen, N20,or mixtures of oxygen or N20with nitrogen (exceptions, see below). Although P40sand P4Ol0

330

CLADE, FRICK, AND JANSEN

are the anhydrides of phosphorous and phosphoric acids, respectively, it is impossible to prepare them by draining the corresponding acids even by treatment with strongly water-withdrawing agents. For the preparation of mixed phosphorus(II1N) oxides, at least two routes are reported in the literature; the first is based on the pyrolysis of P406 and the second, on the reduction of P4010by elemental phosphorus. Some methods for the preparation of pure P407have been published, but synthesis of pure phosphorus(III/V) oxides has continued to be a challenge in preparative chemistry and is a current subject of investigation.

2 . Phosphorus Pentoxide, or Phosphoric Oxide (P,Ol0) a. Historical Aspects. In the last decade of the seventeenth century, R. Boyle (9)reported that the white fumes forming when phosphorus is burnt in air could be collected yielding a white, very hygroscopic powder. Analyses carried out by H. Davy (101,P. L. Dulong ( I I ) , and J. J. Berzelius (1) showed this substance to be an oxide of phosphorus, having the empirical composition “P205.”Since then early investigators like J. J. Berzelius (12),Z. Delalande (13),R. F. Marchand (141,A. G. Grabowsky (15),T. Goldschmidt (16),and G. Pistor (17)have developed methods of preparing P4010in a large scale.

b. Preparation from Elemental Phosphorus. Phosphorus pentoxide has usually been prepared-as was described by H. Davy already in 1821 (IO)-by burning phosphorus in the presence of a sufficient quantity of dry air or oxygen. The crude product-which generally contains small amounts of lower phosphorus oxides-can be purified by sublimation in flowing dry oxygen (18, 19).The use of platinum asbetos as a catalyst in order to complete the oxidation, as recommended by R. Threlfall (20),is rather unsatisfactory, because asbestos adsorbs the pentoxide, thus removing the platinum from the reaction zone (21). On the technical scale, molten phosphorus is burnt in a stream of dry air in a specially designed burner made of stainless steel. The phosphoric oxide vapor is passed into a cooling chamber, where it condenses into a white powder (22).When produced under proper combustion conditions, the product contains only traces of lower oxides, but depending on the conditions of storage, small proportions of polyphosphoric acids may be present. Today, the commercially available phosphorus pentoxide contains 99,8% P4OlO,less than 0,01% P406and traces of As ( 4 0ppm), Fe (2 ppm), and other heavy metals, e.g., Pb (2 ppm) (23). In addition to the production of H3P04(for fertilizers), the following industrial uses of P4010are worthy of note (23):

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

33 1

-Dehydration reactions (e.g., in methyl methacrylate manufacture); -Manufacture of phosphoryl chloride and triethylphosphate; -A vigorous drying agent; and -An admixture for raising the softening point of asphalt. c. Further Ways of Preparing Phosphorus Pentoxide. H. Davy (24) and J. Dalton (25)showed that white phosphorus reacts vigorously with SO3, Cl2O,C102, and NO2, forming P4OI0.Other methods of preparing or PC13 (261, phosphorus pentoxide, e.g., the action of Nz03on the reaction between P0Cl3 and KC103 (271, or heating K3P04, K4P207,or Kurrol’s salt with an excess of SO3 (281, probably do not yield completely anhydrous phosphoric oxide. 3. Phosphorus Trioxide, or Phosphorous Oxide (P,O,) a. Historical Aspects. Already in 1777, B. G. le Sage (29)succeeded in preparing phosphorous oxide by passing a slow stream of dry air over fragments of phosphorus. A. L. Lavoisier (301, and later P. A. Steinacher (31)and H. Davy (10) recognized that the oxide formed was somewhat different from that obtained by oxidizing phosphorus by combustion. Quantitative analyses carried out by P. L. Dulong (11) and J. J. Berzelius ( 1 ) showed the compositions of the oxides to be “P203”and “P206,”respectively. Nevertheless, the existence of an “anhydride of phosphorous acid” remained doubtful (see, for example, Ref. 321, until in 1886 T. E. Thorpe and A. E. Tutton ( 2 ) succeeded in preparing the oxide in a high degree of purity for the first time.

b. Preparation from Elemental Phosphorus. T. E. Thorpe and A. E. Tutton ( 2 )burned phosphorus in one end of a long, horizontally placed combustion tube in a rapid stream of air drawn through the apparatus by a water pump. They condensed the resulting oxide by cooling the other end of the tube with iced water. It was not necessary to dry the air beforehand. Purification of the crude product was carried out by sublimination in a stream of dry COPat 100°C, trapping the purified material in a cooled, U-shaped condenser adapted to the combustion tube. Further development in preparing P406and purification of the crude material (which in most cases contains small amounts of dissolved phosphorus and higher phosphorus oxides) was achieved by Emelbus (33),who combusted phosphorus under reduced pressure, and C. C. Miller (34,35).The latter pointed out that P406,if prepared by the method of Thorpe and Tutton (21,still contains unreacted phosphorus, which cannot be removed completely even by repeated distillation. However, recrystallization from CS2 or light petroleum seemed to be successful. Final traces of dissolved phosphorus could be removed by

332

CLADE, FRICK, AND JANSEN

conversion into the red form via irradiating the crude material and subsequent distillation. If the latter procedure is repeated several times, the preceding treatment with solvents is not necessary. However, D. Heinz (36)has mentioned that even very pure P406slowly decomposes when exposed to sunlight, forming phosphorus(1IW) oxides and elemental phosphorus. The most recent development in the preparation methods of P,06 on a laboratory scale is due to M. L. Walker in 1975 (37).He oxidized yellow phosphorus by air in the specially designed flow system described in Ref. 37. On the industrial scale, several attempts have been made to develop a continuous process for producing P406by combustion of phosphorus under special conditions. Large-scale availability of P406 is of considerable interest because it could replace PCl, as a precursor for the production of phosphorus(II1)compounds on an industrial scale. In 1929, Wolf and Schmager (38, 39) were able to increase the yield of P406 up to 55% by using a mixture of O2 and Nz ( 3 : l), using a pressure of 90-95 Torr and quenching the combustion products at ca. 50°C. In 1960, D. Heinz and E. Thilo applied for a patent (36,40,4 1 ) for the continuous production of P406by oxidation of white phosphorus under reduced pressure. At that time, they used a reaction temperature of 550-6OO0C, used a pressure of 70 Torr, and purified NzO as the oxidizing agent. The yield of P406with respect to P, was up to 7 0 4 0 % . Further improvements of the production of P406are listed in Table 11(see also Ref. 42, especially for thermodynamical and equilibrium considerations).

c. Further Ways of Preparing Phosphorus Trioxide. Many efforts have been made to prepare P,06 from starting materials other than elemental phosphorus, e.g., H3P03 or PCl,. When A. Nacquet (52) mentioned that P406is formed during the reaction of PC1, with H,PO,, many investigators tried to reproduce this finding. A. Gautier (531,A. Besson (54, 551, and L. J. Chalk and J. R. Partington (56) observed yellowish red, insoluble reaction products of varying compositions, the so-called “phosphorus suboxide.” F. Krafft and R. Neumann (57) reported that they were successful in obtaining P406from PCl, and H,PO,, but L. Wolf, E. Kalaehne and H. Schmager (58),as well as W. P. Jorissen and A. Tasman (59), failed t o reproduce F. Krafft’s results. In more recent papers, F. Hossenlopp et al. (60) have pointed out that the reaction between H3P03and PC13 leads to the formation of pyro- and probably polyphosphorous acids that decompose at higher temperatures, and D. Heinz (36)observed a solid red reaction product, from which-by thermal decomposition--P406 could be recovered at least in up to 0.5% yield.

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

333

TABLE I1

DEVELOPMENT OF THE PROCEDURE FOR THE INDUSTRIAL PRODUCTION OF P406 Starting materials

P4/0,/C0, or P4/

02/co2

Molten P , Oz/N2 (1: l), NzO, or N02/air (1: 41 Gaseous P,, 02/ N,, or NO,/air (1.5 : 1) P vapor, electrically excited COZ P vapor, electrically excited COPor NO P vapor/COz (1: 1.5-1: 15)

P vapor/02/N2 (1: ca 3 : >0-4) P vapor/O,/N, (1: ca.3 : >0-25) P vapor/O,/N,

Reaction temperature

Pressure

1800-2800"C, quenching at (500°C 620°C

70 T

650°C

50 T

3000-12,000"C, quenching at <230"C 0-100°C 75O-150O0C, quenching at <350°C 470°C Quenching at 1500-2000"C, or at 1000-1600°C 500-750"C, quenching at <350"c

Special conditions

P406yield

Ref.

Cooling, for example, with H 2 0

0.001-0.1 atm

43

KOH, Na2C03,or Bn(OH)zas catalyst Unreacted P is recycled to the vaporizer Irradiation (2-6 MHz)

80-95%

44

80 or 65%, respectively 36-46%

45

46

Irradiation (1000-5000 MHz) P406/P407 mixture

24-28%

47

90-95%

48

Quenching by addition of Nz Especially designed burner

84%

49

Ading P406,removing solid products

35%

>85%

50, 51 43

Other trials reported for the preparation of P40sfrom PCl, are -Reaction between PCl, and formic acid (yielding H3P03,CO, and HC1) (61); -Action of acetic anhydride on PCl, (yielding acetyl chloride and a poorly characterized residue of varying composition) (61, 62); -Action of PCl, on acetic acid, trichloroacetic acid, and butyric acid (no formation of P,Os observed) (59); -Action of PCl, on sodium formate (yielding NaC1, H3P03, and CO) (59); -Reaction between PCl, and (Et,Sn),O (63);and -Reaction between PCl, and (N(CH3)4)2S03 in liquid SOz (yielding up to 80% P40s)(36, 64).

334

CLADE, FRICK, AND JANSEN

4 . Phosphorus(I1IlV) Oxides

a. Historical Aspects. In 1884, P. Hautefeuille and A. Perrey (65) combusted phosphorus under conditions that may have led to the formation of phosphorus(II1N) oxides, but they believed to have obtained a crystalline variant of P4O10. Two years later, T. E. Thorpe and A. E. Tutton (2) showed that a distinct crystalline oxide of the empirical composition “PO;’ could be obtained by heating the mixed products from the slow combustion of phosphorus in an evacuated sealed tube at 290°C.’ Building on the results obtained by Thorpe and Tutton, C. A. West ( 5 )prepared the oxide by thermal decomposition of P406 at 200-250°C for two to three days in an atmosphere of dry carbon dioxide. The crude material was purified subsequently by repeated sublimation. Due to its low solubility in most organic solvents and its tendency to polymerize or to decompose at higher temperatures, molecular weight determinations attempted by the early investigators failed (see, for example, Ref. 5).Corresponding to dinitrogen tetroxide, N,04, the substance was called “phosphorus tetroxide” or “phosphorosophosphoric oxide,” and the formula “P204)’was assigned to it. The formation of “phosphorus tetroxide” was observed subsequently by a number of researchers: by C. C. Miller (671,as a product of the oxidation of P406in presence of water vapor, by E. Britzke and N. Pestow (681, as well as P. H. Emmett and J. F. Schulz (691,as a product of the reaction between phosphorus vapor and CO,, and by S. Brunauer and J. F. Schultz (701, as a product of the reaction between phosphorus and water vapor at 1000-1200°C. Nevertheless, the true nature of ‘‘PZ04” remained unclear until 1964, when D. Heinz et al. (71, 72) discovered that these samples represent in fact a mixture of two different solid solutions. The orthorhombic a-phase has been reported to have the empirical composition P408.1-9.2 and to consist of P408,P409,and occasionally P4Olo molecules, whereas the monoclinic P-phase has been reported to have the composition P407,7-8,0 and to contain P407and P408molecules in a solid solution.

b. Preparation of Phosphorus(III1V) Oxides. It was assumed that phosphorus(II1N) oxides could not be synthesized as pure solids, forming solid solutions as a rule. Such phases usually were prepared by thermal decomposition of P406 (2, 5 ) . Another method, based on the reduction of P4010by red phosphorus at 450-525”C, has been reported by D. Heinz et al. (71,72).They pointed out that-in both cases-higher Under somewhat similar conditions, E. J. Russell (66) was able to recover the same substance.

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

335

reaction temperatures lead to the formation of oxygen-rich products and vice versa. The reduction of phosphorus(I1W) oxide mixtures with red phosphorus yields oxide mixtures with a lower oxygen content. A few years later, D. Heinz et al. (73) discovered that phosphorus(III/ V) oxides (/?-phase)are also formed by the oxidation of P40sin carbon tetrachloride with oxygen. By this reaction, P407/P408mixtures with an overall composition P 4 0 7 , 0 - 7 , g are obtained. Occasionallyphosphorus(III/V) oxide mixtures are by-products of a process for the production of P406on a technical scale (48,741. Although M. Loeper and U. Schulke (75) have mentioned that P407 sublimes at 50-80°C (100-200 Pa) and P408at 150-160°C from phosphorus(II1N) oxide mixtures, it seemed to be impossible to separate the three oxides from each other even by repeated sublimation. For gas phase equilibria between phosphorus(III/ V) oxides, see Ref. 42. c . Preparation ofpure Phosphorus(IIIlV) Oxides. M. L. Walker and J. L. Mills prepared P407 by heating a solution of P40s in dry diglyme at 135-140°C (371,by heating a solution of P406in Benzene in presence of traces of moisture, or by stirring a solution containing equimolar quantities of P406and triphenyl phosphine oxide in dry tetrahydrofurane at room temperature (76, 77). The observation that the substance partly decomposed upon sublimation indicates that it probably still contained traces of the solvent. A better method for preparingpure P407was discovered by M. Jansen and M. Moebs (791.' They oxidized P406in a closed system and an oxygen-buffered atmosphere. The reaction vessel consisted of a V-shaped glass tube, one end containing P406and the other containing dry silver oxide. The tube was sealed under argon, and the end that contained the silver oxide was placed into a furnace and heated at 190-210°C. Under these conditions, the oxygen pressure above Ag,O is sufficient to oxidize the P40, that is present in the gas phase. According to Gibb's rule, P407 is the only phosphorus oxide that can coexist under equilibrium conditions with an excess of unchanged P,O, as a condensed phase. By this reaction, P407is formed as colorless,transparent, highly lustrous crystals just above the part of the tube that is heated. Further investigations about the synthesis of P407by oxidizing P406 with highly diluted oxygen (obtained from a special designed Pure P407 argodoxygen mixing apparatus) is still in progress (80,811. is also obtained in up to 20% yield by the oxidation of P40s with Pure P407is also formed in the reaction of P406with alkali metal oxides, e.g., K20 or Cs20 (78).

336

CLADE, FRICK, AND JANSEN

FIG.2. Apparatus for the preparation of pure P407(79).

(CH3),N-S03(1,2-dichloroethane, Soxhlet apparatus) or with 1,Cbenzodioxane.SO, (THF). Oxidation of P406with 1,4-dioxane.S03yields mixtures of P407and P408(82). Although crystals of the @-phaseof phosphorus(II1N) oxides with a high content of P408can be obtained (7, 71,72, 73),, nobody has yet been able to isolate the pure oxide. D. Heinz et al. assumed that phosphorus(II1N) oxide mixtures that contain >95% of P408form only amorphous solids. This may be due to the fact that P408has the lowest molecular symmetry of all phosphorus oxides considered here. Further investigation of this subject is still in progress. M. Jansen and B. Luer have been successful in isolating pure P,09 from phosphorus(I1I.N) oxide mixtures by dissolving P40, and P408 with 1,4-dioxane and subliming the residue, which consisted mainly of the less soluble P4og (83). Recent experiments (80) indicate that crude P4og is also accessible by the reaction of P40,, with red or white phosphorus at 450°C and subsequent sublimation from 390°C to room t e m p e r a t ~ r eX-ray-powder .~ diffraction and 31PMAS-NMR data prove that the amount of P408in the product obtained is less than 3%.

By oxidizing P406 with tetrabutylammoniumperiodate in dry CH2C12,occasionally a solution that contains P408as the only phosphorus oxide was obtained, as indicated by 31PNMR spectroscopy. Nevertheless, attempts to isolate the substance have failed, because the product polymerizes when the solvent is drawn off (80). The so-called “phosphorus suboxide” (see, for example, Ref. 84)reduces P,Olo already a t lower temperatures, e.g., 350”C,forming phosphorus(III/V) oxide mixtures with a high content of P,09 (85).

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

337

B. CRYSTAL AND MOLECULAR STRUCTURES 1. Historical Aspects

By the end of the nineteenth century, molecular weight determinations carried out by means of vapor density measurements proved that phosphorus(V)and phosphorus(II1)oxide in the gaseous state consist of molecules of the compositions P40s (3,86)and P4010(87), respectively. However,their molecular structures remained uncertain (see, for example, Ref. 88) until adequate techniques for structure determination such as electron and X-ray diffraction became available (6,7, 72). 2 . Determination of the Molecular Structures of Phosphorus Oxides in the Gaseous State

The structures of P40sand P4OlOwere first investigated by means of electron diffraction. L. R. Maxwell, S. B. Hendricks, and L. S. Deming (89) reported the structure of P40s, but were unable to deduce the molecular structure for P4010.The authors erroneously attributed a lower symmetry than Td.The correct structure of P4010was found by G. C. Hampson and A. J. Stosick (90). Several investigations followed with the goal of improving the accuracy of the bond lengths and angles (see Tables I11 and IV). Concerning the phosphorus(I1W) oxides, no structural data in the gaseous state are available; the only attempt (89) to determine the molecular structure of one of the phosphorus(I1W) oxides, P408,has failed. 3. Determinution of the Molecular Structures in the Solid State a. Crystal Structure of P,OIo. Phosphorus(V) oxide was the first oxide of phosphorus that has been structurally characterized by means of X-ray crystallography. There are three crystalline modifications (95-97), which can be designated as M, R, and S modifications, accord-

TABLE I11 MOLECULAR STRUCTURE OF P406 IN THE GASEOUSSTATE

Ref. 89 P-0 0-P-0 P-0-P

1.67(3)A 128.5(1.5)"

Ref. 90

Ref. 91

1.65(2)A 99(1)" 127.5(1.0)"

1.63") A 99.8(0.8)" 126.4(0.7)"

Note. Standard deviations are given in parentheses.

338

CLADE, FRICK, AND JANSEN

TABLE I V

MOLECULAR STRUCTURE OF P401, IN THE GASEOUSSTATE

p-obr

P-ot Obr-P-Obr Ob,-P-O, P-obr-P

Ref. 90

Ref. 92

Ref. 93

1.62(2) A 1.39(2) A 101.5(1.0)" 116.5(1.0Y 123.5(1.0Y

1.60(1) A 1.40(3) A

1.604(3) hi 1.429(4) A 101.6(8)" 116.5(3Y 123.5(7Y

124.3(1.0Y

Note. Standard deviations are given in parentheses.

ing to a suggestion of E. Thilo and E. Wieker (94). In this chapter, only the structure of the metastable modification, which consists of P4010molecules (M modification), will be considered, because the other modifications consist of polymeric networks of tetrahedral PO4 units connected via common vertices. The M modification is the commercially available form of phosphorus(V) oxide. Its crystal structure was refined by C. D. J. Cruickshank (98) using the experimental data collected by H. C. J. de Decker and C. H. McGillavry (95)(see Table V). Surprisingly this refinement yielded different values for the bond lengths of the two crystallographically independent P(v)-od bonds. These values, 1.41(2) and 1.51(4)A,respectively, should be equivalent within the T,symmetry of the free molecule. Because they differ by an unreasonably large amount, a redetermination of the crystal structure of P4010was undertaken (99). The geometry of P4Ol0as obtained by this latter refinement is consistent with the point group T d and is also in good agreement with the electron diffraction results for the gaseous oxide. TABLE V MOLECULAR STRUCTURE OF P4010IN THE SOLID STATE

Ref. 95 P(V)-O, P(v)-od

1.63 hi 1.41 A

Ref. 98 1.58(2) hi 1.41(2) and 1.51(4) A,

Ref. 99 1.599(3) A 1.441(4) A

respectively o,-P(v)-o, o,-P(V)-od P(v)-o,-P(v)

102.3" 115.6" 121.4"

101.0" 117.0' 124.3"

Note. Standard deviations are given in parentheses.

102.2(2)" 116.1(2Y 122.8(1Y

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

339

b. Crystal Structure of P406. Phosphorus(II1) oxide was the last oxide of phosphorus to be characterized structurally in the solid state (100,101). This is due to its low melting point, which introduces difficulties for the growth of single crystals. These have been met successfully by growing crystals from the melt in Lindemann glass capillaries in situ on a diffractometer. In the course of low-temperature X-ray investigations (101), Guinier photographs indicated a structural phase transition of higher order taking place at -45 2 4°C.The determination of the structure of the low-temperature modification (102) revealed that the molecular geometry of P406does not change at the phase transition, instead the molecular packing is changing slightly (see below). The bond distances in solid P40, [1.653(3) A1 are larger than those determined for gaseous P406 [1.638(3) A]. However, the confidence ranges for the bond length distributions still overlap. c. Crystal Structure of P40,. The crystal structure of P409was first investigated in 1964 (6). This determination was affected by using a mixed crystal of the empirical composition P408,96.Based on the experimental data obtained, a refinement (93)was carried out. A more precise determination (lower standard deviations) for pure P409(83) has been supplied. The bond lengths and angles obtained in these three investigations are listed in Table VI.

d . Crystal Structure ofP,O,. The only structural data on the crystal structure of P40s were obtained again from a mixed crystal with the empirical composition P407,g (7). This indicates that some of the P408 molecules are substituted by P407 molecules, and thus the data obtained are less reliable. A similar caveat holds for the results of a structure refinement (91) based on the same diffraction data.

TABLE VI

BONDLENGTHS(A) IN SOLID P409

P(III)-ob P(v)-ob P(V)-O, P(v)-od

Ref. 6

Ref. 93

Ref. 83

1.66(1) 1.60(1) 1.59(1) 1.44(2)

1.661(13) 1.605(17) 1.590(12) 1.418(14)

1.675(2) 1.573(3) 1.608(2) 1.443(2)

Note. Standard deviations are given in parentheses.

340

CLADE, FRICK, AND JANSEN

TABLE VII BONDLENGTHS(A) IN PHOSPHORUS OXIDES

P(III)--O,, P(III)-ob P(v)-ob P(V)-O, P(v)-od

P406(100)

P407(78)

P408( 9 1 )

P409(83)

P4Ol0(99)

1.653(3)

1.640(9) 1.680(6) 1.590(8)

1.633(10) 1.672(14) 1.576(17) 1.596(9) 1.414(15)

1.675(21 1.573(3) 1.608(2) 1.443(2)

1.599(3) 1.441(4)

1.450(5)

Note. Standard deviations are given in parentheses.

e. Crystal Structure of P407. The structure of P407was determined in 1981 independently by K.-H. Jost and M. Schneider (103)and by M. Jansen and M. Moebs (78, 79), the results being virtually identical. f. Comparison of the Molecular Structures of the Phosphorus Oxides. Figures 1 and 3-6 illustrate the structures of the molecules P40, ( n = 6-10) with a schematicpresentation of the different P-0 linkages; Table VII provides a comparisonof correspondingbond lengths, employing the most accurate data available. Here, distortion of the P,06 cage due to the addition of terminally bonded oxygen atoms is illustrated. Whereas the P406molecule has only one type P-0 linkage, there are three different types in the P40, and P409molecules and even four in P40,. These show considerable variations in their lengths. The distortion of the molecules is reflected most clearly by the P-Oh distances. /---

FIG.3. Molecular structure of P406.See legend for Fig. 1.

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

341

FIG.4. Molecular structure of P407.See legend for Fig. 1.

P(III)-ob, e.g., in P,O~(1.680 A), is 6.5% longer than the P(v)-ob linkage (1.573 A).Generally, the bonds starting from formal pentavalent phosphorus atoms are shorter than those starting from trivalent phosphorus atoms. In addition, a propagation of the distortion introduced by the terminal oxygen atoms throughout the bonds of the P406 cage can be identified. By comparing analogous linkages in the different phosphorus oxide molecules, it can be seen that in general the P406 cage is contracted upon successive oxidation of the phosphorus atoms

FIG. 5. Molecular structure of P409. See legend for Fig. 1.

342

CLADE, FRICK, AND JANSEN

FIG.6. Molecular structure of P 4 0 1 0 . See legend for Fig. 1.

(see also Section 1II.B).Table VIII gives the bond angles in the phosphorus oxide molecules. As a rule, the angles involving P(II1) atoms are always smaller than those involving P(V) atoms because the lone pair of P(II1) requires more space than a P(V)-Od linkage (104).Analogous bond angles in the different phosphorus oxide molecules expand when terminally bonded oxygen atoms are added. Although most of the phosphorus atoms in the oxides are crystallographically independent, the symmetries of the molecules as determined for the crystalline state do not differ significantly from those TABLE VIII BONDANGLES("1 IN PHOSPHORUS OXIDES (78, 83, 91, 99, 101) p40,

99.6

p4°1

100.0 98.3 97.4 103.5

114.9 127.0

128.1 123.9

P408

98.5 97.3 105.0 102.7 115.6 113.6 130.1 124.4 121.2

p409

103.2 100.8 118.7 114.4 126.3 122.4

P40lO

102.2 116.1 122.8

343

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

found or expected for the gaseous state and are expressed by the same point groups (78, 79,82,83,99,101).

g . Molecular Packing of the Phosphorus Oxides in the Solid P408,P409,and P4010show close State. The phosphorus oxides P407, relations in the metrices and symmetries of their space groups (105). K.-H. Jost realized that the lattice constants of P408resemble those of a monoclinic unit cell of rhombohedral P409(only the c axis is half as long). P407has nearly the same unit cell parameters as P,08 (83). Eventually, it was demonstrated that a monoclinic setting with close relationship to the crystal lattices of the other phosphorus oxides can be found for rhombohedral P4Ol0too (Fig. 7) (105). The unit cells of all these oxides can be reduced to one common, approximately body-centered “pseudo-unit cell” with a = b = c and

Y

I

-X

FIG.7. Projection of the crystal structures of p407 (A), p40~(B),p409 (0, and P4010 (D) along the c axis of the monoclinic (A, B)and pseudomonoclinic (C, D) settings (105).

344

CLADE, FRICK, AND JANSEN

a = /3 = y 90" containing two molecules. This body-centered packing of molecules is realized with highest possible structural symmetry (point group, Td; space group, 1Z3rn) for hexamethylenetetramine. By means of a systematic reduction of the space group symmetry, this structure can be correlated with those of the different phosphorus oxides. The relationships among the phosphorus oxides have been displayed in Fig. 8 using a tree-like structure to exhibit the hierarchic order of group-subgroup relationships (see also Ref. 105). The observed intermolecular distances show that there are only van der Waals-type intermolecular forces. Obviously, the structural relations of the phosphorus oxides are due to the endeavor to adopt a n optimally dense packing (see also Section 1II.B).Here the rules formulated by A. I. Kitaigorodski (106),i.e., maximal space-filling structures L-

I

4 3m

k2 I

P43"

R 3c

pZ-1 FIG.8. Group-subgroup relationships among phosphorus oxides (105).

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

345

are created by the matching of concave and convex regions at the surfaces of neighboring molecules, are fulfilled. As a consequence,packings of molecules frequently resemble those of metals or alloys. In case of the phosphorus oxides, the tungsten type of arrangement is adopted. It has been determined that (surprisingly!) P40sadopts neither this packing (101,102) nor an fcc arrangement such as As406. The structures of both modifications are formed of "rods" of P406 molecules aligned along one of their threefold symmetry axes. The rods are parallel to the crystallographic a axis and are closely packed in the bc plane. The P406molecules of the monoclinic high-temperature modification show a crystallographic mirror plane that is lost upon the transition to the triclinic low-temperature modification (102).The mirror plane defined by two phosphorus and two oxygen atoms is retained during phase transition, but loses its special orientation with respect to the crystal lattice, thus lowering the crystal symmetry to triclinic. Table IX gives a summarizing overview of the crystal structures of all five molecular phosphorus oxides. TABLE IX CRYSTALLOGRAPHIC

Oxide

a

DATAOF MOLECULARPHOSPHORUS OXIDES

Crystallographic symmetry

Space group

Monoclinic

P2,lm

Triclinic

pi

Monoclinic

PZ1/n

Monoclinic

C2lc

Trigonal-rhombohedra1

R ic

Hexagonal

R 3c

Standard deviations are given in parentheses.

Lattice constanta a = 6.422(1) A b = 7.877(2) A c = 6.786(3) A p = 106.UlY a = 6.39(1) A b = 7.86(1) c = 6.77(1) A a = 87.7(2)" p = 107.1(2Y y = 91.5(1)" a = 9.807(2) A b = 9.973(4) A c = 6.840(2) A p = 96.83(1Y a = 9.66(4) A b = 10.10(4) A c = 6.93(3) A /3 = 96.8(3)" a = 10.0057(2) A a = 57.947(2Y a = 10.3035(9) A c = 13.5102(19) A

346

CLADE, FRICK, AND JANSEN

TABLE X VIBRATIONAL SPECTRA OF

IR (CS2solution)

285 vw 302 vw 370 w 407 vs 549"w 569" w 587"w 643 vs 687"w 702"w 718 vw 808 vw 832" m 849" w 919 ws 1018 m 1048 w 1180 w 1258 s 1273 s 1313 w 1460 vw 1545 w 1625 w 1880 vw

PdOs (114)

Raman

Assignment

285 vw 302 (10) 370 (variable) 407 (18) 465 (variable)

E FZ Dissolved PA?

569 ( 3 , ~ )

A1 302 f 3021825 = 6041587 A1

613 ( 1 0 0 , ~ ) 643 (55) 691 (1)

FZ

Dissolved P4

FZ E 302 + 407 = 709 ?

919 (5)

407 + 407 = 814 Fl 569 + 285 = 854 F2 407 + 613 = 1020 407 + 643 = 1050 919 + 302 = 1221? 613 + 643 = 1256 643 + 643 = 1286 919 + 407 = 1326 832 + 643 = 1475 919 + 643 = 1562 832 + 832 = 1664? 919 + 919 = 1832??

" Very weak or (probably) absent in the vapor spectrum.

C. SPECTROSCOPY 1 . Vibrational Spectroscopy

a. P406 and P4010. The fundamental work carried out by H. Gerding. H. v. Brederode, and H. C. J. de Decker [Raman spectra and calculation of force constants (107-110)]has been completed by T. A. Sidorov and N. N. Sobolev [IR spectra (111,112)1,by D. H. Zijp (113), and by A. C. Chapman (114)(vibrational assignment and calculation of improved force constants). The frequencies and intensities of the bands observed, and the assignment (as given by Chapman) are listed in Tables X and XI.5 Force constants for the P,06 and P,Olo molecules The gas phase vibrational spectra of P406and P4010are reported in Refs. 115 and 116; for matrix IR spectra, see Refs. 117-119.

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

347

TABLE XI VIBRATIONAL SPECTRA OF P4010(114)

IR

Raman

Assignment

194 vw? 237 vw? 257 s 278 m

F2

355 vw br? 424 s

F2

559 s 576 vw

A, Fz

E

280 330 w; sh

421 510?m;br 573 s 610 vw; sh 670 w; sh 698 w; sh 760 vs; m 820 v;sh 8381 m 846 1010 vs 1110 m 1130 m 1181 w 1242 w 1273 m;w 1390 vs 1628 m ;s 1695 m;s 1720 w 1770 w 1835 vw 1880 vw 1935 vw 2000 vw 2060 vvw 2150 vw

710 vw? 720 s 760 vw 828 vw

F1,E? ? F2

556.2 = 1112 556 + 573 = 1129 760 + 424 = 1184 ?

720 + 556 = 1276 1385 m 1418 s

F2

A1 820 ’ 2 = 1640 846 . 2= 1692 1010 + 720 = 1730 1010 + 760 = 1770 1413 + 424 = 1837 9

1390 + 559 = 1949 1010 * 2 = 2020 1384 + 720 = 2094? 1413 + 760 = 2173

were first calculated by H. v. Brederode and H. Gerding (109) using Bjerrum’s simple valence force field system. Table XI1 gives a list of the improved force constants (obtained by using the extended force field method) given by Zijp (113).

348

CLADE, FRICK, AND JANSEN

TABLE XI1

FORCECONSTANTS(mdyn/A) CALCULATED FOR p40SAND p4°10

P4010

p4°S

f (pobr) f(P0,) d (POb,P) d (ObrPObr)

I

I1

I

I1

4.787

3.825

0.51 0.153

0.38 0.125

5.25 11.0 0.35 0.35 0.61

3.902 10.84 0.42 0.21 0.557 0.11 0.69 0.08

d (O,,POJ 0.11 0.44

Fg" Fa F,

Note. I, simple force field; 11, extended force field (113). A determination carried out by Chapman (114) using a modified Urey-Bradley force field yielded similar results. In the case of P4Ol0,a slight improvement of the force field has been achieved by Muldagaliev and coworkers (120). F,, F,, and Ft are the coefficients for the second-order development of the potential energy, having the form 12

2 V = Fq r &br

12

6 br

+ F,

rgp + Ft

2ria,.

For these calculations, the bond lengths and angles from electron diffraction data (89, 90, 93; see also Section 1I.B) have been used. Considering the empirical correlation between the valence stretching constants and the bond order [according to D. W. J. Cruickshank (123),], a bond order of 1.0625 was derived for P-0 in P,O, (114), already suggesting a slight d-orbital participation. For P4OlO,the bond orders are 1.33 (P-Ob,) and 2.2 (P-0,) ( 1 1 4 ) , respectively. This has been taken as support for the assumption that d-orbital participation in the P-Ob, bonds is much stronger for P,O,,, than that for P406. Although very accurate structural data for the P407and P,O, molecules are available (see Section II.B), no force constants have been reported so far. For the purpose of discussing the relation between molecular geometries and chemical bonding (which is an important topic of this chapter), such calculations would be very desirable. A comprehensive discussion about vibrational spectra and force constants of phosphorus compounds is supplied in Ref. 124.

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

349

TABLE XI11 VIBRATIONAL FREQUENCIES OF PHOSPHORUS (IIIN) OXIDES IN SOLID ARGON(117) p407

P408

p409

Assignment

1379.4

1401.6 1384.7 1004.3 995.9

1406.4 1397.8 1019.4 1011.9 1005.7 706

P=O stretch; Al P=O stretch; B2 or E, respectively P40s cage modes

984.8 976.3 966.4 653.5 649.6 634.6 625.9 537.5 427.6 391.0

683.7 669.3 614.5 556 440.4

563.4 43 1

P406cage modes P406cage modes, A, P=O bending P406cage modes

b. Phosphorus(III/V) Oxides. In 1989, Z. Mielke and L. Andrews ( I 17) recorded matrix-infrared spectra of phosphorus oxides that had

been produced by oxidizing P40s with 03.For P40s and P4010,the results agreed with earlier data [see Chapman (114)l.Table XI11 gives a list of the frequencies observed for the phosphorus(I1I.N) oxides and their assignments. M. L. Walker and J. L. Mills (37),and later M. Jansen and M. Moebs (79),investigated the vibrational spectra of P407(see Table XIV). In Ref. 79, a scheme correlating the Raman-active bands of P407with those observed in the Raman spectra of P,Os and P4O10 is given (see Fig. 9). Because there is no preparative access to pure P408(see Section II.A), no vibrational spectrum of this compound is available. The vibrational spectrum of pure P409shows absorptions at 1379(vs),1137(m),989(vs), 777(m), 751(m), 657(s), 609(w), and 515(w) cm-'. No assignment has yet been supplied (1211. 2. 31PNMR Spectroscopy in Solution

The 31P NMR spectrum of P40s consists of an extremely narrow singlet signal (122); its chemical shift is +112.5 k 0.1 ppm upfield versus 85% &PO,. It has been suggested to use P406as an internal standard in 31PNMR spectroscopy (1221,becuase its resonance falls in a special region convenient for referencing the majority of phosphorus compounds. In 1979, M. L. Walker, D. E. Peckenpaugh, and J. L. Mills

(282.6)*

inactive

A2

missing

E

inactive

285 (9

vw

303

m

E

269 306 (p?)

F2<

407

333

111

E

263

vs

E

278 (317.4)' (370.4)' (388.9)

m inactive missing inactive

m

W

missing

E

392 m

E

in

A1 A1

625 (p)

644 673 (831.7)*

m W

A1

rn F2 -3657 E

inactive

F1 92 1

vh

E 708

<

vw

E A?

inact've missing 935(p?)

vw

Al

960

vw

E

1333 (p?)

ni

Al

E

>

42s 560 576 719

7

5

VS W

S

vw

762(?) (7X1.2)'

missing

(954. I )*

inactive

>

vw

niisbing

F2<

Ill-s

1417

vs

calculated

FIG.9. Correlation of the vibrational absorption bands of P407with those of P40s and P,Olo (79).

