ADVANCES IN PHOTOCHEMISTRY Volume 28 Editors
DOUGLAS C. NECKERS Center for Photochemical Sciences, Bowling Green State ...
77 downloads
1753 Views
2MB 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
ADVANCES IN PHOTOCHEMISTRY Volume 28 Editors
DOUGLAS C. NECKERS Center for Photochemical Sciences, Bowling Green State University, Bowling Green, Ohio
THOMAS WOLFF Technische Universita¨t Dresden, Institut fu¨r Physikalische Chimie und Elektrochimie, Dresden, Germany
WILLIAM S. JENKS Department of Chemistry, Iowa State University, Ames, Iowa
A JOHN WILEY & SONS, INC., PUBLICATION
ADVANCES IN PHOTOCHEMISTRY Volume 28
ADVANCES IN PHOTOCHEMISTRY Volume 28 Editors
DOUGLAS C. NECKERS Center for Photochemical Sciences, Bowling Green State University, Bowling Green, Ohio
THOMAS WOLFF Technische Universita¨t Dresden, Institut fu¨r Physikalische Chimie und Elektrochimie, Dresden, Germany
WILLIAM S. JENKS Department of Chemistry, Iowa State University, Ames, Iowa
A JOHN WILEY & SONS, INC., PUBLICATION
Copyright # 2005 by John Wiley & Sons, Inc. All rights reserved. Published by John Wiley & Sons, Inc., Hoboken, New Jersey. Published simultaneously in Canada. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400, fax 978-646-8600, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008. Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. For general information on our other products and services please contact our Customer Care Department within the U.S. at 877-762-2974, outside the U.S. at 317-572-3993 or fax 317-572-4002. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print, however, may not be available in electronic format. Library of Congress Cataloging in Publication Data: Library of Congress Catalog Card Number: 63-13592 ISBN 0-471-68241-1 Printed in the United States of America 10 9
8 7 6
5 4 3 2
1
CONTRIBUTORS
Y. Alexeev Pacific Northwest National Laboratory 901 Battelle Boulevard P.O. Box 999 Richland, WA 99352 P. K. Chowdhury Biomedical Application Centre Centre of Advanced Technology Indore-452 013 India K. Das Biomedical Application Centre Centre of Advanced Technology Indore-452 013 India M. D. D’Auria Dipartimento di Chimica Universita´ della Basilicata Via N. Sauro 85 85100 Potenza Italy
Rolf Dessauer P.O. Box 3796 Greenville, DE 19807 L. Emanuele Dipartimento di Chimica Universita´ della Basilicata Via N. Sauro 85 85100 Potenza Italy Takashige Fujiwara Department of Chemistry Center for Laser and Optical Spectroscopy The University of Akron Akron, OH 44325 M. S. Gordon Pacific Northwest National Laboratory 901 Battelle Boulevard P.O. Box 999 Richland, WA 99352
v
vi
M. Halder Biomedical Application Centre Centre of Advanced Technology Indore-452 013 India Edward C. Lim Department of Chemistry Center for Laser and Optical Spectroscopy The University of Akron Akron, OH 44325 David C. Moule Department of Chemistry Brock University St. Catherines, ON L2S3A1 Canada
CONTRIBUTORS
J. Park Department of Chemistry University of Pennsylvania Philadelphia, PA 19104 J. W. Petrich Pacific Northwest National Laboratory 901 Battelle Boulevard P.O. Box 999 Richland, WA 99352 R. Racioppi Dipartimento di Chimica Universita´ della Basilicata Via N. Sauro 85 85100 Potenza Italy
PREFACE
Volume 1 of Advances in Photochemistry appeared in 1963. The stated purpose of the series was to explore the frontiers of photochemistry through the medium of chapters written by pioneers who are experts. Teh editorial policy has always been to solicit articles from scientists who have strong personal points of view, while encouraging critical discussions and evaluations of existing data. In no sense have the articles been simple literature surveys, although in some cases they may have also fulfilled that purpose. In the introduction to Volume 1 of this series, the founding editors, J. N. Pitts, G. S. Hammond and W. A. Noyes Jr. noted developments in a brief span of prior years that were important for progress in photochemistry: flash photolysis, nuclear magnetic resonance, and electron spin resonance. A quarter of a century later, in Volume 14 (1988), the editors noted that since then two developments had been of prime significance: the emergence of the laser from an esoteric possibility to an important light source and the evolution of computers to microcomputers in common laboratory use of data acquisition. These developments strongly influenced research on the dynamic behavior of the excited state and other transients. With the increased sophistication in experiment and interpretation since that time, photochemists have made substantial progress in achieving the fundamental objective of photochemistry: elucidation of the detailed history of a molecule that absorbs radiation. The scope of this objective is so broad and the systems to be studied are so many that there is little danger of exhausting the subject. We hope that this series will reflect the frontiers of photochemistry as they develop in the future. Rolf Dessauer’s contribution, ‘‘The Invention of Dylux1 Instant-Access Imaging Materials and the Development of Habi Chemistry—A Personal History,’’ is a departure from previously established patterns for Advances in vii
viii
PREFACE
Photochemistry chapters. Rolf’s personal history is of the development of one of the most important commercial contributions made by photoscientists in the twentieth century. We hope you’ll find his trials and tribulations from invention to commercial success along the path that led to a successful Dylux instructive and inspirational. Bowling Green, Ohio, USA Dresden, Germany Ames, Iowa, USA
Douglas C. Neckers Thomas Wolff William S. Jenks
CONTENTS
Hypericin and Its Perylene Quinone Analogs: Probing Structure, Dynamics, and Interactions with the Environment K. DAS, M. HALDER, P. K. CHOWDHURY, J. PARK, Y. ALEXEEV, M. S. GORDON, AND J. W. PETRICH
1
Thiophosgene: A Tailor-Made Molecule for Photochemical and Photophysical Studies DAVID C. MOULE, TAKASHIGE FUJIWARA, AND EDWARD C. LIM
27
1,2-Cycloaddition Reaction of Carbonyl Compounds and Pentaatomic Heterocyclic Compounds M. D. D’AURIA, L. EMANUELE, AND R. RACIOPPI
81
The Invention of Dylux1 Instant-Access Imaging Materials and the Development of Habi Chemistry—A Personal History ROLF DESSAUER
129
Index
263
Cumulative Index, Volumes 1–28
277
ix
HYPERICIN AND ITS PERYLENE QUINONE ANALOGS: PROBING STRUCTURE, DYNAMICS, AND INTERACTIONS WITH THE ENVIRONMENT M. Halder, P. K. Chowdhury, M. S. Gordon, J. W. Petrich Department of Chemistry, Iowa State University, Ames, IA 50011, USA K. Das Biomedical Applications Centre, Centre of Advanced Technology, Indore-452 013, India J. Park Department of Chemistry, University of Pennsylvannia, Philadelphia, PA 19104-6323, USA Y. Alexeev Pacific Northwest National Laboratory, 901 Battelle Blvd., P.O. Box 999, Richland, WA 99352, USA.
CONTENTS I. II.
Introduction: Excited-State Intramolecular H-Atom Transfer in HypericinLike Perylene Quinones Outstanding Questions Regarding Hypericin-Like Perylene Quinones
Advances in Photochemistry, Volume 28 Edited by Douglas C. Neckers, William S. Jenks, and Thomas Wolff # 2005 John Wiley & Sons, Inc.
1
2
HYPERICIN AND ITS PERYLENE QUINONE ANALOGS
A. Ground-State Heterogeneity 1. Recent Theoretical and Computational Approaches 2. Recent Experimental Approaches B. Are There Multiple H-Atom Transfers in the Perylene Quinones? C. Identifying the H-Atom Translocation III. Concluding Remarks Acknowledgments References
I. INTRODUCTION: EXCITED-STATE INTRAMOLECULAR H-ATOM TRANSFER IN HYPERICINLIKE PERYLENE QUINONES Hypericin and hypocrellin (Fig. 1.1.) are naturally occurring perylene quinones that have generated great interest recently owing to their biological activity [3–22], in particular, their light-induced biological activity [23–27]. The importance of light for their function has motivated our study of the photophysics of hypericin and its analogs [28–38]. By means of H/D substitution, investigation of O-methylated analogs, and complementary studies using both transient absorption and fluorescence upconversion spectroscopies, we have argued that the major primary photophysical process in hypericin and hypocrellin A in organic solvents is excited-state hydrogen atom transfer. We have suggested that the labile protons resulting from the intramolecular hydrogen-atom transfer reactions may be important for understanding the light-induced biological activity of hypericin and hypocrellin A. Notably, hypericin and hypocrellin A acidify their surroundings upon light absorption [39–41]. The role of photogenerated protons takes on significance in the context of the growing body of literature implicating changes in pH with inhibition of virus replication [42], antitumor activity [43, 44], and apoptosis (programmed cell death) [45–47]. Our argument for intramolecular excited-state H-Atom transfer in hypericin is as follows. The deshydroxy analog of hypericin, mesonaphthobianthrone (Fig. 1.1), is nonfluorescent, except in strong acids [29, 38] (e.g., sulfuric acid), where it produces a fluorescence spectrum that has nearly the same shape as that of hypericin in alcohols (Fig. 1.2). These results demonstrate the importance of a ‘‘protonated’’ carbonyl group for producing hypericin-like fluorescence. The hypericin emission spectrum grows in on a 6–12-ps time scale in all solvents, except in sulfuric acid where it is instantaneous. Thus the rise time for the appearance of the hypericin emission is taken as evidence for an excitedstate H-atom transfer [29, 48]. Confirming this interpretation are the fluorescence upconversion measurements of hypericin and O-methyl hypericin analogs [38, 48, 49], which are incapable of executing intramolecular excited-state
3
Figure 1.1. Hypericin (normal form, double tautomer, and monotautomer), mesonaphthobianthrone, hypocrellin A (normal form, double tautomer, and monotautomer), hypocrellin B, hypomycin B [1], and calphostin C [2].
4
HYPERICIN AND ITS PERYLENE QUINONE ANALOGS
Figure 1.2. Comparison of the fluorescence excitation and emission spectra of hypericin (a) and hypocrellin (b) in a 1:1 ethanol/methanol mixture at room temperature. The excitation spectra were measured at 295 K (solid line) and at 77 K (dashed line). The excitation spectra were monitored at 650 nm at 295 K and at 620 nm at 77 K.
H-atom-transfer reactions. Hypericin exhibits a rise in its fluorescence signal, whereas the methylated derivatives do not. Rising components of 10 ps, attributable to intramolecular H-atom transfer are clearly observed in the fluorescence upconversion traces of both hypericin and hypocrellin A (Figs. 1.3 and 1.4, respectively). For simplicity, in the rest of our discussion, we refer to this shorter-lived component as the ‘10-ps component.’ The amplitude of the rising component is emission wavelength dependent and occurs on the blue edge of the emission spectra. The clear and complementary observation in fluorescence of the 10-ps component in both hypericin and hypocrellin A is a crucial link in providing a unified model of the hypericin and
INTRODUCTION: EXCITED-STATE INTRAMOLECULAR H-ATOM TRANSFER
5
Figure 1.3. (Top) Fluorescence upconversion transient for hypericin in ethanol at lem ¼ 576 nm. The fit curve is described by the following equation (with background subtracted): FðtÞ ¼ 0:21 exp ðt=6:5 psÞ þ 1:00 exp ðt=1Þ. (Bottom) At lem ¼ 653 nm, however, there is no rising component in the fluorescence trace. Similar behavior is observed for hypocrellin A (Fig. 1.4). The excitation wavelength was the second harmonic of our unamplified Ti:sapphire oscillator, 414 nm. The panel below the kinetic trace displays the residuals between the fit and the data.
hypocrellin photophysics [34]. In addition to the 10-ps component in hypocrellin A, there is a longer-lived transient, also assigned to H-atom transfer, whose time constant ranges from 50 to 250 ps in the solvents we have studied [33]. The H-atom transfer rate in hypericin has no significant solvent dependence [29]. The H-atom transfer rate for the longer component in hypocrellin A has a strong dependence on the bulk viscosity [33]. The time constant for H-atom
6
HYPERICIN AND ITS PERYLENE QUINONE ANALOGS
Figure 1.4. A series of upconversion traces for hypocrellin A in octanol collected at different emission wavelengths. FðtÞ ¼ 0:10 exp ðt=4:1 psÞ þ 1:00 exp ðt=1Þ; lem ¼ 591 nm. Note that at redder emission wavelengths, the amplitude of the rising component is negligible.
transfer in hypocrellin ranges from 50 to 250 ps in the solvents we have studied. The viscosity dependence is remarkable not only because it is absent in hypericin but also because it is exceedingly well described by a bulk effect and does not require specific consideration of the structural aspects of the solvents, which vary considerably. It is often the case that trends are followed only for solvents of a given kind, for example, alkane or alcohols, primary alcohol or higher degree alcohol, and hydrogen bonding or nonhydrogen bonding. NMR results cited below lead us to suggest that the viscosity dependence on the excited-state transfer process is a consequence of the coupling of the H-atom transfer to conformational changes of the seven-membered ring in hypocrellin [50]. Hypericin does not have such a ring and does not exhibit such viscosity-dependent effects. We now consider two objections that may be raised to the assignment of the excited-state processes in hypericin and hypocrellin as H-atom transfer. These are the absence of a deuterium isotope effect for the 10-ps component of hypericin and hypocrellin and the mirror image symmetry between the absorption and the emission spectra in hypericin and hypocrellin.
OUTSTANDING QUESTIONS REGARDING HYPERICIN-LIKE PERYLENE QUINONES
7
The lack of a deuterium isotope effect [29, 32, 34, 38, 51] may be attributed to the reaction coordinate not being identified with the proton coordinate. There is precedent for this in other systems [52–54]. Requiring the absence of mirror image symmetry between the absorption and the emission spectra assumes that the potential energy surface of the emitting species is significantly different from that of the absorbing species. Such a displacement in the coordinate of the emitting species is clearly evident in the most commonly studied proton transfer systems (Fig. 1.5): methyl salicylate [55], 7-azaindole dimer [56, 57], 2-phenyl-benzotriazole [58], and 3-hydroxyflavone [59–63]. If, however, we consider systems in which the normal and tautomer species are symmetric, or nearly so, this disparity no longer exists or is significantly minimized. 5-Hydroxytropolone [64, 65] presents an excellent example of such a case. Other examples are the double H-atom transfer in naphthazarin [66] and in the 4,9-dihydroxyperylene-3,10-quinone subunit of hypocrellin, producing entirely symmetric structures (Fig. 1.5). We argue that hypericin and hypocrellin A have very similarly symmetric normal and tautomeric forms, as indicated by the highlighted bond systems in Fig. 1.1. That is, regardless of the tautomeric form in which the molecule finds itself, there is always conserved an aromatic core to which is attached a hydroxyl group peri to a carbonyl group. Even in the case of the monotautomerized species, it is possible to draw resonance forms that, upon superposition, restore the aromatic character of the substructure involved in the H-atom-transfer reaction. Consequently, we conclude that the mirror image symmetry observed in hypericin and hypocrellin is not at all surprising. If, on the other hand, the excitedstate reaction were a genuine proton transfer, then the resulting charge-separated species would be expected to exhibit an emission spectrum significantly different from that of the absorption spectrum, as in 3-hydroxyflavone, and the rate of reaction should be very sensitive to solvent polarity, which is not the case for hypericin or hypocrellin.
II. OUTSTANDING QUESTIONS REGARDING HYPERICIN-LIKE PERYLENE QUINONES A. Ground-State Heterogeneity As noted above, a possible objection to our assignment of the excited-state reaction to H-atom transfer in these perylene quinone systems is the observation of mirror image symmetry between the absorption and the emission spectra, which indicates minimal structural changes between the absorbing and the emitting species. Our first attempt to explain this symmetry was to suggest that the
8
HYPERICIN AND ITS PERYLENE QUINONE ANALOGS
Figure 1.5. Proton-transfer and H-atom-transfer species. In the upper half of the figure, the product of tautomerization is structurally and electronically different from the reactant. In contrast, H-atom transfer in 5-Hydroxytropolone results in nearly identical species, and double-tautomers of naphthazarin and 4,9-dihydroxyperylene-3,10-quinone are the ‘‘mirror’’ images of ‘‘normal’’ structures.
OUTSTANDING QUESTIONS REGARDING HYPERICIN-LIKE PERYLENE QUINONES
9
ground state of hypericin was populated with at least one other species, for example, a monotautomer [29]. This seemed to be reasonable, especially given the breadth of the visible absorption spectrum: There are no ‘‘gaps’’ of zero absorbance anywhere between the ultraviolet and 600 nm. Temperature-dependant 1H NMR and 2D-ROESY studies of hypericin, however, indicate that there is only one conformer/tautomer for hypericin in the ground state [50]. On the other hand, the NMR measurements indicate that three ground-state species are significantly populated for hypocrellin A [48, 50], owing largely to the flexibility of the seven-membered ring. That the seven-membered ring plays an important role in determining the populations of conformers and tautomers in the ground state is demonstrated by the NMR study of hypocrellin B. The seven-membered ring of hypocrellin B contains a double bond and is consequently more rigid than that of hypocrellin A. Only one ground-state conformer/tautomer is observed by NMR for hypocrellin B.
1. Recent Theoretical and Computational Approaches Previous quantum mechanical calculations also indicate that for hypericin the ground state is much less heterogeneous than we had believed. Based on ab initio calculations, (RMP2/6-31G(d) level of theory, using geometries obtained with the 3-21G basis and Hartree-Fock wavefunctions) only one hypericin species, the ‘‘normal’’ form, is populated in the ground state for an unionized gas phase species [67] (Fig. 1.6). Here we present more recent calculations using analytical instead of numerical Hessians, as was done in the earlier calculation. The initial structure used for hypericin was taken from previous work [67]. The structure was then reoptimized at the restricted Hartree-Fock (RHF) level theory, using the 6-31G(d) basis set. The Hessian (matrix of energy second derivatives with respect to Cartesian coordinates) of the optimized structure was calculated using a recently developed analytic Hessian method [68]. Diagonalization of the Hessian provides harmonic normal modes and corresponding vibrational frequencies [69]. Transition states and minima are indicated by 1 and no imaginary frequency mode, respectively. Hessian calculations provide vibrational frequencies and a diagnostic for the nature of a stationary point. There are two types of Hessian calculations: seminumerical, using a finite difference of analytic gradients, and fully analytic. The analytic approach employed in our method is usually preferable due to the significantly increased accuracy of the calculated vibrational frequencies as well as its considerable time savings. The relative efficiency and accuracy of analytic Hessians increase with the size of the molecule. All calculations presented were performed with the quantum chemistry program GAMESS [70].
10
HYPERICIN AND ITS PERYLENE QUINONE ANALOGS
Figure 1.6. (Left) Optimized structures for potential-energy minima of hypericin based on the RMP2/6-31G(d) level of theory, using geometries obtained with the 3-21G basis and Hartree-Fock wavefunctions. At the top, two views are shown for the ‘‘normal’’ form (Fig. 1.1), which is here referred to as the minimum energy structure (MIN). M1 and M2 are monotautomers; D1 and D2 are double tautomers. (Right) Estimated relative energies (kcal/mol) of minima and transition states. Energies including zero-point corrections are given in parentheses.