E

351

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

TABLE XIV IR

AND

RAMANABSORBANCES OF P&

Absorbances (37) IR absorbances (79)

cs2

solution (IR)

crystalline (Raman)

CS2 solution

crystalline 1345(s) 965(s;br) 930(sh) 78Uw) 711(m) 695(vw) 674(w)

{:2:

Raman lines (79) 1333(p?;m) 960 (vw) 935(p?;vw) 708(vw)

]

635(sh) 615(m) 554(vw) 532(m) 427(m) 392(vw) 353(vw) 323(w) 314(w) 299(m) 266-277(m)

I657(m)} 625(p;vs) 534(m) 429(p;m) 392(m) Missing 333(w) Missing 306(p?;m) 269(m)

Assignment (79)

A1

651 . 2 = 1302 424 + 629 = 1053 E A1 392 ' 2 = 784 E 323 + 353 = 676 A1 E 314 ' 2 A1 266 . 2 A1 A1 E E E E? A1 E

=

628

=

532

reported their 31PNMR spectroscopical investigations on phosphorus (IIUV)oxide chalcogenides ( 77). They observed the resonances of P407 at +20 ppm (D)and -173 ppm (Q)and those of P408at +11.3 ppm (T) and - 154.7ppm (TI.The shifts were measured relative to P406, where positive values indicate upfield shifts. As expected from the molecular structure (see Section II.B),P407shows an A,X-type spin system, the resonances of the trivalent phosphorus atoms being shifted downfield. is 3 : 1, and the coupling The intensity ratio of the signals (P(III)/P(V)) is 2.4 Hz.P40s shows an A,X,-type spin system; the constant 2Jpp spectrum consists of two triplet signals with an intensity ratio of 1 : 1, the coupling constant being 13.5 Hz.Because of its insufficient solubility in most organic solvents, it took until 1991 to observe the 31P NMR

352

CLADE, FRICK, AND JANSEN

spectrum of P409 in solution for the first time (82).7The AX3-type spectrum shows resonances at -56.3 pprn (Q)and -145.2 ppm (D, relative t o P406),with an intensity ratio of 1 : 3. The coupling constant 2Jpp was found t o be 42.8 Hz. 3. Solid-state NMR Spectroscopy a. P4Ol0. In 1991, A.-R. Grimmer and G.-U. Wolf investigated the various polymorphs of phosphorus pentoxide by means of 31P MAS-NMR spectroscopy (126).The spectrum reported for the volatile modification suggests that in P4010all phosphorus atoms are equivalent (which does not agree with the crystallographic features; see below). The isotropic chemical shift is +46.7 ppm (versus 85% H3P0, as an external standard), and the values for the anisotropic shielding tensor are uI = -51.7 ppm and uII = +244.4 ppm. This means that the shielding tensor has positive axiality along the P=O bond.8 In an investigation (8O), 31PMAS-NMR spectra of P4010,showing two independent isotropic signals at -45.3 and -46.9 ppm, were recorded. They corresponded to the two crystallographically independent types of phosphorus atoms (see Section II.B), with an intensity ratio of 3 : 1. The more intense of these signals has a remarkably large half width and might consist of three different signals. This would mean that all four phosphorus atoms in P,Ol, can be distinguished spectroscopically, whereas three of them are crystallographically equivalent (!). b. P,O,. Because of the low melting point of phosphorus(II1) oxide (22.8"C),solid-state NMR spectroscopy has been performed only on static samples so far (SO).' The chemical shift of the isotropic P(II1) signal is +113.0 ppm; the components of the anisotropic shielding tensor are listed in Table XV. c. PhosphorudZIIW) Oxides. The 31PMAS-NMR spectrum of P,07 (80)shows two isotropic signals, at +128.1 ppm [P(III)I and -53.2 ppm [P(V)],with an intensity ratio of 2 : 1. This means that-in contrast to P4O10 (see above)-the three crystallographically independent [P(III)] atoms cannot be distinguished. The remarkably large half width of the P[III] signal is probably caused by a dynamic effect, e.g., a rotation of M. L. Walker and J. L. Mills (125)have reported the observation of a quartet signal at -6.63 ppm, *JPp = 2.5 Hz,which they assumed to be caused by P409. Reference 82 clearly points out that this attribution is not reliable. The same has previously been observed in phosphoryl compounds (127,128) and some phosphates (129). For relaxation time measurements on solid P406and a discussion of relaxation mechanisms, see Refs. 130,131.

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

353

TABLE XV OBSERVED AND CALCULATED [IGLO METHOD(130)l COMPONENTS OF THE ANISOTROPIC SHIELDING TENSOFS OF PHOSPHORUS OXIDES

-230

- 202 -230 -202 + 120 + 139 -113 -88

-235 -218 -235 -203 + 86 + 124 - 128 - 99

- 150 - 151 - 150 - 151 + 137 + 146 -55 -52

- 35 -41 - 35 -41 +229 +240 +53 +53

-59 -63 -59 -49 +222 + 244 + 35 +44

-52 -49 -52 -49 +244 + 255 +46 +52

the molecule around its C, axis. Therefore only the mean values of the anisotropic chemical shift could be observed. Obviously, this effect did not affect the determination of the crystal structure by means of X-ray diffraction (see Section 1I.B). This suggests that the rotation proceeds via very fast rotational jumps between long stationary times. Low temperature measurements on P407have not been performed yet. The spectrum of P409(80)shows two isotropic signals, at +54.5 ppm [P(III)] and -34.7 ppm [P(V)I. Because all three P(V) atoms are crystallographically equivalent, this agrees with the expectations. No solid-state NMR data for P408have been reported (see Section 1I.A). Graphical interpretations of the spectra [using the Herzfeld-Berger approximation (13311yield the components of the anisotropic shielding tensors (see Table XV). Comparing the isotropic shifts of P(II1) in P,Os and P407,it becomes obvious that mainly uII is influenced by the oxidation of one single phosphorus atom, whereas remains approximately constant. Similarly, by comparing the isotropic shifts of P(V) in P407 to P&g, it is seen that mainly cI is affected. These tendencies are shown in Fig. 10, in which the static spectra of the phosphorus oxides are depicted schematically.

P(II1)

1

300 250 200 150 100 50

0

-50 -100 -150 -200 -250 pprn

-_.-_.....

m.-.,-v-TT

300 250 200 150 100 50

0

-50 -100 -150 -200 -250 pprn

~ ~ . - - - . ~ - . _ - - _ * - . I' , 300 250 200 150 100 50

T

p

0

-50 -100-150-200-250 ppm

0

-50 -100-150 -200-250 ppm

~

p4010

T-

.l-r-v-T-T-p-7T-,

300 250 200 150 loo 50

I..

1 ,

- --1

FIG. 10. Schematic representation of the 31Pshielding tensors of phosphorus oxides (80).

355

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

4 . X-Ray Absorption Spectroscopy

The phosphorus K shell photoabsorption spectra (XANES and EXAFS) of P,06 and P4OlOhave been reported in comparison with those of P(OPhI3and O=P(OPh), (82,134,135).The absorption bands and their assignments are listed in Table XVI, and the full spectra are depicted in Figs. 11 and 12. The first very intense absorption band at low energies (so-called “white line”; attributed to transitions of a P(V)-1selectron to unoccupied molecular or Rydberg-type orbitals) occurs in P 4 O l o at 2151.74 eV, a value that is 4.30 eV higher than the energy of the white line in the spectrum of P406.This shift is explained by the higher oxidation state of phosphorus in P4O10, which implies a stronger attraction of the electrons. The locations of the preedge absorptions and of the shape resonances of both spectra on the energy scale are very similar to those found in triphenylphosphite and triphenylphosphate, respectively. This illustrates that transistions of core electrons are mainly influenced by effects of the first coordination shell. In general, the location of the white line is characteristic for the local environment of the atom under consideration. TABLE XVI

EXPERIMENTAL AND THEORETICAL ABSORFTIONSTRUCTURE FOR P406 AND P4010(135) Term value (eV) Peak No.

Energy” (eV)

Experimental Calculated

P406

1 2 3 4 IP 5 6 P40,Ll 1 2 3 4 IP 5 6 7

2144.57 2147.44 2149.7 2152.09 2153.7 2154.7 2164.1

-9.25 -6.25 -3.98 - 1.60 0 1 9.4

2150.3 2151.7 2153.51 2 154.76 2156.17 2158.1 2161.99 2170.17

-5.87 -4.38 -2.97 -1.07 0 1.93 5.83 14

-6.25 -3.61 -3.6

Assignment

-

? 1s -D 5e t 5a, 1s 6e 1s- 7e

Lorentz width (eV)

Peak area

(N)

(N)

0.82 0.71 0.9 3.09

0.02 1 0.3 0.85

0.81

0.8 0.93 1.74 1.65

0.03 1 0.37 0.3

0.03 1 0.3 0.6

p

density

1 0.27

Shape resonance Shape resonance -5.61 -4.38 -2.75 -2.07

1s1sIs+ 1s-

6a,

6e 8e

9a,

+ 7e

Shape resonance EXAFS oscillation Shape resonance

The energy values have been obtained by “unfolding” the spectra by a least-squares fit using Voigt profiles, the main criterium being the minimization of the number of profiles.

356

CLADE, FRICK, AND JANSEN

L

15 -

~

p406

IP.

3

7

4 5

6

1 10-

1

1

2

m c 0 .c

I'

u

g

v) n

a

p4010

53 IP.

7

4

1

0-

2140

'

I

2150

"

2160

'

I

2170

"

2180

2 90

FIG. 11. XANES spectra of P40s and P4OI0(135).

Ionization limits cannot be inferred from photoabsorption spectra; however, they are accessible from photoelectron spectra (see below). Because there exists a linear correlation between 1s and 2p binding energies, the 1s ionization potentials for P406and P4Olo were estimated using data reported for PC13 (135). In the case of P4Ol0,recent work (138,139) shows that the estimated value is in excellent agreement with the experimentally determined one (see below). The various absorption lines have been assigned with the help of MS-Xa calculations. Because the main influence on the shape of the XANES spectra is from the atoms in the first neighbor shell, only "constituent clusters" have been treated in the simulations, e.g., P033for P40s and PO:for P,Olo. In terms of this approximation, some details of the spectra (e.g., intensity and line widths of the absorptions) could be simulated sufficiently, whereas the predicted term values calculated for transitions near the absorption edge lie too close to each

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

357

20 -

? m

Y

C

0 .c P

o, XI

10-

I

triphenylphosphite

1

Energy [eV]

FIG.12. XANES spectra of (PhO)sP and (Ph0I3PO (135).

other. This points to the difficulties in modeling near-edge unoccupied orbitals (see Section 1I.D). The Fourier-transformed EXAFS spectra (135) (depicted in Figs. 13 and 14) show maxima at 1.7 and 2.95 A (P406)and at 1.6 and 2.9 A (P,Ol0),the first maximum corresponding to the P-0 distance and the second corresponding to the P-P distance (the two different P-0 distances in P4010are not resolved). This is in satisfactory agreement with X-ray structural data (see Section 1I.B). Besides structural information (which is of minor interest in this case), EXAFS spectra reveal experimental scattering phases for the P-0 system and allow us to distinguish between resonances caused by single scattering processes and those caused by multiple scattering processes (so-called shape resonances). If one plots the inverse squared bond lengths vs the term values of the shape resonances, a linear relation results (see Table XVII and Fig. 151, and thus an empirical correlation between the term values of

358

CLADE, FRICK, AND JANSEN

E, 0

c v)

c

E

c

I

1

.

2

I

.

3

I

.

I

.

5

4

]

6

Distance R (A)

FIG.13. Fourier-transformedand phase-correctedEXAFS spectra of P406 (135).

the shape resonances and the interatomic distances seems to exist (135).However, it is hard to find a physically meaningful explanation for the existence of this correlation. The XANES spectrum for P,09 (80,140)shows a preedge absorption band at 2148.2 eV, which is assigned to an electron transition from the P(II1)-1sorbital into an unoccupied MO with main contributions from P(II1) and 0. Above 2150 eV, two broad resonances have been observed, which could be fitted by three Voigt profiles of approximately

I

.

I

1

1

2

.

1

1

3 Distance R (A)

4

.

1

5

.

6

FIG. 14. Fourier-transformedand phase-correctedEXAFS spectra of P4010(135).

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

359

150 135

-

120 -

45

~

0060 0105 0150 0195 0240 0285 0330 0375 0420 0465 0510

1/R2 ( Ang-2) FIG. 15. Term values of the shape resonances versus the inverse squared distances (135).

equal amplitude and half width (at 2150.9, 2152.3, and 2153.2 eV). Comparing this part of the spectrum to the corresponding energy region of the spectrum of P4OlO,these resonances can be assigned to P(V)-ls transitions. The “splitting” (P4010shows only one resonance in this region; see Table XVI) is probably due to the lower symmetry of the molecule. The spectrum of P407is rather similar to that of P40s but shows an additional absorption band at 2152.8 eV, corresponding to the excitation of an P(V)-lselectron. As one would expect, the intensity ratio of the P(III)/P(V) absorption bands is 3 : 1, whereas it is 1: 3 in the case of P40,. TABLE XVII (135) TERMVALUES OF THE SHAPE RESONANCEAND DISTANCES ~~

Molecule p4°S p4°

10

P(OC&,)3 PO(OC&,), p4°S

P(OC&)3 PO(OC&)3 p4°10

Distance P-P P-P P-C P-C P-0 P-0 P=O P=O

(A)

2.95 2.826 = 2.582 = 2.582 = 1.647 = 1.61 = 1.49 = 1.44

= =

lir2

(A-2)

0.115 0.125 0.15 0.15 0.37 0.385 0.45 0.49

Term value [eVl

No.

1 1.9 2.86 2.84 9.4 10.6 12.2 14.0

1 2 3 3 4 5 6 7

360

CLADE, FRICK, AND JANSEN

TABLE XVIII IONIZATION ENERGYDATAFOR P40ti(141) IE [eVl

Assignment

5t2 (P lone pair) 3 q (P lone pair) 2e (P-0 bonding) 2tl (0 lone pair) 4tz (0 lone pair) 3tz (P-0 bonding) Itl (P-0 bonding) 2a, (P 3s) 2tz (P 3s)

10.55 12.79 13.90 15.84 17.98 21.71 [He(II)I

5. Photoelectron and Auger Electron Spectroscopy

UV photoelectron spectra of P406and P4010have been reported (141) and compared with the spectrum of P, (142,143). The He(1) photoelectron spectrum of P,06 contains five well-defined bands; an extra band \ \

F'b,

\ \ \

\

\

\

-6

1.45

1.55

1.65

1.75

P - 0 distances [A]

FIG.16. Correlation of the IS energies of P(V) with the P(V)-Ob bond lengths for P,06 derivates (165).

361

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

TABLE XIX

IONIZATIONENERGY DATAFOR P4O10 (141) IE (eV)

Assignment ~~

13.40 13.92 14.44 (sh) 14.76 (sh) 15.36 16.54 18.37 [He(II)I 20.80 [He(II)I 23.83 [He(II)I

3tl (apical 0 PP) 7tz (apical 0 PP) 3e (apical 0 p ~ ) 6tz (0 lone pair) 6tz (0 lone pair) 2e (P-0 bonding) 4ul (P lone pair) 5tz (0 lone pair) 4tz (P-0 bonding) Itl (P-0 bonding) 3Ul (P 3s) 3tz (P 3.9)

appears at a higher binding energy in the He(I1) spectrum. Due to the more complicated electronic structure of P401,,, its UPS bands are more difficult to assign. However, a clear qualitative resemblance to the spectrum of P,Os in the high-binding-energy region is observed, and the experimental ionization energies of both oxides (see Tables XVIII and XIX) have been correlated with each other (and with those of P,) (141) (see Fig. 16). These experimental results are in excellent agreement with ab initio MO calculations (141). X-ray photoelectron spectroscopy (XPS)has been used to determine phosphorus (and oxygen) core binding energies in P40s(144)and P4OlO (145-147). The energy values obtained are listed in Table XX. As in TABLE XX OF P406AND P4O10 (ev) CORE BINDING ENERGIES

P(ls)

P(2s)

P(2p)

O(l8)

139.87 137.4 Experimental 135.6

539.25(3)

Ref.

~~

p4°6

P40lO

Theoretical 139.3 2150.25

193.25

135.8

534.2 532.5 517.3 515.0

148 147 144

138,139

362

CLADE, FRICK, AND JANSEN

the case of the UPS spectra (see above), the values for P4010are in sufficient agreement with MS-Xa calculations of the electronic structure. The most prominent peak in the Auger electron spectrum of P4010 was observed at 1848.0 eV (KL,,, L2,3; 'D,). With respect to the XPS data listed in Table XX, the Auger parameters a = 1983.8 eV, 5 = 30.65 eV, and R,"" (2p2p) = 6.55 eV are obtained (135,136). 6. Other Techniques

The vacuum-ultraviolet spectrum of P40, (149)shows two bands at -48900 cm-' and 63900 cm-', the first of which exhibits a low-energy shoulder. By analyzing the energies of these bands with respect to the symmetry of the molecule, they can be assigned to fully allowed 'A, + 'T, transitions (mainly involving orbital transitions 5t2 3e and 2t1+ 3e, respectively). The experimentally obtained data are reproduced by extended Huckel and ab initio SCF calculations. The microwave spectrum of the pyrolysis products of CH30PC12in the range of 26.5-39.5 GHz (150)suggests that-in addition to CH,=PCl, HC=P, CH3C1,and HCHO-P407 is present. This assumption is based on the observed rotational constant of the molecule (Bobs= 827.8). Mass spectrometric studies on P4010and P40, have been conducted (151,153). In Ref. 151,thermodynamic properties of P4010are derived and compared with earlier data (for example, Refs. 156-1591, and in Ref. 153,the heat of formation of P40s is derived by measuring the temperature dependence of the equilibria by mass spectrometry ( H , = -405 t 17 kcal/mol). This latter value for the heat of formation of P406seems to be in good accordance with that reported by S. B. Hartley and J. C. McCoubry (1601, whereas the data given by W. R. Koerner and E. Daniels deviate considerably (161).The mass spectra indicate that all three phosphorus(III/V) oxides can exist in the vapor phase; the fraction of the lower oxides increases with higher temperatures.

-

D. THEORETICAL STUDIES 1. General Remarks Only a few theoretical studies on molecular phosphorus oxides have been reported, with broadly differing purposes. Two investigations deal with the explanation of photoelectron data and core binding energies (141,1481, whereas in a third UV, vacuum-UV, and MCD spectra (149) are considered. The results of MS-Xa calculations have been compared to XANES spectra (135),and IGLO calculations served to simulate MAS-NMR spectra (162).The only substances considered were P406 and P4010.Some calculations referring directly to data obtained from

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

363

spectroscopic investigations were mentioned in Section II.C, whereas the literature reviewed in this section is of a more general scope. 2. Bonding Features

The goal of an early publication (163)was to investigate the extent to which the formally empty 3d orbitals of phosphorus participate in the bonding scheme of P406 and of P4OlO.For these calculations of the LCAO SCF CNDO formalism has been used. Population analyses demonstrated that 3d orbital participation should occur to a considerable degree. The d-orbitals are notably occupied, and their populations increase with an increasing number of oxygen atoms in the molecule. It has been suggested that electronegative groups attached to the phosphorus atoms would lead to a contraction of the 3d orbitals and cause them to mix with the inner orbitals ( s , p )to a greater extent. Moreover, P,06 appears to be virtually nonpolar, whereas in P4010the P-0, bonds are highly polarized. They are found to be multiple in nature and to have a considerable 7~ component involving the p and d orbitals on phosphorus. However, subsequent investigations have shown that semiempiric calculations (for example, CND0)-in contrast to ab initio methods-are not reliable for atoms beyond the second period, because the fixing of the parameters for d-AOs is problematic (164).Mulliken population analyses and determinations of the orbital energies in P407 have been compared with those for P406S(see Section 1II.D) (166). An explanation was attempted for the geometrical distortion of the The geometries P406cage when an oxygen or sulfur atom is added (165). of the molecules P406and P407(P406S)were calculated using a DZPbased HF formalism and were found to be in an excellent agreement with experimental data. Moreover, the geometries of other hypothetical P406derivatives [P406+and P406X(X = F+,C1+,H+, N-)] have been determined. The calculated P(v)-ob distances of these molecules are correlated with the effective charges on the P(V) atoms (see Fig. 17).As a measure for the effective charges, the P(V)-1senergy shifts (compared with the 1s energy of phosphorus in P406)have been used. The P(V)-Ob distances in the different P406derivatives vary over a range greater than 15 pm, and in the case of P406N- the bond length would even increase compared with P406. The distances of P(v)-ob have been plotted vs the distance of P(III)-ob. Although the substituent X is changed, the sum of these two bond lengths remains approximately constant. The general conclusion from this work is that the geometrical distortion of the P406cage as a whole can be related to the amount of charge transferred to the terminal substituent (165).

3 64

CLADE, FRICK, AND JANSEN

0

0

0

0

0

1.60

-

0

+

0

P4O6N-

P406 P407 P406S P406H+ P406+ P~OSCI-P4O6F+ P4O6 derivatives

FIG. 17. Bond lengths of P(V)-Ob (t), those of P(III)-O~ (O), and sum of both for P,Os derivates.

(+)

111. Molecular Phosphorus Oxide Sulfides

A. SYNTHESES 1, General Remarks Two general routes for the preparation of phosphorus oxide sulfides are reported in the literature: -Sulfurization of phosphorus oxides (route 1) and -Commutation reaction between P4010and P,Slo (route 2). The preparative challenge in both cases is the exact adjustment of the experimental parameters in order to obtain the desired species as the major product. 2. Phosphorus(V) Oxide Sulfides a. Historical Aspects. Until 1976, P406S4was the only phosphorus oxide sulfide identified unambiguously. It was first mentioned in 1891

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

365

by T. E. Thorpe and A. E. Tutton (3, 41, who heated a sealed tube containing P406and S to 168°C.A very vigorous reaction was observed, and the authors noted that the use of more than 5 g of P406and its equivalent of sulfur led to a violent explosion. In 1897, a compound of the composition P404s6 was reported to form during the reaction of POCl, with H2S (167).Only elemental analyses were available to confirm the composition. P404s6 is also supposed to form by the reaction (168)

6(CH3)3SiSSi(CH3)3 + 4 POCl, = 12(CH3),SiC1+ P404Ss. Elemental analysis (sulfur content) was the only evidence given for the existence of this oxide sulfide, and a speculative structure analogous to P4010with the six bridging atoms of the cage being sulfur atoms, was proposed. However, J. L. Mills and coworkers (77) were not able to reproduce the syntheses as reported and concluded that the structure suggested for P404SG was at least unlikely. In addition, the melting point as reported by E. W. Abel et al. differs from that obtained by G. U. Wolf and M. Meisel, who proved the identity of their product (P404s6)unambiguously (see below).

b. Preparation according to Route 1. In 1973,a patent (169)describing a continuous technical process for the preparation of P406S4from P406and S was applied for. In this process, P406is added to a sulfur melt heated at 190°C, and so the limitations of the method of Thorpe and Tutton were overcome. P40& and P408S2have been detected as by-products in the reaction of P407and P,Slo in o-xylene (77,170).

c. Preparation according to Route 2. The use of the reorganization reaction is the preferred route for preparing P406s4 because the starting materials are easily available. P4010 and P4S10are mixed in a molar ratio of 3 : 2 and then heated at 400°C for a few hours. P406s4 is distilled from the reaction mixture simultaneously with its formation therein (171,172).Other phosphorus(V) oxide sulfides can be prepared by the reaction of P4OlOand P4S10 in a P4Slo/sulfurmelt at 500°C (173).P405S5 and P404s6 are obtained as pure substances by repeated distillation in uamo of the products that continuously distill off the mixture during the reaction. P40S9 can be obtained in a pure state by extraction of the distillation residue with CS2 and subsequent evaporation of the solvent. P408S2,P407S3, P403S7, and P402S8are available only as mixtures and have been identified by 31P NMR spectroscopy in solution (173).

366

CLADE, FRICK, AND JANSEN

3. Phosphorus(III1V) Oxide Sulfides

a. Historical Aspects. The oxidation of P4S3in CS, solution with molecular oxygen has been found to yield an amorphous phosphorus oxide sulfide ofthe composition P&o4 (174). The substance was characterized by elemental analysis, and the authors thought to have obtained a compound analogous t o the phosphorus sulfide P4S7.However, until 1976 (170) no phosphorus(III1V) oxide sulfide was definitely known.

b. Preparation according to Route 1 . Pure P406Sand pure P406S2 are accessible by reacting P406with P4S10 in toluene or o-xylene, respectively (77, 170, 175). P40,S, P407S2,and P408S have been detected spectroscopically(31PNMR) among the product mixtures resulting from the reaction of P407 and P4Slo in o-xylene (77, 170). c. Preparation according to Route 2. Pure P403S6forms analogously to P40S, (see above) by fractional crystallisation from a CS, solution that is obtained by extraction of the solid residual products from the reaction of P4010and P4S10 (173).

d . Preparation ofP406S3. Pure P406S3 can be prepared by desulfurization of P406S4with P(C6H5I3in toluene at room temperature (175).

B.

CRYSTAL AND

MOLECULARSTRUCTURES

1. Crystal and Molecular Structure of

P40S4

The molecular structure of P406s4 in the gaseous state was determined in 1939 by means of electron diffraction (176). Its crystal structure was determined by X-ray diffraction (177) and refined on the basis of diffractometer data in order to obtain state-of-the-art structural parameters (80). The bond lengths and angles obtained in these three investigations are listed in Table XXI. TABLE XXI MOLECULAR STRUCTURE OF P40& ~~

Distance P(W-0, Distance P(V)-S Angle O,-P(V)-O, Angle O,-P(V)-S Angle P(V)-O,-P(V)

~

Ref. 176

Ref. 177

Ref. 80

1.61(2)h 1.85(2)b 101.5(1.0)” 123.5(1.0Y 116.5(1.0F’

1.62(2)8, 1.86(2)b 100.8(9Y 117.2(8)” 124.5(6Y

1.620(2)8, 1.886(1)h 101.7(1)” 116.7(1Y 124.2(3)”

Note. Standard deviations are given in parentheses.

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

367

TABLE XXII BONDLENGTHS(A) IN PHOSPHORUS OXIDE SULFIDES (80)

P(I1I)-Oa P(III)-ob P(v)-ob P(V)-O,

P(V)-S

1.637(2) 1.678(2) 1.596(2) 1.890(1)

1.636(2) 1.671(2) 1.595(2) 1.614(2) 1.885(1)

1.661(2) 1.595(2) 1.611(2) 1.882(1)

1.620(1) 1.886(1)

Note. Standard deviations are given in parentheses.

2. Crystal Structure of P403S6

Among the structurally characterized phosphorus oxide sulfides, P40,S6is the only representative that is not derived from the P406cage (I78,179).Its molecular structure is shown in Fig. 18. The structure is analogous to that of P409 (see above), with three oxygen and three sulfur atoms bridging the four phosphorus atoms in the adamantanelike cage, as well as three terminally bonded sulfur atoms. 3. Crystal Structures of

P,O&

(n = 1-3)

The structures of pure P406S,P406S2, and P406S3in the solid state were determined (166,175).The structural data obtained are listed in Tables XXII and XXIII. The P406cage is made to contract by successive addition of sulfur atoms, as can be seen in Table XXIV, in which the

FIG.18. Molecular structure of P40& (179).

368

CLADE, FRICK, AND JANSEN

TABLE XXIII

ANGLES(“1 IN PHOSPHORUS OXIDE SULFIDES (80, 101 )

Oa-P(I1I)-Oa o,-P(III)-o~ o~-P(III)-o~ ob-P(v)-o~ O,-P(V)-O,

99.6

99.5 98.6 103.1

oc-P(v)-o,

O,-P(V)-S Oc-P(V)-S P~III~-oa-P~III~ P(III)-ob-P(v) P(V)-O,-P(V)

115.3 128.1 124.2

98.8 98.0 103.6 101.5 116.6 115.1 129.0 125.2 123.5

98.4 101.6 100.8 117.9 116.1 126.1 124.2

101.2 116.9

124.2

volumes of the cages of the different substances are compared with each other. By comparing the molecular geometries of the phosphorus oxide sulfides with those of the oxides, it is obvious that the extent of the distortion due to binding terminal chalcogen atoms depends on whether oxygen or sulfur is added. In the P,O& molecule, the P(V)-Ob distance is shorter than that in the corresponding oxide, P,09. This produces a discrepancy in the results obtained from an analogous consideration of the P40, and P,06S moelcules, for which the geometries of the P,O, cages remain virtually identical (166). As with the oxides, the molecules of the phosphorus oxide sulfides in the solid state belong, within the limits of experimental error, to the same point group that would be expected for the free molecules (80, 166,175).

TABLE XXIV

VOLUMESOF THE P40sCAGESOF THE SUBSTANCES P406S, ( n = 0-4) (80) Substance

Volume of the P406cage (A3) 100.7 99.8 98.9 96.9 96.1

369

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES 0

A

FIG.19. Stereoplot of the unit cell fo P406S3 (shifted by 0.2442; 0.0248; 0.3283) (175).

4 . Molecular Packings of the Phosphorus Oxide Sulfides P406S, (n = 0-4) in the Solid State

In contrast to the oxides, the phosphorus oxide sulfides do not show pronounced relationships to each other with respect to their metrices or space group symmetries. The P406Sand P406S3molecules form a severely distorted face-centered cubic packing (166,175).This is shown for P406S3in Fig. 19 (the P406cages are representated by tetrahedra). The section shown has its origin shifted by 0.2442;0.0248;0.3283. For P406S2,no simple description for the molecular packing has been found. P406s4 molecules form a slightly distorted primitive cubic packing, as can be seen in Fig. 20 (80).

FIG.20. Stereoplot of the packing of P406s4 molecules (80).

370

CLADE, FRICK, AND JANSEN

TABLE XXV PACKING OF P&Xn MOLECULES(x = 0 , s; n

Density (g/cm3) Molar volume (cm3) Volume of the molecule (A3) Volume of the unit ceWZ (A3) Efficiency value

2.14 99.3 100.7 164.9 0.61

2.21 (2.34) 114.0 (100.8) 117.15 (103.5) 189.5 (166.3) 0.62 (0.62)

=

0-4)

IN THE SOLID STATE

2.23 127.4 135.9 211.2 0.64

2.10 (2.64) 150.5 (101.5) 152.9 (109.4) 250.6 (168.6) 0.61 (0.65)

(80)

2.05 (2.28) 175.6 (123.0) 171.5 (110.2) 282.7 (207.0) 0.61 (0.53)

The occurrence of completely different crystal structures within the same class of substances, as well as the difference in the structures of the analogous oxides, is surprising, especially if one is aware that van der Waals-type forces are predominant in all cases. Achievement of an effective space-filling is thought to be the main factor determining the type of a molecular van der Waals packing. It has been shown for many structures of organic compounds that the principle of “dovetailing” (“protrusions” into “hollows”) is effective (106). In order to check whether this principle applies in this case, the volumina of the P406X, molecules are compared in Table XXV with the space available in the respective crystal lattice (80),which is obtained by dividing the volume of the unit cell by the number of molecules in it. The ratio of the two volumina is defined as the “efficiency” of the packing. With the exception of P401,,, efficiency values between 0.61 and 0.65 have been obtained. This indicates that, in spite of the differences among the crystal structures of the phosphorus oxides and oxide sulfides, their space-fillings are very similar. C. SPECTROSCOPY

1. Vibrational Spectroscopy a. P406S4. Characterizations of P40&4 by means of vibrational spectroscopy (180)have resulted in a complete assignments of the absorption bands observed, as well as force constant calculations (108). Table XXVI summarizes these early data and compares them to later

371

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

TABLE XXVI VIBRATIONAL ABSORPTION BANDSOF SOLID P4o,js4 ~~

Ref. 113 Ref. 180 Ra

Ra

IR

Ref. 80 Ra

1388(m) 1328(m) 1214(m) 1139(w) 978(vs)

98 1 934(p)

929(w) 898(vw)

930 897 769

700(dp) 668(dp)

713(vw) 698(w) 669(w) 646(vw)

694 666

IR

882(m) 753(w) 712(w) 704(w) 671M 646(m) 569(w) 548(m)

Assignment 113

F2 A1 E

F2 A1 F2

498(p) 460 447(p) 400(dp)

400

354(dp)

353

194(dp)

259 194

150(dp)

148

F2

447(vw) 406(vs) 387(m) 352(m) 328(w) 317(vw) 288(vw) 271(vw)

F2

E F2 170(w) 160W 155(vs) 145(vs) 133(w)

E

95

experimental data (80).The most reliable values for the wave numbers are those from Ref. 80, a study in which a spectrometer with significantly higher resolution than that in the previous investigations was used. Thus in this work many more bands were observed than reported in the older literature; e.g., the Raman line at 150 cm-' [already observed by H. Gerding and H. v. Brederode in 1945 (180)]is split into five bands at 170, 160,155,145,and 133 cm-' due to lattice effects.

372

CLADE, FRICK, AND JANSEN

The bands at 1388 cm-l might be explained by the presence of small amounts of P4010due t o the route of preparation (see Refs. 171 and 172). Force constants for the P406s4 molecule were calculated from Raman spectra and were improved later (113).Table XXVII gives the values obtained using the simple valence force field method (I) or the extended force field method (11) (113);the calculations have been carried out using interatomic distances and angles from electron diffraction data (178).

b. Phosphorus(III1V) Oxide Sulfides. The infrared spectrum of P406Swas recorded in 1979 (77). However, it took until 1993 (80), when each compound of the series P406S,( n = 1-3) became available as a pure substance (see Section II.A), for comprehensive spectroscopical studies. The vibrational absorption bands of these species in a crystalline state are listed in Tables XXVIII and XXIX. Given the molecular symmetries of the phosphorus(III1V) oxide sulfides, 16 absorption bands (both IR and Raman) for P,06S (point group, C 3 J ,25 IR and 30 Raman bands for P40,S, (point group, C2J, and 19 bands (both IR and Raman) for P406s3 (point group, C3J are expected. All IR-active absorptions should be Raman active too. Deviations of the expected spectra from the observed spectra are in most cases due t o the fact that symmetry analyses do not predict the intensity of an absorption band. The experimental spectra of P406S3show more than the 19 absorptions predicted from symmetry analysis. This is an indicaTABLE XXVII FORCECONSTANTS[mdyn/Al FOR THE P&& MOLECULE (113)

f (PO) f (PS) d (POP) d (OPO) d (OPS)

Fg" Fa Ft

I

I1

5.25 5.50 0.35 0.35 0.43

3.848 5.122 0.373 0.187 0.373 0.12 0.69 0.08

' F,, Fa,and F,are the coefficients for the second-order development of the potential energy (see Section 11.0.

373

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

T A B L E XXVIII VIBRATIONAL

Ref. 77 IR

1006(m) 940-986(~) 895(m) 850(m) 687(m) 663(m) 642(m) 626(m)

ABSORPTION BANDSOF P406s

Ref. 80 Ra

958(w) 91Uw) 814(w) 690(w) 653(w) 632(m) 592M

IR 2924(w) 2853(w) 1264(s) 1202(w) 1010(sh) 948b;br) 900(sh) 830(m) 688(m) 668(w) 646(w) 63Uw) 598(m)

Ref. 80

Ref. 77 IR

Ra

528(m) 515(m)

IR 534(w)

505(s)

415(m)

415(m) 344(m) 313(w) 300(m) 186(s) 159(w) 59(s) 32(s)

502(w) 466(w) 451(w) 417(m)

TABLE XXIX VIBRATIONAL

ABSORPTIONS OF P40&

AND

P4O& (80)

446(m) 420(w) 421(s) 367(vw) 354(s) 343(m)

159(vs)

450(vw)

374

CLADE, FRICK, AND JANSEN

FIG.21. Molecular structure of P4s604 (182).

tion that the symmetry has been reduced by the effect of the crystal lattice, and therefore the point group approximation is no longer sufficient. Calculations of force constants of phosphorus(II1N) oxide sulfides are ongoing. In order to gain an experimental basis as broad as possible for the discussion of correlations between molecular geometry and chemical bonding, such calculations are very desirable.

2. 31PNMR Spectroscopy in Solution The 31PNMR spectrum of P406S4 in CS2 solution (181) consists of a singlet signal with a chemical shift of - 103.4ppm relative to internal P40s. For P404s6,H.Grunze (182)has reported an AX,Y-type spectrum, which would be in agreement with the molecular structure shown in Fig. 21. The 31PNMR spectra of the remaining phosphorus oxide sulfides1° were recorded using solutions that in most cases contained mixtures of these compounds (77). The chemical shifts and coupling constants lo The phosphorus oxide sulfide P403S6has been described by G. U. Wolf and M. Meisel (178) shows an A,X-type 31PNMR spectrum with signals at +46.0 ppm [P(III)I and +55.7 ppm [P(V)],the coupling constant being 108 Hz.This correspondsto the molecular structure described in Section 1II.B.

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

375

TABLE XXX 31PNMR SPECTRA OF PHOSPHORUS (III/V)OXIDESULFIDES IN SOLUTION WITH P406AS THE INTERNALSTANDARD (77) P(II1)

P(V)=O

P(V)=S

*JPP

6 (ppm)

6 (ppm)

6 (ppm)

(Hz)

-101.7 (Q) -92.3 (T) -91.8 (D) -88.1

P406S P40sS2 P406S3 P407S

+12.5 (D) +9.2 (T) -27.5 (Q) +11.0

-160.0

P407Sz

-32.9

-155.5

P407S3 P4OBS

-38.9

-159.5 ( Q ) -152.2

-103.2 (D) -87.0

-159.0 (TI

-105.1 (TI

P40RS2

-89.1

12.8 1.9 14.3 9.5 [P(III)-P=OI 0.7 [P(III)-P=Sl 46.5 (P=O-P=S) 28.1 [P(III)-P=Ol 18.7 [P(III)-P=Sl 50.0 (P=O-P=S) 48.0 33.0[P(III)-P=Ol 23.0 [P(III)-P=Sl 55.0 (P=O-P=S) 52.5

for these substances are listed in Table XXX. The spin systems of the spectra coincide with the molecular structure in each case (see Section 1II.B). 3. Solid-state NMR Spectroscopy

The 31PMAS-NMR spectra of the phosphorus oxide sulfides P406S, (n = 0-4) have been reported (80,162).P406s2and P406s3 show more isotropic signals than those obtained from solutions of the same compounds (see above). The spectrum of P406S2comprises two isotropic signals for P(II1) and two for P(V), and the spectrum of P406s3 shows one signal for P(1II) and three signals for P(V). The intensity ratio in both cases is 1: 1: 1: 1. This means that the crystallographically independent (however, within the limits of accuracy of the structure determination, geometrically equivalent) phosphorus atoms produce different signals in the solid-state NMR spectrum. In contrast, the spectrum of P406Sshows only one isotropic signal for P(II1) and one for P(V),the intensity ratio being 3 : 1.This effect is explained by rapid rotational jumps of the molecules around their threefold axis. Lowtemperature measurements (at - 100°C) show that the P(II1) signal splits into three components of approximately equal intensity, as is expected from the crystallographic data. The spectrum of P406S4shows only one isotropic signal because of the crystallographical equivalence

376

CLADE, FRICK. AND JANSEN

of all four phosphorus atoms in the molecule. This signal exhibits a remarkably broad half width and can be interpreted as the superposition of four signals that have approximately equal intensity and have half widths corresponding to those of the signals of the other oxide sulfides (see P,Olo in Section 11.0. The differences between the chemical shifts of the isotropic signals of the solid substances and those obtained from solutions are less than 9 ppm. This shows that lattice effects have only marginal impact. Graphical analyses of the side band patterns [corresponding to Herzfeld and Berger (13311 yield the principal axes of the shielding tensors and the amounts of the anisotropic chemical shifts. These values are listed in Table XXXI. Within the range of experimental error, with one exception, all tensors of the anisotropic chemical shift show positive axialities. For 31Pthere exists only one plausible orientation of the tensors with axial symmetry within molecules of the given structure, namely, with its largest component, uII,along the P=X bond or in the P(II1)-lone-pair direction. The shielding tensor of P(II1) in P,O,S seems to have negative axiality, but this observation can be explained by assuming rotation of the molecule around the threefold axis at room temperature. This interpretation has been confirmed by low-temperature measurements. Figure 22 shows TABLE XXXI

COMPARISON OF THE OBSERVED AND CALCULATED 31PNMR DATAOF PHOSPHORUS OXIDESULFIDES P406S,( n = 0-4) (162) p406

p4°6s

p40&

p4°6s3

- 125 - 141 - 125

-136 - 154 - 136 - 145 + 200 + 233 -23.7 -22.1

- 145

-137

159 - 145 - 151 + 226 + 245 -21.0 -21.2

- 149

-218 -209 -218 - 189 + 90 +131 -115.6 -89.2

176 -171 - 176 -171 + 103 + 138 -83.6 -68.2

141 + 205 + 227 -15.3 -18.2 -

-230 -202 -230 - 202 + 120 + 139 -113 -88.2

-95

209 -95 - 201 - 178 + 133 - 122.6 -92.4 -

-

-

p4°6s4

137 149 + 240 +254 -11.1 - 14.9 -

-

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

377

(3

1

e

4'6' II

1

I

I

+200

t100

0

-100

-200

I -200

I -100

I 0

t100

I

I

I

FIG.22. Schematic representation of the sulfides (162).

I -300

8 [PPml

I

I +300 0 [PPml shielding tensors of phosphorus oxide

+200

schematically the shielding tensors for static samples that have been simulated using the experimental MAS data. For P(V),the value of uIIincreases from +205 ppm (P,O,S) to +240 ppm (P&S4), with an increasing number of terminally bonded sulfur

378

CLADE, FRICK,AND JANSEN

atoms, whereas ( T remains ~ approximately constant at - 135 ppm. This effect is correlated with the decreasing length of the P=S bond, which implies a higher electron density. In contrast, for the P(II1) tensors cII remains approximately constant at + 100 ppm, whereas uI increases from -230 ppm (P406)t o - 176 ppm (P40,S3),with an increasing number of terminally bonded sulfur atoms. The matrix elements of the anisotropic shielding tensors as calculated by the individual gauge for localized orbitals (IGLO) method (130) are compared with the data obtained experimentally in Table XXXI; except for P,O,S, the agreement is convincing. 4 . X-Ray Absorption Spectroscopy The X-ray photoabsorption spectra of the gaseous phosphorus oxide sulfides P,06S, (n = 1-4) have been recorded for the phosphorus K edge (80, 186). The energies of the XANES absorptions are listed in Table XXXII; the spectra are depicted in Fig. 23. The white line in the spectrum of P,06S4 appears at 2150.1 eV. This value is 2.7 eV larger than the corresponding one for P40s, illustrating the dependence of X-ray absorption on the valency of the absorbing atom. In addition, the energy difference between the white lines of P40, and P,06S4is much smaller than that between P406and P401,,; this is obviously due to the lower electronegativity of sulfur with respect to oxygen. Considering the spectra of the phosphorus(II1N) oxide sulfides, it is seen that the absorption at 2148 eV increases in intensity with an increasing number of terminally bonded sulfur atoms, whereas the bands above 2149.5 eV decrease. This effect is explained by the increasing amount of P(II1) and the correspondingly decreasing amount of P(V).This phenomenon is further illustrated by simulated spectra that have been obtained by simply adding the spectra of P406and (PhO),P=S. The striking similarity between the simulated and the observed spectra indicates that the relevant molecular orbitals are highly localized and that it is sufficient to consider the first coordination shells t o explain the main features of the XANES absorptions. At room temperaTABLE XXXII

P(1s)-XANES ABSORPTIONS (eV) OF PHOSPHORUS OXIDESULFIDES ( 8 0 ) P408 p4°6s2 p406s3

P406s4

2147.8 2148.4 2148.2

2149.7 2150.3 2150.1 2150.1

2150.5 2151.6 2151.1 2150.9

2152.9 2152.2 2152.1 2152.5

2153.3 2153.4 2153.6

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

379

Energy [eV]

FIG.23. XANES spectra of phosphorus oxide sulfides (80, 186).

ture, the XANES spectrum of P406S fits well into the trends shown by the other phosphorus oxide sulfides (134).Surprisingly, the intensities of the absorption bands change upon heating the sample. Decomposition of the substance could be definitely excluded, because the effect is reversible. This phenomenon is not understood so far. 5. Photoelectron and Auger Electron Spectroscopy The P U S ) and P(2p) binding energies in P406S, P406S2,and P 4 0 6 s 4 have been determined (80). As it would be expected, the ionization potentials increase with the increasing oxidation state of the phosphorus atom. Astonishingly, they seem to be independent of the electronegativity of the terminally bonded chalcogen atom; e.g., the P ( l s ) and P(2p)binding energies are approximately the same in P4OI0and P 4 0 6 s 4 , respectively. This observation can be explained by the fact that both (+ attraction and T back-donation of sulfur are less effective than those

380

CLADE. FRICK, AND JANSEN

in the case of oxygen. As a result, the electron density of phosphorus in (RO),P=O is approximately the same as that in (RO),P=S. The XPS and P-KLL Auger electron spectra of the phosphorus(III/ V) oxide sulfides P406Sand P406S2(80) do not show any significant splitting as would be expected due to the different oxidation states of phosphorus. This finding has not been understood so far; further experimental and theoretical work is required.

D. THEORETICAL STUDIES The relative stabilities of the structural isomers P406s4 (with a P406 cage) and P4S604 (with a P4s6 cage) have been discussed with reference to the hypothetic adamantane-like 0 1 : + system (183).In this “uniform reference frame,” the bridge atom sites have negative partial charges with respect to the bridgehead atoms, but the locations of highest negative charge density are the terminal or exoatom sites. By replacing 0 atoms with P and S atoms, the “rule of topological charge stabilization,” which was derived in Ref. (184), becomes effective. This rule claims that-in comparison to the reference frame-the heteroatoms are preferentially located in positions where the electron density is accumulated or depleted due to connectivities and electron levels, respectively. Atoms with low electronegativity would prefer positions of low charge density and vice versa. Thus P4s604 should be more stable than P406s4. But, as was discussed in Ref. 182, the experience of synthetic chemistry casts some doubt on this conclusion. Furthermore, the known phosphorus oxide sulfide of the formula P4s604 has the structure shown in Fig. 21 (1821,which is in contradiction to the above theoretical results. The surprising fact that the geometry of the P406cage in P407 and P406Sremains virtually unchanged upon substitution of the terminal oxygen atom through a sulfur atom has been investigated (165) by SCF calculations with the TURBOMOLE program. The Mulliken population analyses of P406S and P407 show that the most pronounced differences appear at P(V), whereas the charges of all other atoms are much less affected. Furthermore, the charge distributions on s, p , and d orbitals (except for the terminal P-0or P-Sbond) remain constant. A more detailed analysis (185)demonstrated that population analysis might not be reliable because of the high polarity of the molecules. Moreover, attempts to compare the differences of the electronic energies in P,06S and in P&, considering either the real or the undistorted geometry of the P406cage, have failed because the resulting differences

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

381

are far smaller than the variation of the core repulsion energies due to the shortening and lengthening of the P-Ob bonds. This failure suggested the use of P(1s) energy as a standard for the effective charges (165) (see Section 1I.D) and thus the correlation of the PUS) energy with the distortion of the cage. As already pointed out in Section II.D, there exists a strong correlation between P(ls) energies and distortion of the P406cage. Here again the geometrical equivalence is reflected by the P(1s) energies, which are approximatelythe same for P406Sand P407.It appears as if sulfur terminally bonded to phosphorus has the same “effectiveelectronegativity” as oxygen attached to the same position, in spite of the fact that oxygen has the higher “standard” electronegativity. On the other hand, the electron density of the P=O bond exceeds that of the P=S bond. Thus one could explain the same effective electronegativities of oxygen and sulfur in this situation by assuming that sulfur has a more pronounced ability t o accomodate additional electron density.

IV.

Comparative Considerations

The synthetic approaches to molecular phosphorus(I1W)oxides and oxide sulfides have some features in common: The compounds P406X (X = 0, S) are prepared from the lowest oxidized member of the P406X, series (P406itself) by the reactions

I0,l

p406

p4°1

and

whereas P406S3is obtained by the reduction of the highest oxidized members of the P406Xnseries (P4OlOand P406s4) by the reactions

and

382

CLADE, FRICK, AND JANSEN

However, much more striking are the general differences in reactivity encountered during the syntheses of oxides and oxide sulfides. Due t o the higher oxidation capability of oxygen compared with sulfur, the reaction conditions for the preparation of pure P407must be fixed much more carefully. As has been pointed out in Sections 1I.A and III.A, the reaction of P406and O2takes place at room temperature, whereas the reaction of P406with elemental sulfur requires elevated temperatures. In addition, it is generally much easier to produce P4?& in a pure state than to produce the corresponding phosphorus oxides. This fact is mainly due to the tendency of the oxides to form solid solutions, which is favored by the close relationships in metrics and symmetries of their crystal structures. These relationships might be a consequence of the fact that differences in the molecular volumes of the members of the P40,0, family are smaller than those of the members of the P406S, family. Further, synthesis and purification of the oxide sulfides are facilitated by the better solubility of the higher sulfurized members in organic solvents. Besides the differences in the molecular packings already mentioned in Sections 1I.Band III.B, the distortions of the P406cages vary depending on whether oxygen or sulfur atoms are added. This is as expected and can be attributed to the different abilities of oxygen and sulfur to withdraw electron density from the cage. The only exception to this rule is the couple P406S and P407,for which the P406cages exhibit approximately the same distortion. As has been discussed in Section III.D, these similarities are a result of two competing effects, which lead to approximately equal “effective” electronegativities of sulfur and oxygen in that special situation. Variations in the electronic properties of the P406X, molecules with respect t o the type of the terminally bonded atoms have been investigated using several different spectroscopic techniques. The isotropic 31PNMR shifts of the oxides and the oxide sulfides differ considerably (oxides, -53 to -35 ppm; oxide sulfides, 11to 25 ppm; reference, H3P04) and are most pronounced for P(V). Using 31PMAS-NMR investigations, it has been demonstrated that these effects mainly are derived from different crl values. Compared with the electron density at the P=S unit, that at the P=O unit seems to be more concentrated along the direction of the bond axis, leading to a higher component in the shielding tensor in this direction. As shown by XPS and especially by XANES spectra, the core binding energies of the phosphorus atoms are significantly affected by the type and number of the terminally bonded atoms. The position of the white line on the energy scale is correlated with the electronegativities of the substituents. Thus the

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

383

white line of P4OI0is shifted by 1.7 eV to higher energies compared with that of P@&. Even the core electrons of the phosphorus(II1) atoms, which are not directly involved in these substitutions, show some sensitivity to the terminally bonded chalcogen atoms. Although the positions of the XANES resonances that are assigned to P(II1)1s --., a*[P(III)-01 transitions are not affected upon substitution of the terminally bonded oxygen atoms through sulfur atoms, their intensities differ significantly. V. Concluding Remarks

The preceding sections reveal the substantial gain that has been realized in our knowledge about phosphorus oxides and oxide sulfides. In striking contrast to earlier findings, each individual member of this family of compounds that is notoriously sensitive to moisture has been synthesized, or seems t o be accessible, as a pure solid. Most of the compounds have been studied thoroughly with respect to their molecular and crystal structures, giving a sound basis for systematic comparisons. The first attempts to correlate geometric features of the molecules and their electronic structures have been undertaken. However, further experimental input, as well as thorough quantum mechanical descriptions including core states, is needed. Besides filling in the missing links, it would be desirable to consider in addition the phosphorus oxide selenides. With their inclusion, the tendencies already detected in the oxides and oxide sulfides should become much more pronounced. Furthermore, one would introduce a n additional probe atom (77Se)for NMR, a powerful analytical technique in this context. A preparative chemistry based on phosphorus oxides or oxide sulfides as educts has not been developed. Especially for the binary oxides, a high potential-both in solid state and in molecular chemistry-is obvious in this regard. Knowing that demands to avoid intermediates containing chlorine are arising, even a broad application of phosphorus oxides in technical chemistry becomes imagineable. For this purpose the yields in producing P40sand especially the phosphorus(I1W) oxides have to be improved considerably, which is a challenge even at the present state of knowledge. ACKNOWLEDGMENTS Financial support by the DFG (SFB 334 and Gottfried Wilhelm Leibniz-Programm) and the Fonds der Chemischen Industrie is gratefully acknowledged.

384

CLADE, FRICK, AND JANSEN

REFERENCES 1 . Berzelius, J. J., Gilbert’s Ann. 53, 393 (1816). 2 . Thorpe, T. E., and Tutton, A. E., J . Chem. SOC.49, 883 (1886). 3. Thorpe, T.E., and Tutton, A. E., J . Chem. SOC.59, 1019 (1891). 4. Thorpe, T. E., and Tutton, A. E., 2. Anorg. Chem. 1, 5 (1891). 5. West, C. A., J . Chem. SOC.81,923 (1902). 6. Jost, K.-H., A& Cryst. 17,1593 (1964). 7. Jost, K.-H., Acta Cryst. 21,34 (1966). 8. Meisel, M., 2. Chem. 23, 117 (1983). 9 . Boyle, R.,Phil. Trans. 18,478 (1694). 10. Davy, H., Phil. Trans. 102, 405 (1821). 1 1 . Dulong, P. L.,Phil. Mug. 48,271 (1816). 12. Berzelius, J. J., in “Lehrbuch der Chemie,” Vol. 11, p. 499 (1825). 13. Delalande, Z.,Ann. Chim. Phys. 1(3), 117 (1841). 14. Marchand, R. F., J . Prukt. Chem. 16(1),373 (1839). 15. Grabovsky, A. G., Sitzungsber. Akad. Wien 52, 170 (1865). 16. Goldschmidt, T,.D.R.P. No. 110,174(1899). 17. Pistor, G., D.R.P. No.426,388(1924). 18. Finch, G. I., Peto, R. H. K., J . Chem. SOC.121,692 (1922). 19. Whitaker, H., J . Chem. SOC.127, 2219 (1925). 20. Threlfall, R.,Phil. Mag. 35(5),1 (1893). 21. Shenstone, W. A., Beck, C. R., J . Chem. SOC.63,475 (1893). 22. Kempf, R.,J . Prakt. Chem. 78(2),201 (1908). 23. “Ullmann’s Enzyklopadie der Technischen Chemie,” Vol. XVIII, p. 303f.Verlag Chemie, Weinheim, 1979. 24. Davy, H., thesis, London, 1800. 25. Dalton, J., Ann. Phil. 9,186 (1807). 26. Geuther, A., and Michaelis, A.,Chem. Ber. 4, 766 (1871). 27. Oddo, G., Atti Accad. Linc. 10(5),452 (1901). 28. Baumgarten, P., and Brandenburg, C., Chem. Ber. 72,555 (1939). 29. le Sage, B. G., Mkm. Acud., 321 (1777). 30. Lavoisier, A. L., MPm. Acad., 65 (1777). 31. Steinacher, P. A., Gehlen’s J . 1, 681 (1803). 32. Reinitzer, B., Chem. Ber. 14, 1884 (1881). 33. Emelbus, H.J., J . Chem. SOC.127, 1362 (1925). 34. Miller, C. C., J . Chem. SOC.,1847 (1928). 35. Miller, C. C., J . Chem. SOC.,1823 (1929). 36. Heinz, D., 2. Anorg. Allg. Chem. 347,167 (1966). 37. Walker, M. L., and Mills, J. L., Syn. Reactivity Znorg. Metal-Org. Chem. 5,29(1975). 38. Wolf, L.,and Schmager, H.,Chem. Ber. 62,771 (1929). 39. Wolf, L., D.R.P. No. 444,664(1927). 40. Heinz, D., and Thilo, E., East German Patent No. 26,660(1960). 41. Thilo, E., and Heinz, D., German Patent No. 1,172,241(1964). 42. Heinz, D., Asijew, R. G., Muller, I. F., and Kauschka, G., Pure Appl. Chem. 52,

825 (1980). 43. Whyte, D. D., Pflaumer, P. F., and Roberts, T. S., French Patent No. 1,531,913

(1968). 44. Heinz, D., Neumann, P., and Crahmer, H., East German Patent No. 1,956,999

(1970).

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

385

Heinz, D., Crahmer, H., and Neumann, P., German Patent No. 2,028,239(1971). Krause, H. H., McNulty, J. S., and Hall, R. E., U.S.P. No. 3,652,213(1972). Mezey, E. J., and Hall, R. E., U.S.P. No. 3,652,211(1972). Hertzog, K., East German Patent No. 139,426(1980). Mustroph, G., Kurze, R., Ober, D., Schneider, H., Hofmann, F.-R., Busch, D., Rothe, H., and Kuhnert, E., German Patent No. 288,139(1991). 50. Hartmann, K., Hebecker, D., Schneider, H., Kurze, R., Busch, D., Mustroph, G., and Otto, J., German Patent No. 292,367(1991). 51. Otto, J., Schneider, H., Hartmann, K., Hebecker, D., and Kurze, R., German Patent No. 292,636(1991). 52. Nacquet, A., in “Grundzuge der modernen Chemie,” p. 218. 53. Gautier, A,, Compt. Rend. 76,49,173(1873). 54. Besson, A., Compt. Rend. 124, 763 (1897). 55. Besson, A,, Compt. Rend. 125, 1032 (1897). 56. Chalk, L. J., and Partingon, J. R., J . Chem. SOC.,1930 (1927). 57. Krafft, F., and Neumann, R.,Chem. Ber. 34, 565 (1901). 58. Wolf, L., Kalaehne, E., and Schmager, H., Chem. Ber. 62, 1441 (1929). 59. Jorissen, W. P., and Tasman, A., Rec. Trav. Chim. Pays-Bas 48,324 (1929). 60. Hossenlopp, F., McPartlin, M., and Ebel, P. J.,Bull. SOC.Chirn, 791 (1960). 61. v. Druten, A., Rec. Tmv.Chim. Pays-Bas 48,312 (1929). 62. BBchamp, A., Compt. Rend. 40, 944 (1855). 63. Anderson, H. H., J . Org. Chem. 19,1766 (1954). 64. Jander, G., Wendt, H., and Hecht, H., Chem. Ber. 77, 698 (1944). 65. Hautefeuille, P., and Perrey, A., Compt. Rend. 99,33 (1884). 66. Russell, E. J., J . Chem. SOC.83, 1263 (1903). 67. Miller, C. C., Proc. R . SOC.Edinburgh 46, 239 (1926). 68. Britzke, E., and Pestow, N., Tr. Nauchn. Znst. Udobr. 69,5 (1929). 69. Emmett, P. H., and Schultz, J. F., Znd. Engr. Chern. 31, 105 (1939). 70. Brunauer, S.,and Schultz, J. F., Znd. Engr. Chem. 33, 828 (1941). 71. Thilo, E.,Heinz, D., and Jost, K.-H., Angew. Chern. 76,229 (1964). 72. Heinz, D.,Z . Anorg. Allg. Chem. 336, 137 (1965). 73. Heinz, D., Rienitz, H., and Radeck, D., Z. Anorg. Allg. Chern. 383, 120 (1971). 74. Ewert, G., Forster, I., Heinz, D., Kauschka, G., and Kurze, R., German Patent No. 294,465(1991). 75. Loeper, M., and Schulke, U., Z. Anorg. Allg. Chem. 500, 40 (1983). 76. Walker, M. L., and Mills, J. L., Znorg. Chem. 16, 3033 (1977). 77. Walker, M. L., Peckenpaugh, D. E., and Mills, J. L., Znorg. Chem. 18,2792 (1979). 78. Jansen, M., and Voss, M., Angew. Chem. 93, 120 (1981). 79. Jansen, M., and Moebs, M., Z. Anorg. Allg. Chem. 514, 39 (1984). 80. Frick, F., thesis, Bonn, 1993. 81, Jansen, M., and Moebs, M., unpublished results. 82. Liier, B., thesis, Bonn, 1991. 83. Jansen, M., and Liier, B., Z. Krist. 197,247(1991). 84. Harnisch, H., Z . Anorg. Allg. Chem. 300, 261 (1959). 85. Clade, J., and Jansen, M., unpublished results. 86. Thorpe, T. E., and Tutton, A. E., J . Chern. SOC.57,545 (1890). 87. Tilden, A., and Barnett, R. E., J. Chem. Soc. 69, 154 (1896). 88. le Bas, G., Chem. News 111, 113 (1915). 89. Maxwell, L. R., Hendricks, S. B., and Deming, L. S., J . Chem. Phys. 5,626 (1937). 90. Hampson, G. C., and Stosick, A. J., J . A m . Chem. Soc. 60,1814 (1938).

45. 46. 47. 48. 49.

386

CLADE, FRICK, AND JANSEN

91. Beagley, B., Cruickshank, D. W. J., Hewitt, T. G., and Jost, K.-H., Tmns. Furuduy. SOC.65,1219 (1969). 92. Akishin, P. A., Rambidin, N. G., and Zasorin, E. Z., Kristullogrufiyu 4,360 (1959); Sou. Phys. Cryst. 4,334 (1959). 93. Beagley, B., Cruickshank, D. W. J., Hewitt, T. G., and Haaland, A., Trans. Furuduy SOC.63, 836 (1967). 94. Thilo, E.,and Wieker, W., Z. Anorg. Allg. Chem. 277, 29 (1954). 95. de Decker, H. C. J., and McGillavry, C. H.,Rec. Truu. Chim.Puys-Bus60,153(1941). 96. de Decker, H. C. J., Rec. Truu. Chim. Pays-Bas 60,413 (1941). 97. McGillavry, C. H., de Decker, H. C. J., and Nijland, L. M., Nature 164, 448 11949). 98. Cruickshank, C. D. J., Actu Cryst. 17,677 (1964). 99. Jansen, M., and Luer, B., Z. Krist. 177, 149 (1986). 100. Jansen, M., Voss, M., and Deiseroth, H.J., Angew. Chem. 93, 1023 (1981). 101. Jansen, M., and Moebs, M., Inorg. Chem. 23,4486 (1984). 102. Jansen, M., and Wipperfurth, P., unpublished results. 103. Jost, K.-H., and Schneider, M., Actu Ciystrullogr. B 37,222 (1981). 104. Gillespie, R. J., in “Molekulgeometrie, Elektronenpaar-Abstoljungund molekulare Struktur.” Verlag Chemie, Weinheim, 1975. 105. Jansen, M., and Moebs, M., Z. Krist. 159,283 (1982). 106. Kitaigorodski, A. I., in “Molekulkristalle.” Akademie Verlag, Berlin, 1979. 107. Gerding, H., v. Brederode, H., and de Decker, H. C. J., Rec. Truu. Chim. Pays-Bas 62,549 (1942). 108. Gerding, H., and de Decker, H. C. J.,Rec. Truu. Chim. Pays-Bus 64,191 (1945). 109. v. Brederode, H., and Gerding, H., Rec. Truu. Chim. Pays-Bas 67,677 (1948). 110. Gerding, H.,J . Chim. Phys. 46, 118 (1948). 111. Sidorov, T.A., and Sobolev, N. N., Opt. Spektr. 2, 710,717 (1957). 112. Sidorov, T.A,, and Sobolev, N. N., Fiz. Sborn. 3, 148 (1957). 113. Zijp, D.H., Adu. Mol. Spectrosc. 1-3, 345 (1962). 114. Chapman, A. C., Spectrochim. Actu A 24, 1687 (1968). 115. Beattie, I. R., Livingston, K. M. S., Ozin, G. A., and Reynolds, D. J., J . Chem. SOC. A , 449 (1970). 116. Konings, R. J. M., Cordfunke, E. H. P., and Booij, A. S., J. Mol. Spectroscopy 152, 29 (1992). 117. Mielke, Z., and Andrews, L., J . Phys. Chem. 93, 2971 (1989). 118. Mielke, Z., and Andrews, L., Inorg. Chem. 29, 2773 (1990). 119. McCluskey, M., and Andrews, L., J . Phys. Chem. 95,3545 (1991). 120. Muldagaliev, Kh. Kh., Mirgorodskii, A. P., and Lazarev, A. N., Neorg. Muter. 8, 81 (1972). 121. Jansen, M., Clade, J., and Frick, F., unpublished results. 122. Chapman, A. C., Homer, J., Mowthorpe, D. J., and Jones, R. T,. Chem. Commun., 121 (1965). 123. Cruickshank, D. W. J., J. Chem. Soc., 5486 (1961). 124. Goubeau, J., Pure Appl. Chem. 44,393 (1975). 125. Walker, M. L., and Mills, J. L., Znorg. Chem. 14,2438 (1975). 126. Grimmer, A.-R., and Wolf, G.-U., Eur. J. Solid StuteZnorg. Chem. 28, 221 (1991). 127. Grimmer, A,-R., and Muller, D., 2. Chem. 16, 496 (1976). 128. Grimmer, A.-R., 2. Chem. 18,233 (1978). 129. Grimmer, A.-R., Resonance Relut. Phen. 20,483 (1979). 130. Schindler, M., and Kutzelnigg, W., J. Chem. Phys. 76, 1919 (1982). 131. Mowthorpe, D. J., and Chapman, A. C., Mol. Phys. 15,429 (1968).

MOLECULAR PHOSPHORUS OXIDES AND OXIDE SULFIDES

387

132. Aksnes, D. W., Acta Chem. Scand. 23, 1078 (1969). 133. Herzfeld, J., and Berger, A., J . Chem. Phys. 73,6021 (1980). 134. Redeker, A., Kuper, G., Hormes, J., Frick, F., Jansen, M., and Muhlhauser, M.,

Phosphorus, Sulfur Silicon 76,239 (1993). 135. Kuper, G., Chauvistr6, R., Hormes, J., Frick, F., Jansen, M. Luer, B., andHartmann,

E.,Chem. Phys. 165,405(1992). 136. Sodhi, R. N., and Cavell, R. G., J . Electron Spectrosc. Relat. Phen. 32,283 (1983). 137. Dake, C. S . , Baer, D. R., and Friedrich, D. M., J. Vacuum Sci. Technol. A 7,

1634 (1989). 138. Franke, R., Chass6, T., Al-Araj, B., Streubel, P., and Meisel, A., Phys. Stat. Solid.

B 160, 143 (1990). 139. Franke, V., Chass6, T., Streubel, P., and Meisel, A., J . Electron Spectrosc. Relat.

Phen. 56, 381 (1991). 140. Hormes, J., and Redeker, A., unpublished results. 141. Egdell, R. G., Palmer M. L., and Findlay, R. H., Znorg. Chem. 19, 1314 (1980). 142. Brundle, C. R., Kuebler, N. A., Robin, M. B., and Basch, H., Znorg. Chem. 11,

20 (1972). 143. Evans, S.,Joachim, P. J., Orchard, A. F., and Turner, D. W., Znt. J.Mass Spectrosc.

Zon Phys. 9, 41 (1972). 144. Tahri, Y., and Chermette, H., J . Electron Spectrosc. 56, 51 (1991). 145. Makarov, L. L., Zaitsev, Yu. M., and Batrakov, Yu. F., Theor.Eksperim Khim. 12,

78 (1976). 146. Makarov, L. L., Zaitsev, Yu. M., and Batrakov, Yu. F., Zzv. Akad. Nauk SSSR,

Ser. Fiz. 40,399 (1976). 147. Fluck, E., and Weber, D., 2. Naturf. B 29,603 (1974). 148. Lee, T. H., Jolly, W. L., Bakke, A. A., Weiss, R., and Verkade, J. G., J . A m . Chem. SOC.102,2633 (1980). 149. Rose, J. L., VanCott, T. C., Schatz, P. N., Boyle, M. E., and Palmer, M. H., J . Phys.

Chem. 93,3504 (1989). 150. Bak, B., Kristiansen, N. A., and Svanholt, H., J . Mol. Struct. 78, 63 (1982). 151. Hashizume, A., Wasada, N., and Tsuchiya, T., Bull. Soc. Chem. Jpn. 39, 150

(1966). 152. Daasch, L.W., Weber, J. N., Ebner, M. A., and Sparrow, G., J. Mass Spectrosc. Zon

Phys. 2,503 (1969). 153. Muenow, D. W., Uy, 0. M., and Margrave, J. L., J. Znorg. Nucl. Chem. 32, 3459

(1970). Southard, J. C., and Nelson, R. A., J . A m . Chem. SOC.69, 911 (1937). Hill, W.L., Faust, G. T., and Hendricks, S. B., J.Am. Chem. Soc. 65,794 (1943). Smits, A.,2. Phys. Chem. B . 46, 43 (1940). Smits, A.; and Rutgers, A. J., J . Chem. Soc. 125,2573 (1924). Frandsen, M., Bur.Standard J . Res. 10,54 (1933). Hoeflake, J. M. A., and Scheffer, F. E. C., Rec. Trav. Chirn. Pays-Bas 45,191 (1926). Hartley, S. B., and McCoubry, J. C., Nature 198,476(1963). Koerner, W.B., and Daniels, E., J . Chem. Phys. 20, 113 (1952). Frick, F., Hoffbauer, W., Grimmer, A. R., Jansen, M., Fleischer, U., and Kutzelnigg, W., Naturwiss. 80, 465 (1993). 163. McAloon, B. J., and Perkins, P. G., Theor. Chim. Acta 24, 102 (1972). 164. Kutzelnigg, W., Angew. Chem. 96,262 (1984). 165. Muhlhauser, M., Engels, B., Marian, C. M., Peyerimhoff, S. D., Bruna, P. J., and Jansen, M., Angew. Chem., 106,578 (1994). 154. 155. 156. 157. 158. 159. 160. 161. 162.

388

CLADE, FRICK, AND JANSEN

166. Frick, F., Jansen, M., Bruna, P. J., and Peyerimhoff, S. D., Chem. Ber. 124,1711

(1991). 167. Besson, A,, Compt. Rend. 124, 151 (1897). 168. Abel, E. W., Armitage, D. A., and Bush, R. P., J. Chem. SOC.Suppl. 1,5584 (1964). 169. Baeker, M., Heinz, D., Kurze, R., and Radeck, D., East German patent No.

101,368(1973). Walker, M. L., thesis, Texas Technical University, 1976. Pernert, J. C., and Brown, I. H., U.S.P. No. 2,577,207(1951). Pernert, J. C., Chem. Eng. News 27,2143 (1949). Wolf, G.-U., and Meisel, M., 2. Anorg. Allg. Chem. 509,101 (1984). Stock, A.,and Friederici, K., Chem. Ber. 46, 1380 (1913). Frick, F.,and Jansen, M., 2. Anorg. Allg. Chem. 619,281 (1993). Stosick, A. J., J.A m . Chem. SOC.61,1130 (1939). Mijlhoff, F. C., Portheine, J., and Romers, C., Rec. Trav. Chim. Pays-Bas 86,257 (1967). 178. Wolf, G.-U., and Meisel, M., 2. Chem. 20, 451 (1980). 179. Palkina, K. K., Maksimova, S. I., and Wolf, G.-U., Neorg. Mat. 16, 1466 (1980). 180. Gerding, H.,and v. Brederode, H., Rec. Truu. Chim. Pays-Bas 64, 183 (1945). 181. van Wazer, J. R., Callis, C. F., Shoolery, J. N., and Jones, R. C., J.A m . Chem. SOC. 78,5715 (1956). 182. Grunze, H., Pure Appl. Chem. 52,799 (1980). 183. B. M.Gimarc, and J. J. Ott, J. Am. Chem. SOC.108,4298(1986). 184. Gimarc, B. M., J. Am. Chem. SOC.105, 1979 (1983). 185. Engels, B., Muhlhauser, M., Peyerimhoff, S. D., and Jansen, M., unpublished results. 186. Pantelouris, A., Hormes, J. Giinther, C., Hartmann, E., Frick, F., and Jansen, M., Chem. Phys., in press. 170. 171. 172. 173. 174. 175. 176. 177.

ADVANCES IN INORGANIC CHEMISTRY, VOL. 41

STRUCTURE AND REACTIVITY OF TRANSFERRINS E. N. BAKER Department of Chemistry and Biochemistry, Massey University, Palrnerston North, New Zealand

I. Introduction 11. Biological Roles 111. Transferrin Structure A. Primary Structure B. Three-Dimensional Structure C. Variations among Transferrins D. Similarities with Bacterial Binding Proteins IV. Properties of the Metal and Anion Sites A. Spectroscopic Monitors of Metal Binding B. Metal Substitution and Spectroscopy C. Anion Binding D. Differences between the Two Sites V. Mechanisms of Binding and Release A. Uptake of Iron B. Release of Iron VI. Recombinant DNA Studies VII. Concluding Remarks References

1. Introduction

When ferric iron (preferably as ferric nitrilotriacetate) is added to egg white, with a little salt, and the mixture is shaken, a vivid red color develops. This simple experiment introducesthe protein ovotransferrin (formerly known as conalbumin) and the most striking property of transferrins, their ability to rapidly and tightly bind iron. The transferrins are a family of metal-binding proteins (Table I), which apparently owe their place in biology t o the need to deal with the solution properties of iron. Under physiological conditions, the favored state of iron is Fe(III),but this is prone to rapid hydrolysis at concentraM,ultimately leading to the precipitation of tions greater than 389 Copyright 8 1994 by Academic Press, Inc. All rights of repduction in any form reserved.

390

E. N. BAKER

TABLE I PROTEINS OF THE TUNSFERRINFAMILY

Protein Serum transferrin

Source Blood

Ovotransferrin Avian egg white (conalbumin) Lactoferrin Milk, tears, saliva, and (lactotransferrin) other secretions White blood cells Melanotransferrin Melanoma cell surface (p97)

Function Iron transport Protection (antibacterial, antioxidant) Growth factor? Protection (antibacterial, antioxidant) Protection (antibacterial, antioxidant) Iron absorption? Growth factor? Iron translocation?

insoluble ferric hydroxides (I 1. Bacteria have evolved iron chelation systems based on low molecular weight chelate compoundscalled siderophores (2);animals instead use proteins of the transferrin family for the solubilization, sequestration, and transport of ferric iron. The transferrin family is typified by serum transferrin, the iron transport protein in blood from which the name is taken (literally “transport of iron”). Serum transferrins appear to be present in all vertebrates so far examined; they have been found in crabs and spiders (3) and appear likely to be widespread also in insects ( 4 ) . A second member of the family, lactoferrin, is widespread in the secretory fluids of higher animals, including milk, tears, saliva, mucosal, and genital secretions (5), and in white blood cells (6);in species that possess both proteins, lactoferrin is distinguishable from serum transferrin by its higher isoelectric point (8-9 for lactoferrin compared with 5-6 for transferrin), its distinct sequence, and different bodily location and functional roles. The protein of avian egg white, referred to as ovotransferrin above, is actually a post-translationally modified form of the avian serum transferrin, differing only in its attached carbohydrate (7). Finally, the most recently recognized member of the transferrin family is a membrane-bound protein, which is present at low levels on the surface of normal cells, but becomes expressed at high levels on melanoma cells (8).This protein, at first called p97 from its apparent molecular weight (-97,0001, is referred to here as melanotransferrin. The transferrins are typically monomeric glycoproteins with a single polypeptide chain of 670-700 amino acids and a molecular weight of -80,000. Their characteristic property is to bind, very tightly but reversibly, two Fe3+ions together with two C0;- ions. The relationship

TRANSFERRIN STRUCTURE AND REACTIVITY

391

between the metal ion and the anion, first recognized in the landmark paper of Schade and coworkers (91, is synergistic in the sense that neither is bound strongly in the absence of the other. The proteins are also characterized by an internal duplication in which their N-terminal and C-terminal halves are homologous (10,11), with each carrying a single iron binding site. Within this general framework, there are variations. Melanotransferrin and the transferrin from the sphinx moth, Munduca sextu, each bind only one iron atom (12, 13) even though they are the same size as the other transferrins, and there have been reports of both smaller [Pyura stolonifera, M, -40,000 (14)land larger [crab, Cancer rnugister, M, -150,000, (15)ltransferrins; it is likely that more variations will be discovered. There have been numerous reviews of transferrin chemistry. Earlier reviews by Aisen and Listowsky ( 1 ) and Chasteen (16) have been followed by extensive recent articles that have particularly emphasized the functional properties (31, the physical biochemistry (171,and the structures (18)of transferrins. Here we will concentrate on the bioinorganic chemistry of these intriguing proteins, with a particular emphasis on their binding properties, as shown by chemical and spectroscopic studies, but all set firmly in a structural perspective. II. Biological Roles

The fundamental role of transferrins, shared by all (except,perhaps, melanotransferrin), is to control the levels of free iron in body fluids by binding, sequestering and transporting Fe3+ ions. This helps to maintain the availability of iron while preventing the deposition of insoluble ferric hydroxide aggregates. It also has two further effects. First, the proteins thus protect against the toxic effects of free iron that might otherwise catalyze the formation of the free radicals that damage cells (19). Second, they have bacteriostatic properties that probably are derived from the fact that in viuo the proteins are largely present in their apo- (iron-free) forms and are then able to bind iron so tightly that it is unavailable for bacterial growth. Breast-fed infants, for example, are protected from stomach upsets by the presence, in human milk, of lactoferrin, which inhibits bacterial growth. Similarly the antibacterial properties of egg white derive from the presence of ovotransferrin. In addition to these general roles, certain of the transferrins have more specific functions. Serum transferrin has the role of transporting iron from sites of absorption to sites of storage and utilization. There

392

E. N. BAKER

the iron-loaded transferrin binds to specific cell receptor molecules, is internalized, releases its iron inside the cell, and is then returned to the cell surface, from where it detaches and reenters circulation as the apoprotein (20). This process is shown diagrammatically in Fig. 1. Whether lactoferrin also has a role in iron absorption is not clear, but the presence of lactoferrin receptors in the gut and elsewhere (21, 22), together with the high bioavailability of the iron in human milk, suggests that it could have. It must be stressed, however, that the iron release properties of lactoferrin are quite different from those of serum transferrin (see Section VI), and their respective receptors are also different. Other metals may also be carried by transferrins, notably aluminium, implicated in Alzheimer’s disease; A13+ is bound by transferrins (23),suggesting a possible therapeutic use ( 2 4 ) .All of the manganese in human milk is reportedly bound to lactoferrin (25),and it is highly likely that levels of other metal ions in biological fluids may be similarly controlled. The fact that the conformation of lactoferrin is the same whether Fe3+ or Cu2+is bound (26; see also Section 1V.B) suggests that transferrins carrying copper, and probably other metals, should bind equally well to receptors as those carrying iron. Moreover] the low degree of iron saturation of transferrins in uiuo (-30% for serum transferrin, -10%for lactoferrin) implies a potential capacity for binding other metal ions.