The structure of hypericin in its ‘‘normal’’ form is presented in Figures 1.7 and 1.8. It was established [67] that this C2 structure is the global minimum on the ground electronic state. The functional groups on carbon atoms 1, 13, and 14 (Figs. 1.1 and 1.7) are of primary interest. In particular, the distances between the oxygen atoms are crucial in understanding the H-atom-transfer kinetics and has been discussed by us in detail in our considerations of a unified picture of the hypericin and hypocrellin photophysics [34, 48, 71]. In the present calculation, ˚, the distances of the hydrogen bonds involving both O-H bonds are about 1.75 A slightly longer than in the previous work in which the 3-21G basis set was used [67]. 2. Recent Experimental Approaches Our most recent attempt to investigate the problem of ground-state heterogeneity experimentally uses tunable pump laser pulses derived from a home-made optical parametric amplifier (Fig. 1.9). A white light continuum is used to seed a two-stage optical parametric amplifier pumped by the second harmonic of a regeneratively amplified Ti-sapphire laser
OUTSTANDING QUESTIONS REGARDING HYPERICIN-LIKE PERYLENE QUINONES
11
Figure 1.7. Structure of hypericin in the ‘‘normal’’ form (7,14-dioxo tautomer). See the text for details.
operating at 815 nm. Our system is based on the design by Greenfield and Wasielewski [72]. A 10% beam splitter is used to divert a small fraction of the compressor output for continuum generation. A combination of a l/2 plate (HWP) and a thin-film polarizer (TFP) are used to control the exact intensity of the red before focusing it into a 1-mm-thick fused silica. The continuum generated is passed through a short-pass filter to eliminate any remaining fundamental
Figure 1.8. Side-on view of the hypericin ‘‘normal’’ form (7,14-dioxo tautomer). See the text for details.
12
HYPERICIN AND ITS PERYLENE QUINONE ANALOGS
Figure 1.9. The optical parametric amplifier (OPA). BS, beam splitter; TFP, thin-film polarizer; HWP, half-wave plate; LWP, dichroic beam splitter or long wave pass filter; F, filter.
and then through a HWP to obtain the right polarization for type II phase matching for the 3-mm-thick type II BBO crystal. The remaining IR is then reduced with a 2 telescope and passed through a 2-mm-thick LBO-I. The blue light generated at 407 nm is split into two parts by a combination of HWP and TFP. We used roughly equal amounts of blue to pump the first and second stage of the OPA. In the first stage, we used a single 500-mm lens to control the spot size of the blue at the crystal. Care was taken to avoid continuum generation by the blue itself, by not focusing it directly at the crystal. We have found that using a single long-focal-length lens in the first stage instead of using a telescope significantly increases the stability of the output of the first stage. The white light and the blue were combined with a dichroic beam splitter and spatially and temporally overlapped in the BBO-II. In the second stage, the residual blue spot size was reduced with a 3 telescope and passed through a HWP to have the matching polarization for type-I phase matching. The parametric output from the first stage and the residual blue were again combined with a dichroic beam splitter and spatially and temporally overlapped at the 2-mm-thick type-I BBO. Finally, a filter is used to select the tunable output in the visible region by rejecting the idler and the residual blue. Typically, the OPA
OUTSTANDING QUESTIONS REGARDING HYPERICIN-LIKE PERYLENE QUINONES
13
gives about 1 mJ of energy and is tunable from 475 to 700 nm. Tunability is achieved by angle tuning of the two BBO crystals simultaneously. This tunable source was used to investigate the transient absorption kinetics of hypericin in DMSO as a function of pump wavelength (Fig. 1.10). The startling result is that using pump wavelengths from 495 to 600 nm, the data can be fit globally by a sum of two exponentials, which except for two traces exhibits the 10-ps component characteristic of H-atom transfer. Fit results are compiled in Table 1.1. Thus, both experimental and theoretical work continue to point toward the remarkable conclusion that hypericin in the ground state exists in only one tautomeric or conformational form, as indicated by both optical and NMR spectroscopies.
Figure 1.10. Transient absorption kinetics for hypericin as a function of pump wavelength. In each panel, the first value corresponds to the pump wavelength; the second, to the probe wavelength. The significant feature of these data is that they can be uniformly fit to two time constants, of which one corresponds to the characteristic H-atom-transfer time of 10 ps. The results of a global fit to these data are compiled in Table 1.1.
14
HYPERICIN AND ITS PERYLENE QUINONE ANALOGS
Figure 1.10. (Continued)
OUTSTANDING QUESTIONS REGARDING HYPERICIN-LIKE PERYLENE QUINONES
15
TABLE 1.1 Absorption Transients for Hypericin in DMSO as a Function of Pump Wavelengtha l Pump (nm) 495 495 495 495 520 520 520 520 600 600
l Probe (nm) 560 600 610 660 560 600 610 660 610 660
a1 0.09 0.23 0.39 0.20 0.15 0.13 0.22 0.20 — —
a2 1 1 1 1 1 1 1 1 1 1
a
The absorption transients are fit globally to a sum of two exponentials: AðtÞ ¼ a1 exp ðt=t1 Þ þa2 exp ðt=t2 Þ. The two time constants are kept fixed for all the kinetic traces at values of 10 and 5000 ps for t1 and t2, respectively. The value of 5000 is approximately the value of the fluorescence lifetime of hypericin, which is long on the time scale used to acquire these kinetic traces and may be considered to be infinite.
B. Are There Multiple H-Atom Transfers in the Perylene Quinones? Given the structure of the perylene quinones of interest here (Fig. 1.1), with either one (hypocrellin A) or two (hypericin), hydroxyl groups peri to a carbonyl on either end of the molecule, one is naturally inclined to inquire how many hydrogen atoms are transferred in the excited state. And, if more than one is transferred, one must necessarily inquire whether the process is stepwise or concerted. (It is useful to note the difference between a concerted and a synchronous reaction. A concerted reaction takes place in a single kinetic step, with no reaction intermediate, in which some of the changes in bonding take place to different extents in different parts of the reaction. A synchronous reaction is one in which all the bond-making and bond-breaking processes take place at the same time and proceed at the same extent during the reaction [73, 74]. It is a common error to assume that concertedness implies synchrony.) The availability of hypomycin B [1], where there is only one peri hydroxyl group and only one intramolecular hydrogen bond, provides an excellent means to investigate these questions. Hypomycin B is unique in that it has only one intramolecular hydrogen bond as opposed to the two in hypocrellin A and the four in hypericin (Fig. 1.1). Picosecond transient absorption data for hypomycin B fail to reveal any stimulated
16
HYPERICIN AND ITS PERYLENE QUINONE ANALOGS
Figure 1.11. Comparison of the kinetic traces of hypomycin B in MeOD (solid lines) and in MeOH (dashed lines). The probe wavelengths are given in the top right hand corner of each panel. A global fit was carried out to fit the decays of the several wavelengths. In MeOH, the global time constant was 82 ps, whereas that in MeOD was 75 ps; lex ¼ 407 nm.
emission, let alone rise time in stimulated emission (unlike hypocrellin and hypericin), which we have interpreted in terms of excited-state H-atom transfer (Fig. 1.11). Furthermore, a global analysis of the excited-state kinetics at four different probe wavelengths in MeOD and MeOH yields no significant difference in the excited-state kinetics: The time constants are 82 and 75 ps, respectively. Transient absorption data for hypomycin B in other solvents, such as DMSO and buffer-DMSO mixtures (data not shown), also failed to reveal any rise time in the stimulated emission. In the context of our previous arguments and criteria for identifying H-atom transfer in hypericin, hypocrellin, and their analogs, one might hastily conclude that hypomycin B does not undergo this process.
OUTSTANDING QUESTIONS REGARDING HYPERICIN-LIKE PERYLENE QUINONES
17
If subsequent experiments do indeed demonstrate that excited-state H-atom transfer does not occur in hypomycin B, then one may draw the conclusion that multiple transfers (either concerted or stepwise) must occur in these perylene quinones and that by frustrating the process in one half of the molecule, the process in the other half is impeded. At this point, such reasoning is speculative and contrary to the growing body of evidence provided by theory and experiment. As indicated above, quantum mechanical calculations indicate that the double-H-atom transfer in hypericin [67] and in the perylene quinone nucleus [75] of hypocrellin is energetically unfavorable compared to the single-transfer event. Experiments for hypericin in which one half of the molecule cannot participate in H-atom transfer owing to protonation of the carbonyl group (or even perhaps complexation with a metal ion) [76] also indicate that the transfer process can still occur. Given the richness of the photophysics of these perylene quinones and their attendant complexity, it is premature to conclude that hypomycin B does not execute an excited-state H-atom transfer. Certainly, much more work is required to reach a proper judgment. This will include fluorescence upconversion experiments with picosecond resolution in order to interrogate the entire time scale of relevance as well as to monitor only emission, which can simplify the analysis [48]. For example, the failure to detect stimulated emission for hypomycin B may simply be the result of the presence of a strongly absorbing species in the same spectral region. Also, experiments must be performed to determine if hypomycin B, like hypericin and hypocrellin A, acts as an excited-state acid. It is clear, however, that hypomycin B is an important system with which to test and refine our current understanding of these naturally occurring perylene quinones. One of the most important questions that the study of hypomycin B might fruitfully address is the elucidation of the reaction coordinate in the H-atom-transfer process in the perylene quinones. As we have discussed in depth elsewhere, despite the similarities in the structures of hypericin and hypocrellin, which are centered about the perylene quinone nucleus, their excited-state photophysics exhibit rich and varied behavior. The H-atom transfer is characterized by a wide range of time constants, which in certain cases exhibit deuterium isotope effects and solvent dependence. Of particular interest is that the shortest time constant we have observed for the H-atom transfer is 10 ps. This is exceptionally long for such a process, 100 fs being expected when the solute H atom does not hydrogen bond to the solvent [62]. That the transfer time is so long in the perylene quinones has been attributed to the identification of the reaction coordinate with skeletal motions of the molecule [48, 50]. We have previously observed that when hypericin is bound to human serum albumin, it no longer undergoes an excited-state H-atom transfer. Assuming that the binding occurs through the interaction of one of the two carbonyl groups of hypericin and the N1-H of the single tryptophan residue (W214), which would
18
HYPERICIN AND ITS PERYLENE QUINONE ANALOGS
necessarily impede H-atom transfer on this half of the hypericin molecule, we suggested that the absence of H-atom transfer in the complex indicated concerted, double-H-atom transfer in the excited state of hypericin [36]. We suggested that H-atom transfer is completely impeded when hypericin binds to HSA because skeletal motion is coupled to the H-atom transfer [48, 50]. Fluorescence anisotropy measurements of the HSA–hypericin complex indicate that the hypericin is rigidly bound and that there is no rapid restricted motion of hypericin relative to the protein. By analogy, one might argue that if H-atom transfer does not occur in hypomycin B, it is not because the process requires that two H atoms be in flight but because the required skeletal motion is restricted by the presence of the O-C-O bond. Although this response is plausible, it is not easy to reconcile it with the observation that hypericin undergoes Hatom transfer in a glass at low temperatures (the energy of activation is 0.05 kcal/mol) [37], where the amplitude of skeletal motion would seem to be less than that in the HSA matrix. These sorts of problems and questions continue to illustrate the need for further elucidation of the reaction coordinate for the H-atom transfer in hypericin and its analogs. To conclude this section, we note that our assignment of excited-state H-atom transfer to the primary photoprocess in hypericin, hypocrellin, and their derivatives has occasioned some objections, to which we refer in the introduction and that we address in detail elsewhere [48, 71]. An additional concern, which has been brought to our attention and which is relevant in the light of the previous discussion, is the following. As we note above, we have measured the energy of activation for the H-atom transfer in hypericin to be 0.05 kcal/ mol (or 20 cm1). The absence of an isotope effect for the hypericin reaction (and for the 10-ps reaction in hypocrellin A) indicates that the reaction coordinate is not the hydrogen atom coordinate (which theoretical and experimental results suggest is 1450 cm1 in the hypericin triplet [77]) and consequently must involve the skeletal motions noted above. It has been suggested by an anonymous colleague that if there is a slow H-atom transfer that is not limited by the H-atom coordinate, ‘‘then it must be that vibrational excitation of oxygen or ring modes is what limits the rate. This would correspond to a significant vibrational barrier and hence large activation energy.’’ We disagree with the last statement. Quantum mechanical calculations [67] (see also www.msg.ameslab.gov/Group/ Supplementary_Material/Hypericin/) indicate that there are four calculated frequencies (unscaled) below 100 cm1: 40 cm1, out-of-plane motion of oxygens and carbons; 48 cm1, oxygen and carbon displacements; 80 cm1, mostly OH oxygen motion; and 84 cm1: mostly OH oxygen motion. There is no dearth of low-frequency vibrations in large biological molecules, as the calculations and many experiments suggest [78–80], and we believe that such motions may indeed be coupled to the H-atom transfer in these perylene quinone systems.
OUTSTANDING QUESTIONS REGARDING HYPERICIN-LIKE PERYLENE QUINONES
19
Our previous results on hypericin indicate that excited-state H-atom transfer occurs even when one of the carbonyls is prohibited from accepting a hydrogen. The presence of such a transfer is apparent under very acidic conditions in AOT reverse micelles and cannot be excluded upon chelation of Tb3þ [76]. There is thus no evidence for a concerted H-atom-transfer mechanism in hypericin. In the present study, contrary to our initial expectations, we are not even able to demonstrate that hypomycin B executes an excited-state H-atom transfer; hence our investigation sheds no light on the general question of how many H atoms are transferred in the perylene quinones and whether the transfer is concerted or stepwise. On the other hand, if further investigation reveals that H-atom transfer does not occur in hypomycin B, the result would have considerable implications for an understanding of the reaction coordinate for the H-atom transfer.
C. Identifying the H-Atom Translocation Above we addressed the absence of a deuterium isotope effect for the 10-ps component of hypericin and hypocrellin and the mirror image symmetry between the absorption and the emission spectra in hypericin and hypocrellin. We recognize, however, that no matter how satisfactory one finds our reasoning, a direct demonstration of an excited-state H-atom transfer is required, which entails measurements of the carbonyl or hydroxyl stretching frequency as a function of time subsequent to laser excitation. A first step in this process is the identification of vibrational modes that indicate translation of the H-atom between the enol and the keto oxygens. We have addressed this problem by employing time-resolved infrared spectroscopy to hypericin and to two of its analogs that cannot undergo excited-state H-atom transfer (Fig. 1.12) [77]. The salient feature of these spectra is in the region between 1400 and 1500 cm1, where a strong band is present for hypericin and is absent in both O-hexamethoxy hypericin, which lacks hydrogen atoms that can coordinate to the carbonyl, and in the reduced analog, which lacks the necessary carbonyls. The ground-state infrared spectrum of hypericin is included at the top of Figure 1.12. Ab initio quantum mechanical calculations, at the Hartree-Fock 3-21G level [77], reveal strong normal modes in this spectral region for the triplet species of the normal form of hypericin and the 6,14- and 7,13-dioxo monotautomers (Fig. 1.1). Calculations were not performed for the triplet species of the double tautomers. Assignments of these modes have been given elsewhere [77]. Our most recent calculations on the ground-state 7,14-dioxotautomer using the analytical Hessian method and the 6-31G basis set provide the results compiled in Table 1.2 and the corresponding theoretical infrared spectrum depicted in Figure 1.13.
20
HYPERICIN AND ITS PERYLENE QUINONE ANALOGS
Figure 1.12. Comparison of the transient infrared spectra on the microsecond time scale of hypericin, O-hexamethoxy hypericin, and a hypericin analog that lacks carbonyl groups (the hexaacetoxy analog). The salient feature of the data is that the latter two compounds, which cannot execute excited state H-atom transfer owing to the absence of either labile protons or appropriate carbonyl groups, lack the feature at 1450 cm1. Ab initio calculations at the Hartree-Fock 3-21G level for the normal and two monotautomeric forms of the hypericin triplets indicate normal modes with substantial O...H...O character in the region 1400–1460 cm1 [77]. While these preliminary results do not demonstrate a time-resolved H-atom transfer, they do clearly point to a region of the spectrum that must be investigated in further studies. Hypericin and hexamethoxy hypericin, solid line, 0–1 m,s and dashed line, 14–18 ms; reduced analog, solid line, 0–0.5 ms, and dashed line, 7–9 ms.
OUTSTANDING QUESTIONS REGARDING HYPERICIN-LIKE PERYLENE QUINONES
21
TABLE 1.2 Calculated Frequencies of the Normal Mode Vibrations of the Ground State 7,14-Dioxotautomer Hypericin Species Freqency (cm1) Scaled Frequencya IR Intensity
a
b
1262
1123
16
1341 1523
1194 1356
20 13
1741
1550
14
1840
1638
53
3835
3413
12
4107
3655
6
Descriptionb C(3)-O-H, C(4)-O-H, C(2)-O, C(5)-O; motions and ring breathing Ring breathing C(1)-O-H, C(6)-O-H, C(8)-O-H, C(13)-O-H; motions and planar ring breathing Peripheral carbons stretch in plane and ring breathing Peripheral carbons stretch up and down in phase O-H stretches on C(1), C(6), C(13) O-H stretches on C(3), C(4)
The calculated frequencies are scaled by a factor of 0.89 [81]. See Figure 1.7 for atom labelling.
The major finding of this study is that a vibrational mode corresponding to H-atom translocation can been identified in hypericin by the joint contributions of synthetic, computational, and spectroscopic methods. Identification of this mode is only a first step in providing a direct demonstration of excited-state intramolecular H-atom transfer in hypericin and its analogs. There is considerable work to be accomplished. As indicated elsewhere [77], ab initio calculations predict that the normal modes in this region of the spectrum are close for the normal and the monotautomeric forms. The direct observation of the formation of the tautomer will require both adequate temporal and spectral resolution. It must be remembered, furthermore, that the identification of the H-atom translocation mode is not equivalent to the identification of the reaction coordinate. We have attributed the absence of a deuterium isotope effect on the excited-state H-atom transfer (for the 10-ps component in hypericin and hypocrellin A) to the zero-point energy in the proton coordinate lying above the barrier, with the H-atom being effectively delocalized between the two oxygen atoms. Consequently, the reaction coordinate for the excited-state H-atom transfer cannot be identified with the proton coordinate, and it must be concluded that other intramolecular motions are in fact responsible for the process. Temperature-dependent measurements indicate that these motions are extremely low amplitude, Ea ¼ 0.05 kcal/mol for hypericin [37]. Because the nature of this motion is not yet identified, we refer to it as the skeleton coordinate [48, 71, 82]. We propose that it is the time scale for this latter conformational change
22
HYPERICIN AND ITS PERYLENE QUINONE ANALOGS
Figure 1.13. Theoretical infrared spectrum of hypericin. The agreement with the experimental spectrum, presented in Figure 1.12, is quite good. Hypericin has 156 vibration frequencies. All these frequencies are presented in this simulated IR spectrum. The calculated frequencies were scaled by a factor 0.89 [81]. Frequencies in the range of 1100–1800 cm1 correspond to C-H stretches and backbone breathing. Toward the high end of the spectrum, from 3200 to 4100 cm1, the frequencies correspond to O-H stretches. The description of highest calculated frequencies is presented Table 1.2. The majority of the vibrations correspond to internal bending and stretching in the backbone of hypericin (ring breathing).
that determines the observed H-atom-transfer time. The exact nature of the conformational changes that are coupled to the H-atom-transfer reaction in hypericin and hypocrellin has yet to be identified [82].