FIG.1. Schematic representation of the cycle of iron delivery to cells by transferrin, showing the uptake of diferric transferrin by cell receptors, internalization, release of iron at the lower intracellular pH, and recycling and release of apotransferrin.

TRANSFERRIN STRUCTURE AND REACTIVITY

393

The above roles all depend on the ability of transferrins to bind metal ions reversibly. The other element in their biological functions is their ability to bind to cells. Both transferrin and lactoferrin have potent growth factor activity (27,281, which depends on binding to cellular receptors (28, 29). Lactoferrin’s ability to bind to a wide variety of cells ( 3 0 , 3 1 )has also brought suggestions of a role in modulating the immune and inflammatory responses (311. The bacteriostatic effects of lactoferrin and ovotransferrin, arising from iron deprivation, also appear to be supplemented by a bactericidal activity that results from direct contact with the bacteria ( 3 2 ) .In the case of lactoferrin, a region of the molecular surface that probably attacks bacterial cell walls has been identified through isolation and characterization of a bactericidal peptide ( 3 3 ) . Finally, the one membrane-bound transferrin, melanotransferrin, although its function has not yet been established, may aid the rapid proliferation of tumor cells, perhaps through an ability to bind and translocate iron. Ill. Transferrin Structure

A. PRIMARY STRUCTURE The large size of transferrins (670-700 residues), with the consequent difficulties of chemical sequencing, meant that it was not until 1982 that the first amino acid sequences, those of human serum transferrin (10)and chicken ovotransferrin (34,351,were established. These were closely followed by that of human lactoferrin (11). The twofold internal repeat in each sequence (see below) was immediately apparent, and comparison of all three sequences then identified conserved tyrosines and histidines that were potential ligands for iron (11). Since that time many more sequences have become available through the advent of recombinant DNA technology and the deduction of amino acid sequences from the base sequences of cloned DNA. At the present time, the primary structures (amino acid sequences) of 14 proteins of the transferrin family have been established. These include seven serum transferrins, from human (10, 36), pig (371, horse (381, rabbit (391, toad (Xenopus Zaeuis) ( 4 0 ) ,sphinx moth (M. sextu) (131, and cockroach (Bluberus discoidulis ( 4 ) ;chicken (34,351 and duck ( 4 1 ) ovotransferrins; four lactoferrins, from human (11,421,mouse (431, pig ( 4 4 )and cattle (45,46);and the human tumor cell melanotransferrin ( 4 7 ) .All of these sequences are available from sequence databases such as EMBL and SWISSPROT.

394

E.N. BAKER

Levels of sequence identity between the proteins are generally high. For example, lactoferrins, which have so far been found only in mammals and are probably a more recent evolutionary development, have 60-70% sequence identity, as shown in Table 11. A similar level of identity is found between the serum transferrins of higher animals and there is still 50-60% identity between these transferrins and lactoferrins. Thus these transferrins of the higher animals form a highly conserved family, likely to have very similar three-dimensional structures. The evolutionarily more distant insect transferrins and the membrane-associated melanotransferrin have diverged considerably further, but their level of identity with the higher transferrins (20-30% for the insect proteins and -40% for melanotransferrin) is still sufficient t o imply that they share the same three-dimensional structure. Of fundamental significance to understanding transferrin structure and function is the two-fold internal amino acid sequence repeat. In each protein, the N-terminal half of the polypeptide is homologous with the C-terminal half, with the level of identity between the two halves ranging from 26-28% in the insect proteins to -40% in higher transferrins and as high as 46% in melanotransferrin. This repeat is expressed TABLE I1 SEQUENCE IDENTITY (%) IN TRANSFERRIN FAMILY" Protein Protein hTf rTf PTf eTf hLf bLf mLf PLf cOTf XTf cTf MsTf MTf

hTf rTf pTf eTf hLf bLf mLf pLf cOTf XTf cTf MsTf MTf 78 78 71 72 61 61 57 61 53 45 32 29 43

72 72 62 61 57 61 52 46 31 27 42

71 72 73 61 61 58 61 53 46 31 27 41

72 72 73 62 62 56 61 53 47 32 27 42

61 62 61 62 69 70 70 52 45 32 27 41

61 61 61 62 69 63 73 53 44 30 28 41

57 57 58 56 70 63 64 49 46 32 29 40

61 61 61 61 70 73 64 51 44 32 28 42

53 52 53 53 52 53 49 51 46 31 25 40

45 46 46 47 45 44 46 44 46 33 23 38

32 31 31 32 32 30 32 32 31 33 45 30

29 27 27 27 27 28 29 28 25 23 45

43 42 41 42 41 41 40 42 40 38 30 27

27

" hTf, human transferrin; rTf, rabbit transferrin; pTf, porcine transferrin; eTf, equine transferrin; hLf, human lactoferrin; bLf, bovine lactoferrin; mLf, mouse lactoferrin; pLf, porcine lactoferrin; cOTf, chicken ovotransferrin; XTf, Xenopus transferrin; cTf, cockroach transferrin; MsTf, M.sexta transferrin; MTf, human melanotransferrin.

395

TRANSFERRIN STRUCTURE AND REACTIVITY

also in the three-dimensional structures of the proteins, which are bilobal, with a binding site in each lobe (Section III.B), and marks the transferrins as a classic example of proteins that have evolved via a gene duplication event. It is assumed that the gene for an ancestral 40-kDa protein with a single binding site has been doubled to give a bilobal, two-sited molecule of twice the size. A hypothetical scheme for transferrin evolution is given in Fig. 2. A number of consequences flow from the structural duplication in transferrins. First, the degree of similarity between the two binding sites assumes considerable importance, both for biological function and for interpreting chemical and spectroscopic studies. Second, it means that stable half-molecules, each with a single binding site, can be prepared, either by limited proteolysis or by recombinant DNA methods. Half-molecules offer the chance to examine the properties of singlesited molecules without the complications of seeing the “averaged” characteristics of the two sites. The most useful half-molecules are those prepared by recombinant DNA methods because the “cleavage” point is deliberately engineered. Two recombinant half-molecules have so far been prepared and characterized, the N-terminal half of human

Single-sited ancestor Gene duplication

Bacterial periplasrnic binding proteins?

transferrins Melanotransferrin

Lactoferrin

I FIG.2. Possible evolutionary development of the transferrin family, showing also the proposed relationship of bacterial periplasmic binding proteins (Section 1II.D).

396

E. N. BAKER

serum transferrin, comprising residues 1-337 (481,and the N-terminal half of human lactoferrin, comprising residues 1-333 (49);both these half-molecules have as their C-terminal residues amino acids from within the connecting peptide that links the two lobes. No C-terminal half-molecule has yet been prepared by these methods, however. Many proteolytic fragments of transferrins have been prepared over the years. They have played a significant part in analyzing the differences between the two sites, but they have the disadvantage that the cleavage sites are fortuitous in the sense that they depend on what enzyme is used and what sites happen to be offered by the protein structure; this is not usually predictable. The “best” half-molecule fragments are probably those prepared from chicken ovotransferrin, for which cleavage occurs in the connecting peptide to give stable N-terminal and C-terminal half-molecules (50-52). The use of immobilized subtilisin offers a particularly good preparative route (52). These halfmolecules bind iron reversibly and also reassociate strongly (53, 5 4 ) to give a noncovalent complex capable of delivering iron to cells via the transferrin receptor ( 5 5 ) .Half-molecule fragments of bovine serum transferrin with very similar, but not identical, spectroscopic properties have also been prepared (56);these and other fragments are well summarized in Ref. 17. The difficulties inherent in proteolytic fragments are illustrated by human lactoferrin, for which trypsin cleavage gives an N-terminal fragment of 30 kDa (residues 1-281) and a C-terminal fragment of 50 kDa (residues 282-691) (57).That is, cleavage occurs not in the connecting peptide but 50 amino acids earlier; the N-fragment thus lacks the last 50 amino acids of the N-lobe and its properties are likely to be different from those of the true half-molecules.

B. THREE-DIMENSIONAL STRUCTURE The first crystallographic studies on transferrins date back more than 20 years (581,and crystals of various transferrins have since been reported. These include the diferric forms of rabbit ( 5 9 ) and human ( 6 0 )serum transferrins, hen (61) and duck ( 6 2 )ovotransferrins, human ( 6 3 )and bovine ( 6 4 )lactoferrins, and the apo- (iron free) forms of human lactoferrin ( 6 5 )and duck ovotransferrin ( 6 2 ) .In spite of all this activity, the crystals in many cases have proved difficult to handle, and the X-ray analyses quite challenging. A low-resolution analysis of rabbit serum transferrin in 1979 demonstrated the bilobal nature of the molecule ( 6 6 ) ,but it was not until 1987, with the publication of the structure of human lactoferrin ( 6 7 ) ,that full details of a transferrin

TRANSFERRIN STRUCTURE AND REACTIVITY

397

structure emerged. The structure of rabbit serum transferrin soon followed (68),and it is likely that structure analyses of some of the other species will reach fruition over the next few years. Complementing the structural studies of the intact transferrins, a number of fragments have also been crystallized, including proteolytic N-terminal half-molecules of rabbit serum transferrin (69)and chicken ovotransferrin ( 70),recombinant N-terminal half-molecules of human lactoferrin (71)and human serum transferrin (721,and a quartermolecule fragment of duck ovotransferrin ( 73).All of these have now led to high-resolution structures ( 74-77). The most detailed description of a complete transferrin molecule is that of human lactoferrin, a t 2.8-A resolution (781,and most of the data in the following sections come from this work and from refinement of the same structure at 2.1-A resolution (79).As would be expected from the high level of sequence similarity, the three-dimensional structure of rabbit serum transferrin (681, although at lower resolution (3.3 A), is completely consistent with that of lactoferrin; the differences are at the level of individual amino acid changes, together with some differences in lobe and domain orientations. These are discussed below (Section III.B.l). Coordinates for human lactoferrin, in both diferric ( 78)and apo- (80) forms, for diferric rabbit serum transferrin (68), and for the three fragment structures, the proteolytic N-lobe of rabbit serum transferrin ( 741, the recombinant N-lobe of human lactoferrin ( 75)and the duck ovotransferrin quarter-molecule ( 761,all in their iron-bound forms, can be obtained from the Brookhaven Protein Data Bank (Brookhaven National Laboratory, Upton, New York). 1. General Organization All transferrins characterized so far consist of a single polypeptide chain of 670-700 amino acid residues. The lactoferrin and serum transferrin structure analyses show that the folding (polypeptide chain conformation) is the same in both proteins and, given their sequence homology, can be assumed to hold for all proteins of the transferrin family. The polypeptide chain is first of all folded into two globular lobes, representing the N-terminal and C-terminal halves of the molecule; these are referred to as the N-lobe and C-lobe, respectively. In human lactoferrin the N-lobe comprises residues 1-333 and the C-lobe, residues 345-691, whereas in rabbit serum transferrin the equivalent lobes comprise 1-328 and 342-676. Each lobe contains a single iron binding site, and each has essentially the same folding (described more fully in Section III.B.2). The two lobes do not have equivalent orientations

398

E. N. BAKER

with respect to the molecule as a whole, however, being arranged "frontto-back" as in Fig. 3. This is one of the factors which may contribute to inequivalence in their functional properties. The two lobes are joined by a short connecting peptide, which is the only covalent link between them. This peptide varies between different transferrins, both in length (7 t o 14 residues-see Section 111.0 and in conformation; in lactoferrin, it is 12 residues long and forms a threeturn helix, whereas in serum transferrin it is 14 residues long and has a much less regular structure. The relative orientations of the two lobes also appear to vary from one transferrin to another. In human lactoferrin the C-lobe can be superimposed on to the N-lobe by a twofold screw axis rotation, a rotation of 180" followed by a translation of 25 A along the rotation axis. In rabbit serum transferrin, however, the super osition requires a rotation of 167",followed by a translation of 24.5 ; i.e., there is a difference of -13" between the lobe orientations in lactoferrin and transferrin (81). Whether this has any functional significance remains to be seen, but we may anticipate more variations of this type between

x

FIG.3. Domain organization of transferrins. The N-terminal lobe (above) is divided into domains N1 and N2,and the C-terminal lobe (below), into domains C1 and C2. The two lobes are related by a screw axis, a rotation of -180", and a translation of -25 A. The two iron sites are identified with closed circles. The connecting peptide that joins the two lobes is helical in lactoferrin (solid line) and less regular in transferrin (dashed line).

TRANSFERRIN STRUCTURE AND REACTIVITY

399

different transferrins-in fact human and bovine lactoferrins have also been shown to differ in this respect (82). Within each lobe there is a further subdivision into two separately folded domains. This is shown schematically in Fig. 3 and can be clearly seen in the ribbon diagram of Fig. 4. The domains in the N-lobe of lactoferrin are labeled N1 (residues 1-90 and 252-333) and N2 (residues 91-251), with the equivalent C-lobe domains being C1 (residues 345-433 and 596-691) and C2 (residues 434-595). This subdivision into domains also has a crucial functional significance as the cleft separating the two domains of each lobe houses the metal binding site, and the domain structure has major implications for mechanisms of binding and release; this is explored further in Sections III.B.2-III.B.4. One feature of the domain structure is that the relative orientations of the two domains in each lobe vary slightly, despite the close structural correspondence. If the N2 and C2 domains of human lactoferrin are overlaid, the other domains do not quite match; the N 1 domain must be rotated a further 6” t o bring it into line with the C1 domain. The difference is that in lactoferrin the C-lobe domains are slightly more closed over their binding site than are the N-lobe domains. The

FIG.4. Ribbon diagram of human diferric lactoferrin, showing the organization of the molecule, with the N-lobe above and C-lobe below. The four domains (Nl, N2, C1, and the C-terminal helix (C) are indicated. C2), the interlobe connecting peptide (H), The glycosylation sites in various transferrins are shown by triangles and numbered (1, human transferrin; 2, rabbit transferrin; 3, human lactoferrin; 4, bovine lactoferrin; 5 chicken ovotransferrin). The interdomain “backbone” strands in each lobe can be seen behind the iron atoms. Adapted from Baker et al. (82), with permission.

400

E. N. BAKER

two lobes of rabbit serum transferrin also appear to differ in this respect, but this time the N-lobe seems more closed. Similar small differences of domain closure are found in bacterial binding proteins (Section 1II.D). A potentially confusing aspect of transferrin structure and function is that, prior to determination of the lactoferrin and transferrin structures, the N- and C-terminal halves of each molecule were often referred to as N- and C-domains. Given the clear subdivisions apparent in Fig. 4, however, the commonly accepted definition of protein domains (separately folded units; see Ref. 83), and the functional importance of these units, it seems preferable to think of the transferrins as proteins comprising two lobes, but four domains. A final point of general organization concerns the carbohydrate. All transferrins so far characterized, except apparently for one fish transferrin (841, are glycoproteins. There is, however, no pattern to the sites of attachment of the carbohydrate chains on different proteins-they appear almost randomly distributed over the protein surface (Fig. 4), strengthening the view that the carbohydrate plays no direct role in function. Rabbit serum transferrin, for example, has one carbohydrate chain, on its C-lobe (residue 490); human serum transferrin has two, both on the C-lobe (residues 416 and 611); human lactoferrin has two, one on each lobe (at residues 137 and 478); and bovine lactoferrin has four, one on the N-lobe (residue 233) and three on the C-lobe (residues 368, 476, and 545). 2 . Polypeptide Folding

As noted above, the two lobes have very similar folding. This is only to be expected given their high (-40%) sequence identity. The differences, at the level of polypeptide folding, are confined primarily to small insertions and deletions in the loops that join secondary structure elements. These are almost all located on the molecular surface and do not disturb the basic structure-indeed 90% of the main chain atoms of the N-lobe of human lactoferrin can be superimposed on equivalent atoms in the C-lobe with a root-mean-square deviation of only -1.2 A. The agreement would be even closer were it not for the small difference in the closure of the two domains, described above. The folding pattern for a typical transferrin lobe or half-molecule is shown schematically in Fig. 5. Each domain is based on a mixed psheet (i.e., a mixture of parallel and antiparallel p-strands) overlaid with a-helices, which pack against the two faces of the sheet. The first -70 residues make a coherent unit of three parallel p-strands (a, b, c) and three helices (1, 2, 31, which form about half of the first domain (N1 or Cl).A fourth parallel strand, made in two parts (d and e), then

TRANSFERRIN STRUCTURE AND REACTIVITY

401

FIG.5. Polypeptide folding pattern found in each lobe of human lactoferrin. Helices (cylinders,) are numbered 1 to 12 and p-strands (arrows) are labeled “a” to “ k as in Anderson et al. (78). The interdomain backbone strands are shaded and the position of the hinge is indicated.

runs behind the iron site to begin the folding of the second domain, crossing over at about residue 90.The next -160 residues form the whole of the second domain (N2 or C2), with five p-strands and a number of helices, before a second long extended strand (again in two parts, i and j ) crosses back behind the iron site to complete the folding of the first domain. Finally, the chain crosses for a third time, finishing with a helix (111 packed against the second domain. Three features of the folding pattern are of particular importance. First, the N-termini of many of the helices are directed toward the central binding cleft; the positive charge they carry should help attract anions into the binding cleft (see also Section V.A). Second, the two extended p-strands running behind the iron site and linking the two domains can be thought of as “backbone”strands (85).A hinge in these

402

E. N. BAKER

two backbone strands is of crucial importance to the conformational change that occurs during binding and release (80,82,86).Third, the iron atom takes its protein ligands from four different parts of the structure, widely spaced along the polypeptide chain-one from domain 1,one from domain 2, and the other two from the two backbone strands. The importance of this latter feature for the design of the metal site is outlined below (Section III.B.4). 3 . Disulfide Bonding Both lobes contain disulfide bonds, the total number varying from 15 in chicken transferrin to 22 in human serum transferrin. Although they clearly contribute to the high stability of transferrins, to heat and denaturants, and hence t o their relatively long lifetimes in circulation, it is not likely that they play any major role in defining the protein folding. Six disulfides [numbers 1-6 in the nomenclature of Williams (87)lare common to both lobes, in all transferrins, and can be regarded as the conserved “core” disulfides. Most are local in nature (Fig. 61, linking secondary structure elements, and none of them bridge between domains. Three more, numbers 7-9, are found in the C-lobes of nearly

Domain I

1

Domatn I I

FIG.6. Pattern of disulfide bridges commonly found in transferrins. Numbering corresponds to that introduced by Williams (87)and used in Table V. Figure taken from Bailey et al. (68),with permission.

TRANSFERRIN STRUCTURE AND REACTIVITY

403

all transferrins. These, in contrast, are long range (i.e., they join Cys residues that are far apart in the sequence) and probably make the C-lobe more rigid than the N-lobe. Number 7, which bridges between the C1 and C2 domains, may be of particular importance in modulating the iron binding and release properties of the C-lobe (80,85;see also Sections III.B.5 and V.B). Others are much more variable in their occurrence (Section 111. C)-some may influence the properties of different transferrins, but most of them probably just reflect natural variations in different species, with little functional effect. 4 . Metal and Anion Sites

Prior to first crystallographicanalyses, many physicochemical studies were directed toward establishing the nature, number, and arrangement of ligands at the iron site. These have been reviewed extensively ( 1 , 3 , 1 6 , 1 7 ) . Four types of ligand were anticipated by these studies, the side chains of the amino acids tyrosine and histidine, the synergistic anion (normally bicarbonate), and water (or OH- ion). The question was finally settled by X-ray crystallography, however, as there is a high degree of unanimity between the various structure analyses of diferric human lactoferrin (78) and rabbit transferrin (68) at resolutions of 2.1 and 3.3 8,respectively, and the iron-bound N-terminal halfmolecules of human lactoferrin (75) and rabbit transferrin (74) at resolutions of 2.0 and 2.3 A, respectively.

a. Iron Coordination. The archetypal transferrin iron binding site is shown schematically in Fig. 7, which is taken from the N-terminal lobe of human lactoferrin. The same organization,however, holds good for both the N-terminal and the C-terminal sites, in both lactoferrin and transferrin. The iron coordination is distorted octahedral, with ligands provided by four amino acid sidechains from the protein, 2 Tyr, 1 His, and 1 Asp, together with the synergisticallybound (bilcarbonateanion. Bond lengths are all around 2.0-2.2 8, (Fig. 81,consistent with EXAFS measurements, which indicate the coordination of six low-Z ligands (0or N)with bond distances of 1.9-2.1 A (88).The bond angles show larger deviations from ideality, largely because of the small chelate “bite” angle of the (bilcarbonate anion, of -65”. The involvement of tyrosine and histidine ligands had been expected from spectroscopy; the electronic spectral band at -465 nm, which is responsible for the orange-red color, is characteristic of iron-phenolate interaction (89, 901, and both EPR (91, 92) and NMR (93) studies indicated metal-histidine coordination.The presence of the Asp ligand

404

E. N. BAKER

" \ /

H

H

FIG.7. Schematic diagram of the characteristic transferrin metal and anion binding site. Numbering is as for the N-lobe of human lactoferrin, but the same arrangement of ligands is found in the C-lobe and in the N- and C-lobes of almost all transferrins (Table 111). For reference, the residue numbers for human lactoferrin and human transferrin are shown in the inset.

was not anticipated, because its spectroscopic silence makes it difficult to detect, but the carboxylate group is an excellent ligand for Fe(II1) and serves as such in many nonheme iron proteins (94). In the transferrins it appears to play a crucial role in the metal site, as it coordinates the metal through one carboxylate oxygen while hydrogen bonding between the two domains with its other oxygen (Fig. 9a). With the anion occupying the two remaining coordination positions, there is no room for water in the immediate coordination sphere. The EXAFS (95) and NMR proton relaxation (96) measurements, which suggested a bound water, must be explained by the presence of several water molecules close to the metal, but not directly coordinated. The two most likely candidates are water molecules hydrogen bonded to the first Tyr ligand and to the His ligand, 3.8 and 6.5 A, respectively, from the iron atom. These are present in both N- and C-sites in all the lactoferrin structures and in the half-molecule of rabbit transferrin. Another water molecule, attached to the active site Arg sidechain and 4.5 from the iron, has also been noted in transferrin ( 74). Although none of these water molecules are directly coordinated, they must contribute to the stability of the binding site, and changes in the water structure may be of importance for iron release.

TRANSFERRIN STRUCTURE AND REACTIVITY N-lohc

C-lobe

Ofio-Fe-092 060-Fe-0192 060-Fe-0I

XH Ihl

OY2-I:r-Ol

92 I s.5

OY2-Fe-02 0192-Fe-NZ.53 0192-Fe-02 N253-Fe-01 N253-Fe-02

405

X4

X4

86

167

0395-Fe-0435 0395-Fe-0528 0395-Fe-01

83 170 92

0435-Fe-01 0435-Fe-02

95 157

0528-Fe-N597 0528-Fe-02

93

N597-Fe-01 N597-Fc-02

106

91 172 1 IS

FIG.8. Bond lengths (A) and angles P) at the iron sites of human lactoferrin.

b THR 117 0,

FIG.9. Hydrogen bonding interactions in the transferrin binding site, with distances as in the N-lobe of human lactoferrin. (a) The Asp ligand, showing its position between two helix N-termini, and the important interdomain hydrogen bond to helix 5. (b) The carbonate ion, hydrogen bonded to the Arg sidechain and the helix 5 N-terminus.

406

E.N. BAKER

b. Anion Binding. The anion occupies a pocket on domain 2 (N2 or C2). The pocket is formed by two positively charged groups, the side chain of an arginine residue and the N-terminus of an a-helix, number 5 in the nomenclature used for both transferrin (68) and lactoferrin (78). Note that helices have a partial positive charge, estimated at 0.5+ to 0.75+, because of their dipolar nature (97).The anion fits beautifully between protein and metal (Fig. 9b), in such a way that the full bonding potential of each oxygen is expressed, either in metal coordination or in hydrogen bonding. The hydrogen bonds are close to optimal in their geometry, with lengths -2.8 A, angles at the hydrogens close to 180",and angles at the oxygens -120", emphasizing the nearperfect fit of the anion. The form of the anion, whether carbonate or bicarbonate, has never been certain. Considerations of the number of protons that are released upon metal binding have suggested bicarbonate (98),but NMR spectroscopic studies (99)favor carbonate. The anion site revealed crystallographically strongly suggests carbonate because of the hydrogen bonding pattern (Fig. 9), the charge of at least 1.6-t on the protein and the bidentate iron coordination ( 78).Therefore the description as carbonate is probably preferable. c. Design of the Metal Site. The design of the metal site is based upon two main requirements, that Fe3+iron should be tightly bound and that binding should be reversible. The first requirement is met by the choice of ligands. The protein provides three anionic oxygen ligands in the form of one carboxylate oxygen and two phenolate oxygens (the Tyr ligands are deprotonated when the metal binds-see Section IV.A), as well as the single neutral imidazole nitrogen. Two further oxygens are provided by the carbonate ion. This arrangement should be highly favorable for Fe3+,with its preference for anionic oxygen ligands, and this is reflected in the low redox potential of around -500 mV (100). The size of the metal binding site is such as to allow all metal-ligand bonds to be -2.0 A, as expected for Fe3+,and the charge also complements a tripositive metal ion perfectly. The charge on the synergistic anion (2-, assuming carbonate) is approximately matched by the positive charge (-1.6+)of its binding pocket, leaving the 3-charge of the protein ligands to be matched by the 3+ charge of the metal ion. The exclusion of water from the coordination sphere also helps to prevent hydrolysis. The requirement for reversibility of binding dictates the overall location and construction of the binding site. As noted before it lies in a deep cleft between two domains. The metal takes its ligands from four different parts of the protein structure, one (Asp) from domain 1, one

TRANSFERRIN STRUCTURE AND REACTIVITY

407

(the second Tyr) from domain 2, and two (His and the first Tyr) from the two polypeptide strands that cross over between the two domains at the back of the iron site (the backbone strands; Fig. 5). This means that when the domains are moved apart, to a more open configuration, through a hinge in the backbone strands, the metal site is necessarily pulled apart, leading to iron release. This is discussed further below (Section III.B.5). The synergistic anion has a particular importance in helping to create the metal site. In the absence of a suitable anion, the positive charge of the Arg side chain and helix 5 N-terminus must inhibit metal ions from binding in the specific site. This presumably accounts for the very weak, nonspecific binding found for metal ions in the absence of suitable anions (101). Once the anion is in place, however, not only is this positive charge neutralized but the anion then offers two ligands for metal coordination. The intimate burial of the anion between metal and protein also suggests it could have a role in iron release-protonation of the carbonate ion, for example, could disrupt the hydrogen bonding pattern and help in the breakup of the binding site (see also Section V.B). Perhaps the presence of this one non-protein ligand, which can be displaced, is an essential part of ensuring reversibility of metal binding. 5. Conformational Change

Transferrins have long been known, from biophysical studies (102, 1031,to undergo a substantial conformational change during iron binding and release. This takes the form that the protein becomes more compact when the metal is bound. Very important progress has recently been made in defining the nature of this conformational change and understanding its significance, primarily from crystallographic studies of apolactoferrin (80), but complemented by solution X-ray scattering measurements (104-106). The apolactoferrin structure analysis was based on protein from which the carbohydrate had been removed enzymatically; this deglycosylated protein had identical properties of iron binding and release and identical spectroscopic parameters, but gave crystals of the apoprotein, which diffracted to high resolution, 2.0 A (65). Two striking results came from the X-ray analysis (80).The N-lobe was shown to have undergone a very large conformational change, whereas the C-lobe remained essentially unchanged even though no metal was bound (Fig. 10). The conformational change in the N-lobe involves an opening of the binding cleft through relative movement of the two domains; the N2 domain rotates 54" relative to N1, about an axis passing through the

408

E. N. BAKER

N2

N1

FIG.10. Ca plot of the “one open, one closed” structure of human apolactoferrin( 8 0 ) . The open N-lobe is at left, the closed C-lobe, at right. The disulfide bridge 7 (residues 483-677), which may restrict the flexibility of the C-lobe and which has no counterpart in the N-lobe, is indicated at bottom right.

two backbone strands that run behind the iron site, connecting the two domains (Fig. 11). This is just as anticipated from the polypeptide folding (85).Analysis of the conformational change (86)shows that it is a rigid body movement in which neither domain shows any significant structural change and in which the hinge is localized to only a very small number of residues, 90-91 and 250-251. A further element is that a helix in the N2 domain (number 5) appears to pivot on another (number ll),which remains associated with the N1 domain (801,and there is a fine balance between the interfaces buried and exposed by this movement (86). The 54” domain movement is one of the largest seen in any protein. It has the effect of opening the binding cleft wide, thus exposing a number of residues previously buried, including several basic sidechains. The position of the hinge neatly splits the two ligands provided by the backbone strands, Tyr 92 and His 253, so that in the “open” form Tyr 92 remains with domain N2 (and Tyr 192), whereas His remains with domain N1 (and the Asp ligand). This division of the ligands has important implications for mechanisms of binding and release (Section V).

TRANSFERRIN STRUCTURE AND REACTIVITY

409

FIG.11. Schematic diagrams of the “open”(left) and “closed”(right) forms, shown for the N-lobe of human lactoferrin. In the conformational change, helix 5 in domain N2 appears to pivot on helix 11. The hinge in the two backbone strands (at Thr 90 and Pro 251) is indicated with an arrow. Taken from Anderson et al. (80).

The discovery of a closed C-lobe in the apolactoferrin crystal structure was unexpected but has given some added insights into the likely dynamics of transferrins. It seemed certain the the C-lobe must be able to open in a similar way to the N-lobe, although it was known to bind iron more strongly (107)and release it more slowly (1081,in transferrin at least, and it appeared to be conformationally less flexible (108). It was suggested that a disulfide bridge, number 7, which makes a link between the C1 and C2 domains and which has no counterpart in the N-lobe, might impose extra rigidity, although it would still be unlikely to totally prevent opening. Because no metal ion, anion, or other species that could be holding the two domains together was bound, it was concluded that the requirements of crystal packing had preferentially selected out one of several conformations existing in solution (80).The implications are of profound importance. Even in the absence of a bound metal ion or anion, both open and closed conformations must be accessible in solution, and there must be very little energy difference between them, given the weakness and small number of intermolecular contacts in the crystals. Solution X-ray scattering measurements, which make use of the power of synchrotron radiation, have now added important elements to the picture. Closed and open structures give significantly different scattering profiles (Fig. 121, especially in the medium angle range (104).These studies have shown that the opening of the N-lobe of apolactoferrin seen in the crystal structure is consistent with the solu-

410

E. N. BAKER

s = 2 sin @/h(A-')

s = 2 sin ~ h ( A - 1 )

FIG. 12. Low-angle X-ray solution scattering profiles for the iron-free (--) and ironloaded (---) forms of (a)human serum transferrin, (b) human lactoferrin, and (c)chicken ovotransferrin. In d, the calculated profiles for closed (-) and fully open (--) lactoferrin are shown. From Grossmann et al. (104),with permission.

tion scattering profile. They further show, however, that both lobes do open in solution, in contrast to the apolactoferrin crystal structure, with its closed C-lobe. There is, however, no real contradiction between the two types of study, which illustrate different aspects of the conformational variability. The solution scattering studies also show that the extent of closure can be affected by the particular metal ion bound (105;see also Section IV.B.5) and by mutations of ligands in the metal binding site (106). The picture of conformational change has now been nicely completed by the analysis of a second crystal form of apolactoferrin in which both lobes were found to be open, just as in the solution studies (Fig. 13). Although the resolution is limited (3.5 A), the opening of the N-lobe is again -50", whereas the opening of the C-lobe is somewhat less, at -15" (109).This adds support to the earlier expectations (80)that the extra disulfide bridge might inhibit opening of the C-lobe and highlights the existence of structural and functional differences between the two lobes. One final structure which merits comment is that of a monoferric form of human serum transferrin. In this structure, determined at 3.0 resolution (110)the N-lobe is metal-free and has an opening of -50" relative to the iron-bound N-lobe of rabbit transferrin; the opening thus

TRANSFERRIN STRUCTURE AND REACTIVITY

411

FIG.13. Stereo diagram of the fully open structure of human lactoferrin, determined crystallographically (109).The N-lobe is upper, the C-lobe, lower.

almost exactly matches that in apolactoferrin. The C-lobe, however, has iron bound, and has the expected closed configuration, essentially identical to that in rabbit transferrin. 6. Minitransferrins: Half- and Quarter-Molecules

The bilobal structure of transferrins means that half-molecules, representing either the N-terminal or C-terminal lobe, can be relatively easily prepared, either by limited proteolysis or by recombinant DNA methods (Section 1II.A). Relatively high-resolution crystal structures have been determined for three such half-molecules, the proteolytic N-lobes of rabbit transferrin (74) and chicken ovotransferrin (77) at 2.3 A and the recombinant N-lobe of human lactoferrin at 2.0-A resolution (75). These show that both the protein structure and the metal and anion binding sites are the same as in the intact parent structures. In fact comparison of the metal and anion sites of the lactoferrin and transferrin half-molecules with each other and with the N-lobe of lactoferrin shows very close correspondence; 92 atoms from the nine residues, plus metal and anion, making up the immediate binding site can be superimposed with an rms deviation of only 0.4 A ( 75). Solution scattering measurements on half-molecules show that they undergo very similar conformational changes, on iron binding or release, to those of the corresponding lobes of intact transferrins (104). This, together with the structural correspondence noted above, makes them valid as models of single-sited transferrins. The lactoferrin half-

412

E. N. BAKER

molecule offers one note of warning, however. The structure analysis (75) showed that in the absence of contacts from the C-lobe, a helix, number 11, at the back of the iron site, had unraveled in the halfmolecule to become an extended strand. This appears to be associated with altered properties of iron release (see also Section V.B). Quarter-molecules (single domains) can also be prepared, although this is relatively straghtforward only for the N2 domain, which is made up of a single continuous piece of polypeptide. Both N1 and C1 domains comprise two noncontiguous sections of polypeptide, and preparation of a separate C2 domain would require that the disulfide bridge, number 7, linking it to C1, be broken first. The crystal structure of the N2 domain (quarter-molecule) of duck ovotransferrin has recently been determined at 2.3-81 resolution ( 76). The anion site (which is formed entirely by residues of the N2 domain) is unchanged, as anticipated, and the iron atom is bound by the two carbonate oxygens and the phenolate oxygens of the two Tyr ligands (Fig. 14). This arrangement models a likely key intermediate in the uptake of iron by transferrins (see Section V.A). In the crystal structure the iron atom is actually six-coordinate, with the two remaining ligands coming from a piece of nonprotein density, which could be a glycine molecule resulting from the proteolytic preparation of the material. The folding within this single-domain fragment is the same as for the N2 domains of intact transferrins, again emphasizing that each domain is a stable entity and consistent with the idea that conformational change in transferrins occurs through rigid body movement of these domains.

c. VARIATIONS AMONG TRANSFERRINS The overall levels of sequence similarity between different transferrins (Section 1II.A)emphasize their family relationship and imply that they share a common three-dimensional structure. The variations in the iron-binding residues, and the residues which form the anion site, are listed in Table 111. These show that the N-terminal site is highly conserved through all species. The Asp ligand and both Tyr are totally invariant, while the His ligand is invariant except in the two insect transferrins, where it appears to be changed to Gln; this latter substitution would result in the replacement of the imidazole nitrogen ligand by an amide oxygen, but should not otherwise alter the iron site significantly, as Gln and His have similar steric properties. In the N-terminal anion site, the Thr, which hydrogen bonds to one carbonate oxygen, is invariant, while the Arg residue is changed only to Lys, which like-

TRANSFERRIN STRUCTURE AND REACTIVITY

413

FIG.14. The “quarter-molecule” Id-kDa duck ovotransferrin structure, showing the iron site at right, with coordination from the C0:- ion and two Tyr residues. [The two remaining coordination positions are occupied by a non-protein ligand, possibly a glycine molecule (not shown)]. Adapted from Lindley et al. ( 76),with permission.

wise carries a positive charge. There do not seem to be any damaging changes at the helix N-terminus (residues 122-124l-thus all species seem likely to have near-identical N-terminal binding sites. In the C-terminal site there is equally strong conservation in most species, but with two striking exceptions. In M.sextu transferrin both Tyr ligands are changed (to Asn and Asp), the His ligand is changed to Arg, which cannot serve as a ligand below pH -11, and the anionbinding Arg is changed to Thr (which has no positive charge). In human melanotransferrin, the Asp ligand is changed to Ser and there are major changes in the anion site, with the hydrogen bond from Thr lost (change to Ala), the positive charge of the Arg lost (change to Ser), and the appearance of a Pro at the helix 5 N-terminus. Both these species, therefore, appear to have defective C-terminal sites, and iron titrations confirm that they each bind only one iron (12,13), presumably in their “normal” N-terminal sites. One other region adjacent to the iron site contains variations that may influence metal binding and release. In the N-lobe of human lacto-

414

E. N. BAKER

TABLE I11 SEQUENCE SIMILARITIES IN

METALAND ANIONBINDING SITES‘ N-lobe residues‘

Proteinb

60

92

192

Human Tf Rabbit Tf Pig Tf Horse Tf Human Lf Bovine Lf Mouse Lf Pig Lf Chicken Otf Xenopus Tf Cockroach Tf M.Sexta Tf Melano Tf

253

117

His His His His His His His His His His Gln Gln His

Thr Thr Thr Thr Thr Thr Thr Thr Thr Thr Thr Thr Thr

121

122

123

Ser Ser Ser Ser Thr

Ala Ala Ala Ala Ala Ala Ala Ala Ala Ala Val Val Val

Ser Ser

Ser Ser Thr Asn Asn Thr

124

210

216

301

Glu Glu Glu Glu Glu Glu Glu Glu Glu Glu (Glu) (Pro) Glu

C-lobe residue‘S‘ Proteinb

395

435

528

597

461

465

466

467 468

546

552

Human Tf Rabbit Tf Pig Tf Horse Tf Human Lf Bovine Lf Mouse Lf Pig Lf Chicken Otf XenopusTf CockroachTf M. SextaTf Melano Tf

Asp Asp Asp Asp Asp Asp Asp Asp Asp Asp Asp Asp Ser

Tyr Tyr Tyr Tyr Tyr Tyr Tyr Tyr Tyr Tyr Tyr Asn Tyr

Tyr Tyr Tyr Tyr Tyr Tyr Tyr Tyr Tyr Tyr Tyr Asp Tyr

His His His His His His His

Thr Thr Thr Thr Thr Thr Thr Thr Thr Thr Thr Ser Ala

Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg Thr Ser

Thr Thr Thr Thr Thr Thr Thr Thr Thr Thr Asn Phe Pro

Ala Ala Ala Ala Ala Ala Ala Ala Ala Ala Ala Ser Ala

Lys Lys Lys Lys Lys Lys Lys Lys Gln Lys Lys Ser Arg

Glu Gln Gln Gln Cln Glu Gln Asp Glu Glu Glu His Asp

His His His His Arg His

Gly Gly Gly Gly Gly Gly Gly Gly Gly Gly Gly Gly Gly

644 Arg Arg Arg Arg Asn Asn Asn Asn Lys Lys Lys Asn Lys

Metal ligands, residues of the anion site (Thr and helix 5 N-terminus),and putative “trigger” residues. Tf, transferrin; Lf, lactoferrin; Otf, ovotransferrin. For references, see text, Section I1X.A. Human lactoferrin numbering. Residues in parentheses are those for which the alignment is uncertain.

ferrin two basic residues are located near the back of the iron site; Arg 210 hydrogen bonds to one Tyr ligand and nearby Lys 301 forms an interdomain salt bridge with Glu216. In transferrins both residues are Lys and there are structural differences in which Lys 206 (equivalent to Arg 210) does not directly hydrogen bond to the Tyr ligand, and the salt bridge is also not formed (see Section V.B).It has been suggested that protonation [a “pH trigger” ( 77)1 or the binding of secondary, nonsynergistic anions [a “salt trigger” (81)I to these lysines could stimulate iron release and that these differences could then contribute t o the characteristic differencesbetween lactoferrins and transferrins ( 77,

415

TRANSFERRIN STRUCTURE AND REACTIVITY

811. These questions are discussed further in Section V.B. There are differences in the C-lobes as well but their significance is less clearfuture mutagenesis studies may help. Two other variations merit comment. In the lactoferrins the connecting peptide is 12 residues long and contains no Pro residues (Table IV). It forms a regular a-helix in human lactoferrin (67) and probably all the other lactoferrins. In the mammalian transferrins it is longer (14 residues), however, contains Pro residues (and one or more Cys residues that form extra disulfide bonds), and has a very irregular conformation (68).These differences may be linked with the different lobe orientations found for the two proteins (Section III.B.l). In other members of the family the connecting peptide can be as short as 7 residues, as in the two insect transferrins (Table IV), or 9 residues, as in melanotransferrin. If differences in the connecting peptide are associated with different orientations and modes of association of the two lobes, then these may in turn be pointers to variations in functional properties. Another variation that may affect the dynamic properties of different transferrins is in the number and distribution of disulfide bonds. Because conformational change is important for metal binding and release (Section V) the restraints imposed by these covalent bridges may be important. The likely distribution of disulfide bonds in different species is given in Table V. While these are deduced from sequence alignments and are thus tentative, several conclusions can be drawn. All species appear to have the same six conserved disulfide bonds in their N-lobe TABLE IV SEQUENCE VARIATIONS IN INTERLOBE CONNECTING PEFTIDE Protein

Sequence"

Human transferrin Rabbit transferrin Pig transferrin Horse transferrin Human lactoferrin Mouse lactoferrin Bovine lactoferrin Pig lactoferrin Chicken ovotransferrin Xenopus transferrin Cockroach transferrin M.sexta transferrin Melanotransferrin

-0-

Y E Y V T A I R N L R E G T C P E A P T D E C K P V KrWC YEYVTAVRNLREGI CPDPLQDECKAVKWC Y Q Y V T A L R N L R E E I S P D S S K N E C K K V R WC

YEYVTAI RNLREDI RPEVPKDECKKVKWC S G Y F T A I Q N L R K S --E E E V A A R R A R V V WC F S Y T T S I QNLNKK--QQDVI AS KARVTWC S R Y L T T L K N L R E T --AE E V K A R Y T R V V WC L P Y L T A I Q G L R E T --AAE V E A R Q A K V V WC F E Y Y S A I Q S M R K D --Q L T P S P R E N R I Q WC S RLFQCI QALKEGV-KEDDS AAQVKVRWC K A N Y T D V I E R D -------T G A P HR F V R F C K A N Y T E V I E R G -------NG APE L VVRL C H E Y L H A M K G L L ---C --D P N R L P P Y L R WC

The one-letter code for amino acids is used (C, Cys; P, Pro; etch Alignment on either side of the connecting peptide is based on the lactoferrin and transferrin 3D structures. The last helix in the N-lobe and first p-strand in the C-lobe are indicated. Cys and Pro residues are in boldface.

416

E. N. BAKER

TABLE V

DISULFIDE BRIDGESIN TRANSFERRINS No." Residuesb hTf rTf 1 9 2 1 9 3115-198 4157-173 5170-181 6231-245 10 134-336 11 160-183 26-283' 91-327' 1 2 3 4 5 6 7 8

9 12 13

348-370 358-371 459-534 493-507 504-517 575-589 483-677 405-686 427-649 342-608 627-632

4 -

5

+

3

+ + + + +

pTf +

6

eTf +

+

+ +

+ +

+

+

+

+ + +

+

+

+

+

+

+

+

hLf

t

bLf

mLf

pLf

cOtf

N-lobe + + + + + + + + + + +

+ + + +

+ + + + +

+ + + +

+

+ + +

XTf +

+

+ +

t

+ +

+

+ + ? +

MsTf

cTf

+ +

+ +

+

+

+ +

+ + + +

MTf + +

+

+ + +

t

+

+

+

+ + + + + + + + + +

+ + + + + + + + + +

+ + + + + + + + + +

+

i

+

~~~

+ + + + + + + + + +

+ + + + + + + + +

+

Globe

+

+ + + + + + + + +

+

~

+ t + + + + + + + +

+ + + + + + + + + +

+ + + + + + + + +

~

+ + + + + + + +

+

+ + + + + + +

+

+

+ + + + +

+

+ + + + + + +

~

" Numbering according to Williams (87). Residue numbering for human lactoferrin. Deduced from sequence and structure alignment.

(numbers 1-6), but several have one or two extra, one of which (number 10) may influence domain opening since it joins part of the N2 domain to the connecting peptide. Xenopus transferrin appears to be an outlier in its N-lobe disulfides. In the C-lobe, 9 disulfides are strongly conserved but interestingly the two species which do not bind iron in their C-lobe, M. sexta transferrin and melanotransferrin, are also the outliers with respect to the disulfide bonding, having fewer than other species.

D. SIMILARITIES WITH BACTERIAL BINDINGPROTEINS A remarkable feature of transferrin structure, discovered when the human lactoferrin structure was determined (67, 851, is the striking similarity with a group of bacterial binding proteins. These proteins, the bacterial periplasmic binding proteins, bind and transport certain small molecules, such as sugars, amino acids and oxyanions, through the periplasmic space before delivering them via specific receptors in the bacterial cell wall (111).They thus share with transferrins the

TRANSFERRIN STRUCTURE AND REACTIVITY

417

properties of strong but reversible binding and receptor-mediated release. The structures of a number of the bacterial proteins have by now been determined (111).All are the same size (300-350 residues) as each lobe of transferrin. All are divided into two domains with the substrate bound in the interdomain cleft, just as for each transferrin half-molecule. In the bacterial proteins, as for transferrins, substrate binding is associated with a large-scale, hinge-bending, conformational change, likened to a "Venus flytrap" motion (112),in which the domains close over the bound substrate. For two of the proteins both closed (substrate-bound)and open (substrate-free) structures have been determined; for the maltodextrin binding protein the domain movement is about 35" (113 1, whereas for the lysine-ornithine-arginine binding protein it is 52" (114), almost exactly as in lactoferrin. Most remarkably, one group of the bacterial binding proteins, which includes the two anion-binding proteins so far analyzed [specific for sulfate (115)and phosphate (116)],has even closer similarity. First, the polypeptide folding pattern in these proteins is almost identical to that in each lobe of lactoferrin (Fig. 15); the central p-sheet of each

FIG. 15. Polypeptide folding patterns for (a) one-half of a transferrin molecule (the N-lobe of lactoferrin) and (b) the bacterial periplasmic sulfate-binding protein. Adapted from Baker et al. (85), with permission.

418

E. N. BAKER

domain has the same topology; the backbone strands, where the hinge is located, have the same arrangement; and many of the helices match. Second, the anion binding site in each protein coincides with the site used to bind the carbonate ion in the transferrins; the largest single contributor to anion binding in each of the proteins is the N-terminus of the same helix (helix 5 in lactoferrin) together with a hydrogen bonding side chain (Arg in the transferrins, Ser in the bacterial proteins) attached at the N-terminus (Fig. 16). Of course there are also differences (Fig. 151,but the similarities in folding, function, and location of the anion binding site argue for a common evolutionary origin. In fact, what sequence similarity there is, is greater between lactoferrin and the sulfate-binding protein (15% identity when aligned by 3D structure) than between the sulfate- and phosphate-binding proteins

(
Whether the two groups of proteins do share a common evolutionary origin is not clear, but they could have evolved from an ancestral anionbinding protein, with the metal-binding capacity of transferrin being added later, prior to the gene duplication that led to the present twosited transferrins. A further connection may be provided by a recently discovered bacterial protein involved in iron transport (1171-perhaps this is the “missing link”? Irrespective of whether there is an evolutionary link, the structural and functional similarities are intriguing and revealing. Perhaps the transferrins could be regarded primarily as anion-binding proteins? The suggestion is not facetious, because the anion apparently binds first ( 118-120), although their physiological roles are clearly based on their metal-binding properties.

FIG.16. Comparison of the anion binding sites in lactoferrin (left) and the bacterial sulfate-binding protein (right).From Baker et al.(82),with permission.

TRANSFERRIN STRUCTURE AND REACTIVITY

419

IV. Properties of the Metal and Anion Sites

A. SPECTROSCOPIC MONITORSOF METALBINDING The orange-red color that develops when ferric iron is added to solutions of metal-free transferrins has traditionally provided the main indicator of iron binding in the two specific sites. The color is due to an intense charge-transfer absorption band at around 465 nm. The precise wavelength of maximum absorption varies slightly from one transferrin to another, in the range 460-470 nm, and it also represents an average of the maxima for the two sites, which differ slightly (Section 1V.D). The similarity of the visible spectra to those of metal-phenolate complexes (90)suggested that they arise from tyrosine-metal interaction, and the charge transfer band is generally agreed to be derived ) metal (d.rr*)transition (121). from a ligand phenolate ( 7 ~ + The rise in the 465-nm absorbance as Fe3+is added to the apoprotein (generally as a ferric nitrilotriacetate or ferric citrate complex) can be used to monitor iron binding and forms the basis of iron titrations that demonstrate the presence of two specific sites per molecule (Fig. 17). The ionization of tyrosine on metal binding (PhOH --$ PhO- + H+) also causes a dramatic increase in absorbance in the UV region, both at 245 nm and -290 nm, due to perturbations of the T-T* transitions of the aromatic ring. This effect has been widely exploited for metal titration studies, using UV difference spectroscopy (122-125). Typical metal titration difference spectra are shown in Fig. 18. These can also be used to determine the number of metal ions that are bound (Fig. 18) and to determine metal binding constants, at least for the weakerbinding metal ions (126, 127). Fluorescence quenching measurements can also be used to monitor metal binding (12,128,129).The fluorescence spectra of transferrins, as for most proteins, are dominated by emission from tryptophan residues, but some of this is quenched if a cation, such as a transition metal ion or lanthanide ion, is bound nearby; one estimate is that energy can be transferred from tryptophan to Fe3+ or Cu2+over distances up to -20 8, (128). Metal binding is thus accompanied by a substantial decrease in transferrin fluorescence (e.g., Fig. 17). The decrease is nonlinear, however, with the binding of the first metal ion contributing a greater decrease than the second (12,128).This is probably because some tryptophan residues lie between the two metal sites and their fluorescence is quenched no matter which metal site is occupied first; binding of the second metal ion would have little further effect on these residues. One study (129)suggested that iron bound at

420

E. N. BAKER I

a

Lf

0'

b

0

I

110

2;o

I

I

I

I

1.0

2.0

mole ratio

I

mole ratio

FIG. 17. Spectroscopic monitors of iron binding. In a, the rise in absorbance at 466 nm as Fe3+is added is used to show that human melanotransferrin (MtO binds only one Fe3+ ion per molecule, compared with two for lactoferrin (LO. In b, the decrease in fluorescence is used to demonstrate the same phenomenon. In each case, the mole ratio is moles Fe3+added per mole of protein. From Baker et al. (12),with permission.

the C-terminal site produces more quenching than iron at the N-terminal site, but attempts t o relate nonlinearity of fluorescence quenching to site binding preferences have tended to be inconclusive ( 17).

B. METALSUBSTITUTION AND SPECTROSCOPY The transferrins are regarded primarily as iron-binding proteins. This is proper, because their affinity for Fe3+is substantially greater than for any other metal ion and their physiological role is primarily concerned with the binding, sequestration, and transport of iron. They can, however, accommodate a wide variety of other metal ions, including most of the first row transition elements, several of the second and third row transition elements, group 13 metal ions, lanthanides, and actinides (Table VI). Most of these are unlikely to be physiologically significant (some exceptions are noted below), but many have been

42 1

TRANSFERRIN STRUCTURE AND REACTIVITY

0.8 0.5

a Q

0.6

a

0.4 0.3

0.4

0.2 0.2

0.1

0

0

250

250

300

x nm

300

A nm

4

1.0

0.0

r

3.0

2.0

0

1

2

r

3

4

FIG. 18. Typical U V difference spectra, showing the increase in absorbance, AA, during the binding of (a) V3+and (b) Yb3+to human lactoferrin. The spectra are shown as a superimposed series of spectra, corresponding to successive increments of the added metal ion. Below each set of spectra is the corresponding titration curve, showing that in each case two metal ions are bound per molecule ( r is the ratio of moles of the metal iron added per mole of protein).

TABLE VI

METALIONS REPORTED TO BINDTO TRANSFERRINS f-block elements

Transition elements Row 1 Sc3+ (130) V3*. V02', VO,' (131-133) CrJ* (134, 135) 1134-136) (137, 138) (139, 140, 134-1361 (137, 140) CU" (89, 135, 136) Zn2+ 189, 127)

Mn2', Mn3' Fe". Fes+ Co2', Co3+ Ni"

Row2

Row3

Group 13 elements

Ru3+ (141)

He' (142)

A13+ (144)

Cd2+ (140) R2' (143)

Ga3' (126) In3+ (145) TI3+(146)

Lanthanides

Actinides

La3+ (147) Th" (151) Ce3+.Ce" (147) h4+ (151)

Pr3' (148) Nd3' (1481 Sm3+ (149) Eu3+ (148) &I3' (150) Tb3' (148) Ho3+ (148)

EP' (148)

422

E. N. BAKER

exploited for their spectroscopic properties in investigations of transferrin structure and function. A feature of fundamental importance to understanding the metalbinding properties of transferrins is the strong preference for cations with a high positive charge. For iron this is shown by the binding constants for Fe2+relative to Fe3+,these being approximately lo3 for Fe2+(137)and lo2' for Fe3+(107,138). Likewise when manganese is added as Mn2+it becomes oxidized to Mn3+(134,135), and all of the non-transition metal ions bound are tri- or tetravalent. Binding of metals other than iron is generally inferred from spectroscopic observations, especially UV difference spectra that monitor binding to tyrosine ligands. It is, however, important to distinguish whether binding occurs at the two specific sites or at nonspecific sites that might be found at various locations on the molecular surface. To this end, Aisen (252)has proposed three criteria that might identify a metal ion as specifically bound, i,e,, (i) no more than two metal ions bound per molecule, (ii) binding to iron-saturated transferrin not observed, and (iii) one (bikarbonate, or other suitable anion, bound with each metal ion. Most often it is the third criterion that has not been demonstrated to be met. The metal ions that have definitely been shown to bind specifically, by these criteria, are Cr3+,V02+,Mn3+,c ~Cu2+, ~ and~ Ga3+, , as well as Fe3+(152), but it is highly probable that most of the species in Table VI do bind in the two specific sites. Even where binding to the specific sites is indicated, however, this does not necessarily mean that all ligands are the same. UV difference and visible charge transfer spectra monitor binding to the tyrosine residues, which, together with the synergistic anion, belong to just one domain (domain 2). Such spectra may not show, therefore, whether the structure is closed (as with Fe3+)or open and whether the Asp and His ligands are also coordinated or not. Crystallographic and biophysical studies (Section IV.B.51, or spectroscopic probes sensitive to the other ligands, are necessary to resolve these questions. Binding constants have been determined for some metal ions, but their determination is complicated by the presence of two very similar sites and by the pH and (bilcarbonate-dependenceof the values. Readers are referred to an excellent discussion by Aisen and Harris ( 17).For Fe3+,at ambient HC03- concentrations, the log K values are estimated as 19.4 and 20.7 for the two sites (107)and it is generally the case for most metal ions analyzed that the two sites differ by approximately one unit in their log K values (Table VII).

423

TRANSFERRIN STRUCTURE AND REACTIVITY

TABLE VII SELECTED BINDING CONSTANTS FOR METALSBOUND TO TRANSFERRINS~ ~

~

Metal ion

"!03-1

Fe3' Fez+ Ga3+

Ambient 5mM Ambient 5mM Ambient

~ 1 3 +

Zn2+ Ni2 Nd3+ Sm3' Gd3 +

+

5mM

Ambient Ambient Ambient

Reference

PH 7.4 1.4 7.4 7.4 7.4 1.4 7.4 7.4 1.4

20.1 3.2 18.1 13.5 5.7 4.1 6.1 7.1 6.8

19.4 2.5 17.1 12.5 4.3 3.2 5.0 5.4 -

107 137 126 153 127 137 149 149 150

'Conditional equilibrium binding constants for the stated pH and [HC03-]( 1 7 ) . 1 . Metals of the First Transition Series

The first row transition elements display a rich chemistry in their binding to transferrins. As noted above the trivalent oxidation states are highly favored, and although no binding constants have been reported, the very stable complexes formed with Cr3+,Mn3+,Co3+, and Cu2+have been extensively studied for transferrin (1341, ovotransferrin (136),and lactoferrin (135, 154). Like Fe3+,the elements Mn3+, Co3+, and Cu2+give intensely colored complexes (brown, yellow, and yellow, respectively) because of the presence in the visible region of the phenolate ( r )+ metal ( d r * )LMCT band (121).Typical spectra are shown in Fig. 19. Cr3+ complexes, on the other hand, are pale blue-gray in color and their visible spectra contain only weak, spinallowed d-d bands ( 1 7). Analyses by Patch and Carrano (121)suggest that the charge transfer band for Cr3+is shifted to -240 nm, where it would be obscured by the intense tyrosine r-r* transitions. Cr3+gives rise to useful EPR spectra, which gave one of the first clear indications of differences between the two sites in transferrins (Section 1V.D). Vanadium can bind to transferrins in three different oxidation states. V(II1) has been reported (131) as giving an air-stable complex with transferrin, probably as the simple cation V3+;the corresponding lactoferrin complex of V(II1) is, however, extremely air sensitive, being oxidized in minutes (1561, and the transferrin result has been questioned (133).V(IV)forms a colorless complex with transferrins, shown by EPR spectroscopy to incorporate the vanadyl (V02') ion (155).Displacement studies show that the V02+ions occupy the specific Fe3+sites

424

E. N. BAKER

co3+

Wavelength

nm

FIG.19. The visible absorption spectra of various metal complexes of human lactoferrin; all spectra are for 1%protein solutions. Very similar spectra are obtained for transferrin and ovotransferrin. Adapted from Ainscough et al. (135),with permission.

and that the synergistic (bikarbonate ion is required (132). Although oxidation to V(V) occurs rapidly in air, the V(IV) complex is stable under nitrogen and its d1electron configuration makes it another extremely useful EPR probe. In its highest oxidation state, V(V), vanadium probably binds t o transferrins as the dioxovanadium(V) ion, VO,+, giving a colorless complex. An interesting feature here is that a synergistic anion is apparently not required (133), with the two cis oxygen atoms perhaps taking the place of two carbonate oxygens. The net charge of V02+ is then the same as Fe3+-CO32- . Most of the divalent transition metal ions bind only weakly to transferrins, the binding constants for Fe2+,Ni2+,and Zn2+being approximately lo3, lo4, and lo5, respectively (Table VII). Cu2+may be an exception. Although no binding constants have been determined for any of its complexes with transferrins, and metal displacement studies indicate that it binds less strongly than A13+ (144)and Mn3+(1541,it does form a very stable complex and has been used more than any other metal ion (apart from Fe3+)for physicochemical studies on transferrins. With its d9 electron configuration, Cu2+has, like V02+,provided an excellent and much-used EPR probe of transferrin structure. It was the EPR spectra of Cu2+-transferrin (911, -0votransferrin (1571,and -1actoferrin (92)complexes that provided the most compellingevidence

425

TRANSFERRIN STRUCTURE AND REACTIVITY

for a histidine ligand in the metal coordination sphere, prior to crystallographic analysis; the triplet superhyperfine splitting (Fig. 20) is characteristic of Cu(I1) interaction with a single nitrogen ligand. Cu2+is also the only metal ion, apart from Fe3+,for which the details of its coordination geometry in a transferrin complex have been determined crystallographically; the structure of dicupric lactoferrin has recently been reported at 2.1-A resolution ( 2 6 ) (see below). Cu2+-transferrin complexes have also been used to examine the relationship between metal and anion binding (Section 1V.C). Because most of the first transition series elements are essential in biological systems there is considerable interest in the possibility that transferrins may be involved in their binding and translocation. This is certainly a distinct possibility, because neither transferrin nor lactoferrin is more than -30% saturated with iron in body fluids. There is as yet little hard evidence, however. Transferrin has been reported to be the main carrier for manganese in blood (158),just as lactoferrin is in milk (25), and thermodynamic studies suggest that transferrin is also capable of competing with serum albumin for zinc under the

r

1

J

V

i ,

CCI

1 1 1 1 1 1 2600 2620 2640

I 2600

I

2660 2680

I 2800

I

2700

I

I

3000

I 3200

1

I 3400

Magnetic field (gauss)

FIG.20. The EPR spectrum of 65Cu2+-transferrin-carbonate.The inset shows a lowfield hyperfine line (at an eightfoldincrease in gain) that displays superhyperfine splitting resulting from the coordination of one nitrogen ligand in each site. Adapted from Zweier and Aisen (91), with permission.

426

E. N.BAKER

conditions that prevail in circulation (127).Despite the biological importance of copper, however, surprisingly little is known of the likely role of transferrins in binding it in uiuo.

2. Group 13 Metals The group 13 metals are of particular interest because they have a characteristic 3+ charge, like Fe3+,and because several of them are important medically. Aluminium is widely consumed in drinking water and foods and has been linked to various diseases, including dialysis encephalopathy and Alzheimer's disease (24,159). Gallium is of pharmacological importance because 67Ga3+is an effective radiotracer, used as an imaging agent in diagnostic medicine (1601,and two radioactive isotopes of indium, "'In3+ and '131n3+,have similar applications. Thallium, on the other hand is a potent poison. Transferrins are likely carriers of all of these species, and the stability and nature of the metal-transferrin complexes are important for several reasons: in determining where the metals are likely to accumulate and how the interaction is likely to be affected by competitive chelators. The chemistry of the group 13 metals makes binding studies difficult, as careful control of pH and bicarbonate concentration is necessary to prevent formation of species such as Al(OH),- and Ga(OH),-. Nevertheless, UV difference spectra have shown that A13+,Ga3+,In3+,and T13' all form transferrin complexes with two metal ions per molecule (126, 144-1461, A& values imply the ionization of two tyrosines per bound metal ion, as for other specifically bound metals. NMR studies using 13C-enrichedbicarbonate show virtually identical spectra for A13+and Ga3+,implying equivalent metal-anion environments (99). Of the four metal ions, Ga3+is most like Fe3+in size, with an ionic radius of 0.62A compared with 0.65 A for Fe3+(1611, and it is, accordingly, the most strongly bound. The binding constants for Ga3+-transferrin are 1017and 10l8 for the two sites, almost comparable to those for Fe3+and higher than those for any other metal ion for which data are available. Ga3+-transferrin is recognized by transferrin receptors, providing the means by which Ga3+enters tumor cells (162).Its diamagnetism and similar chemistry have also led Ga3' to be used in place of Fe3+in 'H NMR studies of transferrin (163). A13+ (ionic radius, 0.54 A) is smaller than Fe3+ but still forms a stable transferrin complex, with binding constants 10'2.5and 10.'3.5; A13+is displaced by Fe3+(144)but the A1,Tf complex is stable enough that it has been shown to bind to cell surface receptors (164).Smallangle X-ray scattering experiments, however, indicate that the A13+-transferrin complex is conformationally distinct from the diferric

TRANSFERRIN STRUCTURE AND REACTIVITY

427

protein, implying perhaps an altered domain closure and altered receptor interactions (105). In spite of its larger size, with an ionic radius of 0.80A (161), In3+ appears to bind to transferrin with an affinity close to that of Fe3+ (145).In3+displaces Cu2+from copper-saturated ovotransferrin, and the In3+even remains bound in the presence of an added twofold excess of Fe3+. Indium-transferrin also migrates indistinguishably from iron-transferrin (145)and gives the same “closed” conformation, as judged by small-angle X-ray scattering (105). Finally, thallium has been used as an NMR probe of transferrin (146).Although thallium is bound in through the binding of 205T13+ both metal sites, they are distinguishable spectroscopically perhaps because the presence of a larger cation (radius of T13+,0.89 A) accentuates the differences between the two sites (see also Section 1V.D). 3. Lanthanides

The lanthanide ions resemble Fe3+ in their charge (3+), but are substantially larger, ranging in ionic radius from 0.86 A for Lu3+t o 1.03A for La3+(161). There have been a number of studies of lanthanide binding to transferrins, mostly aimed at using the metal ions as spectroscopic probes. Luk (14x1used UV difference titrations to demonstrate binding of a series of lanthanide ions to human transferrin. The four smaller lanthanides used (Eu3+,Tb3+,Ho3+,and Er3+ gave results indicative of binding at both sites, whereas two larger ions (Pr3+and Nd3+)bound only at one site; it was suggested that one of the specific transferrin sites was too small to accommodate the larger lanthanides. Harris (149)used UV difference titrations to determine binding constants for Sm3+and Nd3+(Table VII). He concluded that two metal ions were bound in each case and highlighted problems inherent in the bicarbonate dependence of specific metal binding. In the case of the lanthanides, this results in competition with the precipitation of insoluble lanthanide carbonates. Contradictory results have been obtained for Gd3+binding to human transferrin. In a careful study, Zak and Aisen (150)found only one Gd3+ion bound per molecule, showed this to be in the C-terminal site, and determined its binding constant (Table VII). In contrast, O’Hara and Koenig (165)obtained convincing evidence for two Gd3+bound per molecule. The difficulties appear to arise from the relative weakness of lanthanide binding making it sensitive to competitive effects. For lactoferrin, which binds metal ions more strongly than transferrin (166),all the lanthanide ions, from Yb3+ to La3+,bind with two metal ions per molecule (1471,with the log K2values ranging from 6.4

428

E.N.BAKER

for Yb3+to 4.8for La3+;binding at the second site thus becomes weaker as the ionic size increases. One striking result for lactoferrin is that Ce3+ initially forms a colorless Ce2Lfcomplex, like the other lanthanides, but that on exposure to air the complex rapidly turns deep red. Ce3+ is oxidized to Ce4+and a strong visible absorption band at 442nm appears, with an extinction coefficient ( E = 4640 M-' cm-l) reminiscent of the transition metal ion LMCT bands (147). The smaller ionic radius of Ce4+compared with that of Ce3+(0.87 A compared with 1.01 A) may help to promote formation of this complex. One other point of controversy concerns the AE values obtained in UV difference titrations of lanthanide ions. These are significantly higher than the corresponding values for transition metal ions (7400-8700M-l cm-' compared with 5000-6600 M-' cm-'1, leading to debate about the number of tyrosines coordinated (167). Although it is possible that lanthanide binding could recruit a third Tyr ligand ( 781, careful comparisons of the AE values with those of complexes with a small-molecule phenolate chelate analog suggests that the number is two per metal ion in all cases (167). Lanthanide complexes of transferrin have been used for several purposes. Gd3+-transferrin gives a characteristic EPR signal at g = 4.96, quite unlike the spectra for other Gd3+complexes (165);Eu3+has been used to probe differences between the two sites by Eu(II1) excitation spectroscopy (168); and the luminescence of excited Tb3+ ions bound in one site of mixed-metal Tb3+-Mn3+and Tb3+-Fe3+transferrin complexes has been used t o determine the intersite distance (169). The value obtained, 35.5 A, compares well with the value of 42 A later obtained from the lactoferrin crystal structure (67). 4 . Actinides and Other Metal Ions

Very few metals of the second and third transition series have so far been shown to bind to transferrins. Ru(II1) has been shown to bind to human transferrin, giving a deep red complex, when added as a chelate complex with nitrilotriacetate (1411, and similar behavior has been found for Pt(I1)when PtC1,2- is added (143).Cadmium has been shown by UV difference spectroscopy to bind in both sites of ovotransferrin (140),and Cd20Tfcomplexes have been used to investigate anion binding by '13Cd and 13C NMR spectroscopy using bicarbonate (170) and oxalate (171 ) as synergistic anion. Hafnium has been of particular interest because of its ability to model the binding of plutonium. HP' has a similar size to Pu4+(0.83 A compared with 0.86A) and behaves similarly in its metabolism and interactions with biochemical ligands (142). 18Hf also gives access to the spectroscopic method of perturbed

TRANSFERRIN STRUCTURE AND REACTIVITY

429

angular correlation of y rays (142). The complex of Hf4+-transferrin has hafnium bound at both sites (172),but appears not to have the same “closed”configuration as transferrin complexes of smaller cations (105). The actinides plutonium, neptunium, protoactinium, and thorium (151,173) bind to transferrin. The larger Th4+ion (radius, 0.94 A) still binds to both sites, although binding to the second site (probably the N-terminal site) is significantly weaker than that to the first and apparently involves only one Tyr ligand compared with two Tyr in the other (151).Although UV difference spectra for P u are ~ equivocal ~ (174),it seems likely that two Pu4+are bound. The likely carrier properties of transferrin for Pu4+makes the design of competitive chelators of some importance (151).

5. Structural Aspects of Metal Substitution The metal ions that bind to transferrins are very diverse in size and in their coordination preferences, raising the question of what degree of induced fit there is in the metal sites. When the protein domains close over the bound metal ion (Section III.B.51, does the metal simply fit into a prepared site or are the details of the binding site moulded to some degree by the stereochemical preferences of the metal? This question is of great importance for correlating structure with spectroscopy. The only metal ion other than Fe3+for which crystallographic data are available is Cu2+.In a high resolution study of Cu2+-substituted lactoferrin, Smith et al. (26)showed that although the overall structure of the molecule is not changed at all, compared with the diferric protein, the metal sites are subtly different. Small movements in the metal ion position (1.0 in the N-terminal site, 0.4 A in the C-terminal site) bring about a change in the metal coordination geometry and increase the nonequivalence of the two sites. In the N-site the copper is five coordinate, square pyramidal, with a long (2.8 A) apical bond to one Tyr ligand (Tyr 92) and a monodentate (bilcarbonate ion (Fig. 21); the monodentate coordination of the latter results primarily from the copper movement but is also facilitated by a slight (20”)rotation of the anion. In the C-site, however, the metal ion remains six-coordinate but more distorted from regular octahedral geometry than is the case for Fe3+.Corroboration for these crystallographic observations comes from an EXAFS study of Cu2+binding to ovotransferrin, which is indicative of one five-coordinate site and one six-coordinate (175). The degree to which a given metal can stabilize the “closed” configuration seen for the diferric proteins (67,68)has been addressed by small angle X-ray scattering studies (105).These show that Cu,OTf is

430

E. N. BAKER

FIG.21. Metal and anion binding in the crystal structure of Cuzt-substitutedhuman lactoferrin. (a) The five-coordinateN-lobe copper site, in which either the anion or Tyr 92 may be protonated, and (b) the six-coordinate C-lobe copper site. In each case the axial ligands, Asp 60 and Tyr 192 in the N-lobe and Asp 395 and Tyr 528 in the C-lobe, are present but not shown. Adapted from Shongwe et al. (192),with permission.

conformationally indistinguishable from the diferric protein, in accord with the crystallographic results for lactoferrin. The binding of In3+ also gave the same closure of the domains over the metal ions, as did AP+, although with some difference in magnitude in the latter case. In contrast, Hf4+ did not give the same domain closure, apparently remaining open as for the apoprotein, even though its ionic radius is only slightly greater than that of In3+(0.83 A compared with 0.80 A). This may not represent an absolute size cutoff, however, for the closed structure. The crystallographic results for Cu,Lf (26) show there is some room for adjustment in the metal sites, and whether a particular metal is able to stabilize the closed ferric-like structure must depend on a subtle interplay between size, geometrical preferences, and binding strength. It seems likely that most of the smaller metal ions, especially transition metal ions such as Cr3+,Mn3+,and Co3+,will favor the closed state, whereas most ofthe larger ions may not. The distinction is critical for the metabolic fate of transferrin-bound metals, because presumably only the closed structure is recognized by transferrin receptors (105). The concept that transferrin complexes of larger metal ions may not exhibit the same closed structure as that of Fe3+ is not incompatible with spectroscopic studies. UV difference spectra reflect tyrosine coordination, and it is known that the “open”, metal-free structure has both tyrosines close together, adjacent to the (bikarbonate site on domain 2 (80), but far (8 to 9 A) away from the remaining Asp and His ligands

TRANSFERRIN STRUCTURE AND REACTIVITY

43 1

on domain 1. Thus a lanthanide could bind to both Tyr and the anion, in the open configuration, and water molecules or other species could bind to the other coordination sites instead of the Asp and His ligands. This should not greatly affect the UV difference spectra, although it could perhaps explain the differences in A&values. Ultimately crystallographic studies of other metal-substituted transferrins are needed to define the coordination of some of the larger metal ions.

C. ANIONBINDING Ever since the pioneering studies of Schade and coworkers (9)and Warner and Weber (176) the interdependence of metal and anion binding has been recognized as a hallmark of transferrin chemistry. The relationship is synergistic in the sense that neither is bound strongly in the absence of the other; the log K value for the binding of bicarbonate to apotransferrin, measured from UV difference spectroscopy, is only 2.73 (1771, whereas Fe3t binds only very weakly and nonspecifically in the absence of a suitable anion (101).Together, however, they give the strong specific binding characteristic of transferrins. Anions that are able to promote such binding are referred to as synergistic anions (178) to distinguish them from other, secondary anions that bind more weakly, at different sites, and have a modulatory effect on the specific sites (Section IV.C.3). 1 . Characteristics of Synergistic Anions

In physiological media, carbonate, derived from dissolved carbon dioxide, fills the role of the synergistic anion. (Note that the form of this anion is variously given as carbonate or bicarbonate, but that a description as carbonate is now probably preferable, for the reasons given in Section III.B.4.) Some other anions can, however, substitute under carbonate-free conditions. A milestone toward understanding their characteristics was a study of some 30 anions, both organic and inorganic, using four different synthetic routes to the Fe3+-aniontransferrin complexes, and monitoring binding by the appearance of the Fe3+ charge transfer band (178). This study, by Schlabach and Bates, showed that many organic anions could substitute for carbonate, but that none of the inorganic anions, including sulfate, nitrate, and phosphate could do so. The latter observations confirmed earlier conclusions by Aisen et al. (179). A selection of synergistic anions is shown in Fig. 22. The common features of all such anions are the presence of a carboxylate group and, one or two carbon atoms removed, a second (proximal) electron donor

432

E.N. BAKER

Carbonate

Oxalate 0

Glyoxylate

Glyoolate OH

Thioglycolate

Lactate OH

Glycine

FIG.22. A selection of synergistic anions bound by transferrins.From Schlabach and Bates (178).

group that has the potential to act as a metal ligand. This second group may, for example,be a second carboxylate (oxalate, malonate), hydroxyl (glycolate),thiol (thioglycolate),amino (glycine),carbonyl (glyoxylate), etc. Some surprisingly large anions, such as the dye xylenol orange (1801, can fill this role, but there are some steric restrictions; e.g., lactate can act as a synergistic anion but methyl-lactate cannot (178). Much attention has been paid to citrate because it is used in chelate competition experiments but it appears it is not a synergistic anion for transferrins.

TRANSFERRIN STRUCTURE AND REACTIVITY

433

2 . Metal-Anion Znteractions Prior to the crystallographic demonstration that carbonate binds directly to the iron atom, in bidentate mode, in diferric lactoferrin (78),much debate surrounded the nature of the metal-anion-protein interactions and the functional role of the synergistic anion. Experiments directed at elucidating these questions,together with the crystallographic results, are now beginning to give a much clearer picture. Both electronic absorption spectra and EPR spectra are sensitive t o the particular anion used in Fe3+-anion-transferrin complexes. The characteristic charge transfer band in the visible spectrum varies conas the proximal electron donor group siderably in intensity and,,A on the anion is changed (Fig. 231,implying direct metal-anion bonding, and the EPR spectra of the carbonate and oxalate complexes of diferric transferrins are extremely different (Fig. 24).Many other spectroscopic studies have also suggested direct bonding of the anion to the metal, including electron-spin-echostudies of Fe3+-, Cu2+-, and V02+-transferrins with either carbonate or oxalate as synergistic anion (181-184),