III. CONCLUDING REMARKS We presented an overview of what we consider to be the current outstanding problems in understanding the photophysics of hypericin and its analogs. Many questions remain unanswered, and a wealth of theoretical and experimental techniques will be required to address them. It is surprising that given the richness of the physical, chemical, biological, and physiological behaviors of hypericin and
REFERENCES
23
the class of molecules to which it belongs, their importance has not been widely appreciated.
ACKNOWLEDGMENTS This work has been supported by NIH grant GM57351 and NSF grants BIR9413969 and CHE-9613962. We are grateful to J. Toscano and P. Jardon for their assistance and their helpful discussions. We thank W.-Z. Liu and H.-Y. Zhang for providing us with hypomycin B. M. Gordon thanks the Air Force Office of Scientific Research for its support.
REFERENCES 1. P. K. Chowdhury, K. Das, A. Datta, W. Z. Liu, H. Y. Zhang, and J. W. Petrich, J. Photochem. Photobiol. A Chem. 2002, 154, 107. 2. A. Datta, P. Bandyopadhyay, J. Wen, J. W. Petrich, and M. S. Gordon, J. Phys. Chem. A 2001, 105, 1057. 3. D. Meruelo, G. Lavie, and D. Lavie, Proc. Natl. Acad. Sci. USA 1988, 85, 5230. 4. J. Lenard, A. Rabson, and R. Vanderoef, Proc. Natl. Acad. Sci. USA 1993, 90, 158. 5. J. B. Hudson, J. Zhou, J. Chen, L. Harris, L. Yip, and G. H. N. Towers, Photochem. Photobiol. 1994, 60, 253. 6. W. T. Couldwell, R. Gopalakrishna, D. R. Hinton, S. He, M. H. Weiss, R. E. Law, and M. L. Apuzzo, Neurosurgery 1994, 35, 705. 7. L. Anker, R. Gopalakrishna, K. D. Jones, R. E. Law, and W. T. Couldwell, Drugs Future 1995, 20, 511. 8. W. Zhang, R. E. Law, D. R. Hinton, and W. T. Couldwell, Cancer Lett. 1997, 120, 31. 9. K. Linde, G. Ramirez, C. D. Mulrow, A. Pauls, W. Weidenhammer, and D. Melchart, Br. Med. J. 1996, 313, 253. 10. O. K. Suzuki, M. Oya, S. Bladt, and H. Wagner, Planta Medica 1984, 50, 272. 11. I. N. S. Takahashi, E. Kobayashi, H. Nakano, K. Suziki, and T. Tamaoki, Biochem. Biophys. Res. Commun. 1989, 165, 1207. 12. A. Andreoni, A. Colasanti, P. Colasanti, M. Mastrocinique, and P. Riccio, Photochem. Photobiol. 1994, 59, 529. 13. C. Thomas, R. S. MacGill, G. C. Miller, and R. S. Pardini, Photochem. Photobiol. 1992, 55, 47. 14. A. L. Vandenbogaerde, E. M. Delaey, A. M. Vantieghem, B. E. Himpens, W. J. Merlevede, and P. A. deWitte, Photochem. Photobiol. 1998, 67, 119. 15. P. Agostinis, A. Donella-Deana, J. Cuveele, A. Vandenbogaerde, S. Sarno, W. Merlevede, and P. deWitte, Biochem. Biophys. Res. Commun. 1996, 220, 613.
24
HYPERICIN AND ITS PERYLENE QUINONE ANALOGS
16. A. L. Vandenbogaerde and P. A. deWitte, Phytother. Res. 1996, 10, S150. 17. A. L. Vandenbogaerde, J. F. Cuveele, P. Proot, B. E. Himpens, W. J. Merlevede, and P. A. deWitte, J. Photochem. Photobiol. B Biol. 1997, 38, 136. 18. A. Mirossay, L. Mirossay, J. Tothova, P. Miskovsky, H. Onderkova, and J. Mojzis, Phytomedicine 1999, 6, 311. 19. L. Mirossay, A. Mirossay, E. Kocisova, I. Radvakova, P. Miskovsky, and J. Mojzis, Physiol. Res. 1999, 48, 135. 20. A. Mirossay, J. Mojzis, J. Tothova, M. Hajikova, A. Lackova, and L. Mirossay, Phytomedicine 2000, 7, 471. 21. A. Mirossay, H. Onderkova, L. Mirossay, M. Sarissky, and J. Mojzis, Physiol. Res. 2001, 50, 635. 22. A. Mirossay, L. Mirossay, M. Sarissky, P. Papp, and J. Mojzis, Physiol. Res. 2002, 51, 641. 23. N. Duran and P.-S. Song, Photochem. Photobiol. 1986, 43, 677. 24. J. W. Lown, Can. J. Chem. 1997, 75, 99. 25. Z. Diwu, Photochem. Photobiol. 1995, 61, 529. 26. G. A. Kraus, W. J. Zhang, M. J. Fehr, J. W. Petrich, Y. Wannemuehler, and S. Carpenter, Chem. Rev. 1996, 96, 523. 27. H. Falk, Angewandte Chem. Int. Ed. 1999, 38, 3117. 28. F. Gai, M. J. Fehr, and J. W. Petrich, J. Am. Chem. Soc. 1993, 115, 3384. 29. F. Gai, M. J. Fehr, and J. W. Petrich, J. Phys. Chem. 1994, 98, 8352. 30. F. Gai, M. J. Fehr, and J. W. Petrich, J. Phys. Chem. 1994, 98, 5784. 31. K. Das, D. S. English, M. J. Fehr, A. V. Smirnov, and J. W. Petrich, J. Phys. Chem. 1996, 100, 18275. 32. K. Das, D. S. English, and J. W. Petrich, J. Phys. Chem. A 1997, 101, 3241. 33. K. Das, D. S. English, and J. W. Petrich, J. Am. Chem. Soc. 1997, 119, 2763. 34. K. Das, A. V. Smirnov, M. D. Snyder, and J. W. Petrich, J. Phys. Chem. B 1998, 102, 6098. 35. K. Das, E. Dertz, J. Paterson, W. Zhang, G. A. Kraus, and J. W. Petrich, J. Phys. Chem. B 1998, 102, 1479. 36. K. Das, A. V. Smirnov, J. Wen, P. Miskovsky, and J. W. Petrich, Photochem. Photobiol. 1999, 69, 633. 37. K. Das, K. D. Ashby, J. Wen, and J. W. Petrich, J. Phys. Chem. B 1999, 103, 1581. 38. D. S. English, K. Das, K. D. Ashby, J. Park, J. W. Petrich, and E. W. Castner, J. Am. Chem. Soc. 1997, 119, 11585. 39. M. J. Fehr, M. A. McCloskey, and J. W. Petrich, J. Am. Chem. Soc. 1995, 117, 1833. 40. F. Sureau, P. Miskovsky, L. Chinsky, and P. Y. Turpin, J. Am. Chem. Soc. 1996, 118, 9484. 41. R. Chaloupka, F. Sureau, E. Kocisova, and J. W. Petrich, Photochem. Photobiol. 1998, 68, 44. 42. L. H. Pinto, L. J. Holsinger, and R. A. Lamb, Cell 1992, 69, 517.
REFERENCES
25
43. K. J. Newell and I. F. Tannock, Cancer Res. 1989, 49, 4447. 44. K. J. Newell, P. Wood, I. Stratford, and I. Tannock, Br. J. Cancer 1992, 66, 311. 45. R. A. Gottlieb, E. Skowronski, and B. M. Babior, Proc. Natl. Acad. Sci. USA 1996, 93, 654. 46. J. E. Li, J. Biol. Chem. 1995, 270, 3203. 47. M. A. Barry, J. E. Reynold, and A. Eastman, Cancer Res. 1993, 53, 2349. 48. A. V. Smirnov, K. Das, D. S. English, Z. Wan, G. A. Kraus, and J. W. Petrich, J. Phys. Chem. A 1993, 103, 7949. 49. D. S. English, W. Zhang, G. A. Kraus, and J. W. Petrich, J. Am. Chem. Soc. 1997, 119, 2980. 50. A. Smirnov, D. B. Fulton, A. Andreotti, and J. W. Petrich, J. Am. Chem. Soc. 1999, 121, 7979. 51. D. S. English, K. Das, J. M. Zenner, W. Zhang, G. A. Kraus, R. C. Larock, and J. W. Petrich, J. Phys. Chem. A. 1997, 101, 3235. 52. W. Frey, F. Laermer, and T. Elsaesser, J. Phys. Chem. 1991, 95, 10391. 53. J. L. Herek, S. Pedersen, L. Banares, and A. H. Zewail, J. Chem. Phys. 1992, 97, 9046. 54. L. Peteanu and R. A. Mathies, J. Phys. Chem. 1992, 96, 6910. 55. J. Goodman and L. E. Brus, J. Am. Chem. Soc. 1978, 100, 7472. 56. C. A. Taylor, M. A. El-Bayou, and M. Kasha, Proc. Natl. Acad. Sci. USA 1969, 63, 253. 57. S. Takeuchi and T. Tamara, Chem. Phys. Lett. 1997, 277, 340. 58. Y. R. Kim, J. T. Yardley, and R. M. Hochstrasser, Chem. Phys. 1989, 136, 311. 59. T. C. Swinney and D. F. Kelley, J. Phys. Chem. 1991, 95, 10369. 60. G. A. Brucker, D. F. Kelley, and T. C. Swinney, J. Phys. Chem. 1991, 95, 3190. 61. D. McMorrow and M. Kasha, Proc. Natl. Acad. Sci. USA 1984, 81, 3375. 62. B. J. Schwartz, L. A. Peteanu, and C. B. Harris, J. Phys. Chem. 1992, 96, 3591. 63. A. J. G. Strandjord and P. F. Barbara, J. Phys. Chem. 1985, 89, 2355. 64. R. Croteau and R. M. Leblanc, Photochem. Photobiol. 1978, 28, 33. 65. F. A. Ensminger, J. Plassard, T. S. Zwier, and S. Hardinger, J. Chem. Phys. 1993, 99, 8341. 66. S. Mazzini, L. Merlini, R. Mondelli, G. Nasini, E. Ragg, and L. Scalioni, J. Chem. Soc., Perkin Trans. 1997 2, 2013. 67. J. W. Petrich, M. S. Gordon, and M. Cagle, J. Phys. Chem. A 1998, 102, 1647. 68. Y. A. Alexeev, M. W. Schmidt, T. L. Windus, and M. S. Gordon, unpublished data. 69. W. D. Gwinn, J. Chem. Phys. 1971, 55, 477. 70. M. W. Schmidt, K. K. Baldridge, J. A. Boatz, S. T. Elbert, M. S. Gordon, J. H. Jensen, S. Koseki, N. Matsunaga, K. A. Nguyen, S. Su, T. L. Windus, M. Dupuis, and J. A. Montgomery Jr., J. Comp. Chem. 1993, 14, 1347. 71. J. W. Petrich, Int. Rev. Phys. Chem. 2000, 19, 479. 72. S. R. Greenfield and M. R. Wasielewski, Opt. Lett. 1995, 20, 1394.
26 73. 74. 75. 76. 77.
78. 79. 80. 81. 82.
HYPERICIN AND ITS PERYLENE QUINONE ANALOGS
M. J. S. Dewar, J. Am. Chem. Soc. 1984, 106, 209. Y. Chen, F. Gai, and J. W. Petrich, J. Am. Chem. Soc. 1993, 115, 10158. D. Z. Chen, D. P. Wang, H. Y. Zhang, and B. Tang, Chem. Phys. Lett. 2002, 353, 119. J. Park, A. Datta, P. K. Chowdhury, and J. W. Petrich, Photochem. Photobiol. 2001, 73, 105. B. M. Showalter, A. Datta, P. K. Chowdhury, J. Park, P. Bandyopadhyay, P. K. Choudhury, S. Kesavan, Y. Zeng, G. A. Kraus, M. S. Gordon, J. P. Toscano, and J. W. Petrich, Photochem. Photobiol. 2001, 74, 157. L. Zhu, P. Li, M. Huang, J. T. Sage, and P. M. Champion, Phys. Rev. Lett. 1994, 72, 301. M. H. Vos, J. C. Lambry, S. J. Robles, D. C. Youvan, J. Breton, and J. L. Martin, Proc. Natl. Acad. Sci. USA 1991, 88, 8885. U. Liebl, G. Lipowski, M. Negrerie, J. C. Lambry, J. L. Martin, and M. H. Vos, Nature 1999, 401, 181. A. P. Scott and L. Radom, J. Phys. Chem. 1996, 100, 16502. A. Datta, A. V. Smirnov, J. Wen, G. Chumanov, and J. W. Petrich, Photochem. Photobiol. 2000, 71, 166.
THIOPHOSGENE: A TAILOR-MADE MOLECULE FOR PHOTOCHEMICAL AND PHOTOPHYSICAL STUDIES David C. Moule Department of Chemistry, Brock University, St. Catharines, ON L2S3A1, Canada Takashige Fujiwara and Edward C. Lim* Department of Chemistry and the Center for Laser and Optical Spectroscopy The University of Akron, Akron, OH 44325-3601
CONTENTS I. II.
III.
Introduction Electronic States A. Molecular Orbitals and Electron Configurations B. S0, Ground Electronic State C. S1 (n,p*), First Excited Singlet State D. T1, First Triplet Electronic State E. S2, Second Singlet Excited State Photophysical Properties A. S1 Thiophosgene B. S2 Thiophosgene
*Holder of the Goodyear Chair in Chemistry at The University of Akron. Advances in Photochemistry, Volume 28 Edited by Douglas C. Neckers, William S. Jenks, and Thomas Wolff # 2005 John Wiley & Sons, Inc.
27
28
THIOPHOSGENE
IV. Concluding Remarks Acknowledgments References
I. INTRODUCTION Thiophosgene is a tailor-made molecule for studying the spectroscopic and photophysical properties of medium-size systems. The double bonding of the S, sulfur to the carbon atom, in the form of the thiocarbonyl chromophore C has the effect of red shifting the positions of the electronic transitions to longer wavelengths, where the spectra are often discrete (sharp features in the structure). Moreover, the electronic transitions lie in the visible and near-ultraviolet regions of the optical spectrum and are easily accessible by tunable dye lasers. The molecule also exhibits highly unusual photophysical properties in that both the first (S1) and the second excited singlet (S2) states exhibit fluorescence with very high efficiencies. The understanding of the spectra, the assignments, and analyses is aided by the high-molecular symmetry of Cl2CS. With only degrees of freedom of vibrational freedom, the assignments of the vibrational structure is a tractable though nontrivial problem. These size and symmetry factors allow for high level ab initio Hartree-Fock calculations from which the spectroscopic and dynamical properties can be inferred. Some of these unique spectroscopic and photophysical properties stem from the lack of a vibrational mode of a2 symmetry that is needed to couple the two ~ 1A2(n,p*) vibronically and ~ 1A1(n2) and S1: A lowest electronic states, S0: X allows for the S1 ! S0 internal conversion [1]. This lack of a radiationless pathway leads to the alternative fluorescence deactivation of S1 with high quantum yield. But the unusual feature is that this same a2 internal motion (rotation about ~ 1A1 (p,p*) electronic state S bond) is needed to couple the higher S2: B the C with the S1 state, such that the a2 vibrational deficiency also prohibits the radiationless decay from the S2 state to the S1 state [1]. The net result is that both the S2 and S1 singlet electronic states fluoresce with high quantum yields. This fluorescence from the second excited electronic state is in violation of Kasha’s rule [2]. It is this intense fluorescence from the S2 and S1 singlet electronic states that is used as a powerful diagnostic tool for elucidating the excited-state relaxation rates and mechanisms. This feature of the photophysics/spectroscopy allows not only for direct one-photon S2 ! S0 and S1 ! S0 laser-induced fluorescence (LIF) spectroscopies but also for two-photon optical-optical double resonance (OODR) experiments involving pump–probe experiments between the S0 (ground state) and intermediate S1/T1 states and the target S2 state, S2 S1 S0, or S2 T1 S0, respectively. The differing selection rules for the one- and sequential two-photon vibronic transitions allow for spectral
ELECTRONIC STATES
29
comparisons, a valuable tool for the assignment and understanding of the observed levels. In addition, the energies of these three singlet states are positioned such that the spectroscopies can be explored with dye lasers that operate efficiently in the visible region of the spectrum. Aside from the vibrational wavenumber intervals, Ev, the other tools that are available as an aid in the assignment are (1) the vibronic band intensities within the envelope of the band progressions (Franck-Condon factors); (2) the g ¼ Cl(35,35) Cl(35,37) built-in chlorine isotope effects; (3) ¼ () (þ), the inversion doubling splittings that are created by the quantum penetration of the molecular barriers; and (4) the differing selection rules for the one- and two-photon processes. The final and equally significant advantage is that Cl2CS is the only small thiocarbonyl species that is commercially available, and it does not have to created and detected in a flash pyrolysis system, as do the other thiocarbonyl transients. For these reasons, thiophosgene has been the subject of extensive spectroscopic and photophysical interest in both the gaseous and the condensed phases. Two reviews provide a useful background. The first, by Clouthier and Moule [3], describes the spectroscopy of small carbonyl, thiocarbonyl, and selenocarbonyl molecules and compares their structural and spectroscopic properties. A more recent review of the photophysics and physical photochemistry was given by Maciejewski and Steer [4], who treated a wide range of thiocarbonyl systems in the vapor and condensed phases. In this chapter, we present recent developments in the experimental and theoretical studies of the structure and excited-state dynamics of gaseous thiophosgene. A major emphasis of the chapter is the OODR and other experiments in supersonic jet that probe and characterize dark electronic states and their coupling with the bright states. The paper is organized as follows: We begin with a brief discussion of molecular orbitals and energy-level dispositions of electronic states. We next describe the spectroscopy of individual electronic states in sequence: ground (S0) state, first excited singlet (S1) state, first triplet (T1) state, and second excited singlet (S2) state. In addition to the standard one-photon spectroscopy, the S2 X S0 OODR spectroscopy involving S1 and T1 as an intermediate (X) state is included in the presentation. We conclude the chapter by describing photophysical properties of the S1, S2, and T1 thiophosgene. Nonradiative transitions to dark states, bound as well as dissociative, are presented in this section.
II. ELECTRONIC STATES A. Molecular Orbitals and Electron Configurations Thiophosgene possesses the planar molecular structure of Figure 2.1 in the ground electronic state and can be classified by the symmetry operations of
30
THIOPHOSGENE
Figure 2.1. The molecular structures and symmetry axes for the planar S0 (ground) and S1 (first excited) electronic states of thiophosgene.
the C2v point group. With the symmetry axes taken to conform to the Herzberg notation [5] (z-axis directed along the C S bond; x-axis in the out-of-plane direction), the symmetry properties can be classified according to one of the four irreducible representations, A1, A2, B1, and B2 (Table 2.1). The starting point for the analyses of the various electronic spectra and the corresponding photophysical decay processes begins with a consideration of the electronic states and the electronic configurations in the excitation processes. A schematic view of the molecular orbitals [6] is given in Figure 2.2, along with the symmetry species of the individual orbitals and their approximate descriptions. In order of increasing energy, the outermost orbitals associated only with the C S chromophore are (s), (n0 ), (p), (n), (p*), and (s*), where n is the thiocarbonyl nonbonding orbital, and n0 a second nonbonding orbital, (chlorine orbitals are excluded). The electronic configuration of the ground state can be written (s)2(n0 )2(p)2(n)2(p*)0(s*)0, where each of the inner orbitals are doubly occupied by an electron pair and the outer orbitals (virtual) are empty. The symmetry species of each of these MOs can be inferred from the orbital diagram of Figure 2.2 and the transformation properties of Table 2.1 The s and s* orbitals lie in the molecular plane, have the full structural symmetry of the
TABLE 2.1 C2v
E
A1 A2 B1 B2
þ1 þ1 þ1 þ1
Character Table for the C2v Point Group C2 (z) þ1 þ1 1 1
s (x,z)
s (y,z)
þ1 1 þ1 1
þ1 1 1 þ1
Tz Rz Tx, Ry Ty , Rx
ELECTRONIC STATES
31
~ ~ 1A1(n2), A Figure 2.2. The p, n, p*, and s* molecular orbitals that create the X 1 1 1 ~ ~ A2(n,p*), B A1(p,p*), and C B2(n,s*) states along with the first three low-energy singlet-singlet transitions. (Reprinted with permission from ref. [6].)
static molecule, belong to the totally symmetric A1 representation. The p and p* orbitals project from the plane, transform as translations in the x direction, and are of B1 species. The last of these orbitals, the nonbonding n orbital resides on the sulfur atom and is antisymmetric with respect to a reflection in the xz plane. It belong to the B2 representation. The symmetry of the ground state electronic wavefunction is derived as the direct product of the representations of the occupied one electron orbitals. A1 ðsÞ A1 ðsÞ B1 ðpÞ B1 ðpÞ B2 ðnÞ B2 ðnÞ ¼ A1 ~ 1A1(n2) in the Herzberg This state is labeled S0 in photophysical notation or X notation [5]. There are two notations for labeling the electronic states: In the notation used by photochemists, the singlet and triplet states are labeled S0, S1, S2, . . . and T1, T2, T3 . . . in order of increasing energy. The alternative nota~, B ~ . . . for the singlet states and ~a, ~b, ~, A ~, C tion, by Herzberg, uses the symbols X ~c . . . for triplet states. In this review, we will use both notations as appropriate. The first excited electronic state is created through an electron promotion from the least binding n nonbonding orbital to the lowest antibonding orbital p*, . . . (p)2(n)1(p*)1(s*)0. As the n orbital is of B2 symmetry and p* of B1 symmetry, it follows that the S1 state is of B2(n) B1(p*) ¼ A2 electronic symmetry. ~ 1A2(n,p*). The second S2 This state then bears the spectroscopic designation A excited electronic state is the result of an electron promotion from the p bonding orbital to the p* antibonding orbital to give an S2 state of B1(p) B1(p*) ¼ A1 ~ 1A1(p,p*). The final state for consideration, S3, is brought about by symmetry, B ~ 1B2(n,s*). n to s* excitation, B2(n) A1(s*) ¼ B2, and is labeled C
32
THIOPHOSGENE
Figure 2.3. The Walsh correlation diagram illustrating the changes in the p, n, and s* orbital energies as thiophosgene is distorted from a planar to a pyramidal conformation.
The change in molecular structure that comes about when an electron is lifted from the n orbital and placed into the p* orbital can be understood from the Walsh correlation diagram [7] (Fig. 2.3). The diagram shows the changes in orbital energy as the molecular structure deforms from the plane (C2v symmetry) to a 90 pyramid (Cs symmetry). When the molecule distorts, the B1(p) atomic orbital on the carbon center that forms the p* antibonding orbital mixes with the A1(sp2) s orbitals to create totally symmetric a0 orbitals of the Cs point group. This mixing of p and sp2 orbitals to form an sp3 hybrid reduces the energy of the p* antibonding orbital and destabilizes the planar structure of the ground state to a pyramidal conformation. Walsh’s rules [7] work on the basis that sp3 hybrid orbitals are more stable than the p–sp2 orbital combination. Thus it is the presence of the electron, which is promoted to the antibonding p* orbital, that is responsible for nonplanar structure of S1 thiophosgene. This diagram can also be ~ 1A1(p,p*) state. In this case, the structure is expected to applied to the S2 B be further deformed because the p bonding orbital is destabilized (upward sloping one electron energy curve) while the p* orbital is stabilized (downward ~ 1B2(n,s*) without an sloping curve) during the bending process. The S3 state, C occupation of the p* orbital would assume a planar conformation. The other major change is in the bonding between the sulfur and the carbon atoms. These changes can be neatly summarized by an inspection of the potential energy curves shown in Figure 2.4. This is a CASSCF ab initio study [6] of the first four electronic states, in which a slice of the potential surfaces is plotted against the CS bond length (all other coordinates are fixed). The plots for the first
ELECTRONIC STATES
33
~ 1A2(n,p*), B ~ 1A1(n2), A ~ 1A1(p,p*), and Figure 2.4. The potential energy of the X 1 ~ C B2(n,s*) electronic states of thiophosgene as a function of the CS bond distance. ~ state equiInset: The height of the barrier to molecular inversion along rCS from the B ~ intersection. (Reprinted with permission from ref. [6].) ~ C librium geometry to the B
~ 1A2(n,p*), show that upon excitation, the CS bond ~ 1A1(n2) and A two states, X ˚ . This increase can be understood in terms length increases from 1.64 to 1.79 A of the bond order concept. A s bond, created through the sharing of a pair of s electrons has a bond order of 1, whereas a double bond formed from a s and p combination has a bond order of 2. In the n to p* excitation process, the loss of the electron from the n orbital does not alter the bond order, whereas the place~ 1A2 (n,p*) ment of an electron into the p* orbital contributes 12. Thus, the A state has a CS bond order of 1.5. A lower order reduces the strength and results in a longer bond. A weaker bond also would be more elastic with a smaller force constant. As the curvature of the molecular potential is controlled by the force constant, the potential wells are broader and shallower. These differences in the two potential functions manifest themselves in different fundamentals for the CS stretching mode: n001 ¼ 1139:0 cm1 for the S0 ground electronic state and n01 ¼ 907:4 cm1 for the upper S1 state. The equilibrium position for the S2 third electronic state is further extended. The model applies here as well, because the loss and gain of electrons from and to the p and p* orbitals each contribute 12
34
THIOPHOSGENE
Figure 2.5. Energy level diagram for Cl2CS and its dissociation products. The state at about 30,000 cm1 is a dark, doubly excited state. (Reprinted with permission from ref. [8].)
to the bond order, leaving the S2 state with a nominal single CS bond of order 1. Thus it is not too surprising that the v01 frequency drops to 505.1 cm1 for this state. The final state, S3, derived from n to s* electron excitation does not have a calculated potential minimum and it appears that this state could be dissociative. The energy level diagram [8], which includes predissociation energies and dissociation limits, is shown in Figure 2.5.
B. S0, Ground Electronic State Early on, a molecular structure for the ground electronic state was derived from electron diffraction studies by Pauling and co-workers [9], who demonstrated that the molecule possessed a planar C2v equilibrium structure. Precise structural parameters (bond lengths and bond angles) were determined much later from the microwave spectra of four separate isotopomers of thiophosgene at their natural abundance [10]. The structure may be regarded as a set of three heavy atoms: S þ 2 Cl placed at the corners of an equilateral triangle with the lighter C atom at the center (Table 2.2). This conformation creates an oblate ellipsoid of revolution with the symmetric top c-axis projecting from the molecular plane in
35
ELECTRONIC STATES
TABLE 2.2 The Structural Parameters for Thiophosgene in the Lower S0, S1, T1, ˚ , degrees, and cm1) and S2 Electronic States (in A State
a(ClCCl) y(oop) Barrier
r(CS)
r(CCl)
Reference
1.601
1.729
111.19
0
0
rz microwave [10]
1.605
1.736
111.1
0
0
MP2/6-311G(d) [11]
1.73
1.745
112.3
27.2
—
1.704
1.751
119.2
32
1.729
1.723
112.3
32.2 640.6
UMP2/6-311G(d) [11]
—
—
—
32
726
fit to V(Q4) [14]
—
—
—
45
1650
S0 Band contour [12]
S1 622
Franck–Condon [13]
T1 S2
fit to V(Q4) for n1 ¼ 0 [15]
the B1 (x)-symmetry direction. Thus the moments of inertia are so close to a pure oblate limit that a substitution of the 35 for 37 isotopes of chlorine is sufficient to switch the a and b principal axes. For the planar ground state, the vibrational representation decompose as v ¼ 3A1 B1 2B2 , where the A1 modes transform as a translation along the z (C S) axis and the B1 mode as the single out-of-plane displacement along the x-direction. The last two modes lie in the molecular s (y,z) plane and are antisymmetric to the s (x,z) plane. Figure 2.6 gives a schematic view of the
Figure 2.6. A schematic representation of the six vibrational modes of planar thiophosgene.
36
THIOPHOSGENE
TABLE 2.3 Fundamental Vibrational Frequencies in the S0, S1, T1, and S2 Lower Electronic States of Thiophosgene (cm1) State S0 S1 T1 S2
T0
S) Q1(C a
— 1139.0 18,716.3a 907.4a 17,499e 923e g 34,277.3 505.1g
Q2(C Cl) Q3(ClCCl) b
497 480.0a 474f 495.0g
c
288.5 245.0a 247f 248g
Q4(oop) a
471.5 279.6a 298f 341.5g
Q5(C Cl) Q6(ClCS) 811d — — —
302d 189 193 212g
a
Visible absorption [18]. S2 fluorescence [19]. c S1 fluorescence [13]. d Raman [17]. e OODR [20]. MRS [25]. f T1 absorption [14]. g OODR [15]. b
vibrational modes along with an approximate description of the associated internal coordinates. The most noteworthy aspect of these vibrational modes is the absence of a mode of a2 symmetry. This representation transforms as a rotation about the z axis, Rz. The lack of a mode of a2 species creates a vibrational deficiency in thiophosgene, which has profound implications for the decay characteristics of the lower electronic states. The lower vibrational levels have been studied by infrared (IR) and Raman spectroscopies [16]. Normal coordinate analyses based on force constants transferred from other molecules (Urey-Bradley type) or from ab initio HF calculations have played a part in the construction of the vibrational assignments [17]. The observed fundamental frequencies are given in Table 2.3. Resolved S0 level structure at 15,000 cm1 of vibrational excitation has been observed from the band progressions contained in the fluorescence emission from single vibronic levels of the S1 and S2 electronic states [13, 21]. Figure 2.7 shows the resolved fluorescence along with selected regions recorded under much higher resolution with simulated emission pumping (SEP) [19]. Many of the eigenstates retain their identity at 10,000 cm1, which is somewhat surprising considering the onset of (intramolecular vibrational energy redistribution (IVR)) at this high level of vibrational excitation. These levels, along with the vibrational fundamentals, have been reduced to a set of second-order Dunham constants.
C. S1 (n,p*), First Excited Singlet State ~ 1A2(n,p*) or S1 state of thiophosgene comes about through the occuThe A pation of the p* orbital by an electron extracted from the nonbonding n
ELECTRONIC STATES
37
Figure 2.7. Resolved fluorescence spectra from selected vibronic levels of the 410 1 ~ 1A1(n2) vibrational structure. Insets: SEP spectra A1(p,p*) state, showing details of the X recorded under higher resolution. (Reprinted with permission from ref. [19].)
orbital. The direct product of these outermost electron orbitals transforms as ðnÞ ðp Þ ¼ B2 B1 ¼ A2 . For a purely allowed electric dipole transition, the transition moment MS01 ;S0 must be nonzero, cS1 jmjcS0 6¼ 0. This relationship requires that the product of the corresponding symmetry species transform as the totally symmetric representation, ðcS1 Þ ðmÞ ðcS0 Þ ¼ A1 . As the cS0 and cS1 electronic wavefunctions are of A1 and A2 species, it follows that (m) ¼ A2. This result creates a problem for thiophosgene since, for an allowed
38
THIOPHOSGENE
transition, the electric dipole moment operator must transform as a translation along one of the x, y, or z symmetry directions (B1, B2, or A1, respectively). S) axis, and The A2 representation transforms as a rotation about the z (C as a result the n ! p* electron promotion is forbidden as a purely electric dipole process. By this same reasoning, the magnetic dipole transition that transforms as rotation about the C S bond is fully allowed. A weak absorption system was observed in the visible region of the thiophosgene by Burnelle [18]. From its low oscillator strength in hexane, f ¼ 1.2 104, it was assigned to the n ! p* process. To account for this forbidden system, it is necessary to go beyond the symmetry properties of the electronic wavefunctions and consider higher-order effects brought about by distortions in the nuclear framework. When the molecular structure is displaced during the course of an antisymmetrical vibration, the symmetry of the C2v point group in Table 2.1 reduces to the Cs group with a0 and a00 representations. This relaxation in structure lowers the symmetry restrictions and allows for vibronic (electronic þ vibrational) interactions. In the Herzberg-Teller method [22] for accounting for induced vibronic intensities, the transition moment Me00 e00 between the ground e00 and excited e0 electronic states is expanded as a Taylor’s series in the normal coordinate Q for anitsymmetric mode k qMe0 e00 1 q2 Me0 e00 Qk þ Q2 þ ð1Þ Me0 e00 ¼ Me00 e00 þ qQk 0 2 qQ2k 0 k For small-amplitude vibrations in which the transition moments are relatively smooth, the expansion can be terminated at the linear second term. A vibronic transition between a set of e0v0 e00v00 electronic/vibrational levels is defined by the transition moment qMe0 e00 0 Me0 v0 e00 v00 ¼ hcv0 jMe0 e00 jcv00 i ¼ Me0 e00 hcv0 jcv00 i þ hc 0 jQk jcv00 i ð2Þ qQk 0 v If the electronic transition is allowed, Me00 e00 is nonzero and the first term dominates the expression. This term can be viewed as a product of the electronic transition moment and the vibrational overlap integral, hcv0 jcv00 i, connecting the two vibrational wavefunctions, cv, in electronic states e0 and e00 . The Franck-Condon factors, which are the square of the vibrational overlap integral, determine the intensity distribution among the vibrational bands. The relative intensities of the band members within a vibrational progression is, therefore, given by the ratio of the Franck-Condon factors. If, through a symmetry restriction, the transition moment Me00 e00 vanishes, as in the present case, the band activity in the spectrum comes from the second term. When Qk is a nontotally symmetric
ELECTRONIC STATES
39
normal coordinate, the selection rules v ¼ 0, 2 govern the vibrational overlap integral, hcv0 jcv00 i, whereas the rules v ¼ 1, 3 are needed for a nonzero vibronic integral hcv0 jQk jcv00 i. Since the dipole moment operator, m, depends only on the electronic coordinates,
qMe0 e00 q hcv0 jQk jcv00 i ¼ hce0 jmjce00 i hcv0 jQk jcv00 i qQk qQk 0 qce00 qce0 þ m c 00 hcv0 jQk jcv00 i ¼ c e0 m qQk qQk e X qce00 qce0 c hci jmjce00 i þ ¼ hce0 jmjci i ci qQk qQk i i hcv0 jQk jcv00 i ð3Þ
To apply this result to thiophosgene, we consider the vibronic interaction induced by the activity of the antisymmetric out-of-plane vibrational mode Q4. The summation over states in Eq. (3) in the simplest case can be reduced to a single term, the state S2. q qMS1 ;S2 S2 hS2 jmjS0 ihcv0 jQ4 jcv00 i hc 0 jQk jcv00 i ¼ S1 qQk 0 v qQ4
ð4Þ
This equation can be best understood by the three-state diagram in Figure 2.8, where S0 is the ground electronic state, S1 (n,p*) the first excited state, and S2 a state adjacent to S1 of appropriate symmetry for vibronic interaction. Before intensity can be induced into the forbidden electronic process, the symmetry requirements of the three factors of Eq. (4) need to be satisfied. Consider the first factor, hS1 j q=qQ4 j S2 i. This is a matrix element that describes the coupling between the optically forbidden S1 and the strongly allowed S2 excited states. For a nonzero coupling element, (S1) (q/qQ4) (S2) ¼ A1, and since (S1) ¼ A1 and (q/qQ4) ¼ B1, then (S2) ¼ B2. The second factor, hS2 j m j S0 i, represents the transition moment between the S0 and the S2 electronic states, with m the electric dipole operator. Again, the requirement is (S2) (m) (S0) ¼ A1, and since (S2) ¼ B2 and (S0) ¼ A1 then (m) ¼ B2. Thus, if the vibronic coupling is to a single electronic state, then S2 could be identified as the higher energy B2 (n,4s) Rydberg. The forbidden S1 S0 transition is said to borrow (or steal) oscillator strength from the highly allowed S2 S0 transition and hence is y (b2) polarized. Thus it is the proximity of the highly allowed S2 S0 transition that is responsible for inducing vibronic strength into the S1 S0 system.
40
THIOPHOSGENE
Figure 2.8. Energy level diagram illustrating the vibronic coupling between the 41 level of the S1(n,p*) state and the S2(n,4s) Rydberg state, which induces intensity into the S1 S0 transition.
The last factor in Eq. (4), cðQ04 Þj Q4 j cðQ004 Þ , represents the vibrational overlap between the cðQ04 Þ and the cðQ004 Þ wavefunctions of the S0 and S1 states. For a 0 nonzero integral, cðQ4 Þ ðQ4 Þ cðQ004 Þ ¼ A1 and for the transition from the vibrationally cold zero point level (v ¼ 0), cðQ004 Þ ¼ A1 . Since the distortion is in the out-of-plane direction, cðQ04 Þ ¼ B1 . The vibrational levels in the Qv manifold alternate between A1 and B1 species as v increases from v ¼ 0, 1, 2. . . . . Thus the vibronic transition that begins on the v ¼ 0 level of the S0 state terminates on the first quantum (v ¼ 1) of Q4 in the S1 excited state. This transition is given the notation, 410 , (Mab denotes a vibronic transition involving ‘‘a’’ vibrational quanta in the upper state and ‘‘b’’ quanta in the lower state, where M is the vibration mode number; 0ab with a ¼ 0 and b ¼ 0 would be the notation for the vibrationless 0–0 origin transition). To complete the analysis, it is necessary to consider distortions along the y symmetry direction. In this case, it would be the two modes Q5 and Q6 of B2 species that would mediate the vibronic coupling. By arguments similar to those given previously for x symmetry distortion, a Q5 symmetry distortion would lead to states of ðcðS1 ÞÞ ðcðQ05 ÞÞ ¼ A2 B2 ¼ B1 vibronic symmetry. The schematic arrangement is shown in Figure 2.9. Before entering into a discussion of the assignments and analysis of the observed band structure it is necessary to return to a consideration of the molecular geometry. Recall that Walsh’s rules predicted that the electron occupation of the p* orbital would render the S1 state nonplanar. Such a conformation would have two distinct equilibrium configurations that are related to each other
ELECTRONIC STATES
41
~ 1A2(n,p*) states ~ 1A1(n2) and A Figure 2.9. A schematic energy level diagram for the X of thiophosgene showing the vibronic transitions that give rise to the a, b, and c rotational band types.
by an inversion of the nuclei. The potential functions for an out-of-plane nuclear inversion would contain a double minimum with the height of the central barrier providing a measure of the stability brought about by the nonplanar distortion. Figure 2.10 shows the form of the potential function. This diagram
Figure 2.10. Potential energy diagrams showing the correlation between the vibrational energy levels as thiophosgene is bent from a rigid-planar to a rigid-pyramidal conformation.