0.6

$

0.4

3 %

9 0.2

0.0 Wavelength (nm)

FIG.23. Visible absorption spectra for diferric transferrin complexes utilizing various synergistic anions, showing the variation in A,, for the charge transfer band. Anions are 1, nitrilotriacetate; 2, carbonate; 3, salicylate; 4, thioglycolate; 5, glycine; 6, glyoxylate; and 7, glycolate. From Schlabach and Bates (1781,with permission.

434

E. N. BAKER

I

750

.

.

.

.

I

.

1000

.

.

.

, . . . . , . .

1250

1500

.

. +

1750

Magnetic field (gauss)

FIG.24. EPR spectra for diferric lactoferrin with (a) carbonate and (b) oxalate as synergistic anion. Adapted from Shongwe et al. (1921, with permission.

EPR studies of V02+-transferrins (1851, I3C NMR studies of Fe3+-, A13+-, Ga3+-, and Zn2+-transferrins (99, 186, 1871, and resonance Raman and EXAFS studies of anion binding to Fe3+-ovotransferrin (188). A focal point in the interpretation of anion-binding studies was the far-sighted “interlocking sites” model of Schlabach and Bates ( I 78). In this model (Fig. 25) the anion was proposed to bind to the metal ion through the proximal electron donor group Land to a positively charged region of the protein through its carboxylate group. In this way it bridged between metal ion and protein. Support for the involvement of a cationic group, probably lysine or arginine, was later found from chemical modification (189) and NMR (187) studies. The crystallographic results for lactoferrin ( 75, 78), transferrin (68, 741, and ovotransferrin (77) confirm that the carbonate ion does bridge between metal ion and protein, binding directly, in bidentate mode, t o the metal and hydrogen bonding to a positively charged region of the protein. The latter comprises an arginine sidechain and a helix N-terminus (Section III.B.4). The role of the anion is then to

TRANSFERRIN STRUCTURE AND REACTIVITY

435

FIG.25. A schematic representation of the “interlocking sites” model for the binding of anions to transferrins. From Schlabach and Bates (1781, with permission.

(i) neutralize the positive charge on the protein that might otherwise inhibit the binding of a cation nearby, (ii) partially prepare the site for metal binding by providing two potential ligands, and (iii) occupy two coordination sites on the metal ion, leaving no space for the attack of water.

In addition, protonation of the anion may play a role in the breakup of the metal site, leading to metal ion release (Section V.B) Extension of the carbonate binding mode to other anions does not follow automatically. Resonance Raman and EXAFS studies on the synergistic anions thioglycolate and 2,3-dihydroxybenzoate (188)confirmed that the proximal electron donor group L was bound to the metal, as in the interlocking sites model, but gave no indication whether the carboxylate group might also be coordinated. 13CNMR spectra showed that the two carbon atoms of oxalate displayed different chemical shifts indicative of different environments; either only one carboxylate was coordinated or both were coordinated but had different protein environments (99).On the other hand, both modeling (190) and electron-spinecho (183, 191) studies favored 1,2-bidentate coordination of oxalate, through both carboxylates. The latter study included a detailed analysis of 18 anions and showed convincingly that binding should occur through both carboxylate and proximal (L)groups (191). Confirmation has come through crystallographic studies of oxalate binding to lactoferrin. The crystal structure of a hybrid complex of copper-lactoferrin, at 2.2-A resolution, has carbonate in the N-terminal site and oxalate in the C-terminal site (192). The oxalate ion is bound to the metal ion (Cu2+in this case) in 1,2-bidentate mode, as anticipated (190,1911, i.e., through both carboxylates (Fig. 26a). One

436

E. N.BAKER \ 587

\ 481

/ 461

FIG.26. Binding modes for anions other than carbonate. In (a) the mode of binding of oxalate to human lactoferrin, as determined crystallographically (192,193),is shown. In b is a generalized model for synergistic anion binding to transferrins, based on EPR studies (191)and the crystallographic results for oxalate. From Shongwe et al. (I92), with permission.

carboxylate coordinates the metal and is hydrogen bonded (like carbonate) to the helix 5 N-terminus and Thr 461; the other carboxylate also coordinates the metal and is hydrogen bonded to Arg 465. The same arrangement is found in a second crystal structure, that of diferricdioxalato-lactoferrin at 2.3-81resolution (1931,this time with oxalate in both lobes and Fe3+as the metal ion. The oxalate binding mode is thus independent of the metal ion. In the diferric case, however, oxalate binding in the C-lobe is symmetric, with two Fe-0 bonds of 2.0 81, but in the N-lobe it is asymmetric with bonds of 2.0 and 2.6 A, emphasizing the slight differences between the sites, especially when “nonnative” metals or anions are bound (Section 1V.D). The question of how relatively large anions can be accommodated is answered in part by the lactoferrin structures. One oxalate carboxyl group occupies the carbonate site. To accommodate the other carboxyl group the Arg sidechain must be displaced 2-3 A away from the metal, but this is possible because the interdomain cleft contains a large solvent-filled cavity next to the binding site and this allows room for movement (78,82). There is thus some inner flexibility in the binding cleft. A very important caveat, however, is that anions have been defined as synergistic if the characteristic charge transfer spectrum is obtained when iron is bound. This monitors tyrosine binding but does not prove that the fully closed structure is adopted, because both Tyr and the anion are associated with domain 2 only. Thus it is conceivable

TRANSFERRIN STRUCTURE AND REACTIVITY

437

that some very large anions, such as xylenol orange (1801,may not lead to the same closed protein structure as carbonate and oxalate. Both the EPR (191) and crystallographic (192, 193) analyses lead to an elaboration of the Schlabach-Bates model, shown in Fig. 26b. The carboxylate group of synergistic anions is bound to the helix N-terminus and the Thr sidechain and can simultaneously coordinate the metal through one oxygen. The electron donor group L occupies a second coordination position (giving bidentate binding) and the Arg sidechain is pushed aside. The failure of phosphate and sulfate t o act as synergistic anions must be due to their tetrahedral structure. Both anions bind to the apoprotein, more strongly than carbonate in fact (177), but are presumably unable t o provide the right geometry for the specific metal-anion-protein complex. Nitrate, although isostructural with carbonate, cannot act as a synergistic anion, probably because of its lesser charge (1-,compared with 2- for C03'-), which would still leave a net positive charge at the anion site, and because of weaker hydrogen bonding ability, arising from the lesser polarity of N-0 bonds compared with C-0 bonds. Nitrate does not appear to bind to the apoprotein at all (177). Carbonate appears to have a higher affinity than any other anion, probably because of its perfect fit between metal ion and protein, and the need for other anions to displace the arginine sidechain. The affinity is, however, metal dependent, emphasizing the synergistic relation between metal and anion binding. With Cu2+as the metal ion, carbonate can be displaced from the C U ~ ( C O , ) complex ~L~ by an added large excess of oxalate; no displacement occurs, however, when Fe3+ is the metal ion, even with a very large (200-fold) excess of oxalate (192). Further, competition experiments, using equal concentrations of oxalate and carbonate with lactoferrin, show that under these conditions Fe3+binds carbonate preferentially over oxalate in both sites, whereas Cu2+prefers oxalate to carbonate in its C-terminal site, but carbonate to oxalate in its N-terminal site. The differences have been attributed to the different copper geometries supported in the two sites (192). 3. Nonsynergistic Anions

Nonsynergistic anions, bound at secondary sites, have attracted attention because of their possible role in metal ion release (Section V.B). The ability of nonsynergistic anions to perturb the EPR spectra of transferrins was first shown by Price and Gibson (194), in a study of the known chaotropic ion ClO,-. This effect was subsequently exploited by Folajtar and Chasteen (195), using EPR difference spectra to study

438

E. N. BAKER

the binding of a series of such anions to human transferrin; their binding followed the trend SCN- > C104- > HP20:- > ATP3- > C1-, whereas other anions (HPO:-, AMP2-, SO4’-, F-, BF4-, HC03-) had negligible effects. The order SCN- > C104- > C1- follows the lyotropic series for the binding of anions t o positively charged sites on proteins (196) but the binding appears to be relatively specific, rather than simply chaotropic, at least for a given transferrin. Two C1- ions are bound to each lobe of human transferrin, giving four in all. Moreover they are bound in pairwise fashion, with strong positive cooperativity. Two possible sites for the binding of nonsynergistic anions to transferrins may be identified from the crystal structures (Fig. 271, although there could be others given the large number of positively charged residues (lysine and arginine) on such large proteins. The first is the “essential” arginine at the synergistic anion site of each lobe. Binding to the other side of this arginine could perturb the interactions it makes with the synergistic anion and thus indirectly perturb the metal site. The second may involve the positively charged side chains behind the iron site, near the hinge region. In the N-lobes of serum transferrin and ovotransferrin, a pair of lysines are hydrogen bonded ( 77,811,and disruption of this unusual arrangement by anion binding could perturb the iron site (811. If this is the secondary anion site, however, it must be differentbetween different transferrins and between N- and C-lobes, because of the sequence differences that exist (Section 1II.C).

FIG.27. Possible sites for the binding of secondary, nonsynergistic anions, which may perturb the iron site and modulate release. Residue 121 is the “essential” arginine. Residues 210 and 301 are located behind the iron site, near the Tyr ligands and the hinge region. Numbering is as for the N-lobe of lactoferrin. The identities of these residues, which vary in the two lobes and between different transferrins, can be seen in Table 111.

TRANSFERRIN STRUCTURE AND REACTIVITY

439

An effect that is almost certainly related is that of salts on the relative stabilities of the two metal binding sites and on the kinetics of metal ion release. Increasing concentrations of salts, such as NaF, NaC1, NaBr, NaI, NaN03, Na2S04,and NaClO,, increase the stability of the N-terminal site relative to the C-terminal site of human transferrin (197).In the presence of an accepting chelator, salts also accelerate iron release from the C-site (197,198),thus reversing the normal order (N-site faster than C-site). This is discussed further in Section V.B. Caution should be exercised, however, in extrapolating results from one transferrin to another, however, because sequence differences are likely to alter these weak, secondary binding sites. 4 . Conformational Differences Associated with Anion Binding

Spectroscopic studies have consistently demonstrated the existence of multiple conformational states for the metal sites in transferrins, especially when using metal ions other than Fe3+ and anions other The differences are not necessarily related to intrinsic than CO:-. geometrical differences between the two sites in each molecule, but also reflect changes dependent on pH, the nature of the synergistic anion, or salt effects. For Co2+-substitutedovotransferrin, for example, not only are the N- and C-sites distinguishable by CD spectroscopy, when oxalate is the anion, but 'H NMR spectra reveal the existenceof conformers(139). For V02+-substitutedtransferrin, the EPR spectra were examined using 16 different anions (185),and the resultant spectra could be grouped into two classes, A and B, which were anion dependent. Anions with one carboxylate and a nonionized electron donor group L gave only class B spectra, whereas dicarboxylate anions gave both class A and B spectra. For the latter anions, transition between class A and class B spectra was associated with ionization of a protein group of pK -10.0. A pH-dependent change is also seen in the EPR spectrum of Cu2+substituted ovotransferrin, with carbonate as the associated anion, this time associated with ionization of a group ofpK -9.5 (157).EPR spectra of monoferric transferrins have also shown that each site (N or C) exhibits two types of spectrum and that the equilibrium between the two is affected by added NaCl (191); this equilibrium is presumably the cause of the salt-induced EPR perturbations noted by Folajtar and Chasteen (195). Some tentative conclusions about the nature of these conformational differences may be drawn from the crystallographic studies of Cu2+ and oxalate-substituted lactoferrins (26,192,193). Anions which gave class A spectra with V02+-substitutedtransferrins are those that can

440

E. N. BAKER

interact with the “essential” arginine as in Fig. 26, i.e., dicarboxylate anions. It may therefore be the ionization of this arginine that determines the conformer, perhaps by determining the degree of asymmetry in the bidentate anion binding. In the Cu2+complexes, the geometry in the N-lobe of copper-lactoferrin suggests that either the anion or Tyr 92 is protonated (Fig. 21), and it may be the loss of this proton at higher pH that causes the change to a bidentate carbonate, as seen in the C-lobe, and gives the EPR change seen for Cu2+-ovotransferrin (157). In general, given that symmetric bidentate, asymmetric bidentate, and even monodentate anion configurations are seen in the various lactoferrin structures, it seems likely that it is changes of this nature that are detected spectroscopically.

D. DIFFERENCESBETWEEN THE Two SITES Proteins of the transferrin family share a common evolutionary history, which has resulted in the presence of two homologous halves to each molecule (Sections 1II.A and III.B.l). Except for the two outliers, melanotransferrin and M. sextu transferrin, each also has two metal binding sites. This raises a number of questions, which have been the subject of much debate over the years (1,3,16,17). Why are there two sites? How similar are they, and do the differences between them have any functional or physiological significance?Is there any cooperativity between them?

1. Structural Comparison The four metal-binding amino acid residues (2 Tyr, 1Asp, 1His) are present in both N- and C-sites of all transferrins so far sequenced, apart from melanotransferrin and the insect proteins (Table 111).The same is true of the anion-bindingArg and Thr residues, and the residues at the N-terminus of the anion-binding helix are also strongly conserved. Superposition of the 81 common atoms of these residues, plus metal and anion, shows that their rms deviation in the N- and Csites of diferric human lactoferrin is only 0.3 A. This close structural similarity is reflected in their spectroscopic properties. Where these have been compared, with the “physiological” Fe3+ and C032- ions bound, they are so similar as to be virtually identical (107, 56, 199). Nevertheless, there are a number of factors that can potentially lead to inequivalence in properties: (i) Outside the immediate binding site there are sequence differences between the two lobes, e.g., in the basic residues behind the metal site,

TRANSFERRIN STRUCTURE AND REACTIVITY

44 1

in the residues that line the binding cleft beyond the “essential” arginine (Fig. 71, and in the pattern of disulfides. (ii) The “front-to-back”packing of the two lobes (Fig. 3) means that the two binding clefts have different environments with respect to the molecule as a whole (the N-terminal cleft is more exposed and accessible). (iii) The fact that each binding site is created by closure of two proteins domains over metal and anion and that there is considerable “empty space” in the interdomain cleft (i.e., filled only by solvent) gives potential for 3D structural differences, especially when different metal ions and anions are bound. 2 . Differences in Properties

Where chemical or physical differences can be detected between the two sites, there remains the problem of distinguishing which site is which. For serum transferrin this is helped immensely by the ability to prepare monoferric forms, loaded in either the N- or C-site (198, 2001, and to be able to separate them by electrophoresis in 6 M urea, the Makey-Seal method (201).This enabled the so-called A and B sites, differentiated in earlier studies, to be identified with the C-and N-terminal sites, respectively (2021. Comparisons of the diferric proteins with N- and C-loaded monoferric transferrins or (more recently) recombinant half-molecules have by now revealed a number of inequivalences. Both kinetic and thermodynamic effects differentiate the two lobes of transferrins. Aisen et al. (107) have shown that C-terminal site of transferrin binds iron more strongly than the N-terminal site, with their effective binding constants differing by a factor of about 20. The C-terminal site also appears to be the more strongly binding site for other metal ions, for example, in lanthanide binding (149, 150).Iron release also differs, with the rate of iron release being faster for the N-terminal site (108). These two effects, tighter binding and slower release from the C-lobe, may be linked to its lesser flexibility (85),as seen in thermodynamic measurements (108)and inferred from the “one opedone-closed” apolactoferrin structure (80, 82) (see Section III.B.5). The reduced flexibility of the C-lobe may arise from the presence of an extra disulfide bridge [483-677, lactoferrin numbering, or number 7 in the nomenclature of Williams (8711. This disulfide, which has no equivalent in the N-lobe of any transferrin, adds an extra constraint between the two domains of the C-lobe (Fig. 10). Predictions (80) that it would inhibit opening of the C-lobe have been born out by the lesser opening of this lobe seen in the fully open apolactoferrin structure (109).

442

E. N. BAKER

The two sites also differ in their pH stability towards iron release. Experiments on serum transferrin showed that one site loses iron at a pH near 6.0, and the other at a pH nearer 5.0 (203,2041, giving a distinctly biphasic pH-induced release profile (Fig. 28). The acid-stable A site was later shown to be the C-terminal site (202).It is this differential response to pH, together with kinetic effects (below),that enables N-terminal and C-terminal monoferric transferrins to be prepared (200). Although the N-terminal site is more labile, both kinetically and to acid, the reasons are not necessarily the same; the acid stability may depend on the protonation of specific residues (Section V.B) and is likely to differ somewhat from one transferrin to another in response to sequence changes. The biphasic acid-induced release of iron seen for transferrin is not shared by lactoferrin. Although biphasic release from lactoferrin, in the presence at EDTA, has been reported (2051, under most conditions both sites release iron essentially together at a pH(2.5-4.0) several units lower than that for transferrin (Fig. 28). The two sites (in transferrin, at least) also show differences in iron loading behaviour. In uitro, when Fe3+is added as a chelate complex, there are differences in which site is preferentially loaded, depending on the nature of the chelate ligand; these differences are apparently kinetically determined and differ from one transferrin to another (17).

P

5 n 4)

LL

hp

i

"

2.0

3.0

4.0

5.0

6.0

7.0

8.0

PH FIG.28. The pH dependence of iron release from human serum transferrin (TO, human lactoferrin (Lf),and the recombinant N-terminal half-molecule of human lactoferrin (LfN).Also shown is a plot (dashed line) for the release of cerium from Ce4+-substituted lactoferrin, demonstrating the increased difference between the two sites for metal ions other than Fe3+.

TRANSFERRIN STRUCTURE AND REACTIVITY

443

In uiuo, a study of fresh serum showed that the N-site of human transferrin was more highly occupied, the average distribution in 22 samples being 39% apo-Tf, 23% Fe,-Tf, 11%Fe,-Tf, and 27% Fe2-Tf(206). 3. Metal and Anion Substitution

Differences between the two sites become more pronounced for metal ions other than Fe3+ and anions other than C032-. The differences are most pronounced for larger metal ions such as lanthanides. For transferrin some of the larger lanthanides appear to bind in only one of the two sites (Section IV.B.3), and for lactoferrin, although binding occurs in both sites, the second metal ion binds much more weakly, as shown by the curvature of the UV difference titration graph (Fig. 18); the biphasic release of Ce4+from lactoferrin contrasts with that of Fe3+ (Fig. 28). Even metal ions of the first transition series, of similar size to Fe3+,enhance the differences between the two sites. When Cr3' is bound to either transferrin (134) or lactoferrin (1541, different EPR signals are seen for the two sites, and one C P ion is much more readily displaced by Fe3+than the other. Likewise, the EPR spectra of V02+substituted transferrin indicate different metal configurations in the two sites (2071,as do 13CNMR studies of Co2+-substitutedovotransferrin (139).In these cases one metal ion is also released much more readily than the other as the pH is lowered. The crystal structure of copper-lactoferrin (26)shows the kind of differences that may occur. In one site the coordination geometry is six-coordinate, distorted octahedral, whereas in the other it is fivecoordinate, square pyramidal. One could suggest that the sites are and that there is an element optimized for the binding of Fe3+and C0:of misfitting when a different metal ion, with different size, stereochemical requirements, or both, is bound. The sites can adjust, with small movements, but these are different in the two sites. O:are Distinct differences are also seen when anions other than G used. The crystal structure of oxalate-substituted diferric lactoferrin shows differences in the anion binding in the two sites; in the C-site the oxalate is symmetric bidentate, whereas in the N-site it is asymmetric (193). When Cu2+is the metal ion the oxalate binding differences become even more pronounced. Copper-transferrin binds oxalate only in its N-terminal site (91). Copper-lactoferrin and copper-ovotransferrin each bind two oxalate ions but binding occurs preferentially in the C-lobe (157,192).These different affinitiesmean that hybrid complexes can be prepared with oxalate in one site and carbonate in the other (92, 157, 192).The use of oxalate as synergistic anion gives rise to spectroscopically distinct sites for other metal ions also (1711.

444

E. N.BAKER

The X-ray structural studies on lactoferrin show that it is not simply a question of how much room there is for a larger anion or cation in a given site. The N-terminal site in lactoferrin apparently has more room than the C-terminal site, yet it is the C-terminal site that is preferentially occupied by oxalate (I92); perhaps the explanation is that the more favorable square pyramidal copper geometry in the Nsite (with monodentate anion) makes it less amenable to substitution of oxalate. In general, the two sites have enough flexibility that the precise structure adopted depends on the particular metal and anion, emphasizing the synergistic relationship between the two and making the result of a given substitution rather hard to predict. 4 . Functional Aspects

The small but significant difference in iron release from the two sites of transferrin led Fletcher and Huehns (208) to suggest that they might have different biological functions. It was suggested that one site might be involved in iron transport and release and the other, more as a storage site (for antibacterial or iron defense purposes). Although the idea has remained controversial (e.g., see discussion in Ref. 31, several recent observations have led to renewed interest. The apolactoferrin structure (80) suggested a distinct difference in flexibility between the two lobes, which should also apply to transferrin. Studies of transferrin-receptor interactions (209) have also shown that the receptor specifically acts on the C-lobe, prying it open to release the iron, whereas the N-lobe loses iron by the long-proposed acid-mediated release (20). This neatly explains the observation that circulating transferrin has more iron in its N-lobe, whereas the less facile C-lobe release would have been expected to lead to a buildup in the C-site. These observations come together with the evolutionary comparisons, which show that the N-lobe is highly conserved, through all species, whereas the C-lobe has become diversified. In serum transferrins, ovotransferrins, and lactoferrins it releases iron less readily than the N-lobe, because of its lesser flexibility, whereas in melanotransferrin and hornworm transferrin it no longer binds iron at all. Perhaps C-site binding has only remained where a receptor mechanism exists to extract iron from this site? Finally the question of whether there is any cooperativity between the sites remains to be addressed. Although evidence of cooperativity from solution studies has mostly been equivocal (31,there are certainly structural interactions between the two lobes, involving helices from each (78, 851, and studies of a half-molecule fragment of lactoferrin have shown that separating the N-lobe from the C-lobe gives it altered

TRANSFERRIN STRUCTURE AND REACTIVITY

445

properties of iron release (49). Thermodynamic studies have shown that binding at one site is signaled to the other, presumably through changes in interlobe interactions (210) and it seems likely that the binding properties of each lobe are modified by the presence of the other. V. Mechanisms of Binding and Release

The most striking feature of transferrin chemistry is that iron is bound with extraordinary avidity, yet it can be released without any denaturation and the protein can be recycled through many cycles of uptake and release. The mechanisms by which this is done are of fundamental importance to understanding biological transport processes. A. UPTAKEOF IRON 1. Mechanism of Binding

In uiuo uptake of iron by transferrins usually involves its addition as a ferric-chelate complex, to prevent hydrolytic attack on the ferric ion (2111. Complexes such as ferric citrate and ferric nitrilotriacetate are commonly used. Under these conditions, kinetic schemes for the uptake of iron by transferrins have identified five steps in the formation of the specific metal-anion-transferrin ternary complex (120). These may be summarized as follows. 1. Binding of the (bilcarbonate anion to apotransferrin. 2. Detachment of one or more ligands from the added metal chelate.

3. Formation of a quaternary transferrin-anion-metal-chelate complex. 4. Loss of the chelate ligands(s1. 5. Conformational change to the final specific transferrin complex.

Evidence that the anion binds first comes from kinetic data (119) and from spectroscopic results, in which both 'H NMR (118) and UV difference (177) spectra indicate that the anion binds to the apoprotein. Strong support comes from the 3D structural data; the positive charge at the anion site should deter metal binding until it is neutralized by a suitable anion ( 78,85). Suggestions that nitrilotriacetate transiently occupies the anion site when iron is added as a Fe3+-NTA complex (212) may imply that the early steps can vary depending on the form of the added iron, but the key point probably remains that the anion site must be occupied as a first step. Spectroscopic evidence for the

446

E. N. BAKER

quaternary complex envisaged in Step 3 has been obtained from studies using ferric-acetohydroxamate in iron uptake experiments (120). The above mechanism is totally consistent with the crystallographic results from the various forms of lactoferrin and transferrin (Section 1II.B).These lead to a structural model of binding shown pictorially in Fig. 29. In the first step the synergistic anion (usually carbonate) is bound in the specific site on domain 2 of each lobe. Binding may be preceded by electrostatic attraction from the exposed helix N-termini and several basic sidechains in the open interdomain cleft. With the anion bound, four of the six iron ligands are in place on

+o

+a

FIG.29. A structural model of the steps involved in the in uitro uptake of iron by transferrins, shown for one lobe. (0)Iron; (A)carbonate; Y, Tyr ligands; H, His ligand; D, Asp ligand; (0) chelate ligands. The positive charge at the anion site is due to the helix N-terminus and the Arg side chain. (Note that this is for the case in which Fe3+ is added as a chelate complex.)

TRANSFERRIN STRUCTURE AND REACTIVITY

447

domain 2 (two carbonate oxygens and two Tyr sidechains); the metal then binds to these groups, possibly with some of its chelating ligands still attached, to give the quaternary complex. A model for such an intermediate is provided by the 18-kDa domain 2 fragment of duck ovotransferrin, whose crystal structure has been determined by Lindley et al. (76). In this structure the iron atom is bound to the bidentate carbonate ion and both Tyr residues, with the remaining two coordination sites occupied by a non-protein ligand, possibly a glycine molecule (Fig. 14). The final step in binding involves the closure of the two domains over the metal ion. Any remaining chelate groups are expelled as the metal ion binds t o the Asp and His ligands t o complete its coordination. The closed configuration is locked together by the Asp ligand, which plays a critical role in the metal-bound structure (78). Not only does the Asp carboxylate bind to the metal ion, but also it is involved in a strong hydrogen bonding interaction between the two domains. Its nonligated carboxylate oxygen atom receives a hydrogen bond from the NH group of residue 122 (466 in the C-lobe) in domain 2 as well as from the NH of residue 62 (397) in domain 1.The strength of these interactions is probably enhanced by the fact that both NH groups are at positively charged helix N-termini, helix 3 and helix 5 (Fig. 9a), and their importance is emphasized by the fact that when the Asp ligand is mutated to Ser which has only a single hydroxyl group, no stable closed structure appears to be formed (106). 2 . Importance of Dynamics

Protein dynamics clearly plays a crucial role in metal binding and release. With respect to metal binding, the open structure described in Section III.B.5 should not be taken to imply that in the absence of a bound metal ion the cleft is always open. In fact the closed but metalfree C-lobe seen for one form of apolactoferrin (80) suggests that in the absence of a metal ion very little energy separates the open and closed states, and in fact there may be a dynamic equilibrium between them (82); at the very least the closed configuration may now and again be sampled. Similar conclusions have been drawn for bacterial binding proteins for which closed but ligand-free structures have also been observed (213). The importance of this model is that in the intermediate in which the metal ion is bound to domain 2 (step 3 in Fig. 29) it would be some 8 to 9 away from the Asp and His ligands on domain 1, as judged by the extent of opening of the apolactoferrin (80) and C-terminal monoferric transferrin (110) structures. How then does it “find” these

448

E. N.BAKER

two ligands to complete its coordination? Only if the dynamics of an equilibrium allows it to explore the closed structure, at which time the ligands will be able to lock on to the metal. Such dynamics are common in proteins, especially where the movement of domains (214)or flexible “lids” or “flaps” (215) are concerned. Finally, the Asp ligand has been described as a “trig er” associated with domain closure (106).However, at a distance of 9 from a metal ion bound to domain 2 it is hard to see how it could induce closure. Rather, it should be seen as a lock that holds the closed structure in place once the protein dynamics have brought the domains close together.

1

B. RELEASE OF IRON Iron release can be stimulated by a number of factorsthat can operate individually or together. In uitro, these include reduction of Fe3+to the much more weakly bound Fe2+,the use of competitive chelators, and the acid liability of the two sites, which results in iron release at low pH. In uiuo, receptor interactions are of fundamental importance. Added to these are the modulatory effects of ionic strength (216)and nonsynergistic anions such as C1- and C10,- (198). 1 . Kinetics

Most kinetic studies of iron release have focused on pathways involving the use of chelate ligands such as EDTA (2171, pyrophosphate (218-2201, phosphonates (220,221 1, catecholates (108,2161, hydroxamates (1201, and nitrilotriacetate (2211. In many cases, simple saturation kinetics are observed, and interpreted in terms of the formation (120,122 ). The of a quaternary complex, 1igand-Fe-tran~ferrin-CO~~failure t o observe this complex spectroscopically [in contrast to iron uptake studies (120)] has been explained in terms of a rate-limiting conformational change, giving a basic three-step mechanism, which is essentially the reverse of that given for iron uptake (Section V.A.l). 1. A rate-limiting conformational change of the diferric protein from closed to open configuration. 2. Rapid attack of the chelator to give a quaternary complex. 3. Rapid decay of the quaternary complex to products. The kinetics are, however, considerably more complex. Both pH and salt affect the two sites differently (Section IV.D.2), so that iron release kinetics are very different at pH 6 compared with those at pH 7.4, for example. The kinetics are also dependent on the particular chelator

TRANSFERRIN STRUCTURE AND REACTIVITY

449

used; this is not surprising because most of the ligands used are anionic, many of them able to bind quite strongly to transferrins even in the absence of iron (177). There is now accumulating evidence of more than one pathway for release. Harris and coworkers (221,223,224) identified two parallel pathways, operative at either site, one being the saturation pathway envisaged above, the other being first order with respect to chelate concentration.Their relative importance can depend both on the nature of the chelator and on concentrations. Using pyrophosphate, iron removal from the N-site follows only the saturation pathway, whereas for the C-site both pathways operate (223);using nitrilotris(methy1enephosphonate) (NTP)the saturation pathway is favored,but using nitrilotriacetate (NTA)the first-order pathway operates at both sites (2211. The first-order pathway may be associated with interactions involving the chelate anion; it could either substitute for carbonate in the specific site (223) or bind to some other allosteric effector site, such as that occupied by C1- or C10,- (219). Much attention has been paid to the significance of salts (217) and nonsynergistic anions (195,198) in promoting or modulating iron release, especially given the important observation that iron release extrapolates to zero as the ionic strength of the medium nears zero (216). A variety of anions accelerate the first-orderpathway (2201,with Clodbeing the most active species; this has been attributed to the presence of cationic groups near the metal binding site, which form a specific effector or allosteric anion-bindingsite (219,220). The complex kinetics may result from competition between chelate anions and effector anions for such sites. Egan et al. (225) have further analyzed the release of iron from C-terminal monoferric transferrin to pyrophosphate in terms of the existence of a kinetically significant anion binding (KISAB) site. Here it is envisaged that two pathways operate, with either pyrophosphate or an added anion occupying this site. Comparisons of release from free and receptor-complexed transferrin also show that the release-promoting effects of the receptor and of the anion (in this case C1- are independent of each other (226). Many studies have noted weak cooperativity between the sites during iron release (3). One recent analysis used mixed-metal transferrins, with kinetically inert Co3+in one site and Fe3+in the other (221,224). With pyrophosphate, release of iron from the C-site was accelerated by the presence of a metal in the N-site, but no corresponding effect was seen for iron release from the N-site. The cooperative effects were also weaker and somewhat different for different chelators (221). Finally it is important to realize that most studies of iron release

450

E. N. BAKER

have focused on human serum transferrin. Many of the finer details may be dependent on the interactions of chelators, salts, etc., with residues that are within the binding cleft but outside the immediate iron site; these residues tend to vary from one species to another and from one transferrin to another, and it is likely that kinetic details will also.

2 . Structural Aspects of Iron Release The conformational change from closed to open configuration is a key feature of models for iron release. The nature of this change can now be inferred from crystallographic and solution studies (Section III.B.5). What is less clear is how the conformational change is triggered. In uiuo, receptor binding is clearly involved, but pH and salt effects can also play a part. In uitro, reduction of the pH is in itself sufficient. In attempting to understand iron release in structural terms one must look for potential protonation and anion binding sites and seek more knowledge of the interactions made with transferrin receptors. The effect of pH differs for the two sites of transferrin and differs between transferrin and lactoferrin. When titrated with acid, in the absence of chelators, serum transferrin loses iron over the pH range 6.0 to 4.0; release is biphasic (Fig. 281, with iron lost from the more acid-labile N-lobe site first (Section IV.D.l). Lactoferrin, on the other hand, is distinctly more stable in acid, with release occurring 2 pH units lower, over the pH range 4.0 to 2.5, and the two sites losing iron essentially together. Several explanations for the effect of pH on iron removal have been put forward. Protonation of the carbonate ion could cause repulsion between it and the anion-binding Arg residue (121 in the N-lobe, 465 in the C-lobe), or promote a change from bidentate to monodentate coordination, as seen in the N-lobe of copper-lactoferrin (26). Either effect could then be the first step in the breakup of the Fe3+-transferrin complex. An alternative site where protonation could stimulate iron release is at the back of the iron site, in the hinge region. It is here that distinct differences between lactoferrin and transferrin, involving ionizable residues, are found (Fig. 30). In the N-lobes of both rabbit transferrin (81) and chicken ovotransferrin (771, a pair of lysine residues (206 and 296, transferrin numbering) are in hydrogen bonded contact, implying that one is in its neutral form; protonation of this lysine would break this interdomain interaction and could destabilize the closed structure. The pair of lysines has been referred to as a “dilysine trigger” ( 77). In the C-lobe of transferrin a different combination of charged residues is found (a salt bridge Lys * * * Asp * - Arg), which

TRANSFERRIN STRUCTURE AND REACTIVITY

45 1

.. ..

SER

FIG. 30. Residues at the back of the iron site, near the hinge region, that may be implicated in the stimulation or modulation of iron release. The interactions present in human lactoferrin and rabbit transferrin are compared. Where the conformations are different, lactoferrin residues are shown with solid bonds, transferrin, with open bonds. Where the residues differ in identity or number, those for transferrin are in parentheses.

would be less easily protonated. In the N-lobe of human lactoferrin the interactions are different again; one Lys is changed to Arg (Table 111) and a conformational difference leads to a Glu - - - Lys ion pair in place of the Lys . - .Lys pair. This could certainly account for the greater acid stability of the lactoferrin N-lobe site. (The situation may not be so simple, however, because the sequences of both bovine and porcine lactoferrins have both lysines.) Crystallographicand mutagenesis studies will be required to disentangle these effects. Studies of half-molecule fragments also suggest that the region a t the back of the iron site, near the hinge in the interdomain connecting strands, could be the site where protonation stimulates release. The recombinant N-terminal half-molecule of human lactoferrin releases iron over the pH range 6.0 to 4.0 (Fig. 281, approximately 2 pH units higher than that of intact lactoferrin, but very similar t o transferrin (49,205).The crystal structure shows that this decreased acid stability is associated with the loss of stabilizing contacts normally made by the C-lobe, leading to unwinding of a helix at the back of the iron site and

452

E. N. BAKER

increased solvent exposure of both this region and the hinge (75). A proteolytic fragment of lactoferrin that lacks even more of the structure in this region is correspondingly more acid labile (227). Potential anion-binding sites that could stimulate iron release have been discussed in Section IV.C.3 and are shown in Fig. 27. Binding to sites near the hinge, such as the Lys . . . Lys pair in the N-lobe of serum transferrin, has the potential either t o promote domain opening, by disrupting interdomain interactions or the nearby hinge, or to perturb the iron site via the iron ligands; the Lys * Lys pair in transferrin and Arg 210 in lactoferrin are hydrogen bonded to the Tyr ligands (75, 77, 78). Binding to the “essential” Arg residue could disrupt the metal-synergistic anion interaction, leading to a change to monodentate coordination or complete displacement of the synergistic anion. Defined structural pathways for iron release may exist. If some chelate anions bind to cationic groups on the protein first, as suggested (2211, the metal could be passed to this site en route to the outside. If binding was to the “essential” arginine, which also helps hold the synergistic anion, the metal might just transfer from one to the other. An intriguing observation concerns interactions with cyanide. The iron in transferrin is usually high-spin Fe3+,but it can be converted to low spin in the presence of CN- (228); flash-freezing then traps an intermediate in which CN- is exchanged for some of the normal transferrin ligands in the C-lobe. X-ray absorption studies of this intermediate indicate that it involves at least two Tyr ligands, but that at least one of them may be different from those normally ligated (81). This could implicate a Tyr-mediated pathway. Studies of the transferrin receptor indicate that it acts preferentially on the less flexible C-lobe, to stimulate release from this site (209), and that the receptor and anion-binding effects are independent (226). The interdomain strands that contain the C-lobe hinge are highly exposed to the external environment and are rich in charged residues; they could be involved in binding the receptor, simple anions (such as C1-), or both. The problem at present is that so many potential sites exist. What is required is detailed knowledge of transferrin-receptor interactions, either through the crystallographic studies that have already been initiated (229)or by mutagenesis. VI. Recombinant DNA Studies

The past six years have been an explosion of new results from X-ray crystallographic studies, which have added a new dimension to our

TRANSFERRIN STRUCTURE AND REACTIVITY

453

understanding of transferrin chemistry. The next few years should increasingly see another powerful approach, that of recombinant DNA studies, brought to bear. These techniques allow the production of mutant transferrins, in which single, selected amino acids are changed. Alternatively, chimeric transferrins in which part of one molecule is substituted into another can be constructed. Mutagenesis can be carried out either on the whole transferrin molecules or on half-molecules, because half-molecules of defined length can be made. The first successful expression of any transferrin was reported in 1990 when Funk et al. (48)isolated the cDNA for human serum transferrin and introduced a stop signal following the codon for Asp 337; the resulting DNA construct thus coded for the N-terminal half-molecule. This recombinant human transferrin N-terminal half-molecule, hTf/2N, was expressed in a cell culture system incorporating baby hamster kidney cells, after unsuccessful attempts at expression in Escherichia coli. Similar methods have also been used to obtain expression of full-length human serum transferrin (230) and both the N-terminal half-molecule (49) and the whole molecule (231) of human lactoferrin. The use of an animal cell culture expression system has advantages in ensuring correct folding of the recombinant proteins and allowing glycosylation to occur, but is more laborious and gives lower levels of expression than are usually possible withE. coli. All the same, expression levels of 20-40 mg per liter of culture medium have been reported (48,49),ample quantities for characterization. (The recombinant proteins are secreted into the culture medium.) A recent report of the expression of human serum transferrin in E.coli (232)suggests that higher levels may be attainable although little characterization of the protein obtained has yet been carried out. The recombinant whole molecules are both expressed in glycosylated form, although the glycosylation patterns differ from the proteins isolated from natural sources. The recombinant human transferrin binds to receptors both in its glycosylated form and as a nonglycosylated mutant, showing that the carbohydrate is not required for receptor binding (230).Recombinant human lactoferrin shows identical spectroscopic properties and shows an identical profile of pH-dependent iron release when compared with human milk lactoferrin (231). The half-molecules differ somewhat, although this is not unexpected, because the properties of the whole molecules are an amalgam of those of their two slightly different sites. Moreover, the lactoferrin halfmolecule shows that its iron release properties are changed as a result of the loss of interactions from the other lobe (Ref. 49; see also Section V.B.2 and Fig. 28). The visible A,, for the transferrin half-molecule

454

E. N . BAKER

is increased from 465 nm (native transferrin) to 473 nm (2331,whereas that for the lactoferrin half-molecule is reduced from 465 to 454 nm. The reasons are not clear, but presumably have to do with the fine detail of the two metal sites, at a level that probably even high resolution X-ray analyses may not explain. A number of site-specificmutants have already been prepared, both of transferrin (233)and of lactoferrin (234).In all cases the mutations have been made in transferrin or lactoferrin half-molecules and the targets have been amino acids in and around the metal binding sites. Characterization has been limited so far to their visible spectra and some measures of iron binding properties; for lactoferrin the pH dependence of iron release has been determined (2341, whereas for transferrin the strength of iron binding has been inferred from the migration of the mutant proteins on urea gels (233).In neither case have binding constants yet been determined. Already some intriguing observations have resulted, however, giving a glimpse of what may be to come with fuller characterization. Most mutations alter the value of A,, (Table VIII). This is true even of residues outside the immediate binding site, such as Lys 206 and His 207 in transferrin (233), showing that even changes some distance away can perturb the metal site. Lys 206 is one of the two lysines that are hydrogen bonded together behind the iron site and that are potential sites for protonation and salt effects (Section V.B.2).Interestingly, mutations of Lys 206 to Gln and of His 207 to Glu, both mutations that reduce the positive charge, appear to increase the strength of iron binding. Mutation of the Asp ligand to Ser, as in the C-lobe of melanotransferTABLE VIII PROPERTIES OF RECOMBINANT TRANSFERRINS~ Transferrim* N-lobe (Tf,) Asp 63 Ser-TfN Asp 63 Cys-TfN Gly 65 Arg-TfN Lys 206 Gh-TfN His 207 Glu-TfN

,,,A 473 420 440 468 460 484

Iron binding Strong Weak Weak Weak Strong Strong

Lactoferrins N-lobe (LfN) Asp 60 Ser-LfN Arg 121 Ser-LfN Arg 121 ASp-LfN Asp 60 Ser Arg 121 Ser

ILfN

hmax

Iron release (pH)

454 434 454 472

5.5-4.0 7.0-5.0 5.5-4.0 >7.0

472

>7.0

Notation for mutants: Asp 63 Ser means that Asp 63 is mutated to Ser. Asp 63 in transferrin corresponds to Asp 60 in lactoferrin, and similarly Gly 65, to Gly 62; Arg 124, to Arg 121; Lys 206, to Arg 210; and His 207, to Glu 211.

TRANSFERRIN STRUCTURE AND REACTIVITY

455

rin, does not abolish iron binding, but reduces A,, considerably (the protein is yellow) and weakens binding considerably. The transferrin mutant (Asp 63 Ser) loses iron on urea gels, traveling as the apoprotein (2331, and low-angle X-ray solution scattering studies (106)suggest that it does not form the usual closed structure when iron is bound. The lactoferrin mutant (Asp 60 Ser) loses iron below pH 7, considerably more readily than the wild-type protein. Mutations of Asp 63 to Cys in transferrin, and of the nearby Gly 65 to Arg, have similar effects, again possibly inhibiting domain closure. Mutation of the anion-binding Arg 121 to Ser in lactoferrin does not alter Amax or significantly perturb the pH dependence of iron release, implying that the Arg residue is not essential for anion binding and arguing against a repulsion between it and a protonated anion as an important factor in iron release. Intriguingly, the double mutation Asp 60 to Ser and Arg 121 to Ser in the lactoferrin half-molecule (matching mutations in the C-lobe of melanotransferrin) weakens but does not abolish iron binding. VII. Concluding Remarks

Historically, our understanding of transferrin chemistry has depended to a large extent on spectroscopic and other physicochemical approaches. Crystallographic studies over the past few years have added a new structural dimension. The iron ligands have been established definitively and the open and closed forms of the protein have been defined. Other groups in proximity to the metal and anion sites, with possible modulatory roles, can also be identified. It is thus an opportune time to reexamine the complementary role of spectroscopy. For example, UV difference spectra principally monitor the binding of the Tyr ligands to a metal ion, and the same is true, to a first approximation, of the visible charge transfer spectra. Because the Tyr ligands are associated with only one domain (domain 21, these techniques indicate only binding to this domain and do not show whether an open or a closed structure is adopted for a particular metal ion (or anion). Binding to the ligands of the other domain may be indicated by techniques sensitive to the His ligand, e.g., NMR. Thus combinations of approaches, including techniques such as low-angle solution scattering, and the ultimate power of X-ray crystallography can be used to address important questions. Does binding a particular metal ion or anion result in an open or closed structure? If so, what are the implications for the transport of such species?

456

E. N. BAKER

A second major challenge and opportunity for bioinorganic chemists is to use the power offered by recombinant DNA technology. To be able to compare spectroscopicand other results from mutants in which single amino acids have been changed gives a unique possibility to really understand such complex systems. This will also require complementary X-ray structure analyses in order to disentangle the effects of structure and chemistry. (Interpretations of mutagenesis experiments are seldom straightforward.) At the functional and physiological levels, perhaps the greatest need is a better understanding of transferrin-receptor interactions. Mutagenesis experiments again will contribute to this, but the crystal structure of a transferrin-receptor complex would be the ultimate prize. Although this chapter has concentrated on structure, as well as metal and anion binding properties, in the end it is because of their physiological roles, actual or possible, that we study transferrins and find such fascination in their chemistry. ACKNOWLEDGMENTS

I thank Heather Baker, Catherine Day, Rob Evans, Peter Lindley, Clyde Smith, John Tweedie, and Harmon Zuccola for access to their unpublished data; Clyde Smith for help with illustrations; and Heather Baker and Andrew Brodie for their critical reading of the manuscript. I owe a particular debt to Peter Lindley, Rob Evans, and members of the Birkbeck College transferrin group for the free and rewarding collaborative interactions we have always had and to Phil Aisen for being a constant source of inspiration to all in the transferrin field. I gratefully acknowledge financial support for my own research on lactoferrin, over a number of years, from the U S . National Institutes of Health (Grant HD-20859), the Wellcome Trust, the Health Research Council of New Zealand, the New Zealand Dairy Research Institute, and Massey University. I am grateful also to the Howard Hughes Medical Institute for an International Research Scholar award, which has helped support this work.

REFERENCES 1. Aisen, P., and Listowsky, I., Annu. Rev. Biochem. 49, 357 (1980). 2. Neilands, J. B., Annu. Rev. Biochem. 50, 715 (1981). 3. Brock, J. H., in “Metalloproteins” (Harrison, P. M., Ed.), Vol. 2, pp. 183-262. MacMillan, London, 1985. 4. Jamroz, R. C., Gasdaska, J. R., Bradfield, J. Y.,and Law, J. H., Proc. Natl. Acad. Sci. USA 90, 1320 (1993). 5 . Masson, P. L., Heremans, J. F., and Dive, C., Clin. Chirn. Acta. 14, 735 (1966). 6. Baggiolini, M., DeDuve, C., Masson, P. L., and Heremans, J. F., J.Exp. Med., 131, 559 (1970).

TRANSFERRIN STRUCTURE AND REACTIVITY

457

7 . Williams, J., Biochem. J. 83,355 (1962). 8. Brown, J . P., Hewick, R. M., Hellstrom, I., Hellstrom, K. E., Doolittle, R. F., and Dreyer, W. J., Nature 296, 171 (1982). 9 . Schade, A. L., Reinhart, R. W., and Levy, H., Arch. Biochem. Biophys. 11,9(1946). 10. MacGillivray, R.T. A,, Mendez, E., Sinha, S. K., Sutton, M. R., and Brew, K., Proc. Natl. Acad. Sci. USA 79,2504 (1982). 11. Metz-Boutigue, M. H., Jolles, J., Mazurier, J., Schoentgen, F., Legrand, D., Spik, G., and Jolles, P., Eur. J.Biochem. 145,659 (1984). 12. Baker, E. N., Baker, H. M., Smith, C. A,, Stebbins, M. R., Kahn, M., and Hellstrom, I., FEBS Lett. 298,215 (1992). 13. Bartfeld, N. S., and Law, J. H., J.Biol Chem. 265,21684 (1990). 14. Martin, A. W., Huebers, E., Huebers, H., Webb, J., and Finch, C. A., Blood 64, 1047 (1984). 15. Huebers, H. A., Huebers, E., Finch, C., and Martin, A. W., J. Comp. Physiol. B 148, 101 (1982). 16. Chasteen, N. D., Adv. Znorg. Biochem. 5, 201 (1983). 17. Aisen, P., and Harris, D. C., in “Iron Carriers and Iron Proteins” (Loehr, T. M., Gray, H. B., and Lever, A. B. P., Eds.), pp. 239-371. V. C. H., New York, 1989. 18. Baker, E. N., in “Perspectives on Bioinorganic Chemistry” (Dilworth, J. R., Hay, R. W., and Nolan, K. B., Eds.), Vol. 2, pp. 1-40. J. A. I. Press, 1993. 19. Baldwin, D. A,, Jenny, R. R., and Aisen, P., J.Biol. Chem. 259, 13391 (1984). 20. Octave, J.-N., Schneider, Y . J . , Trouet, A,, and Crichton, R. R., Trends Biochem. Sci. 8,217 (1983). 21. Davidson, L. A,, and Lonnerdal, B., Am. J.Physiol. 254, 580 (1988). 22. Hu, W.-L., Mazurier, J., Sawatzki, G., Montreuil, J., and Spik, G., Biochem. J.249, 435 (1988). 23. Martin, R. B., Savory, J., Brown, S., Bertholf, R. L., and Wills, M. R., Clin. Chem. 33,405 (1987). 24. Cowburn, J. D., and Blair, J. A., Lancet, 99 (1989). 25. Lonnerdal, B., in “Iron Nutrition in Infancy and Childhood (Stekel, A., Ed.), pp. 95-117.Vevey/Raven, New York, 1984. 26. Smith, C. A., Anderson, B. F., Baker, H. M., and Baker, E. N., Biochemistry 31, 4527 (1992). 27. Hashizume, S.,Kuroda, K., and Murakami, M., Biochirn. Biophys. Acta 763, 377 (1983). 28. Shimo-Oka, T.,Hagiwara, Y., and Ozawa, E., J.Cell. Physiol. 126, 341 (1986). 29. Mazurier, J., Legrand, D., Hu, W.-L., Montreuil, J., and Spik, G., Eur. J.Biochem. 179,481 (1989). 30. Van Snick, J. L., and Masson, P. L., J.Exp. Med. 144, 1568 (1976). 31. Birgens, H. S., Hansen, N. E., Karle, H., and Ostergaard Kristensen, L., Br. J.Haematol. 54,383 (1983). 32. Dalmastri, C., Valenti, P., Visca, P., Vittorioso, P., and Orsi, N., Microbiologica 11, 225 (1988). 33. Bellamy, W., Takase, M., Yamauchi, K., Wakabayashi, H., Kawase, K., and Tomita, M., Biochim. Biophys. Acta 1121,130 (1992). 34. Williams, J., Elleman, T. C., Kingston, I. B., Wilkins, A. G., and Kuhn, K. A., Eur. J.Biochem. 122, 297 (1982). 35. Jeltsch, J.-M., and Chambon, P., Eur. J.Biochem. 122,291 (1982). 36. Yang, F., Lum, J . N., McGill, J. R., Moore, C. M., Naylor, S. L., van Bragt, P. J., Baldwin, W. D., and Bowman, B. H., Proc. Natl. Acad. Sci. USA 81, 2752 (1984).

458

E. N. BAKER

37. Baldwin, G. S., and Weinstock, J., Nucleic Acids Res. 16, 8720 (1988). 38. Carpenter, M. A., and Broad, T. E., Biochim. Biophys. Acta 1173, 230 (1993). 39. Banfield, D. K., Chow, B. K.-C., Funk, W. D., Robertson, K. A., Umelas, T. M., Woodworth, R. C., and MacGillivray, R. T. A., Biochim. Biophys. Actu 1089, 262 (1991). 40. Moskaitis, J. E., Pastori, R. L., and Schoenberg, D. R., Nucleic Acids Res. 18, 6135 (1990). 41. Bajaj, M., Faik, P., and Evans, R. W., manuscript in preparation. 42. Stowell, K. M., Rado, T. A., Funk, W. D., and Tweedie, J. W., Biochem. J . 276, 349 (1991). 43. Pentecost, B. T., and Teng, C . T., J . Biol. Chem. 262, 10134 (1987). 44. Alexander, L. J., Levine, W. B., Teng, C. T., and Beattie, C . W., Anim. Genet. 23, 251 (1992). 45. Pierce, A,, Colavizza, D., Benaisser, M., Maes, P., Tartar, A., Montreuil, J., and Spik, G., Eur. J . Biochem. 196, 177 (1991). 46. Mead, P. E., and Tweedie, J. W., Nucleic Acids Res. 18, 7167 (1990). 47. Rose, T. M., Plowman, G. D., Teplow, D. B., Dreyer, W. J., Hellstrom, K. E., and Brown, J. P., Proc. Natl. Acad, Sci. USA 83, 1261 (1986). 48. Funk, W. D., MacGillivray, R. T. A., Mason, A. B., Brown, S. A., and Woodworth, R. C., Biochemistry 29, 1654 (1990). 49. Day, C. A., Stowell, K. M., Baker, E. N., and Tweedie, J. W., J . Biol. Chem. 267, 13857 (1992). 50. Williams, J., Biochem. J . 141, 745 (1974). 51. Williams, J., Biochem. J . 149, 237 (1975). 52. Keung, W.-M., Azari, P., and Phillips, J. L., J . Biol. Chem. 257, 1177 (1982). 53. Ikeda, H., Banuchi, J., Nakazato, K., Tanaka, Y., and Satake, K., FEBS Lett. 182, 305 (1985). 54. Williams, J., and Moreton, K., Biochem. J . 251, 849 (1988). 55. Brown-Mason, A., and Woodworth, R. C., J . Biol. Chem. 259, 1866 (1984). 56. Brock, J. H., and Anabe, F. R., FEBS Lett, 69, 63 (1976). 57. Legrand, D., Mazurier, J., Aubert, J.-P., Loucheux-Lefebvre, M.-H., Montreuil, J., and Spik, G., Biochem. J . 236, 839 (1986). 58. Magdoff-Fairchild, B., and Low, B. W., Arch. Biochem. Biophys. 138, 703 (1970). 59. Al-Hilal, D., Baker, E., Carlisle, C. H., Gorinsky, B., Horsburgh, R. C., Lindley, P. F., Moss, D. S., Schneider, H., and Stimpson, R., J . Mol. Biol. 108, 255 (1976). 60. DeLucas, L. J., Suddath, F. L., Gams, R. A., and Bugg, C. E., J . Mol. Biol. 123, 285 (1978). 61. Abola, J. E., Wood, M. K., Church, A., Abraham, D., and Pulsinella, P. D., in “The Biochemistry and Physiology of Iron” (Saltman, P., and Hegenauer, J., Eds.), pp. 27-34. Elsevier, New York, 1982. 62. Rawas, A., Moreton, K., Muirhead, H., and Williams, J.,J . Mol. Biol. 208,213 (1989). 63. Baker, E. N., and Rumball, S. V., J . Mol. Biol. 111, 207 (1977). 64. Norris, G. E., Anderson, B. F., Baker, E. N., Baker, H. M., Gartner, A. L., Ward, J., and Rumball, S. V., J . Mol. B i d . 191, 143 (1986). 65. Norris, G. E., Baker, H. M., and Baker, E. N., J . Mol. Biol. 209, 329 (1989). 66. Gorinsky, B., Horsburgh, C., Lindley, P. F., Moss, D. S., Parkar, M., and Watson, J . L., Nature 281, 157 (1979). 67. Anderson, B. F., Baker, H. M., Dodson, E. J., Norris, G. E., Rumball, S. V., Waters, J. M., and Baker, E. N., Proc. Natl. Acad. Sci. USA 84, 1769 (1987). 68. Bailey, S., Evans, R. W., Garratt, R. C., Gorinsky, B., Hasnain, S., Horsburgh, C.,

TRANSFERRIN STRUCTURE AND REACTIVITY

459

Jhoti, H., Lindley, P. F., Mydin, A., Sarra, R., and Watson, J. L., Biochemistry 27,

5804 (1988). 69. Sarra, R., and Lindley, P. F., J. Mol. Biol. 188, 188 (1986). 70. Mikami, B., and Hirose, M., J. Biochem. 108,907 (1990). 71. Day, C. L., Norris, G. E., Tweedie, J . W., and Baker, E. N., J. Mol. Biol. 228, 973

(1992). 72. Wang, Y.,Chen, J., Luo, Y., Funk, W. D., Mason, A. B., Woodworth, R. C., MacGillivray, R. T. A., and Brayer, G. D., J. Mol. Biol. 227, 575 (1992). 73. Jhoti, H., Gorinsky, B., Garratt, R. C., Lindley, P. F., Walton, A. R., and Evans, R. W., J. Mol. Biol. 200,423 (1987). 74. Sarra, R., Garratt, R., Gorinsky, B., Jhoti, H., and Lindley, P., Actu Cryst. B 46,

763 (1990). 75. Day, C. L., Anderson, B. F., Tweedie, J. W., and Baker, E. N., J. Mol. Biol. 232,

1084 (1993). 76. Lindley, P. F., Bajaj, M., Evans, R. W., Garratt, R. C., Hasnain, S. S., Jhoti, H., Kuser, P., Neu, M., Patel, K., Sarra, R., Strange, R., and Walton, A., Actu Cryst. D 49,292 (1993). 77. Dewan, J. C., Mikami, B., Hirose, M., and Sacchettini, J. C., Biochemistry, 32,

11963 (1993). 78. Anderson, B. F., Baker, H. M., Norris, G. E., Rice, D. W., and Baker, E. N., J. Mol. Biol. 209, 711 (1989). 79. Haridas, M., Anderson, B. F., and Baker, E. N., Actu Cryst. D , submitted for publication. 80. Anderson, B. F., Baker, H. M., Norris, G. E., Rumball, S. V., and Baker, E. N., Nature 344,784 (1990). 81. Baker, E. N., and Lindley, P. F., J . Inorg. Biochem. 47, 147 (1992). 82. Baker, E. N., Anderson, B. F., Baker, H. M., Haridas, M., Jameson, G. B., Norris, G. E., Rumball, S. V., and Smith, C. A., Int. J. Biol. Mucromol. 13, 122 (1991). 83. Wetlaufer, D. B., Proc. Natl. Acad. Sci. USA 70, 697 (1973). 84. Stratil, A., Bobak, P., and Valenta, M., Comp. Biochem. Physiol. B 74,603(1983). 85. Baker, E. N., Rumball, S. V., and Anderson, B. F., Trends Biochem. Sci. 12, 350

(1987). 86. Gerstein, M., Anderson, B. F., Norris, G. E., Baker, E. N., Lesk, A. M., and Chothia, C., J. Mol. Biol., 234, 357 (1993). 87. Williams, J., Trends Biochem. Sci. 7, 394 (1982). 88. Garratt, R. C., Evans, R. W., Hasnain, S. S., and Lindley, P. F., Biochem. J. 233,

479 (1986). Warner, R. C., and Weber, I., J. Am. Chem. Soc. 75,5086 (1953). Gaber, B. P., Miskowski, V., and Spiro, T. G., J.A m . Chem. SOC.98, 6868 (1974). Zweier, J. L., and Aisen, P., J. Biol. Chem. 252, 6090 (1977). Ainscough, E. W., Brodie, A. M., McLachlan, S. J., and Ritchie, V. S., J. Inorg. Biochem. 18, 103 (1983). 93. Alsaadi, B. M., Williams, R. J . P., and Woodworth, R. C., J. Inorg. Biochem. 15,

89. 90. 91. 92.

l(1981). 94. Howard, J . B., and Rees, D. C., Adu. Protein Chem. 42, 199 (1991). 95. Hasnain, S.S., Evans, R. W., Garratt, R. C., and Lindley, P. F., Biochem. J. 247,

369 (1987). 96. Koenig, S. H., and Stillinger, W. E., J. Biol. Chem. 244, 6520 (1969). 97. Hol, W. G. J., Van Duijnan, P. T., and Berendsen, H. J. C., Nature 273,443 (1978). 98. Gelb, M. H., and Harris, D. C., Arch. Biochem. Biophys. 200,93 (1980).

460

E. N . BAKER

99. Bertini, I., Luchinat, C., Messori, L., Scozzafava A,, Pellacani, G., and Sola, M.,

Inorg. Chem. 25, 1782 (1986). Z. E., and Raymond, K. N., Biochim. Biophys. Acta 956, 85 (1988). 101. Bates, G. W., and Schlabach, M. R., J . Biol. Chem. 250,2177 (1975). 102. Rossenau-Motreff, M. Y. F., Soetewey, R., Lamote, R., and Peeters, H., Biopolymers 10, 1039 (1971). 103. Kilar, F., and Simon, I., Biophys. J . 48,799 (1985). 104. Grossmann, J. G., Neu, M., Pantos, E., Schwab, F. J., Evans, R. W., Townes-Andrews, E., Lindley, P. F., Appel, H., Thies, W-G., and Hasnain, S. S., J . Mol. Biol. 225,811 (1992). 105. Grossmann, J. G., Neu, M., Evans, R. W., Lindley, P. F., Appel, H., and Hasnain, S. S., J . Mol. Biol. 229, 585 (1993). 106. Grossmann, J. G., Mason, A. B., Woodworth, R. C., Neu, M., Lindley, P. F., and Hasnain, S. S., J . Mol. Biol. 231, 554 (1993). 107. Aisen, P.,Leibman, A,, and Zweier, J., J . Biol. Chem. 253, 1930 (1978). 108. Kretchmar, S. A., and Raymond, K. N., J . A m . Chem. SOC.108,6212 (1986). 109. Faber, H. R., Anderson, B. F., and Baker, E. N., submitted for publication. 110. Zuccola, H. A,, Ph.D. thesis, Georgia Institute of Technology, 1993. 111. Quiocho, F.A,, Phil. Trans. R . SOC.London 326, 341 (1990). 112. Mao, B., Pear, M. R., McCammon, J. A., and Quiocho, F. A., J . Biol. Chem. 257, 1131 (1982). 113. Sharff, A. J., Rodseth, L. E., Spurlino, J. C., and Quiocho, F. A., Biochemistry 31, 10657 (1992). 114. Oh, B-H., Pandit, J., Kang, C-H., Nikaido, K., Gokcen, S., Ames, G.F-L., and Kim, S-H., J . Biol. Chem. 268, 11348 (1993). 115. Pflugrath, J. W., and Quiocho, F.A., Nature 314,257 (1985). 116. Luecke, H., and Quiocho, F. A., Nature 347, 402 (1990). 117. Berish, S.A., Mietzner, T. A,, Mayer, L. W.,Genco, C. A., Holloway, B. P., and Morse, S. A., J.Ezp. Med. 171, 1535 (1990). 118. Zweier, J. L., Wooten, J . B., and Cohen, J. S., Biochemistry 20, 3505 (1981). 119. Kojima, N., and Bates, G. W., J . Biol.Chem. 256, 12034 (1981). 120. Cowart, R.E., Kojima, N., and Bates, G. W., J . Biol.Chem. 257,7560 (1982). 121. Patch, M. G., and Carrano, C. J., Inorg. Chim. Actu 56, L71 (1981). 122. Tan, A.T., and Woodworth, R. C., Biochemistry 8,3711 (1969). 123. Luk, C. K., Biochemistry 10,2838 (1971). 124. Gelb, M. H., and Harris, D. C., Arch. Biochem. Biophys. 200, 93 (1980). 125. Harris, W. R., Carrano, C. J., Pecoraro, V. L., and Raymond, K. N., J. A m . Chern. SOC.103,2231 (1981). 126. Harris, W. R., and Pecoraro, V. L., Biochemistry 22, 292 (1983). 127. Harris, W. R., Biochemistry 22, 3920 (1983). 128. Lehrer, S.S., J. Biol. Chem. 244, 3613 (1969). 129. OHara, P., Yeh, S. M., Meares, C. F., and Bersohn, R.,Biochernistry 20,4704 (1981). 130. Ford-Hutchinson, A. W., and Perkins, D. J., Eur. J . Biochem. 21,55 (1971). 131. Bertini, I., Canti, G., and Luchinat, C., Znorg. Chim. Actu 67,L21 (1982). 132. Chasteen, N. D., White, L. K., and Campbell, R. F., Biochemistry 16,363 (1977). 133. Harris, W. R., and Carrano, C. J., J . Znorg. Biochem. 22, 201 (1984). 134. Aisen, P., Aasa, R., and Redfield, A. G., J . Biol.Chem. 244, 4628 (1969). 135. Ainscough, E.W., Brodie, A.M., and Plowman, J. E., Znorg. Chim. Acta 33, 149 (1979). 100. Kretchmar, S. A,, Reyes,

TRANSFERRIN STRUCTURE AND REACTIVITY

46 1

136. Tomimatsu, Y., Kint, S., and Scherer, J. R., Biochemistry 15,4918 (1976). 137. Harris, W. R., J . Inorg. Biochem. 27,41 (1986). 138. Aasa, R., Malmstrom, B. G., Saltman, P., and Vanngard, T., Biochim. Biophys.

Actu 75, 203 (1963). 139. Bertini, I., Luchinat, C., Messori, L., Monnanni, R., and Scozzafava, A., J . Biol.

Chem. 261, 1139 (1986). 140. Tan, A. T., and Woodworth, R. C., Biochemistry 8, 3711 (1969). 141. Kratz, F.,and Messori, L., J . Inorg. Biochem. 49,79 (1993). 142. Then, G. M., Appel, H., Duffield, J., Taylor, D. M., and Thies, W-G., J . Inorg.

Biochem. 27,255 (1986). 143. Stjernholm, R., Warner, F. W., Robinson, J . W., Ezekiel, E., and Katayama, N.,

Bioinorg. Chem. 9,277 (1978). Donovan, J. W., and Ross, K. D., J . Biol. Chem. 250, 6022 (1975). Evans, R. W., and Ogwang, W., Biochem. Soc. Trans. 16, 833 (1988). Bertini, I., Luchinat, C., and Messori, L., J . A m . Chem. SOC.105, 1347 (1983). Smith, C.A,, Ph.D. Thesis, Massey University, 1992. Luk, C. K., Biochemistry 10, 2838 (1971). Harris, W. R., Inorg. Chem. 25,2041 (1986). Zak, O.,and Aisen, P., Biochemistry 27, 1075 (1988). Harris, W. R., Carrano, C. J., Pecoraro, V. L., and Raymond, K. N., J . A m . Chem. Soc. 103,2231 (1981). 152. Aisen, P., in “Iron in Biochemistry and Medicine” (Jacobs, A., and Worwood, M., Eds.), Vol. 11, pp. 87-129. Academic Press, London, 1980. 153. Harris, W. R., and Sheldon, J., Inorg. Chem. 29, 119 (1990). 154. Ainscough, E. W., Brodie, A. M., Plowman, J. E., Bloor, S. J., Loehr, J. S., and Loehr, T. M., Biochemistry 19,4072 (1980). 155. Smith, C. A., Ainscough, E. W., and Brodie, A. M., submitted for publication. 156. Cannon, J. C., and Chasteen, N. D., Biochemistry 14, 4573 (1975). 157. Zweier, J . L., J . Biol. Chem. 255,2782 (1980). 158. Davidsson, L.,Lonnerdal, B., Sanstroem, B., Kunz, C., and Keen, C. L., J . Nutr. 119,1461 (1989). 159. H. Sigel (Ed.), “Metal Ions in Biological Systems,” Vol. 24.Dekker, New York, 1988. 160. Parker, D.,Chem. Brit. 26,942 (1990). 161. Shannon, R. D., Actu Cryst. A 32,751 (1976). 162. Larson, S.M., Rasey, J . S., Allen, D. R., and Nelson, N. J., J . Nucl. Med. 20, 837 (1979). 163. Kubal, G., Mason, A. B., Patel, S. U., Sadler, P. J., and Woodworth, R. C., Biochemistry 32,3387 (1993). 164. Morris, C. M., Candy, J. M., Court, J . A., Whitford, C. A., and Edwardson, J. A., Biochem. Soc. Trans. 15,498 (1987). 165. O’Hara, P. B., and Koenig, S. H., Biochemistry 25, 1445 (1986). 166. Aisen, P., and Leibman, A., Biochim. Biophys. Acta 257, 314 (1972). 167. Pecoraro, V. L., Harris, W. R., Carrano, C. J., and Raymond, K. N., Biochemistry 20,7033 (1981). 168. O’Hara, P. B., and Bersohn, R., Biochemistry 21,5269 (1982). 169. O’Hara, P., Yeh, S. M., Meares, C . F., andBersohn,R.,Biochemistry20,4704(1981). 170. Sola, M., Inorg. Chem. 29, 1113 (1989). 171. Sola, M., Eur. J . Biochem. 194, 349 (1990). 172. Taylor, D. M., Thies, W-G., Appel, H., Neu-Muller, M., Schwab, F. J., and Weber, H. W., in “Metal Ion Homeostasis: Molecular Biology and Chemistry” (D.H. Hamer 144. 145. 146. 147. 148. 149. 150. 151.

462

E. N. BAKER