42
THIOPHOSGENE
illustrates the changes in the level positions as the symmetry of the system changes from C2v for the rigid planar case (left side) to Cs for the pyramid case (right side). At the left of the diagram, the potential energy function is quadratic (harmonic), and the levels are equally spaced and are labeled by the quantum numbers 0, 1, 2, . . . . At the right side, the same quantum number description is applied; but in this case, the levels are degenerate. In the intermediate case, the levels split into doublets and are given the additional designation þ or , depending on whether the vibrational wavefunctions are even or odd. For a small doublet splitting, the even or odd levels of the zero point may be designated 0þ or 0 (high barrier notation), whereas the equally useful low barrier notation would be 0 and 1 for these same levels. Figure 2.11 combines the Herzberg-Teller coupling scheme of Figure 2.9 with the level patterns and symmetries expected for the double minimum potential of Figure 2.10 and presents an overall view of the inversion levels, their vibronic symmetries, and the rotational band types. For the lower S0 state, the equispaced v400 manifold of levels (00, 41, 42, . . .) bear the vibronic symmetries A1, B1, A1, . . . whereas the corresponding levels in the S1 state (00, 41, 42, . . .) are A2, B2, A2, . . . . The transitions between the ground state zero point level, 00,
Figure 2.11. An overall view of the vibrational level patterns, their vibronic symmetries, ~ 1A2(n,p*) state. ~ 1A1(n2) with the pyramidal A and the transitions connecting the planar X
ELECTRONIC STATES
43
and the 00, 41, 42, . . . levels of the S1 state result in a, b, a, . . . rotational band types. On other hand, transitions to the inversion levels in combination with quanta of v5 and v6 give rise to c-type and a-type bands. Thus it is possible that bands of all three polarizations are present in the spectrum. Transitions between the vibrational levels in the S0 and S1 combining states are governed by vibronic selection rules by which totally symmetric A1 levels of the S0 ground state (v ¼ 0, 2, 4,. . .) are forbidden to combine with the A2 (v ¼ 0, 2, 4) levels of the S1 state, whereas combinations are possible with the odd levels (v ¼ 1, 3, 5,. . .). These transitions require moments that are y-axis directed and are polarized along the b-principal axis. With this background, we now turn to the assignment and analyses of the visible spectrum of thiophosgene. Early on, the vibrational assignments were taken to a high level by Brand et al. [23], who photographically recorded the visible spectrum with a long (6-m) grating spectrograph. Under this dispersion, the individual 35Cl2CS and 35Cl37ClCS isotopomers of thiophosgene (abundance ratio of 9:6) were resolved as line-like doublets [23]. The starting point for the analysis came from the identification of ‘‘hot’’ bands (transitions from thermally excited vibrational levels in the S0 ground state). From the observed wavenumber intervals between the hot bands and associated higher energy cold bands, they were able to assign the hot structure to the activity of Q004 . These intervals were larger than the v004 ¼ 471.0 cm1 fundamental that had been established by IR measurements, and it was clear that the differences were related to the so-called inversion doubling splittings. The key to the vibrational analyses can be understood from the five critical bands illustrated as stick spectra in Figure 2.12. These stick spectra portray the band structure that would be anticipated for a vibronically allowed, but electronically forbidden, transition, where the planar ground electronic state is
Figure 2.12. A stick representation of the six bands in the S1 key to the vibrational assignments.
S0 spectrum that are the
44
THIOPHOSGENE
combined with a pyramidal excited state. The obvious feature of this diagram is the absence of the 000 band origin as would be expected for an allowed S1 (v ¼ 0) S0 (v ¼ 0) transition. The attachment of an odd quantum of the antisymmetric Q4 mode to the S1 state, S1(v04 ), would provide the vibronic symmetry for the Herzberg-Teller coupling, Ae2 Bv1 ¼ Bev 2 . The resulting vibronic e A originating from the vibrationless S0 (v ¼ 0) state would transition Bev 1 2 create a B2 (y directed) polarized band structure. It would bear the label 410 . The corresponding hot band transition that comes from the thermally populated S0 (v ¼ 1) and terminates on the S1 (v ¼ 0) level, 401 , would have a similar transition moment, although the strength would be reduced by the Boltzman factor for v004. These hot and cold bands allow the inversion doubling splittings to be evaluated by the method of combination differences. When the first cold band, 410 , is combined with the companion hot band, 401 , along with the v004 vibrational fundamental, the inversion splittings becomes 0 ¼ 0 0þ ¼ 41 40 ¼ 410 401 þ v004 ¼ 0:42 cm1 . The next doublet by this method, is 1 ¼ 1 1þ ¼ 43 42 ¼ 430 421 þ v004 ¼ 12:9 cm1 . This order of increase in the inversion doubling splitting with vibrational energy is the result of the low central barrier in the potential governing the Q4 normal mode. The shape of the potential for the out-of-plane motion can be extracted from the inversion splittings and the manifold of energy levels. A useful model potential for the double minimum potential function consists of a combination of quadratic and Gaussian terms: V ðQ4 Þ ¼ A Q24 þ B exp ðC Q24 Þ, where A, B, and C are shaping parameters. The quadratic first term creates the outer walls of the function, while the Gaussian describes the central barrier separating the double minima. Energy levels for the model potential are then calculated by a variational method through a diagonalization of the corresponding Hamiltonian matrix. The final stage of this process is a refinement of the parameters in the model function through a fitting of the observed and calculated levels. In the present case, the fitting process yielded a barrier height of 622 cm1 and an out-of-plane angle of 32.0 for the S1 state of thiophosgene. The adjustments in molecular structure that occur on S1 S0 excitation can be inferred from the band intensity relationships. The vibrational bands form patterns that can be grouped into progressions for each of the totally symmetric normal modes according to their intensity distribution. The lengths and strengths of these progressions relate to the differences in molecular structure in the combining electronic state defined by the multidimensional Franck-Condon principle. Very loosely, this principle states that the length (number of active members) of a progression is controlled by the changes in the location and the directions of the normal coordinates in the two electronic states. Thus, for example, the pyramidalization of the S1 state relative to the planar S0 state occurs along the Q4 normal coordinate (to a first approximation). As expected, the bands in the Q4 progression can be followed for several members, 410 , 430 , 450 , 401 , 421 , etc. consistent with the nonplanar structural displacement. These bands
45
ELECTRONIC STATES
that mark the transitions to the levels in Q4 act as a set of pseudo-origins for progressions in the remaining totally symmetric Q1, Q2, and Q3 modes. As every level serves as a suborigin for every other level, the level structure quickly becomes complex. It is this rapid buildup in the density of vibrational states that often is the limiting factor for establishing the band assignments at higher vibrational energies. The introduction of antibonding electron density that reduces the carbon sulfur bond order from 2 to 1.5 in the n ! p* process, results in an increased CS bond length and a decreased stretching force constant. Thus the quanta of 907 cm1 observed throughout the band spectrum are assigned to the activity of Q01 , the CS stretching mode in keeping with the corresponding ground state frequency v001 ¼ 1139 cm1. The totally symmetric modes Q02 and Q03 are also active in the spectrum and create lengthy progressions in 489 and 245 cm1, respectively. Structural differences between the S0 and the S1 states of ˚ rðCClÞ ¼ þ0:02 A ˚ and yðClCClÞ ¼ þ1:1 have been rðCSÞ ¼ þ0:13 A, obtained from a rotational band contour analysis [12]. A more accurate set of S1 structural parameters can be obtained from a fit of S0 absorption. Recently, Fujiwara et al. [51] the rotationally resolved S1 S0 ð1 A1 Þ fluorescence excitation spectrum recorded high resolution S1 ð1 A2 Þ of thiophosgene in supersonic free jet with a pulse-amplified ring laser system. A line-to-line analysis of rotational fine structures of a selected group of bands yields a refined structure of the molecule in the 1 A2 ðn; p Þ state [51]. The change ˚ for CS bond and – 0.007 A ˚ in the bond length upon S0 to S1 excitation is 0.094 A
for CCl bond. The out-of-plane ClCCl angle is about 24 , which is significantly smaller than the value of 32 derived from inversion splittings. Table 2.4 compares the structural parameters obtained from rotational analyses with those from Franck-Condon analyses and inversion splittings.
TABLE 2.4 A Comparison of the Structural Changes that Occur on Excitation S0 , from Rotational Analyses, Inversion Splittings, and Franck-Condon S1 Analyses
r CS r CCL yClCCl yoop a
S0a
S1b
c
Id
IId
1.600 1.727 111.2 0
1.694 1.720 117.6 23.9
0.094 0.007 6.4 23.9
0.103 0.022 8.0 32e
0.057 0.09 10.8 32e
From microwave spectrum [10]. Fujiwara et al. [51]. c rðS1 Þ rðS0 Þ. d From Franck-Condon analyses [23]. e From inversion splittings. b
46
THIOPHOSGENE
~ 1A2(n,p*) X ~ 1A1(n2) spectrum obtained as (a) the Figure 2.13. A comparison of the A one-photon S1(v0 ) S0(00) LIF excitation spectrum with (b) the two-photon S2(00) S1(v0 ) S0(00) OODR spectrum. (Reprinted with permission from Ref. [24].)
A different perspective of the vibrational structure of the S1 electronic state is illustrated in Figure 2.13b. This is an OODR that was obtained by sequentially exciting Cl2CS with two photons of different colors. In this experiment, a photon from the first laser (the pump photon) induces a S1 S0 vibronic transition that is followed after a short time delay by a second S2 S1, probe photon that carries the excitation to the S2 state. The pump laser is advanced to the blue and interrogates the bands of the S1 S0 system while the probe laser is scanned at the same rate to the red such that the total energy matches a selected vibrational level of the S2 state. In this way, an excitation spectrum of the vibrational band structure of the S1 state is constructed by monitoring the fluorescence that originates from the S2 state. What is not too surprising is that the one-photon LIF spectrum (Fig. 2.13a) and two-photon OODR spectra (Fig. 2.13b) are similar, since these spectra sample the same S1 level structure. The major differences between these panels lies in the intensity relationships of the bands within the progressions. These differences can be understood by recognizing that the OODR is a sequential process where a substantial time delay is introduced between the pump and the probe photons. Thus the Franck-Condon factors for the S2 S1 S0 process is a
47
ELECTRONIC STATES
product of the factors of the individual S1 S0 and S2 S1 transitions. The transition probabilities for the S1 S0 first stage can be inferred from the band intensities of the lower trace while the probabilities for the second S2 S1 stage must be estimated from a comparison of the upper and lower traces. The first photon pumps the system from a planar S0 to a nonplanar S1 state and is followed by the second photon, which lifts the nonplanar S1 to a nonplanar S2 state. Because the first stage is Franck-Condon forbidden (a nonplanar–planar transition) and the second stage is Franck-Condon-allowed (nonplanar–nonplanar) the ODDR spectrum has more of the characteristics of a Franck-Condon-allowed transition than does the one-photon LIF spectrum. The much stronger 410 pseudo-origin band and the shorter band progressions in the OODR spectrum relative to the corresponding bands in the one-photon LIF spectrum are a direct consequence of the Franck-Condon factors of the one- and two-stage OODR processes. New features do appear in the OODR spectra, but they are submerged into the background noise. Here the magnetic dipole band 420 can be identified at 1 1þ ¼ 43 42 ¼ 12:9 cm1 to red of the 430 band. Likewise, in this same region the 610 appears weakly. The position of this band can be fixed from the 611 sequence transition that is prevalent in the hot-jet spectrum leading to v06 ¼ 189 cm1.
D.
T1, First Triplet Electronic State
Singlet-triplet transitions are forbidden by the selection rule S ¼ 0, which states that the spin of the electron cannot change direction (spin flip) during an electronic transition. Intensity, however, can be introduced into the spin-forbidden transition by allowing the three levels of the T1 triplet state to mix with adjacent singlet states, S1, and thereby acquire the singlet character that is needed for an electric dipole transition. A mixing of the triplet and singlet ^ SO j Si , wavefunctions is accomplished by matrix elements of the form T1 j H ^ SO is the spin-orbit operator that couples the electron spin, S; with the where H ^ SO / L S. The spinorbital L (or spatial) part of the electronic wavefunction, H forbidden T1 S0 transition has a transition moment MT1 ;S0 that is given by perturbation theory as MT1 ;S0 ¼
X Si
X 3 1 ^ E1 E1 T1 ;Si cT1 j HSO j cSi mSi ;S0 þ T1 ;S0
1
^ SO j 3 cT mTi ;T1 cS0 j H i
Ti
ð5Þ
The first term in this expression represents the coupling between the T1 state in question and the higher Si singlet excited states. The second term is a mixing of
48
THIOPHOSGENE
~ 1A1(n2) Figure 2.14. The coupling scheme by which the spin-forbidden a˜ 3A2(n,p*) X 1 2 ~ ~ transition borrows oscillator strength from the B A1(p,p*) X 1A1(n ) electric dipole allowed transition.
the ground singlet S0 state with the higher triplet levels Ti. As the ground electronic state is energetically removed from the higher triplet states, the energy difference ET1 ;S0 ¼ ET1 ES0 (perturbation gap) is large and the second term ^ SO operator couples is usually disregarded. For the special case in which the H to a specific singlet excited state, S2, the summation can be reduced to the single term ^ SO j 1 cS mS2 ;S0 MT1 ;S0 ¼ ET11 ;S2 3 cT1 j H 2
ð6Þ
This expression can be understood from the coupling scheme of Figure 2.14 the three elements that control the transition strength are depicted: Here, 3 ^ SO j 1 cS , the coupling matrix element; ET1 ;S2 the perturbation cT1 j H 2 energy gap; and mS2 ;S0 , the transition moment of the singlet-singlet transition that supplies the induced intensity. Symmetry dictates that the representations of the direct product of the factors ^ SO j 1 cS transform under the group operations accordin the integral 3 cT1 j H 2 ^ SO spining to the totally symmetric representation, A1. The spin part of the H orbit operator converts triplet spin to singlet spin wavefunctions and singlet functions to triplet wavefunctions. As such, the spin function does not have a ^ SO . Rather, the control is embedded bearing on the symmetry properties of H in the orbital part. The components of the orbital angular momentum, (Lx, Ly , ^ SO have symmetry properties of rotations about the x, y, and z and Lz) of H symmetry axes, Rx, Ry , and Rz. Thus, from Table 2.1, the possible symmetry
ELECTRONIC STATES
49
^ SO are (Lx) ¼ B2, (Ly) ¼ B1 and (Rz) ¼ A2. For a sigrepresentations for H nificant coupling, the matrix elements must have electronic wavefunctions that S bond (z-axis) rotate into each other. Rz transforms as a rotation about the C and creates a coupling element of the type hnj Rz j pi. This operator is said to rotate the n orbital that y-projects from the side of the CS bond to the p (px) ^ SO ) ¼ A2, and for hS2 j ^ SO, (H out-of-plane orbital. With this choice for H ^ SO j S1 i 6¼ 0, then (S2) (H ^ SO ) (S1) ¼ A1 and (S2) ¼ A1, providing H the symmetry identity of the coupled singlet state. The allowed transition moment, mS2 ;S0 , which donates oscillator strength to the forbidden transition, would connect states of A1 species. The unique candidate for this donor S2 state would have the p,p* electron configuration and combines with the ground elec~ 1A1(n2) system not ~ 1A1(p,p*) X tronic state with high dipole strength. The B 3 1 2 ~ only donates strength to the ~a A2(n,p*) X A1(n ) transition, but it also provides the z-polarization (transition moment direction). The selection rules that govern the band patterns are those of an electronically allowed system. Transitions from the vibrationless level of the lower S0 (v ¼ 0) state terminate on the (þ) lower member of the inversion doublet rather than on the upper () level as in the case of the S1 S0 transition. The vibrational analysis of the T1 S0 system developed slowly from a series of independent experiments: the direct absorption at long path lengths [14], magnetic rotation spectra (MRS) [25], and cavity ring-down (CRD) [11]. OODR studies have provided the clearest view of the band structure. A roomtemperature, one-color, two-photon study by Clouthier et al. [26] showed that the T1 state could be mapped by the OODR process with T1 as the intermediate state, S2 T1 S0. A clearer view of the triplet vibrational band structure was obtained under jet-cooled conditions in a two-color, two-photon OODR study by Steer and co-workers [20]. This same spectra has been replicated by Fujiwara et al. [24] under improved conditions. In this experiment, the total energy of the two lasers (pump þ probe) was set at the vibrationless, v ¼ 0, level of the S2 state and monitored by fluorescence to the S0 state. The pump laser was scanned upward through the levels of the T1 state, while the probe laser was tuned in the opposite direction. This S2 (00) T1 (x) S0 (00) scan-up, scan-down spectrum is shown in Figure 2.15 and is compared to the analogous singlet–singlet S2 (00) S1 (x) S0 (00) spectrum. ~ The ~a 3A2(n; p ) electronic state may be regarded as a companion to the A 1 A2(n; p ) state since they are derived from the same electron configurations. Thus the triplet state would adapt somewhat the same pyramidal conformation and have an elongated carbon sulfur bond. The forgoing symmetry analysis of the spin-orbit coupling predicted that the vibronic selection rules would be those of a fully allowed electronic transition, with the 000 transition forming the origin band. The 410 transition with quantum addition of the antisymmetric Q4 mode would be forbidden by spin-orbit selection rules though allowed by the
50
THIOPHOSGENE
~ 1A2(n,p*) Figure 2.15. A comparison of the vibrational structure of the a˜ 3A2(n,p*) and A 0 0 0 0 states through (a) the S2(0 ) S1(v ) S0(00) and (b) S2(0 ) T1(v ) S0(00) OODR processes. (Reprinted with permission from Ref. [24].)
vibronic-spin-orbit process. The inversion doubling splittings, , can be derived from the ground state combination differences, viz., 1 ¼ 1 1þ ¼ 43 42 ¼ 431 420 þ v004 . Thus the v004 hot band intervals in the triplet state are smaller than v004 by 1 (in the S1 state they are larger by the corresponding value). This reversal in the direction of the splittings is often useful in establishing the assignment of band multiplicity. The assignments of the vibrational band structure can be quickly established by comparing the OODR (T1) and OODR (S1) spectra of Figure 2.15. The freS stretching frequency and T0 values are collected in Table 2.3. The v1 C quency at 923 cm1 is found to be slightly higher than the 907 cm1 value recorded for the S1 state. This frequency difference is the result of an avoidance by the two electrons of parallel spins (Fermi correlation) that reduces the p* antibonding repulsions (T1 lies below the S1 state by 1217 cm1). This reduction in antibonding character is reflected in a slightly stronger C S bond and higher force constants. The differences in v4 in the two states is also instructive because these intervals, along with the inversion splittings, define the shape of the double minimum potential functions. The higher vibrational frequencies and the smaller
ELECTRONIC STATES
51
inversion doubling splittings must be associated with steeper outer walls and a higher central barrier of 726 cm1 for the T1 state (Table 2.2). The photophysical properties of T1 thiophosgene are somewhat unexpected. As the molecular prototypes thioacetone, thioacetaldehdye, and thioformaldehyde all show strong phosphorescence emission from their lowest triplet electronic states [4], it should be reasonable to expect that thiophosgene should also be emissive. A number of experiments have been specifically designed to directly detect T1 phosphorescence. Clouthier and Moule [25] recorded an excitation spectrum over the requisite spectral region and compared the observed band positions to those of the MRS. From the lack of coincidences between the band patterns in the two spectra, they concluded that the 3A2(n,p*) state was nonemissive. Likewise, Bernath et al. [27] were unable to excite phosphorescence with a tunable dye laser or with 528.7-nm line of a CW argon ion laser. In an extension to this work, Kawasaki et al. [28] pumped the 420 and 310 420 transitions. After integrating more than 30,000 laser shots, they were unable to detect a useful signal. They also tested the possibility of intersystem crossing by pumping into levels of the S1 state without success. In the most recent of these experiments, Moule et al. [11] compared the excitation spectrum, under jet-cooled conditions, with the absorption spectrum recorded as a CRD spectrum, as shown in Figure 2.16. None of the T1 S0 bands in the CRD spectrum appeared in the excitation spectrum of total emission. The conclusions from all of these studies is that T1 thiophosgene decays almost exclusively by T1 ! S0 intersystem crossing (ISC).
Figure 2.16. Part of the CRD absorption spectrum and the excitation spectrum LIF/LIP in the T1 S0 and S1 S0 regions of jet-cooled thiophosgene. (Reprinted with permission from Ref. [11].)
52
THIOPHOSGENE
Figure 2.17. The V(Q4) potential energy functions for the S0 and T1 electronic states of thiophosgene along with the v ¼ (þ), even, inversion-wagging energy levels. (Reprinted with permission from Ref. [11].)
The T1 nonradiative lifetime (i.e., the inverse of the T1 ! S0 ISC rate) was estimated by using the T1 S0 spin-orbit coupling and the Franck-Condon factors for the T1 ! S0 ISC based on the ab initio equilibrium structures, as shown in Figure 2.17, and vibrational frequencies for the T1 and S0 states, computed at the MP2/6-31G(d,p) level of theory [11]. The nonradiative lifetime, calculated as the 1/e of the survival probability, is 20 ps for barrier heights of 770–845 cm1 and out-of-plane angles of 32.69 . (The thiophosgene adopts a pyramidal conformation with the C S bond bent from the ClCCl plane by about 32 .) The computed lifetime is comparable to the lower-limit lifetime of 50 ps obtained from the quantum yields P 103 F with F 1.0 and the measured S1 fluorescence lifetime of 4 ms [11]. The short T1 nonradiative lifetime of thiophosgene can be attributed to the large matrix elements of the T1 S0 spin-orbit coupling (V ¼ 150 cm1) and the strong out-of-plane deformation of the T1 state relative to the planar S0 state, that provides the large Franck-Condon factors for the T1 S0 ISC [11]. An estimate of the T1 lifetime can be made from the measured line width of the T1 S0 absorption. Figure 2.18 presents the S2(00) T1(00) S0(00) OODR spectrum measured under medium-frequency resolution ( 0.05 cm1) [29]. The rotational structure, depicted in Figure 2.18 consists of a prominent band with two satellites (50 times weaker) at 6 cm1 to lower energies. From
ELECTRONIC STATES
Figure 2.18. The 000 origin band of the T1 (Reprinted with permission from Ref. [29].)
53
S0 transition under etalon resolution.
a consideration of the spin-orbit couplings that induce strength into the singlettriplet transition, the strong band can be assigned to the C S bond polarized Tz spin component. A preliminary band contour analysis reveals that the two lower energy components are the parallel and perpendicularly polarized Tx and Ty spin components [29]. From the point of view of the spectroscopy of the T1 state, the spin and rotational constants are found to lie within their expected ranges. On closer inspection, the rotational structure of the Tz component is observed to possess a greater line width ( 0.1 cm1 under etalon resolution) than that of the Tx and Ty components ( 0.06 cm1). This line broadening must be associated with a reduced lifetime for the Tz component of 150 ps. These initial observations on the Tx, Ty , and Tz rotational structure go some distance toward understanding the unusual photophysics of the T1 state. The lack of a one-photon LIP signal, even under enhanced signal-to-noise conditions, now can be attributed to the picosecond radiationless deactivation of the Tz level through the ISC process. Transitions to the longer-lived Tx and Ty spin levels would not play a significant role in the direct phosphorescence emission of the LIP method because of their reduced oscillator strength. The only concrete information about the decay characteristics of T1 comes S0. from time delay measurements in the two-stage OODR process, S2 T1
54
THIOPHOSGENE
Figure 2.19. Temporal profile of fluorescence trains from the S2(00) level via the S2(00) T1(31) S0(00) OODR probe as functions of time delay and spatial distance between the pump–probe lasers, showing the long trail contour of 4 ms indicates a longlived T1 state due to coupling to higher energy levels in the S0 state. Inset: Pump–probe laser configuration including a molecular jet. (Reprinted with permission from Ref. [30].)
Figure 2.19 shows the intensity of the fluorescence emitted from the S2 monitor state as a function of the pump–probe delay time in which the pump and probe beams have been spatially displaced from each other [30]. Two observations demonstrate that the probe S2(00) T1 transitions emanate from the rotational levels of the T1(00) electronic state rather than from the isoenergetic density of S0 vibrational levels, S2 S0 (high v). The first factor is the up-down wavelength tracking of pump and probe laser beams that are needed for OODR method. These two tuned laser beams create conditions that are too stringent for probe transitions that use high-energy vibrational S0 levels as initial states. Second, the vibrational band spectra do not undergo significant changes when the pump–probe time delay is increased to several microseconds. The results of these OODR measurements, therefore, indicate that the T1 lifetime is several microseconds [30]. The lower limit of the quantum yield can be established by combining the observed and radiative lifetimes. In absorption, the T1 S0
ELECTRONIC STATES
55
transition has about one-sixth the strength [14] of the higher energy S1 S0 companion transition that has a measured [18] oscillator strength of f ¼ 1.2 104. From this ratio, the oscillator strength of the triplet-singlet transition can be estimated at 2.0 105 for a radiative lifetime of 231 ms. The loss in phosphorescence signal is then expressed as the quantum yield for phosphorescence, P 0.01. To understand the conundrum raised by the separate observations of two lifetimes for the Tz spin state that differ by five orders of magnitude (150 ps from the line widths and several microseconds from the OODR measurements with the pump–probe time delay), we turn to the high-energy vibrational structure of the S0 ground state observed by Strickler and Gruebele [31]. From the resolved S2 ! S0 dispersed fluorescence spectrum, these workers were able to show that half of the six-dimensional vibrational quantum space was highly structured at energies beyond the S0 dissociation limit of 20,000 cm1. MoreS stretch) and Q4 (out-of-plane over, they observed that quanta of the Q1 (C wag) were conserved as good quantum numbers throughout the spectrum. On the other hand, the levels derived from the remaining modes were found to be highly fragmented and susceptible to IVR. On Franck-Condon grounds, the S and the out-of-plane coordinates lead T1 S0 structural changes in the C to dominant progressions in Q1 and Q4 in the corresponding spectrum. If the Franck-Condon factors that control the spin-orbit coupling between the T1 and the S0 states can be approximated by the band patterns in the S2 ! S0 fluorescence spectrum, then the molecular eigenstate spectrum representing the coupling of T1(00) with the S0 (high v) levels could be grouped into two parts. The first part would be defined by the highly active Franck-Condon coupling of T1(00) to the levels in Q1 and Q4 and their combinations. The second part would consist of the coupling to the much weaker and denser fragmented background levels. The Fourier transform of this molecular eigenstate spectrum would result in a decay curve that would display nonexponential behavior [30]. The strong spin-orbit coupled lines derived from the good quantum numbers in Q1 and Q4 would Fourier transform to form the initial picosecond part of the decay curve, while the weaker background levels would transform to create longer microsecond tail [30]. In summary, the source of the anomalous photophysical behavior of the Tz spin component of the T1 electronic state can be found in the observation of structured vibrational states (Q1 and Q4) embedded in a dense set of fragmented background levels in S0 (high v) at the T1 S0 point of crossing, 17,499 cm1. The lack of phosphorescence from the T1 state, among other factors, is the result of the favorable Franck-Condon factors brought about by the planar– nonplanar nature of the ISC process. This behavior is to be contrasted to the related thiocarbonyl species, CHSCH3 and (CH3)2CS, which are emissive with measured lifetimes of 2 ms [29]. The pyramidal distortions that occur on
56
THIOPHOSGENE
excitation for these thiocarbonyl prototypes are less severe than for thiophosgene, and the coupling elements for the ISC process are substantially smaller due to the lack of heavy atom.
E. S2, Second Singlet Excited State ~ 1A1(p,p*) X ~ 1 A1(n2) or S2 S0 room-temperature Early studies of the B absorption spectrum were carried out by Farnworth and King [32], and then later in emission by Steer and co-workers [21]. Detailed analyses of the hot band structure were made by Judge and Moule [33], who located the origin for the system and identified several band progressions. In the course of recording the S1 S0, LIF excitation spectrum, Clouthier et al. [26] observed a UV emission that could be traced to the S2 state. This was attributed to a two-photon one-color OODR transition between the S2 and the S0 states, where the S1 and T1 levels played the roles of intermediates: S2 S1 S0 and S2 T1 S0. These experiments were followed by a two-color pump–probe study by Dixon and Western [34], who located a group of vibrational bands to the red of the origin that suggested the system origin lay to lower wavenumbers. This controversy was settled by a one-photon S2 S0, jet-cooled, LIF experiment by Steer and co-workers [35], who were able to assign a Q4 (out-of-plane wagging) progression that converged on the origin established earlier. The height of the barrier in v(Q4) extracted from these Q4 data was much lower than that suggested by the observed hot band activity in Q4, indicating that there were unresolved problems with the assignments. Fujiwara et al. [15] recorded one-photon and two-photon spectra under jet-cooled conditions and have reopened the question of the barrier height to molecular inversion. The electron configuration of the S2 state is created through a p to a p* electron excitation and with the direct product of these outermost orbitals of B1 species transforming as the A1 representation. The second singlet state can ~ 1A1(p,p*) X ~ 1A1(n2) or ~ 1A1(p,p*). The transition, B then be written as B S S2 S0, is allowed by optical selection rules and is polarized in the z (C bond) direction. Thus the vibronic transitions that form strong progressions in the excitation spectra will terminate on the (þ) lower member of the inversion doublets. From the Walsh correlation diagram (Fig. 2.3), this state should be nonplanar since the p* orbital energy decreases (stabilizes) with the pyramidal deformation, while the p orbital energy simultaneously increases (destabilizes). Both these changes favor a pyramidal structure. On this basis, the barrier should be higher in the S2(p,p*) state than in the S1(n,p*) excited state. Figure 2.20 shows a representation of the out-of-plane potential for the S0, S1, and S2 states, the barrier heights, and their effects on the inversion splittings. The other anticipated structural change occurs in the CS bond. Removal of an electron from the
ELECTRONIC STATES
57
Figure 2.20. Potential energy functions and vibrational levels for the out-of-plane displacements for the first three singlet states of thiophosgene. (Reprinted with permission from Ref. [15].)
p and its placement in an antibonding p* orbital has the effect of reducing the C S bond. This change in bond order S double bond order to that of a single C induces a large increase in the CS bond, reduces the stretching frequency, and results in the formation of a dominant progression in Q1 (CS stretching mode). These structural and dynamical changes are clearly illustrated in a recent ab initio treatment by Strickler and Gruebele [6], who developed a complete molecular potential for the S2 state from a series of CASSCF ab initio calculations. A segment cut from their sixfold potential is illustrated in Figure 2.21 as a plot of energy against increments in the CS bond and the out-of-plane angle. This three˚ and an outdimensional plot demonstrates that increases of r(CS) ¼ 0.39 A
of-plane displacement of 45 are needed to achieve the equilibrium position (bottom of the well). But what was not anticipated was that at extended CS bond lengths, the S2 electronic system encounters a new excited electronic state. This S3 state is attributed to s* n excitation, and as (s*) (n) ¼ B2 ~ 1B2(n,s*). The plot shows that with increasing A1 ¼ B2, this state is labeled C
58
THIOPHOSGENE
Figure 2.21. The potential energy calculated at the ab initio CASSCF level illustrating the intersection of the pyramidal and planar surfaces. (Reprinted with permission from Ref. [6].)
CS distance, the potential energy of the system increases from its S2 pyramidal ˚ , whereupon the strucequilibrium position to the point of S2–S3 crossing, 0.80 A ture converts to a planar conformation. The large distortions in the r(CS) and y (out-of-plane) directions that are needed to reach the equilibrium position of the S2 state result in long progressions in Q1 and Q4 modes that weaken the 000 band by about two orders of magnitude from the prominent bands at higher energies. Thus, in the room temperature absorption spectrum, the 000 origin band is not directly observed and its position must be established from the attached hot bands, 411 and 402 . This difficulty can be overcome in the one-photon LIF jet-cooled spectrum because (1) the quantum yield for fluorescence at the lowest vibrational energies, Ev ¼ 0, is very high, F 1; (2) the oscillator strength is that of a fully allowed electric dipole transition; and (3) under jet conditions the hot bands that strongly interfere with the clarity of the band patterns in the origin region are suppressed. The vibrational assignments are also greatly aided by the presence of the 35 Cl2CS and 35Cl37ClCS isotopomers that form doublets in the spectrum. Combination differences, 0 ¼ 0 0þ ¼ 41 40 ¼ 411 000 þ v004 and 1 ¼ 1 1þ ¼ 43 42 ¼ 431 420 þ v004 , extracted from hot band intervals are also invaluable in this regard. The vibrational assignments given in Figure 2.22 can be summarized as follows. The progression responsible for the bulk of the intensity comes from the activity of Q1. This is not unexpected and is the direct consequence of the changes in bond order within the CS bond. This progression is observed to
ELECTRONIC STATES
59
Figure 2.22. The one-photon LIF excitation spectra of Cl2CS recorded under (a) cooled (P ¼ 4 atm) and (b) warm (P ¼ 400 Torr) jet conditions. Leaders, denote the progression in quanta of the Q4 out-of-plane wagging mode; 0, 2, and 4, refer to the 000 , 420 , and 440 transitions attached to quanta n of the Q1 mode, 110 ; 1 and 2, designate the first and second hot band transitions 411 and 402 . (Reprinted with permission from Ref. [15].)
take the form of band clusters involving hot and cold transitions in Q004 and Q02 . The reduction in frequency from v002 ¼ 1139 to v02 ¼ 505 cm1 is consistent with the adjustment in C S bonding. The small differences in the C Cl stretching frequency, v002 ¼ 497 and v02 ¼ 495 cm1 would suggest that the activity in Q2 results from a mixing of the internal coordinates, rather than from displacements in the CCl bond. For the 00 (0þ zero point level) the observed hot band intervals in Ev ¼ 41 and 42 give a inversion splitting that is smaller than the measuring error allowing the inversion splitting 0 ¼ 0 0þ to be set to 0.00 cm1. The next two quanta of the Q4 mode, 42 and 44 at 341.4 and 682.4 cm1, display very small splittings, 1 ¼ 0.03 and 2 ¼ 0.16 cm1, respectively. Thus within the manifold of Q4 levels that attach to the vibrational zero point, the level splittings are very small, and it is clear that the barrier to inversion must be very high. Bands involving combinations of Q4 with Q1, however, do show appreciable splittings.
60
THIOPHOSGENE
The 110 transition forms a strong suborigin band along with an attached cold band progression in Q4, 110 , 110 420 , and 110 440 . While 110 by itself has a very small inversion splitting, 0 ¼ 0.07 cm1, a clear splitting of 2 ¼ 6.53 cm1 can be derived from the cold 110 440 band and its hot 110 451 partner band. When this same process is repeated for v4 built on the second quanta of n1, 120 , a 0 of 9.85 cm1 is observed for 110 420. This behavior suggests that while both Q1 and Q4 are involved in the inversion process, neither is truly an inversion mode. The correct picture should use, as a starting point, the potential energy surface shown in Figure 2.21. The high fluorescence yield from the S2 (v ¼ 0) level creates an opportunity to observe three unique excitation spectra: the one-photon S2 S0 and the twophoton processes S2 S1 S0 and S2 T1 S0. The selection rules for these transitions are different and the three spectra reveal different aspects of the structure and dynamics. In the case of the two OODR spectra (Fig. 2.23), it should be appreciated that the pump–probe time delay of 150 ns reduces the two sequential transitions in the OODR process to a single step. Although it is the function of the pump photon to initially populate the intermediate S1 or T1 states, the properties of the OODR spectra are defined by the second step, S2 S1, or S2 T1. Thus we label the two photon spectra OODR(S1) or OODR(T1), depending on whether the intermediate step is the singlet or the triplet intermediate. The OODR(S1) spectrum may be regarded as a stepwise singlet–singlet transition, S2 S1, between the 1A1 and the 1A2 electronic states, which is followed, after a fixed time delay, by the pump transition S1 S0 connecting the 1A2 and 1A1 states. The first transition excited by the probe photon in the OODR(S1) process is electric dipole forbidden, which gains its strength through vibronic coupling to the B2(n,4s) Rydberg state. The pump photon uses this same B2(n,4s) coupling mechanism for its strength, and it might be expected that the transition probabilities of the pump and probe processes should be about the same. The selection rules for the OODR(S1) pump, S1(41()) S0(00(þ)), and probe, S1(41()), yield an overall transition S2(00(þ)) S0(00(þ)), S2(00(þ)) with the result that only the S2(þ) manifold of levels are interrogated. If, on the other hand, the hot band transition is pumped, S1(00(þ)) S0(41()), then the probe transition will terminate on a manifold of S2() symmetry, S2(41()) S1(00(þ)). In this way, it is possible to build up an excitation spectrum for the S2 state in which the either the (þ) or () manifold of states is selected. When the two spectra are scaled to the same electronic origin, the two manifolds may be directly compared as illustrated in Figure 2.24. Here, the two sets of inversion spectra have similar appearances; and as the inversion doubling splittings come from the displacements of one set relative to the other, it is clear that the inversion splittings within the S2 vibrational manifolds are very small indeed.
ELECTRONIC STATES
61
Figure 2.23. A comparison of the vibrational structure of the S2 state from OODR spectra recorded with the S1(2141) and T1(21) states as intermediates. (Upper panel) S2(v0 ) S1(2141) S0(00). (Lower panel) S2(v0 ) T1(21) S0(00). (Reprinted with permission from Ref. [15].)