and D. R. Winge, Eds.), UCLA Symposia on Molecular and Cellular Biology, New Series, Vol. 98,179-188. Alan R. Liss, New York, 1988. 173. Taylor, D. M., Seidel, A., Planas-Bohne, F., Schuppler, U., Neu-Muller, M., and Wirth, R. E., Inorg. Chim. Acta 140,361 (1987). 174. Duffield, J. R., and Taylor, D. M., Inorg. Chim. Acta 140,365 (1987). 175. Garratt, R. C., Evans, R. W., Hasnain, S. S., and Lindley, P. F., Biochem. J . 280, 151 (1991). 176. Warner, R. C., and Weber, I., J . A m . Chem. SOC.75, 5094 (1953). 177. Harris, W. R., Nesset-Tollefson, D., Stenback, J. Z., and Mohamed-Hani, N., J . Inorg. Biochem. 38, 175 (1990). 178. Schlabach, M. R., and Bates, G. W., J . Biol. Chem. 250,2182 (1975). 179. Aisen, P., Aasa, R., Malmstrom, B. G., and Vanngard, T., J . Biol. Chem. 242, 2484 (1967). 180. Harris, D. C., and Gelb, M. H., Biochim. Biophys. Acta 623, 1 (1980). 181, Zweier, J. L., Aisen, P.,Peisach, J.,andMims, W. B.,J. Biol. Chem. 254,3512(1979). 182. Zweier, J. L., Peisach, J., and Mims, W. B., J . Biol. Chem. 257, 10314 (1982). 183. Eaton, S.S., Dubach, J., More, K. M., Eaton, G. R., Thurman, G., and Ambruso, D. R., J . Biol. Chem. 264,4776(1989). 184. Eaton, S. S.,Dubach, J., Eaton, G. R., Thurman, G., and Ambruso, D. R., J . Biol. Chern. 265,7138(1990). 185. Campbell, R. F., and Chasteen, N. D., J . Biol. Chem. 252, 5996 (1977). 186. Harris, D. C., Gray, G., and Aisen, P., J . Biol. Chem. 249, 5261 (1974). 187. Zweier, J. L., Wooten, J. B., and Cohen, J. S., Biochemistry 20, 3505 (1981). 188. Schneider, D.J., Roe, A. L., Mayer, R. J., and Que, L., J . Biol. Chem. 259, 9699 (1984). 189. Rogers, T. B., Borresen, T., and Feeney, R. E., Biochemistry 17, 1105 (1978). 190. Baker, E. N., Anderson, B. F., Baker, H. M., Haridas, M., Norris, G. E., Rumball, S. V., and Smith, C. A,, Pure Appl. Chem. 62, 1067 (1990). 191. Dubach, J., Gaffney, B. J., More, K.,Eaton, G. R., and Eaton, S. S., Biophys. J . 59, 1091 (1991). 192. Shongwe, M.S., Smith, C. A., Ainscough, E. W., Baker, H. M., Brodie, A.M., and Baker, E. N., Biochemistry 31,4451 (1992). 193. Baker, H. M., Brodie, A. M., Smith, C. A., and Baker, E. N., submitted for publication. 194. Price, E. M., and Gibson, J. F., J . Biol. Chem. 247, 8031 (1972). 195. Folqjtar, D. A., and Chasteen, N. D., J . A m . Chem. SOC.104,5775 (1982). 196. Record, M.T., Anderson, C. F., and Lohman, T. M., Quart. Rev. Biophys. 11, 103 (1978). 197. Williams, J., Biochem. J . 201,647 (1982). 198. Baldwin, D. A., and de Sousa, D. M. R., Biochem. Biophys. Res. Comm. 99, 1101 (1981). 199. Zak, O.,Leibman, A., and Aisen, P., Biochim. Biophys. Acta 742,490 (1983). 200. Harris, D.C., Biochemistry 16,560 (1977). 201. Makey, D. G., and Seal, US.,Biochim. Biophys. Acta 453, 250 (1976). 202. Evans, R. W., and Williams, J., Biochem. J . 173,543 (1978). 203. Lestas, A. N., Br. J . Huematol. 32, 341 (1976). 204. Princiotto, J. V., and Zapolski, E. J., Nature 255, 87 (1975). 205. Mazurier, J., and Spik, G., Biochim. Biophys. Acta 629,399 (1980). 206. Williams, J., and Moreton, K., Biochem. J . 185,483 (1980). 207. Chasteen, N. D., White, L. K., and Campbell, R. F., Biochemistry 16,363 (1977).

TRANSFERRIN STRUCTURE AND REACTIVITY

463

208. Fletcher, J., and Huehns, E. R., Nature 218, 1211 (1968). 209. Bali, P. K., and Aisen, P., Biochemistry 30, 9947 (1991). 210. Lin, L-N., Mason, A. B., Woodworth, R. C., and Brandts, J. F., Biochemistry 30, 211. 212. 213. 214. 215. 216. 21 7 . 218. 219. 220. 221. 222. 223. 224. 225. 226. 227. 228. 229. 230.

231. 232. 233. 234.

11660 (1991). Bates, G. W., and Schlabach, M. R., J . Biol. Chem. 248,3228 (1973). Bates, G. W., and Wernicke, J., J . Biol. Chem. 246,3679 (1971). Mowbray, S.,unpublished data. Bennett, W. S . , and Huber, R., Crit. Rev. Biochem. 15,291 (1984). Gerstein, M., and Chothia, C., J. Mol. Biol. 220, 133 (1991). Kretchmar, S.A., and Raymond, K. N., Inorg. Chem. 27, 1436 (1988). Baldwin, D. A., Biochim. Biophys. Actu 623, 183 (1980). Cowart, R. E., Swope, S., Loh, T. T., Chasteen, N. D., and Bates, G. W., J . Biol. Chem. 261,4607 (1986). Bertini, I., Hirose, J., Luchinat, C., Messori, L., Piccioli, M., and Scozzafava, A., Inorg. Chem. 27,2405 (1988). Harris, W. R., and Bali, P. K., Inorg. Chem. 27,2687 (1988). Bali, P.K., Harris, W. R., and Nesset-Tollefson, D., Inorg. Chem. 30, 502 (1991). Carrano, C. J., and Raymond, K. N., J . A m . Chem. SOC.101,5401 (1979). Harris, W. R., Rezvani, A. B., and Bali, P. K., Inorg. Chem. 26, 2711 (1987). Bali, P. K., and Harris, W. R., J. Am. Chem. SOC.111, 4457 (1989). Egan, T. J., Ross, D. C., Purves, L. R., and Adams, P. A,, Inorg. Chem. 31, 1994 (1992). Egan, T. J., Zak, O., and Aisen, P., Biochemistry 32,8162 (1993). Legrand, D., Mazurier, J., Colavizza, D., Montreuil, J., and Spik, G., Biochem. J. 266,575 (1990). Swope, K. S., Chasteen, N. D., Weber, K. E., and Harris, D. C., J.A m . Chem. SOC. 110,3835 (1988). Borhani, D. W., and Harrison, S. C., J . Mol. Biol. 218, 685 (1991). Mason, A. B., Miller, M. K., Funk, W. D., Banfield, D. K., Savage K. J., Oliver, R. W. A., Green, B. N., MacGillivray, R. T. A., and Woodworth, R. C., Biochemistry 32,5472 (1993). Stowell, K. M., Rado, T. A., Funk, W. D., and Tweedie, J. W., Biochem. J. 276, 349 (1991). Ikeda, R. A., Bowman, B. H., Yang, F., and Lokey, L. K., Gene 117,265 (1992). Woodworth, R. C., Mason, A. B., Funk, W. D., and MacGillivray, R. T. A., Biochemistry 30, 10824 (1991). Day, C. L., Baker, E. N., and Tweedie, J. W., unpublished data.

This Page Intentionally Left Blank

INDEX

A Absorption, transient, class I1 mixed-valence complexes, 303 Absorption spectra, see also Visible absorption spectra; X-ray absorption spectroscopy class I1 mixed-valence complexes, 294-295 Actinides, transferrin binding, 428-429 Aluminum, transferrin binding, 426 Aluminum hydrides, 221-226 IR spectra, 223 matrix isolation, 222 synthesis, 222 vibrational spectra, 223-224 Amide protons, in diamidedithiols, 65 Anions binding, transferrins, 406 conformational differences associated with, 439-440 interlocking sites model, 434-435 metal-anion interactions, 433-437 nonsynergistic anions, 437-439 Schlabach-Bates model, 436-437 synergistic anions, 431-433 nonsynergistic, 437-439 sites lactoferrin, 418 transferrins, 403-407 sequence similarities, 412-414 synergistic, characteristics, 431-432 Antiferromagnetic exchange, charge transfer model, 305-307 Aqua complexes, technetium(III), 31 Arene complexes, technetium(I1, 13 Arsine technetium(I1) complexes, 23-24 technetium(II1) complexes, 39-43 technetium(1V) complexes, 52-53 [As2Br812-,241 [As2Br913-,249-250 Ad,, 236

[ A s ~ I ~ ~253 I~-, [As&I4-, 260 AsPh4[TcNC14],83, 85-86, 84-85 (ASP~~)~[TC~N~(O)~(OX)~I, 89 Auger electron spectroscopy molecular phosphorus oxides, 360-362 molecular phosphorus oxide sulfides, 380

B Bacterial binding properties, similarities with transferrins, 416-418 BH4 SOUP, 214-215 [BiBr41-,264 [BiBr6I3-,248-249 [Bi2Brl1I5-,255-256 [Bi4Brl6I4-,244 BiCI3, 235,237 [BiC14]-,264 [{BiC14},l"-, 242 [BiC16I3-, 266-267 [Bi4CIl8l6-,255-256 [Bi8C13016~-, 260-261 BiF,, 234-235 BiI,, 236-238 [BiI6I3-,266 [Bi3Il2I3-,253 Bismuth halides, 267 Bismuth tribromide, 235-237 Bispentafluorosulfanylamine, 151 Bis(pentafluorosulfanyl)bis(trifluoromethyl)hydrazine, 150 Bis(pentafluorosulfany1)perfluoroalkylamines, 149-150 Bleaney-Bowers expression, 308 [Bu4"NI~[Sb2ClJ,241 C

Carbonyl complexes technetium(I), 7-12 technetium(III), 27-29

465

466

INDEX

Carboxylato complexes, technetium(III),

32-33 Charge transfer model, antiferromagnetic exchange, 305-307 Charge-transfer transition dipole moment, 276 [CSH,HN]JSb&lg], 251-252 C=N, functionality, 130 C - 0 bond, 28 Comproportionation constants, class I1 mixed-valence complexes, 290-292 Comproportionation equilibrium,

280-281 [Co(NH&I[Sb2Fgl, 250 Copper, transferrin binding, 424-425 Creutz-Taube ion, 281-282 derivatives, 286 physical characterization, 288-289 [Cs(l8-cro~n-6)[TcNC1~1, 85-87 Cs[TcN(02)2CI],97 Cyano complexes mononuclear [Tc0I3+,55-56 technetium(I), 13-14 technetium(II), 30-31 Cyclopentadienyl complexes, technetium(I), 12

D P-Diketonato complexes, technetiumUII),

33 Dimethylgallane, 192-194 Dimethylgallium tetrahydroborate,

188-189 Dinitrogen complexes, technetium(I), 14 Dioximes, technetium(II1) complexes,

33-36 Disulfide bonding, transferrins, 402-403 Disulfide bridges, transferrins, 415-416 Dithiolene complexes, technetium(VI),

93-94 DNA, recombinant, transferrin studies,

452-455

E Electroabsorption spectroscopy, 279 class I1 mixed-valence complexes, 289,

291,294-297 [ { ( N H ~ ) ~ R U } ~ ( ~ -294, ~ Y Z296 )I~',

Electron diffraction gallaborane, 214 gallane vapor, 204-207 2-galla-arachno-tetraborane, 218 gallium hydrides, 185-188 Electron exchange, 304-313 charge transfer model, antiferromagnetic exchange, 305-307 constant J, experimental evaluation,

308 exchange coupled ruthenium dinuclear complexes, 308-313 magnetic, 304-305 superexchange, 307-309 Electronic coupling, between donor and acceptor wave functions, 278 Electron paramagnetic resonance spectra, see EPR spectra Element(II1) halogenoanions, 238-264 cis E-X bonds, 267 coordination geometries, 265-266 [EX,]-, 239-240 [{EX4},]"-,242-244 [EX5]'-, 246 l{EX,},lz"-, 247-248 [EX,13-, 248-249 [{E,X,},l"-, 260-261 [E2Xs]", 240-241 [EZXgl3-,249-251 [{E2Xg},I3''-, 251-252 [EzXioI4-, 247 [EzX11]5-,255-256 [{E3X1o},In-, 261-264 [E3XLIl2-,256-257 [E3XI2l3-,252-254 [E,X,,},I"-, 263-264 [E4XI6l4-,244-246 [E4X1J-, 255-256 [E5XI8l3-,254-255 [E&2I4-, 257-258 [E8XzsI4-,258-260 [EBX3Ol6-, 260-261 extended Huckel molecular orbital analysis, 266-267 Jahn-Teller distortion, 266-267 potential energy diagram, 278 structural chemistry, 264 VSEPR theory, 265-266 Element trihalides, 234-238 primary and secondary bond, 236-237

467

INDEX

EPR spectra, diferric lactoferrin,

433-434

trapping experiments, 180-181 vapor transfer and sampling, 177,

179-180 history and chemical background,

F

173-177 hydridogallium bis(tetrahydroborate),

First transition series, substitution, transferrins, 423-426 Fluorescence quenching, metal binding of transferrins, 419-420 FSSCHCFZOSO,, 159-161

G

189-192 monochlorogallane, 194-196 physical methods of detection and analysis, 181-188 electron diffraction, 185-188 NMR spectroscopy, 184-185 vibrational spectroscopy, 182-184 Group 13 metals, transferrin binding,

426-427 Ga2D6,vibrational properties, 203,205 Ga2H6,183-184 vibrational properties, 203,205 Ga-H bond, 210,214 Gallaborane, 211-216 electron diffraction, 214 infrared spectra, 212-213 synthesis, 215 Gallane, 172-173,196-198 chemical analysis, 199,201 chemical properties, 208-209 complexes, 175-178 electron diffraction, 204-207 gallaborane, 211-216 2-galla-arachno-tetraborane,216-220 'H NMR spectrum, 207-208 IR spectrum, 200-202 physical properties, 199 search for, 173-175 synthesis, 198-199 vibrational spectra, 200-215 2-Galla-arachno-tetraborane, 216-220 llB NMR spectrum, 217,219-220 decomposition of vapor, 216-217 electron diffraction, 218 structure, 217-218 Gallium, transferrin binding, 426 Gallium hydrides, 171-227;see also Pentafluorosulfanyl hypohalites dimensions, 203-204 dimethylgallane, 192-194 dimethylgallium tetrahydroborate,

188-189 handling, 177,179-181 chemical analysis, 181

H Hafnium, transferrin binding, 428 Halide complexes and clusters mononuclear [Tc013', 56-59 [TcN13-, 82 technetium(II), 17-22 mononuclear and binuclear, 17-20 polynuclear, 20-22 technetium(III), 31-32,45-47 HGa(BH&, 191 Histidine, involvement in transferrin iron binding, 403-404 H3P.GaH3, 208,210 [HTcCpzI, 29 Hiickel molecular orbital analysis, extended dinuclear ruthenium complexes,

311-312 element(II1) halogenoanions, 266-267 Hush model, 274-280 class I1 mixed-valence complexes, parameters, 293 intervalence band properties relationships, 276 wave functions, 277 Hydrazido complexes technetium(V), 78-79 technetium(VI), 92-93 Hydridogallium bidtetrahydroborate),

189-192 Hydrogen, bonding interactions, in transferrin binding site, 404-405

468

INDEX

2-Hydroxyl-l-(pentafluoro-h6-sulfanyl)1,2,2-trifluoroethanesulfonic acid sultone. 157-158

I Imido complexes technetium(V), 78-79 technetium(VI1, 92-93 technetium(VII), 97-99 Indium, transferrin binding, 427 Indium hydrides, 225-227 Infrared spectra aluminum hydrides, 223 gallaborane, 212-213 gallane, 200-202 Interlocking sites model, 434-435 Iron coordination, transferrins, 403-405 release, transferrins functional aspects, 444 kinetics, 448-450 pH dependence, 442,450 structural aspects, 450-452 transferrin uptake binding mechanism, 445-447 dynamics importance, 447-448 Isonitrile complexes technetium(I), 13-14 technetium(II), 31 technetium(III), 45

J Jahn-Teller distortion, element(II1) halogenoanions, 266-267

K

L Lactoferrin, 390 anion sites, 418 biological role, 392-393 conformational change, 407-411 connecting peptide, 415 Cuzt-substituted, 439-440 metal and anion binding, 429-430

diferric, EPR spectra, 433-434 domain organization, 398-400 half-molecules, 411-412, 453-454 iron release, 450-452 lanthanide binding, 427-428 oxalate binding, 435-436 polypeptide folding, 417-418 pattern, 400-401 proteolytic fragments, 396 recombinant, 453 ribbon diagram, 399 sequence identity, 393-394 site-specific mutants, 454 three-dimensional structure, 397 visible absorption spectra, 423-424 Lanthanides, transferrin binding, 427-428

M Magnetic susceptibility, temperature dependence, 310 Mass spectrometry, molecular phosphorus oxides, 362 Melanotransferrin, 390-391 biological role, 393 recombinant DNA studies, 454-455 sequence identity, 394 Me,N-GaH, molecule, molecular scattering intensity pattern, 186-187 Metal-anion interactions, transferrins, 433-437 Metal-to-metal charge transfer, 275 antiferromagnetically exchange coupled system, 306-307 Creutz-Taube ion, 286 class I1 M, 290-292 Ru(II)-Ru(III) dinuclear complexes, 313 solvent effects, class I1 mixed-valence complexes, 297-300 thermochromism, 300 Metal-metal coupling class I11 mixed-valence complexes, 284 crown ether addition, 286-288 dependence on bridging ligand size, 302 Metals binding, transferrins, spectroscopic monitors, 419-420

469

INDEX

carried by transferrins, 392 site, transferrins, 403-407 design, 406-407 sequence similarities, 412-414 M-H-M bridge, 182-183 Microwave spectrum, molecular phosphorus oxides, 362 Minitransferrins, 411-413 Mixed-valence complexes, 274-304 class 11, 289-304 activation free energy for energy transfer, 301-302 bridging ligand nature, 301-303 comproportionation constants,

290-292 distance between ruthenium ions,

300-302 electroabsorption spectroscopy, 289,

291,294-297 Hush model parameters, 293 metal-to-metal charge transfer,

290-292 solvent effects, 297-300 molecular electronics and materials,

303-304 Raman and electronic absorption spectroscopy, 297 transient absorption studies, 303 class I11 bridging ligand nature, 284-286 Creutz-Taube ion, 288-289 criteria, 282-284 metal-metal coupling, 284 ruthenium ion nature, 286-288 comproportionation equilibrium,

280-281

[{(NH,),RU}~(~L-PYZ)I~~, 288-289 absorption and electroabsorption spectra, 294-296 "H&[SbCI51, 246 "H,I2[SbCl,F21, 248 Nitrido complexes technetium(V), 72-78 technetium(VI), 81-92 dimeric and polymeric [TcN13', 85-92 monomeric [TcN13', 81-85 technetium(VII), 97-99 Nitrogen ligands technetium(1) complexes, 15 technetium(I1) complexes, 22-23 technetium(II1) complexes, 33-38 technetium(1V) complexes, 52 Nitrosyl complexes technetium(I), 15-16 technetium(II), 25-27 technetium(III), 44 NMR spectroscopy IlB, 2-galla-aruchno-tetraborane, 217, 219-220 gallium hydrides, 184-185 'H, gallane, 207-208 31P

molecular phosphorus oxides, 349,

351-352 molecular oxide phosphorus sulfides,

374-377 solid-state molecular phosphorus oxides,

352-354 molecular phosphorus oxide sulfides,

375-378

Hush model, 274-280 PKS model, 281 potential energy-configuration diagram, 275 potential energy curves, distortion, 275 Molecular electronics, class I1 mixed-valence complexes, 303-304 Monochlorogallane, 194-196

N Na[SbF41,243 [{(NH3)5R~}2(pDi~yd)14t, 312-313

[ N T c ( ~ - O ) ~ T Cdimers, N ] ~ ~ 90-91

0 Organometallic complexes, technetium(II), 17 Oscillator strength, theoretical expression, 276 Ovotransferrin, 390 biological role, 393 CO'+-substituted, 439 half-molecules, 396,411 quarter-molecule, 412-413

470

INDEX

Oxalate binding, lactoferrin, 435-436 0x0-bridged complexes technetium(IV), 47-52 binuclear complexes, 48-51 mononuclear complexes, 47-48 phosphato complexes, 51-52 technetium(VI), 80-81 technetiumWII), 94 Oxygen ligands, technetium(1V) complexes, 47-52 binuclear complexes, 48-51 mononuclear complexes, 47-48 phosphato complexes, 51-52

P [PBrJ, 239-240 [PCIJ, 239 Pentafluorosulfanylalkanes, 128-129, 132-138 Pentafluorosulfanylalkenes, 132-138 Pentafluorosulfanylalkynes, 132-138 Pentafluorosulfanylamine, 144-145 N-Pentafluorosulfanyl chloroimine, 152-154 Pentafluorosulfanyl compounds, 125-161 bispentafluorosulfanylamine, 151 bis(pentafluorosulfanyl)bis(trifluoromethyl)hydrazine, 150 bis(pentafluorosulfany1)perfluoroalkylamines, 149-150 FSSCHCFzOS02, 159-161 halides, 126-130 2-hydroxyl-1-(pentafluoro-h6-sulfanyl)1,2,2-trifluoroethanesulfonic acid sultone, 157-158 pentafluorosulfanylalkanes,alkenes, and alkynes, 132-138 pentafluorosulfanylamine,144-145 pentafluorosulfanyl N,N-dichloroamine, 145-146 pentafluorosulfanyl N, N-difluoramine, 146 N-pentafluorosulfanyl haloimines, 152-154 pentafluorosulfanyl hypohalites, 130-132 pentafluorosulfanyliminodihalosulfanes, 155-157

pentafluorosulfanyl perfluoroalkylamines, 146-147 pentafluorosulfanyl-p-sultones and sulfonic acids, 157-161 SFsN(CF312, 147 SFSN(Cl)RI, 149 (SFS)ZNX, 151-152 SFSN(X)CF3, 147-148 sulfur cyanate pentafluoride, 142-143 sulfur cyanide pentafluoride, 143 sulfur isocyanate and isothiocyanate pentafluorides, 138-142 sulfur isocyanide pentafluoride, 143-144 tetrakis(pentafluorosulfany1 )hydrazine, 150 tris(pentafluorosulfanyl)amine,150 Pentafluorosulfanyl N,N-dichloroamine, 145-146 Pentafluorosulfanyl N,N-difluoramine, 146 N-Pentafluorosulfanyl fluoroimine, 154 Pentafluorosulfanyl halides, 126-130 N-Pentafluorosulfanyl haloimines, 152-154 Pentafluorosulfanyl hypohalites, 130-132 SFEOF, 130-131 Pentafluorosulfanyliminodichlorosulfane, 155-157 Pentafluorosulfanyliminodifluorosulfane, 155 Pentafluorosulfanyliminodihalosulfanes, 155-157 Pentafluorosulfanyl perfluoroalkylamines, 146-147 [PhCH2Me2N.AlH,12,225-226 Phosphine complexes bidentate, technetium(III), 39-43 monodentate, technetiumUII), 38-39 technetium(I1, 14-15 technetium(II), 23-24 Phosphine ligands, technetium(1V) complexes, 52-53 Phosphito complexes, technetium(I), 15 Phosphorus molecular oxides bonding features, 363-364 compared to molecular phosphorus oxide sulfides, 381-383

471

INDEX

crystal and molecular structures, 337-345 bond angles, 343 comparison of molecular structures, 340-343 crystallographic data, 345 in gaseous state, 337-338 group-subgroup relationships, 344 molecular packing, 343-345 in solid state, 337-345 mass spectrometric, 362 microwave spectrum, 362 molecular packings, 382 photoelectron and auger electron spectroscopy, 360-362 31PNMR spectroscopy, 349, 351-352 solid-state NMR spectroscopy, 352-354 structure, 328 synthesis, 329-336 phosphorus (IIIN) oxides, 334-336 phosphorus pentoxide, 330-331 phosphorus trioxide, 331-333 theoretical studies, 362-364 vacuum-ultraviolet spectrum, 362 vibrational spectroscopy, 346-349 X-ray absorption spectroscopy, 355-359 molecular oxide sulfides bond angles, 368 compared to molecular phosphorus oxides, 381-383 crystal and molecular structures, 366-370 force constants, 372 molecular geometries, 368 molecular packings, 382 solid state, 369-370 photoelectron and Auger electron spectroscopy, 379-380 31PNMR spectroscopy, 374-375 solid-state NMR spectroscopy, 375-378 synthesis, 364-366 theoretical studies, 380-381 vibrational spectroscopy, 370-374 X-ray absorption spectroscopy, 378-379 Phosphorus(III/V) oxides crystal structure, 339

IR and Raman absorbances, 351 molecular structure, 328 31PNMR spectroscopy, 351-352 solid-state NMR spectroscopy, 352-353 synthesis, 334-336 vibrational spectroscopy, 349-351 Phosphorus pentoxide bonding features, 363 core binding energies, 361-362 crystal structure, 337-338 solid-state NMR spectroscopy, 352-353 synthesis, 330-331 vibrational spectroscopy, 346-348 X-ray absorption spectroscopy, 355-359 Phosphorus trioxide bonding features, 363 core binding energies, 361-362 crystal structure, 339 ionization energy data, 360-361 31PNMR spectroscopy, 349, 351 solid-state NMR spectroscopy, 352 synthesis, 331-333 vibrational spectroscopy, 346-348 X-ray absorption spectroscopy, 355-359 Photoelectron spectroscopy molecular phosphorus oxides, 360-362 molecular phosphorus oxide sulfides, 379-380 PKS model, 281 P407,crystal structure, 340-341 P409crystal structure, 339 Polymerization, pentafluorosulfanyl halides, 128 Polypeptide chain, transferrins, 397-398 folding, 400-402, 417-418 [Pr4"N12[Sb2C181, 241

R Raman spectroscopy, class I1 mixed-valence complexes, 297 Reorganization parameter, inner and outer sphere, 279 Ruthenium, dinuclear complexes, 273-319; see also Mixed-valence complexes antiferromagnetic superexchange, 309 electron exchange, constant J, 308

472

INDEX

exchange coupled, 308-313 extended Huckel molecular orbital calculations, 311-312 future studies, 313-314 ligand structures, 314-319 magnetic electron exchange, 304-305 magnetic susceptibility, temperature dependence, 310 properties, 273-274 superexchange, 307-309 Ruthenium (III), coordination sphere, 299-300 Ruthenium(III/II) couples, 282-283 Ruthenium ions, 286-288 distance between, 300-302

S

/

/

SbBr3, 235 SbC13, 235 [SbCIJ, 264 [SbCI5l2-,247-248 [Sb4C116I4-,244-245 [SbzC13F613-,252 SbF3, 234 [Sb*FTI-, 251 [{Sb,Fl,},]"-, 262-263 [(Sb4F13},In-,263-264 [Sb4F16I4-,245-246 SbIB, 236 [{Sb217}n]"-,260-261 [Sb2Iz2l4-, 265 [{Sb311O},]"-, 261-262 [Sb,I1,I2-, 256-257 [Sb5Il8I3-,254, 254-255 [Sb&2I4-, 257-258 [Sb81z814-, 258-259 Schiff base technetium(II1) complexes, 36-37 technetium(1V) complexes, 52 Schlabach-Bates model, 436-437 Serum transferrins, 390 biological role, 391-392 half-molecules, 396 recombinant, 453 structure, 397 SF5Br, 126-130 S F S C g H , 137 S F S C M F Z , 135-136 SFSCH2COOAg, 133-134

SFSCI, 126-130 SFSN(CF&, 147 SF,N(Cl)R,, 149 (SFS)zNX, 151-152 SFSN(X)CF3, 147-148 SFSOC1, 131-132 SF,OF, 130-131 Solution X-ray scattering measurements, transferrins, 409-410 Stark effect, 279 class I1 mixed-valence complexes, 289, 291,294-297 Sulfido complexes, technetiumWII), 94 Sulfur, electrochemical fluorination, 133 Sulfur cyanate pentafluoride, 142-143 Sulfur cyanide pentafluoride, 143 Sulfur isocyanate pentafluoride, 138-141 Sulfur isocyanide pentafluoride, 143-144 Sulfur isothiocyanate pentafluoride, 140-142 Sulfur ligands technetium(I1)complexes, 24-25 technetium(II1) complexes, 43-44 technetium(1V) complexes, 53-54 99mT~ physical properties, 3-4 use in diagnostic nuclear medicine, 3-4

T [Tcz(bdt),I, 53-54 [TcBr6I2-,45-47, 47-48 [T~Br(dmgH)~(dmg)BBu], 34-35 Tc-CI, 83 TcCl,, 45-46 [TCC16]2-,48, 69 [T~*C181~-, 18-19 [TC~(CO)~OI, 5-7 dimeric carbonyl complexes, 11-12 [TC(CO)~] core, carbonyl complexes, 9-10 [TdCO),] core, carbonyl complexes, 8-9 [Tc(CO),X12, dimeric and polynuclear carbony1 complexes, 10-11 [TcH,12-, 99 Tc-N, bond distance, 22-23, 37-38, 75-76, 93 T c N , 85 bond distance, 73

INDEX

[TcNI3~,technetium(V1) complexes dimeric and polymeric, 85-92 monomeric, 81-85 [Tc(NAr),Il, 98-99 [TcNClJ, 73-74, 74-76, 81-82,85 [TcNCl,(AsPh,),l, 76-77 [{T~NC12}2(/~-0)2]~-, 90 [ T C N C ~ ~ ( Pcomplexes, P ~ ~ ) ~ ] 74-75 [TcN12-core, 72-73 TcN{N4}complexes, technetium(V),75 T c - N - 0 angle, 25-26 [{TcN(OH~)~}~(/L-O)~]~*, 88-89 TcVN{S4} complexes, 74 Tc-O,66-67,80-81 bond distances, 47-48, 94-95 TFO, 69-70 [Tc0l3-, mononuclear complexes, 54-68 based on other TcO mixed ligand cores, 67-68 based on TcO{N4},TcO{N,-,,O,,},and Tc){N2P2} cores, 62-64 based on TcO{N,_,S,} cores, 64-67 based on TcO(04},TcO{S4},TcO{04_,,Sn},and TcO{Se,} cores, 59-62 cyano and thiocyanato, 55-56 halide, 56-59 truns-[TcO2I+cores, technetium(V) complexes, 68-70 TcO,-, 80 Tc(p-02)Tccomplexes, 49-50 [ T C ~ O ~ 0x0-bridged, ]~', technetiumW complexes, 70-72 [TcOBr,]-, 57-58 [TcOC14]-,48-49, 53, 56-58, 66, 67-68, 69-71,78 [TcOCI~L],96 [ ( T ~ O ) ~ ( e d71-72 t)~l, [TcO,F], 95-96 Tc--OH2, distance, 83 [TcOI,]-, 57-58 [TcO(MAG3)lZ-,66-67 TcO{N,}, mononuclear [Tc013' complexes, 62-63 TcO{N,-,O,}, mononuclear [Tc013+complexes, 63-64 TcO{N2P2},mononuclear [Tc013+complexes, 64 TcO{N,-,S,} cores, mononuclear [Tc013+ complexes, 64-67 Tc0{0,}, mononuclear [Tc0I3 complexes, 59-61 +

473

tr~ns-[TcO(OH)1~+ cores, technetium(V1 complexes, 68-70 TcO{O,_,S,}, mononuclear [TcO13*complexes, 59-61 TcO(S4},mononuclear [Tc0l3+complexes, 59, 61 TcOISe,}, mononuclear [Tc0I3' complexes, 59, 62 Tc-0-Tc bridge, 70-71, 94 TcOzTc bridge, 91 Tc-P, bond distances, 23-24 Tc-S, bridging distance, 98 T& angles, 25 [TcS13+complexes, technetium(V), 72 Tc-Tc bond, 2 9 , 5 0 4 1 , 54 bond distance, 18-21, 31-32, 90-91, 93 [Tdtdt),], 93 [T~(tdt)(dmpe)~IPF~, 41-42 Technetium, 1-99 isotopes, 2 Technetium(-l), 5 Technetium(O), 5-7 TechnetiudI), 7-17 carbonyl complexes, 7-12 dimeric and polynuclear, 10-12 mononuclear, 7-10 cyano and isonitrile complexes, 13-14 cyclopentadienyl and arene complexes, 12-13 dinitrogen, phosphine, phosphito, and related complexes, 14-15 nitrogen ligand complexes, 15 nitrosyl and thionitrosyl complexes, 15-17 Technetium(II), 17-27 complexes nitrogen ligands, 22-23 sulfur ligands, 24-25 halide complexes and clusters, 17-22 binuclear, 18-20 mononuclear, 17-18 polynuclear, 20-22 nitrosyl and thionitrosyl complexes, 25-27 organometallic complexes, 17 phosphine, arsine, and related complexes, 23-24 TechnetiumUII), 27-44 aqua, halide, and related dimeric cornplexes, 31-32

474

INDEX

carbonyl complexes, 27-29 carboxylato and P-diketonato complexes, 32-33 complexes bidentate phosphine, arsine, and related ligands, 39-43 dioximes, Schiff bases and other nitrogen ligands, 33-38 monodentate phosphines and related ligands, 38-39 sulfur ligands, 43-44 cyano, isonitrile, and thiocyanato complexes, 30-31 cyclopentadienyl complexes, 29-30 nitrosyl and thionitrosyl complexes, 44 nonreducible cations, 40-41 organohydrazine chemistry, 37 Technetium(IV), 45-54 complexes oxygen ligands and 0x0-bridged complexes, 47-52 binuclear complexes, 48-51 mononuclear complexes, 47-48 phosphato complexes, 51-52 phosphine and arsine ligands, 52-53 Schiff base and other nitrogen ligands, 52 sulfur ligands, 53-54 halide and related complexes, 45-47 isonitrile and thiocyanato complexes, 45 Technetium(V), 54-79 complexes not containing multiply bonded ligands, 79 fr~ns-[TcO(OH)1~' cores, 68-70 imido and hydrazido complexes, 78-79 mononuclear [Tc013+complexes, 54-68 based on other TcO mixed ligand cores, 67-68 based on TcO{N4},TcO{04-"On},and TcO{N2P2}cores, 62-64 based on TcO{N4-"Sn}cores, 64-67 based on TcO{04},TcO{S4},TcO{04-"Sn},and TcO{Se,} cores, 59-62 cyano and thiocyanato, 55-56 halides, 56-59 nitrido complexes, 72-78 0x0-bridged [Tc2O3I4+ and other binuclear complexes, 70-72

synianfz isomerism, 64-65 [TcS13+complexes, 72 Technetium(VI), 80-94

dithiolene and related complexes, 93-94 imido and hydrazido complexes, 92-93 nitrido complexes, 81-92 dimeric and polymeric [TcNI3+, 85-92 monomeric [TcNI3', 81-85 0x0 complexes, 80-81 Technetium(VI1). 94-99 complexes not containing multiply bonded ligands, 99 nitrido and imido complexes, 97-99 0x0 and sulfido complexes, 94-97 Tetrakis(pentafluorosulfanyl)hydrazine, 150 Thallium hydrides, 225-227 Thermochromism, metal-to-metal charge transfer, 300 Thiocyanato complexes mononuclear [Tc013-,55-56 technetium(II), 31 technetium(III),45 Thionitrosyl complexes technetium(I), 16-17 technetium(II), 26-27 technetium(III), 44 Transferrins, 389-456; see also specific transferrins

anion binding conformational differences associated with, 439-440 interlocking sites model, 434-435 metal-anion interactions, 433-437 nonsynergistic anions, 437-439 Schlabach-Bates model, 436-437 synergistic anions, 431-433 biological roles, 391-393 diferric complexes, visible absorption spectra, 433 differences between metal and anion sites, 440-445 differences in properties, 441-443 functional aspects, 444-445 structural comparison, 440-441 substitution, 443-444 iron release kinetics, 448-450 structural aspects, 450-452

475

INDEX

iron uptake binding mechanism, 445-447 dynamics importance, 447-448 lactoferrin, 390 melanotransferrin, 390-391 metal binding, spectroscopic monitors, 419-420 metal substitution and spectroscopy, 420-431 actinides and other metal ions, 428-429 binding constants, 422-423 first transition series, 423-426 group 13 metals, 426-427 lanthanides, 427-428 structural aspects, 429-431 ovotransferrin, 390 possible evolutionary development, 395 properties, 390-391 proteins, 390 recombinant DNA studies, 452-455 serum, see Serum transferrins site-specific mutants, 454 solution X-ray scattering measurements, 409-410 structure, 393-418 conformational change, 407-41 1 disulfide bonding, 402-403 domain organization, 397-398 general organization, 397-400 half- and quarter-molecules, 395-396,411-412 metal and anion sites, 403-407 polypeptide folding, 400-402, 417-418 primary, 393-396 similarities with bacterial binding proteins, 416-418 three-dimensional, 396-412 variations among, 412-416 UV difference spectra, 419, 421

Transient absorption studies, class I1 mixed-valence complexes, 303 Trapping experiments, gallium hydrides, 180-181 Tris(pentafluorosulfanyl)amine,150 Tyrosine, involvement in transferrin iron binding, 403-404

U UV difference spectra, transferrins, 419, 42 1

V Vacuum-ultraviolet spectrum, molecular phosphorus oxides, 362 Vanadium, transferrin binding, 423-424 Vapor transfer, gallium hydrides, 177, 179-180 Vibrational spectroscopy aluminum hydrides, 223-224 gallane, 200-215 gallium hydrides, 182-184 molecular phosphorus oxides, 346-349 Visible absorption spectra diferric transferrin complexes, 433 lactoferrin, 423-424 VSEPR theory, element(II1) halogenoanions. 265-266

W Wave functions, 217 A

X-ray absorption spectroscopy molecular phosphorus oxides, 355-359 molecular phosphorus oxide sulfides, 378-379

This Page Intentionally Left Blank

CONTENTS OF PREVIOUS VOLUMES

VOLUME 31 Preparation and Purification of Actinidine Metals J . C. Spirlet, J R . Peterson, and L . B . Asprey Astatine: Its Organonuclear Chemistry and Biomedical Applications J . Brown Polysulfide Complexes of Metals A . Miiller and E . Diemann Iminoboranes Peter Paetzold Synthesis and Reactions of' Phosphorus Rich Silphosphanes G . Fritz INDEX

VOLUME 32 Dynamics of Spin Equibria in Metal Complexes James K Beattie Hydroxo-Bridged Complexes of Chromium(III), Cobalt(II), Rhodium(III), and Iridium(II1) Johan Springborg Catenated Nitrogen Ligands Part 11 Transition Metal Derivatives of Trizoles, Tetrazoles, Pentazoles, and Hexazine David S Moore and Stephen D Robinson

The Redox Chemistry of Nickel A . Graham Lappin and Alexander McAuley Nickel in Metalloproteins R . Cammack Nitrosyl Complexes of Iron-Sulfur Clusters Anthony R . Butler, Christopher Glidewell, and Min-Hsin Li INDEX

VOLUME 33 1.6-Disubstituted Triptycenes Alan G . Massey Cysteine-Containing Oligopeptide Model Complexes of Iron-Sulfur Proteins Akira Nakamura and Norikazu Ueyama Reduction Potentials Involving Inorganic Free Radicals in Aqueous Solution Daijid M . Stanbury The Nitrogen Fluorides and Some Related Compounds H. J . Emeleus. Jean'ne M. Shreeve, and R . D. Verma Higher Oxidation State Manganese Biomolecules John B . Vincent and George Chrisfou Double Bonds between Phosphorus and Carbon R . Appel and F. Knoll INDEX

417

478

CONTENTS OF PREVIOUS VOLUMES

VOLUME 3 4

VOLUME 36

Homoleptic Complexes of 2,2'-Bipyridine E . C CoristablP

Inorganic Chemistry and Drug Design Peter J . Satiler

Compounds of Thorium and Uranium in Low (c:IVl Oxidation States Isabel Santos. A . Pirrs de Matos, and Alfred G . Maddork

Lithium and Medicine: Inorganic Pharmacology N. .J. Birch and J . D . Phillips

Leaving Groups on Inert Metal Complexes with Inherent or Induced Liability Geoffrey A Lnwrancr The Coordination of Metal Aquaions G. W . Neilson and I E .Enderbv An Appraisal of Square-Planar Substitution Reactions R . J Cross Transition Metal Nitrosyl Complexes D . Michael P. Mingos and Durren J . Shermari

The Mo-. V-, and Fe-Based Nitrogenase Systems of Azobacter Robert R . Eadv T h e Extraction of Metals from Ores IJsing Bacteria U . Keith E w a r f and Martin N Hughes

Sol id -State Bioi norganic Chemist rq' : Mechanisms and Models of Biomineralization Stephen Mann and Carole C . Per? Magnetic Circular Dichroism of Hemoproteins M R . Cheesman. C . Greenic~ood.arid A . J Thomson

INDEX

VOLUME 35 Chemistry of Thioether Macrocyclic Complexes Alexander J . Blak<Jand Martin Schrdder Vanadium: A Biologically Kelevant Element Ron Wever and Kenneth K i d i n Structure, Reactivity, Spectra, and Redox Properties of CobaltlIII) Hexaarnines Philip Hendry and Andreas 1,udi The Metallic Face of Boron Thomas P. Fehlner Developments in Chalcogen-Halide Chemistry Burnt Krebs and Frank-Peter Ahler-s Interaction between Optical Centers and Their Surroundings: An Inorganlc Chemist's Approach G. Blasse INI)EX

Flavocytochrome b, Stephen K . C h a p m a n , Scott A . White. and Graeme A . Reid X-Ray Absorption Spectroscopy and t h e Structures of Transition Metal Centers in Proteins C David Garner Direct Electrochemistry of Proteins and Enzymes Liang-Hong Guo and H . Allen 0 . Hill Active-Site Properties o f te Blue Copper Proteins A . G. Sykes The Uptake, Storage, and Mobilization of Iron and Aluminum in Biology S . Jemil A . Faterni, Fahini H A . Kadir, and D U LJ .~Williamson, mid Geoffrey R Moore Probing Structure-Function Relations in Ferritin and Bacterioferritin P . M . Harrison, S . C . Andres. P . J . A r t y m i u k , G. C . Ford, J . R . Guest, J . Hirzmann. D M L O U J S O I I .

CONTENTS OF PREVIOUS VOLUMES

479

Dynamic Electrochemistry of Iron-Sulfur Proteins Frascr A . Armstrong VOLUME 37 On the Coordination Number of the Metal in Crystalline Halogenocuprates(1)and Halogenoargentates(1) Susan Jagner and Goran Helgesson Structures of Organonitrogen-Lithium Compounds Recent Patterns and Perspectives in Organolithium Chemistry Karma Gregorv. Paul uon RagiLe Schleyer. and Ronald Snaith Cubane and Incomplete Cubane-Type Molybdenum and Tungsten 0 x 0 1 Sulfido Clusters Takashi Shibahara Interactions of Platinum Amine Compounds with Sulfur-Containing Biomolecules and DNA Fragments Edwin L M Lempers and J a n Reedijk Recent Advances in Osmium Chemistry Peter A Lay and W Dean Harman Oxidation of Coordinated Diimine Ligands in Basic Solutions of Tristdiimine)iron(III), ruthenium( 1111, and -0smiurntIII1 0 M ~ n s t e dand G Nord INIIEX

EPR Spectroscopy of Iron-Sulfur Proteins Wilfred R . Hagen Structural and Functional Diversity of Ferredoxins and Related Proteins Hiroshi Matsubara a n d Kazuhlko Saeki Iron-Sulfur Clusters in Enzymes: Themes and Variations Richard C a n m a c k Aconitase: An Iron-Sulfur Enzyme Mary Claire Kennedy a n d C . David stout Novel Iron-Sulfur Centers in Metalloenzymes and Redox Proteins from Extremely Thermophilic Bacteria Michael W . W . A d a m s Evolution of Hydrogenase Genes Gerrit Voordoww Density-Functional Theory of Spin Polarization and Spin Coupling in Iron-Sulfur Clusters Louis Noodleman and David A . Case INDEX

VOLUME 39

VOLUME 38

Synthetic Approach to t h e Structure and Function of Copper Proteins Nobumasa Kitajima

Trinuclear Cuboidal and Heterometallic Cubane-Type Iron-Sulfur Clusters New Structural and Reacticity Themes i n Chemistry and Biology R H Holm

Transition Metal and Organic RedoxActive Macrocycles Designed to Electrochemically Recognize Charged and Neutral Guest Species Paul D . Beer

Replacement of Sulfur by Selenium in Iron-Sulfur Proteins Jacques Meyer, Jean-Marc Moulis. Jacques Gaillurd. and Marc Lutz

Structures of Complexes in Solution Derived from X-Ray Diffraction Measurements Georg Johanssan

480

CONTENTS OF PREVIOUS VOLUMES

High-Valent Complexes of Ruthenium and Osmium Chi-Ming Che arid Vivian Wing W a h Yam Heteronuclear Gold Cluster Compounds D Michael P Mingos and Michael J Watson Molecular Aspects on the Dissolution and Nucleation of Ionic Crystals in Water Hitoshr Ohlakc INDEX

VOLUME 40 Bioinorganic Chemistry of Pterin-Containing Molybdenum and Tungsten Enzymes ./ohti H . Etteirtark a i d Charles G . Yo/tr/g

Structure and Function of Nitrogenase Douglas C . Rees. Michael K . Chan. arid Jorlgsurl Ki1,l Blue Copper Oxidases A . Messerschrnidl Quadruply Bridged Dinuclear Coniplexes of Platinum, Palladium. and Nickel Keisrike Uiiinkoshi arid Yoichi Sasaki Octacyano and 0x0- and Nitridotetracyano Complexes of Second and ‘Third Series Early Transition Metals Johatiri G . Leipoldt. Stephen S . Bassotr. and Aitdreas Rood1 Macrocyclic Complexes a s Models for Nonporphine Metalloproteins Vichie McKee Complexes of Sterically Hindered Thiolate Ligands J . R . Dilic,orth and J . H u INDEX