The OODR(T1) transition may be viewed in somewhat the same way as the OODR(S1) process in that the vibronic operator is replaced by a spin-orbit operator connecting T1 triplet levels with adjacent singlet states of an appropriate symmetry. Thus the pump transition T1 S0 is forbidden by the S ¼ 0 rule, and gains its intensity by borrowing from the adjacent S2 S0 transition through a T1 $ S2 coupling. This one-photon process has the vibrational symmetry properties of a z-polarized transition connecting sets of (þ) levels, T1(00 (þ)) S0(00(þ)). The probe transition S2 T1 is also spin forbidden, with T1 spin-orbit coupled to the S0 ground singlet state. Both the pump and
62
THIOPHOSGENE
Figure 2.24. A comparison of the vibrational structure of the S2 state from OODR spectra recorded with the S1(00) and S1(41) states as intermediates. (a) S2(v0 ) S1(41) S0(00). (b) S2(v0 ) S1(00) S0(41). (Reprinted with permission from Ref. [15].)
the probe transitions gain their strength from a borrowing from the same S2 S0 transition. As the perturbation gaps and the coupling elements are of the same magnitude, each step in the two-photon process should have similar transition strengths. The vibronic levels in combination will be of (þ) inversion symmetry, S2(00(þ)) T1(00(þ)) S0(00(þ)). A simple comparison of the two panels in Figure 2.23 reveals that the OODR(S1) spectra has twice the band density as the OODR(T1) spectra. This difference must be associated with the differing selection rules for the two cases. In the OODR(T1) process, the vibrational activity of single quanta of the Q4(B1) or Q5/Q6 (B2) modes requires a second-order vibronic-spin-orbit coupling mechanism, and these bands are expected to be weak. The strong band in the OODR(S1) spectrum at Ev ¼ 212 cm1, which is absent in the OODR(T1) spectrum, is then given the assignment v6. With this assignment, the next band at Ev ¼ 248 cm1 is then attributed to the low-frequency mode v3 by process of elimination.
ELECTRONIC STATES
63
Figure 2.25. Temporal profiles of polarization-dependent OODR transitions to (a) the 00 and (b) the 61 levels in the S2 state, S2(00) S1(41) S0(00) and S2(61) S1(41) S0(00), showing the differences in intensity of the emission from the S2 state with the electric vectors (Es) of the pump and probe lasers that are parallel (E1 // E2) and perpendicular (E1 ? E2).
To add further credence to these critical assignments, the polarization-dependent OODR(S1) spectrum was recorded for the 000 and 610 bands with implement of photoelastic modulator (PEM). This device alters the polarizations of the probe and pump beams with each laser pulse in a shot-by-shot fashion. For the 000 transition, the two beams would have parallel polarizations (pump and probe transition are both y (B2) polarized). For the 610 transition, the two photons would be perpendicularly polarized to each other, with y and z polarization, respectively. The temporal profiles of polarization-dependent OODR spectra in Figure 2.25 show the expected behavior, where the signals of the parallel and perpendicular polarizations for the S2(00) S1(41) S0(00 ) and S2(61) S1(41) S0(00 ) are plotted against the delay time. The effects of reversing the polarizations is not too pronounced in these experiments, as the rotational coherence lifetimes are very short and the overall rotation of the molecule quickly scrambles the polarizations of the signals. At very short delay
64
THIOPHOSGENE
times, the two traces do change in the expected directions, giving credence to the assignments. The concern about these critical assignments relates to the crossing of the ~ 1B2(n,s*). Figure 2.4 shows that stable S2 state is crossed ~ 1A1(p,p*) and C B ˚ . As these states are of A1 and B2 symby an unstable S3 state at r(CS) ¼ 2.44 A metry, they would not interact with each other in the planar C2v conformation. These states, however, would engage in a vibronic interaction induced by distortions in the B2 y-direction. Such a distortion could be brought about by the Q6 mode. The intensity of the 620 band in the OODR(S1) spectrum is very low, and the Ev ¼ 421 cm1 interval is twice the fundamental, 212 cm1, indicating that the V(Q6) potential is nearly harmonic. This observation is in accordance with the calculated vibronic interaction Vev of <0.2 cm1 [6]. This low value may be related to the lack of a common molecular orbital between the interacting state configurations p,p* and n,s*.
III.
PHOTOPHYSICAL PROPERTIES A. S1 Thiophosgene
In the so-called statistical limit of radiationless transitions (where the molecule undergoes an irreversible, exponential decay), the rate constant knr of nonradiative decay from the initial electronic state jsi to the final electronic state j li is given by [36] knr ¼
2p j hsj ^ v j li j2 rl h
ð7Þ
where ^v represents the perturbation responsible for the radiationless decay and rl is the state density of j li. To the first order, ^v is the nuclear kinetic energy operator (T^N ) for internal conversion (a radiationless transition between two electronic ^ SO ) for intersysstates of the same multiplicity) and spin-orbit coupling (H tem crossing (radiationless transitions between states of different spin multiplicity). Within the Condon approximation, which assumes that the electronic matrix elements are a slowly varying function of coordinates, the the nuclear matrix elements for the internal conversion sj T^N j l between adiabatic Born-Oppenheimer states, j si fs ðq;QÞws ðQÞ and j li fl ðq; QÞwl ðQÞ, can be written in the form [37] h2 Vsl ¼
E X q q Y D fl w w fs w j w sui suj lvj qQ q Q lvi i
i
i
j6¼i
ð8Þ
PHOTOPHYSICAL PROPERTIES
65
where f and w represent electronic and vibrational wavefunctions respectively, u and v are the vibrational quantum for modes i and j. Qi represents the massweighted normal coordinate for mode i. Note that the matrix elements responsible for the internal conversion are composed of two factors: an electronic integral hfs j q=qQi j fl i, representing nonadiabatic coupling between the two electronic states; vibrational integrals containing a nuclear integral hws j q=qQi j wl i; and an overlap integral hws j wl i. Following Lin and Bershon [38], the vibrations for which the electronic matrix elements are finite are designated as ‘‘promoting modes,’’ whereas those for which overlap integrals are nonvanishing are classified as ‘‘accepting modes.’’ Since the Franck-Condon factors (which are the squared sum of the vibrational overlap integrals) decrease with increasing Es-Ee energy gas, Eq. (8) accounts for the decrease in the measured rate constant with increasing energy difference between the initial and the final electronic states, viz., the energy-gap law. To the lowest order, the electronic matrix elements in Eq. (8) are determined by the vibronic coupling integral [37, 38] qUðq; QÞ f f ðq;QÞ ðq;QÞ s qQ l q ð9Þ fs ðq;QÞ fl ðq;QÞ ¼ El ðQÞEs ðQÞ qQ where U is the potential energy of interaction between the electrons and nuclei and E represents the energy of the electronic state. Eq. (9) forms the basis for selection rules for radiationless transitions. The vibronic coupling operator qU/qQ transforms as the nuclear coordinate Q, (qU/qQ) ¼ (Q); and for internal conversion (IC) between the initial electronic state j si and the final electronic state j li, the requirement is that the direct product (j li) (Q) (j si) contain the totally symmetric representation of the point group. Despite the important implications, this selection rule for the electronic matrix elements has not been put to a rigorous experimental test. The reason is that, for polyatomic molecules with relatively low symmetries, the normal coordinates usually encompass all irreducible representations of the symmetry group, and a separation of the mode structure into promoting and accepting modes is not feasible. A critical test of the selection rules requires a rather special molecule with very special energy level dispositions and normal coordinates that do not span the entire irreducible representation of the symmetry group. The term vibrationally deficient was suggested by Gardner and Kasha [39] for the class of molecules that fail to span the vibrational irreducible representations. The six normal modes for thiophosgene (Fig. 2.6) show the decomposition of the six normal modes into the three in-plane totally symmetric modes, 3A1, the one out-of-plane mode, B1, and the two in-plane, antisymmetric modes, 2B2:
66
THIOPHOSGENE
(v) ¼ 3A1 B1 2B2. Lacking in this collection is a mode of A2 species. Somewhat fortuitously this vibrationally deficient mode has the correct symmetry to create the selection rules for the IC process. Thus if we return to Eq. (9) and identify the initial state with the singlet n,p* excited state and the final state with the ground state, j si ¼ S1, and j li ¼ S0, then for the S1 to transfer energy to the S0 state, the direct product (S0) (Q) (S1) ¼ A1 (Q) A2 ¼ A1; and as a result, (Q) ¼ A2, identifying the promoting mode as a mode that transforms as a rotation about the z-axis (C S bond). While such modes are available for larger thiocarbonyl C2v , systems, such as thioacetone or thiocyclobutanone, they are not found in tetra-atomic molecules, such as thiophosgene [1,40]. Without a promoting mode, the normal IC process is symmetry restricted, allowing for photon energy deactivation through the alternative pathway of fluorescence emission. The extent to which the IC process is blocked by vibrational deficiency can be ascertained by the quantum yield for fluorescence, F. Such information can be extracted from the measured fluorescence lifetimes, tF, of the vibronic levels of the S1 state, providing the oscillator strength of the S1 S0 transition is known. The radiative lifetime, based on the transition strength, tr, can calculated from the oscillator strength through the expression, tr 1.5/fv2. With a measured oscillator strength of f ¼ 1.2104 and an average transition frequency of v ¼ 19,000 cm1 the radiative lifetime is estimated to be tr ¼ 35 ms. This value is identical to the room-temperature fluorescence lifetime of McDonald and Brus [41]. With equal values of the calculated and observed lifetimes, the quantum yield for fluorescence is unity, F ¼ 1, demonstrating that internal conversion is completely blocked by the absence of a promoting mode, and it does ˜ 1A2(41) level. Thus the not play a role in the photophysical behavior of the A strong fluorescence from the first excited state of thiophosgene can be directly attributed to the absence of a vibrational mode of A2 species within the set of irreducible representations that span the C2v point group. In summary, the high fluorescence quantum yield demonstrates the critical role of the promoting mode and the Gardner-Kasha supposition [39] that vibrational deficiencies may create selection rules that hinder radiationless transitions. These early lifetime results, obtained from bulb conditions, (room temperature) were compared to similar data recorded under jet-cooled conditions. Analyses of the decay characteristics of the 41 level fluorescence at Ev ¼ 0.42 cm1 yielded a lifetime of 1.5 ms and was found to be internally consistent with the value derived from the OODR pump–probe method with beam displacement [42]. Both of these lifetimes are considerably shorter than the 35 ms value accepted for the S1 state [41]. Since it is highly unlikely that nonradiative decay, unimportant at room temperature, would become significant under the lowtemperature conditions of the supersonic jet, the much longer lifetime at room temperature must have its origin in a reduced radiative decay at elevated
PHOTOPHYSICAL PROPERTIES
67
Figure 2.26. Quantum beat–modulated fluorescence decay and its Fourier transform for the 11213141 level of S1 thiophosgene. (Reprinted with permission from Ref. [42].)
temperatures. It has been proposed that this temperature dependence arises from a Coriolis coupling that mixes the Q4 active mode with the other modes of the S1 manifold of vibrational levels [42]. It is interesting that thiophosgene exhibits quantum beats in emission [42] when it is excited into the vibronic levels lying in the vicinity of 20,340 cm1. Figure 2.26 displays the beat-modulated fluorescence decay of the S1(11213141) level. Fourier analysis reveals major frequency components at 6.8 and 12.0 MHz. Similar, but more complex beat patterns were also observed for higher vibronic levels. This quantum beat structure was attributed ˜ 1A2(n,p*) singlet state with the ~b 3A1(p,p*) to singlet-triplet interaction of the A triplet state, from the estimate of 20,160 cm1 based on the electron-impact
68
THIOPHOSGENE
energy loss spectrum [43]. As the radiative transition T2 ! S0 is nearly forbidden due to the lack of T2S2 spin-orbit coupling and since T2T1 internal conversion is symmetry forbidden owing to the absence of an A2 promoting mode, the perturbation of the S1 state by the T2 state is expected to have a negligible effect on the S1 lifetime. Based on the threshold energy for the observation of beat-modulated fluorescence decay, the ~ b 3A1(p,p*) state was placed at about 1 20,340 cm , in excellent agreement with ab initio predictions [44]. Decay curve lifetime measurements were extended for number of the higher S1 vibronic levels. Levels 41, 43, 2141, 1141, 112141, 11223141—levels at (Ev, cm1) 0.42, 292.5, 480.4, 907.8, 1387.4, and 1632.2—were found to have lifetimes, respectively, of (tf, ms ), 4.8, 4.3, 3.9, 3.5, 2.6, and 2.9. What is interesting is that the lifetimes shortens uniformly with increasing vibrational excitation throughout the S1 S0 system until the fluorescence extinguishes at Ev ¼ 3484 cm1. From a free-radical scavenging experiments Okabe [45] attributed this loss of fluorescence to the onset of chlorine free-radical production ~ ) process withbrought about by the fragmentation Cl2CS ! Cl(2P3/2) þ ClCS(X out an impeding dissociation barrier. A wide spectral view of the time-resolved fluorescence excitation spectrum of thiohosgene, ranging from the visible (18,700 cm1) to the UV (37,700 cm1), at different gate widths is presented in Figure 2.27. This LIF spectrum encompasses the transitions from the S0 ground state to both the S1 and the S2 states as well as mapping the intervening spectral region. The upper spectrum (Fig. 2.27a) recorded with short-gate width (0–20 ns), is identical to the LIF S2 S0 system in Figure 2.22a. The spectrum in Figure. 2.27b, recorded with the longer gate width (140–440 ns), is lacking the shorter-lived S2 S0 spectrum, although it now contains the complete S1 S0 system at lower energies. Somewhat surprising, is the emergence of a new excitation spectrum in the intermediate region. This new spectrum was assigned to an extension of the first singlet-singlet ~ 1A1(n2) from considerations of the dis˜ 1A2(n,p*) X transition, S1 S0, A persed fluorescence and fluorescence lifetime measurements. Figure 2.28 shows the extreme contrasts in the decay curves that are observed for the three regions of the LIF spectrum. These decay curves from exciting the S2(00), S1(v), and S1(41) levels at (Ev, cm1) 0.00, 12,658, and 0.42 have lifetimes of 50 ns, 1.6 ms, and 1.5 ms, which mirror the behavior of the variable gate-delay data depicted in Figure 2.27. Indeed, the lifetimes of the two low-energy systems fluctuate by only 2–5 ms over excitation energies that range to Ev ¼ 18,984 cm1 and allow the level patterns to be classified as members of the same ˜ 1A2(n,p*) vibrational manifold. The second aid to the electronic assignment A is shown in Figure 2.29 as a comparison of the dispersed fluorescence from the S1(2141) level at 19,196 cm1 with the excited vibronic levels of the intermediate state at 29,106 cm1. What is immediately apparent is the complexity of the resolved vibrational structure in the fluorescence of the lower panel. The length
69
Figure 2.27. Time-resolved S1 S0 and S2 S0 fluorescence excitation spectra of jet-cooled thiophosgene in the energy range 17,700– 37,700 cm1, recorded with different gate time arrangements: (a) First 20 ns, (b) starting at 140 ns with a 300 ns width, (c) all time integrated. Total emissions were monitored and normalized to input laser powers. The broad unstructured feature at about 37,000 cm1 is believed to be the S2 S0 transition of the thiophosgene dimer, as its intensity is much weaker in warmer supersonic beams. Inset: Schematic energy level diagram of thiophosgene. (Reprinted with permission from Ref. [42].)
70
THIOPHOSGENE
~ (S2) state Figure 2.28. Fluorescence decay from a low-lying vibronic level (00) of the B ~ (S1) state. (Reprinted with and that from the 41 and high-lying vibronic levels of the A permission from Ref. [42].)
of this spectrum is partially the result of a long-wavelength tail that contains features that closely mimic resonance fluorescence from the 2141 level. This correspondence can be accounted for if the excited S1 level undergoes extensive IVR [46] to populate levels that contain low degrees of excitation in the optically modes (v4, in particular). The dispersed fluorescence from other high lying vibronic levels also bears the spatial characteristics of levels that undergo extensive IVR. The most striking feature of the excitation spectra of Figure 2.27 is the breaking off of the S1 ! S0 fluorescence at Ev ¼ 3484 cm1 above the S1 origin and its reappearance at about 9284 cm1 [42]. Thus, over a range of about 5800 cm1, there is no detectable emission from the S1 state. This behavior is consistent with the occurrence of predissociation, leading to the production of Cl (2P1/2) and ~ ) (Fig. 2.27, inst). The characteristic feature of the predissociation, as ClCS (X
PHOTOPHYSICAL PROPERTIES
71
Figure 2.29. Comparison of the low-resolution dispersed fluorescence spectra of S1 obtained after excitation of the 2141 vibronic level and a high-energy vibronic feature at 29,106 cm1. (Reprinted with permission from Ref. [42].)
probed by the S1 ! S0 fluorescence is that there is a region of Ev in which the emission intensity (or lifetime) would is extremely weak (or extremely short) due to the transition to the repulsive exit channel. For Ev far above or far below the intersection region, the fluorescence intensity (or lifetime) is expected to be normal due to very small nonadiabatic interactions between the bound and the repulsive states (Fig. 2.27, inset). To state it somewhat differently, below the crossing point, the molecule has insufficient energy to dissociate, whereas above the crossing point the area occupied by the dissociation channel within the six-dimensional vibrational hypersurface is sufficiently small that vibrational structure is preserved. Based on the observed onset of the fluorescence breakoff, we place the curve crossing (leading to predissociation) at 3484 cm1 above the electronic origin of the S1 state, in excellent agreement with the existing assignment of 3454 cm1. The occurrence of the intensity maximum in the S1 S0 excitation spectrum ˜ 1A2(41) is consistent with this at about 12,000 cm1 above the false origin A 1 1 assignment for two reasons. First, the B2 A2 coupling which induces the S1 S0 transition strength is expected to more efficient for the higher lying vibronic levels of the S1 state because of the smaller 1B2 1A2 perturbation energy gaps. Second, owing to the nonplanarity of the S1 state and the change
72
THIOPHOSGENE
in the C S bond length relative to the ground state, vertical (adiabatic) transitions would likely occur to high-lying vibrational levels of the excited state as a consequence of the Franck-Condon overlap factors. The essential identity of the fluorescence lifetime at very small and very large Ev indicates that the S1 state in thiophosgene does not exhibit significant radiationless decay to the bound S0 ground electronic state, even when the molecule is endowed with very high degrees of vibrational excitation. The lack of S1(A2) ! S0(A1) internal conversion from the vibrationally highly excited S1 state implies that the higher order promoting modes, Q4 with Q6, (b1 b2) with a2 symmetry, are not efficient in inducing a radiationless process. The near constancy of the fluorescence lifetime also indicates that S1 ! T1 intersystem crossing from the manifold of S1 is also inefficient. This behavior is in total contrast to the T1 ! S0 ISC process, discussed earlier, and can be attributed to the small Franck-Condon factor for the radiationless transition that arises from the small S1 T1 energy gap and the similar geometries for the two electronic states. The forbidden nature of the direct S1 T1 spin-orbit coupling is also a factor. As intersystem crossing from the S1 state to the higher lying T2 state is not possible for collision-free thiophosgene, as a consequence of the much smaller vibrational density of the interacting states, the only available decay channel of S1 thiophosgene is the radiative transition (fluorescence) to the ground S0 state.
B. S2 Thiophosgene ~ 1A1(p,p*) 1A1(n2) photographic absorption spectrum of thiophosThe S2 B gene at room temperature consists of a series of sharp lines in complex patterns that quickly broaden and become diffuse at shorter wavelengths as the result of an early onset of a molecular predissociation. Under the emissive requirements of excitation spectroscopy this line broadening can be correlated to the diminution of the fluorescence signal strength with excess vibrational energy. The jet-cooled LIF spectra in Figure 2.27 with different gate widths clearly shows the variation in the temporal behavior associated with the line broadening at higher energies. The strength of the fluorescence emission, as measured by the quantum yield, can be judged from a comparison of the observed lifetimes of single vibronic levels, with the lifetime extracted from the measured oscillator strength. This procedure yields the remarkably high quantum yield of unity (F ¼ 1.0) for the 00 level (Ev ¼ 0.0) from the measured lifetime (tf ¼ 39.8 ns). That the deactivation of the v ¼ 0 level is completely radiative is a testimony of the law of vibrational deficiency. The first and second electronic states are in perfect juxtaposition, and it is only their electronic symmetries that prevent the onset of an S2 S1 IC. The matrix element connecting these electronic states, hS2 ðA1 Þj QP j S1 ðA2 Þi, however,
PHOTOPHYSICAL PROPERTIES
73
requires a promoting mode Qp to bear the symmetry A2 level. Within the C2v group, this representation is not present as one of the six normal modes (v ¼ 3A1 B1 B2), and it is the absence of such a mode that is responsible for blocking the radiationless IC route as a pathway for photon deactivation. It should be noted that the lack of a suitable promoting mode closes the two major IC pathways, S2 S1 and S1 S0 and is responsible for the fluorescence from the S2 state with high quantum yield, in violation of Kasha’s rule. For the higher vibronic levels, the fluorescence quantum yield was found to decrease from F ¼ 1.0 at zero excess energy, Ev ¼ 0.0, to F 0.5 at 835 cm1 and remain at <1.0 until the excess energy reaches 1300 cm1, at which point the S2 S0 fluorescence breaks off. The loss of LIF signal at Ev ¼ 1300 cm1 of vibrational energy has been interpreted in terms of C Cl bond dissociation [45]. Warsylewicz et al. [47] investigated the temporal characteristics of selected bands within the system under picosecond temporal resolution. They found that the fluorescence decay curves for the levels below the 1300 cm1 threshold consisted of a exponential component (30–40 ns) superimposed on a fast nonexponential decay (15 ns). From the area under the decay curve, they established that the fraction that decayed at the radiative rate varied from nearly unity (0.96) at Ev ¼ 0.0 cm1 to being completely absent at Ev ¼ 1211 cm1. The reduction in quantum yield of fluorescence and the onset of the picosecond nonradiative component over the short Ev energy span cannot be attributed to the excess energy dependence of an IC process between the S2 and the S1 states. Rather, the observation of the fast-decaying component suggests that an alternative radiationless process is responsible for depleting the bright states populated by the S2 S0 excitation. Such a process must be slow relative to the decay at zero excess energy and becomes equal to the radiative decay at 835 cm1. Thus the temporal characteristics of the fluorescence decay consist of one component that remains optically bright and gives rise to the exponentially decaying radiative fluorescence (30–40 ns) and a second component that evolves over a period of a few nanoseconds into a set of bound but dark states. ~ , that is Steer et al. [8, 48] have proposed that the dark state is a new state, C formed when two electrons are lifted from the n orbital and placed into the p* orbital in a doubly excited configuration, (n)0(p*)2. Multireference CASSCF calculations of Strickler and Gruebele [6] indicate, however, that this phantom state ~ 1B2(n,s*) electronic state, formed from the one electron is very likely the C ~ state is planar and intern ! s* excitation. Their calculations show that the C 1 ~ S bond distances ( 2.5 sects the nonplanar B A1(p,p*) state at increasing C ˚ ). The barrier separating the B ~ state ranges from 300 cm1 (y ¼ 0) ~ from the C A of the transition dipole moments to about 1650 cm1 (y ¼ 45 ). Calculations ~j m j X ~ , element is about three orders along the C S bond shows that the C ~ state ~j m j X ~ and provides the rationalization of C of magnitude smaller than B darkness.
74
THIOPHOSGENE
Figure 2.30. Schematic energy-level diagrams depicting the pump–probe (a) OODR and (b) OODR–fluorescence depletion (OODR–FD) experiments. The y and z directions denote the polarization vectors of the various transitions. (Reprinted with permission from Ref. [49].)
~ state and the identiSome progress has been made on the detection of the C fication of the associated electron configuration. As this state is dark, a nonradiative absorption technique rather than a fluorescence method is required. Figure 2.30 depicts the energy schematics of such an experiment where the optical–optical double resonance process is combined with fluorescence depletion (OODR–FD). The first step in this method, involves the excitation of jet~ state through a pumping ~ to the A cooled thiophosgene from the ground state X 1 0 3 of the 40 , 41 , or 40 vibronic transitions. This is followed by the time-delayed ~ states. The depletion of ~ or C probe laser, which induces transitions to the B ~ ~ ~ the A population signaling a C A transition is monitored by the resulting ~ !X ~ fluorescence. Under these conditions, the pump laser– reduction in A induced emission appears as a background of essentially constant intensity, on which the depletion induced by the probe laser is superimposed as a negativegoing signal. Figure 2.31 shows a four-panel comparison of the OODR–FD and OODR spectra [49]. The signals are the result of transitions that stem from the
PHOTOPHYSICAL PROPERTIES
75
~ (41) and A ~ (43) levels and terminate on one of the X ~ states. A ~, B ~ , or C populated A major problem with these spectra is to separate the signals and, by elimination, ~ ~ transitions that lead to the dark state. Those transitions to B ~ identify the C A that form strong dips at the high frequency end of in Figure 2.31b, d can be unscrambled and assigned by comparing the positive OODR and negative OODR–FD signals (Fig. 2.31a, c). Pumping the 41 and 43 levels yields the (þ) and () inversion components and reveals the inversion splittings (þ and ~ states. The assignments to the B ~ , and C ~ state were confirmed ~, B ) in the X by their negligibly small inversion doubling intervals. Finally, the FD transitions ~ state must be disentangled from the downward transitions to the upward to the C ~ state. It is important to point out that this is the third major source of fluoresX ~ and C ~ ~ ~ cence depletion in the OODR scheme and is unrelated to the B A A ~ absorptions. This interference is SEP (stimulated emission pumping) of the A state, which leads to depletion of the spontaneous emission. The SEP positions can be calculated from the effective Hamiltonian of Gruebele [19]. The locations of these bands are shown by leaders along with quantum number assignments. The two candidate electronic states lead to different predictions about the direc~ ~ transition moment absorption relative to the moment of the tion of the C A ~ ~ state of (n)0(p)2 configura~ A X absorption. As illustrated in Figure 2.30, a C tion would require pump and probe beams of parallel polarization, whereas perpendicularly polarized beams would be needed to create the (n)1(s )1 configuration. The unaccounted bands that remain after the elimination process ~ ~ transition, although are labeled with asterisks. They are ascribed to the C A their vibronic assignments have yet to be ascertained. ~ X ~ A ~ ) is con~ internal conversion (relative to C The importance of the C ~ firmed by the absence of fluorescence from the A state following excitation to ~ state. It is well known that the dominant accepting the higher levels of the B mode for the S1 S0 internal conversion in aliphatic carbonyls (or thiocarbonyls) is the C O (or C S) vibration [50]. The special role of the carbonyl stretching vibration is a consequence of its lower frequency in the S1 state relative to S0. The frequency and structural differences between the S1 and the S0 states render large Franck-Condon factors for the S1–S0 internal conversion. A simple model calculation has demonstrated that the internal conversion, which is already efficient at low excess energies, becomes even more pronounced with increasing vibrational excitation of the carbonyl stretching mode in the S1 state ~ X ~ IC to the ground electronic state to be [50]. One would expect the direct C ~ state as a significantly more important than a competing IC deactivation to the A ~ ~. result of the unusually long C S bond length in the C state relative to X Combined with the very small radiative decay rate resulting from the low oscil~ X ~ state essentially ~ transition, this would make the C lator strength of the C nonradiative.
76
CONCLUDING REMARKS
77
IV. CONCLUDING REMARKS The combination of supersonic-jet-laser spectroscopy and high-level ab initio calculations provides a powerful method for probing the geometrical structure and photophysical properties of small molecules. In this chapter, we described the experimental and theoretical characterizations of geometrical structures and elementary photoprocesses in supersonic free jet. The major focus of the review was on the OODR spectroscopy, which is designed to identify and characterize dark electronic states of the molecule. The main conclusions that emerge from the review can be summarized as follows. First, the equilibrium geometry of the molecule varies widely among the elec^ ) stats are planar, whereas ~ ) and the ns* singlet (C tronic states. The ground (X ~ ~ ) states are nonplanar. the np* singlet (A), np* triplet (~a), and pp* singlet (B 1 1 ~ ~ 1A2(S1) ! ~ Second, the absence of significant B A1(S2) ! A A2(S1) and A 1 ~ X A1(S0) internal conversion, as evidenced by the high quantum yields ~ !X ~!X ~ and A ~ fluorescence, is due to the lack of the pro(F 1.0) of the B ~ !X ~!X ~ and A ~ moting mode of a2 symmetry, which is required for the B radiationless transitions. ~ ) state, ~ ) thiophosgene to the S1 (A Third, one-photon excitation of the S0 (X followed by photodepletion of the S2 ! S0 fluorescence (fluorescence deple^ ) S 0 (A ~ ) excitation, by another photon, reveals the presence ~, C tion) by Sn (B ^ ~ state. The polarization as well as the of the dark C state in the vicinity of the B ^ inversion splitting patterns of the OODR–FD spectrum are consistent with the C 1 state being the planar B2 (ns*) state predicted by an ab initio calculation. Fourth, the S1 ! S0 fluorescence exhibits loss of emission at low S1 excess vibrational energy and reappearance at very large excess energies, consistent with predissociation. A quantum beat–modulated S1 fluorescence decay was observed at energies corresponding to the expected position of the T2 (p,p*) state. Fifth, the absence of significant phosphorescence is entirely consistent with ~ (S0) spin-orbit coupling and the nonplanar geometry of the ~a the large ~a (T1) X ~ state), which leads to highly efficient ~a ! X ~ interstate (vis a` vis the planar X system crossing. 3
~ !X ~ emission by the probe laser, Figure 2.31. Depletion of the pump laser–induced A ~ ~ ~ ~ ð41 Þ and (d) A ~ ð43 Þ. The B ~ ~ which induces the B A and C A transitions: (b) A 0 0 ~ ð43 Þ ~ )(v0 ) A ~ ð00 Þ OODR–FD spectrum is displayed with energy offset of X (or C ~ 1A2(n,p*): 2(43 –41 ). Also 2 291.62 cm1, which represents two quanta of v4 in the A 0 0 0 1 3 ~ ~ ~ shown for comparison are the B(v ) Að4 ; 4 Þ X ð00 Þ OODR fluorescence excitation ~ !X ~ emission: (a) and (c). The leaders represent the expected band positions of the of B OODR–FD features due to SEP, whereas the asterisks in the lower-energy region of the OODR–FD spectrum (upper panel) denote the stronger features of the bands assigned to ~ ~ transition. (Reprinted with permission from Ref. [49].) the C A
78
THIOPHOSGENE
~ ) ! T1 (~a) transitions to the three spin sublevels of the T1 state Finally, S0 (X were observed by fluorescence-detected S2 T1 S0 OODR experiments, and confirmed by rotational contour analyses. The transition to the Tz spin sublevel is much stronger and broader in linewidth, than the transitions to the Tx and Ty spin sublevels, consistent with the theoretical expectations. Highly differing T1 S0 interaction strengths for the three spin sublevels leads to T1 decay, which appears to be ‘‘biexponential’’ with vastly different lifetimes. Thiophosgene is one of the simplest and best-studied prototype systems for which accurate potential energy surfaces can be obtained through experiment and high-level ab initio calculations. As such, thiophosgene is a molecule tailormade for the fundamental understanding of electronic relaxation in polyatomic molecules. The results summarized in this chapter confirm the special role thiophosgene plays in the field of molecular photophysics and photochemistry.
ACKNOWLEDGMENTS The work performed in the authors’ laboratories was supported by the Office of the Basic Energy Sciences of the U.S. Department of Energy and the National Sciences and Engineering Research Council of Canada.
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.
D. C. Moule and E. C. Lim, J. Chem. Phys. 1999, 110, 9341. M. Kasha, Disc. Faraday Soc. 1950, 9, 14. D. J. Clouthier and D. C. Moule, Top. Curr. Chem. 1989, 150, 167. A. Maciejewski and R. P. Steer, Chem. Rev. 1993, 93, 67. G. Herzberg, Molecular Structure and Molecular Spectra, vol. 1, Van Nostrand Reinhold, New York, 1945. B. Strickler and M. Gruebele, Chem. Phys. Lett. 2001, 349, 137. D. C. Moule and A. D. Walsh, Chem. Rev. 1975, 75, 67. R. P. Steer, Rev. Chem. Int. 1981, 4, 1. L. O. Brockway, Y. J. Beach and L. Pauling, J. Am. Chem. Soc. 1935, 57, 2693. J. H. Carpenter, D. F. Rimmer, J. G. Smith, and D. H. Whiffen, J. Chem. Soc., Faraday Trans. 1975, 71, 1752. D. C. Moule, I. R. Burling, H. Liu, and E. C. Lim, J. Chem. Phys. 1999, 111, 1. J. R. Lombardi, J. Chem. Phys. 1970, 52, 6126. J. C. D. Brand, J. L. Hardwick, and K-E Teo, J. Mol. Spectrosc. 1975, 57, 215. D. C. Moule and C. R. Subramaniam, J. Mol. Spectrosc. 1973, 48, 336. T. Fujiwara, E. C. Lim, and D. C. Moule, J. Chem. Phys. 2003, 119, 7741.
REFERENCES
16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51.
79
A. J. Downs, Spectrochim. Acta. 1963, 19, 147. C. A. Frenzel, K. E. Blick, C. R. Bennett, and K. Neidenzu, J. Chem. Phys. 1970, 53, 198. L. Burnelle, Acad. R. Belg. Classe Sci. Mem. 1958, 30, 7. R. Bigwood, B. Milam, and M. Gruebele, Chem. Phys. Lett. 1998, 287, 333. B. Simard, V. J. MacKenzie, P. A. Hackett, and R. P. Steer, Can. J. Chem. 1994, 72, 745. T. Oka, A. R. Knight, and R. P. Steer, J. Chem. Phys. 1975, 63, 2414. R. M. Hochstrasser, Molecular Aspects of Symmetry, W. A. Benjamin, New York, 1966. J. C. D. Brand, J. H. Callomon, D. C. Moule, J. Tyrrell, and T. H. Goodwin, Trans. Faraday Soc. 1965, 61, 2356. T. Fujiwara, E. C. Lim, and D. C. Moule, unpublished data. D. J. Clouthier and D. C. Moule, J. Mol. Spectrosc. 1981, 87, 471. D. J. Clouthier, A. R. Knight, R. P. Steer, and P. A. Hackett, J. Chem. Phys. 1980, 72, 1560. P. F. Bernath, P. G. Cummins, J. R. Lombardi, and R. W. Field, J. Mol. Spectrosc. 1978, 69, 66. M. Kawasaki, K. Kasatani, and H. Sato, Chem. Phys. 1985, 94, 179. T. Fujiwara, D. C. Moule, E. C. Lim, and R. H. Judge, unpublished data. T. Fujiwara, D. C. Moule, and E. C. Lim, unpublished data. B. Strickler and M. Gruebele, unpublished data. E. R. Farnsworth and G. W. King, J. Mol. Spectrosc. 1973, 46, 419. R. H. Judge and D. C. Moule, J. Mol Spectrosc. 1980, 80, 363. R. N. Dixon and C. M. Western, J. Mol. Spectrosc. 1986, 115, 74. M. Ludwiczak, D. R. Latimer, and R. P. Steer, J. Mol. Spectrosc. 1991, 147, 414. P. Avouris, W. M. Gelbart, and M. A. El-Sayed, Chem. Rev. 1997, 77, 793. S. H. Lin, J. Chem. Phys. 1996, 44, 3759. S. H. Lin and R. Bersohn, J. Chem. Phys. 1968, 48, 2732. P. J. Gardner and M. Kasha, J. Phys. Chem. 1969, 50, 1543. D. C. Moule and E. C. Lim, J. Phys. Chem. A 2002, 106 3072. J. R. McDonald and L. E. Brus, Chem. Phys. Lett. 1972, 16, 587. T. Fujiwara, D. C. Moule, and E. C. Lim, J. Phys. Chem. A 2003, 107, 10223. W. M. Flicker, O. A. Masher, and A. Kupperman, Chem. Phys. Lett. 1978, 57, 978. M. Hackey, F. Grein, and R. P. Steer, Can. J. Chem. 1993, 71, 112. H. Okabe, J. Chem. Phys. 1977, 66, 2058. P. M. Felker and A. H. Zewail, Adv. Chem. Phys. 1988, 70, 265. A. M. Warsylewicz, K. J. Falk, and R. P. Steer, Chem. Phys. Lett. 2002, 352, 48. A. Maciejewski, M. Szymanski, and R. P. Steer, Chem. Phys. 1992, 165, 101. T. Fujiwara, D. C. Moule, and E. C. Lim, Chem. Phys. Lett. 2004, 389, 165. G. D.Gillispie and E. C. Lim, Chem. Phys. Lett. 1975, 34, 513. T. Fujiwara, E. C. Lim, J. Koset, R. H. Judge, and D. C. Moule, in preparation for publication.
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS AND PENTAATOMIC HETEROCYCLIC COMPOUNDS M. D’Auria, L. Emanuele, and R. Racioppi, Dipartimento di Chimica, Universita` della Basilicata, Via N. Sauro 85, 85100 Potenza, Italy
CONTENTS I. II.
III.
Introduction The Paterno`-Bu¨chi Reaction on Furan Derivatives A. Furan B. Benzofuran C. 2,3-Dihydrofuran D. Synthetic Applications E. Asymmetric Reactions The Paterno`-Bu¨chi Reaction on Pentaatomic Heterocycles Different from Furan A. Thiophene B. Pyrrole C. Selenophene D. Imidazole E. Thiazole, Isoxazole, Isothiazole F. Indole
Advances in Photochemistry, Volume 28 Edited by Douglas C. Neckers, William S. Jenks, and Thomas Wolff # 2005 John Wiley & Sons, Inc.
81
82
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
G. Azaindole H. Possible Explanation of the Observed Reactivity IV Conclusions References
I. INTRODUCTION The Paterno`-Bu¨ chi reaction is a milestone in organic photochemistry (Scheme 3.1). Paterno` and Chieffi [1] reported obtaining oxetanes from the photocycloaddition of ketones to olefins in 1909, but this reaction was recognized as an important synthetic reaction only after the work of Bu¨chi et al. [2]. hv O
+
O
Scheme 3.1
After these pioneering works, several reviews cover the enormous number of papers published in this field [3–7]. The synthesis of oxetanes can have a great importance in the development of compounds with relevant biological properties: oxetane ring is present in the skeleton of taxol (1), an important drug used in the treatment of ovarian cancer [8]; in merrilactone A (2), a new sesquiterpene dilactone with neurotrophic activity [9]; and in several antiviral oxetanes, such as 3, 4, and 5, which have been described in literature (Scheme 3.2) [10–12].
O Ph
AcO
O OH
NH O
Ph
HO
O
O
1
2 O
NH2 N
O 3
NH
N N
RO
ON
NH2
O O
O 4
Scheme 3.2
O
O
O H OHOBz OAc
OH
HOHO N
O
O
5
CO2H
` -BU ¨ CHI REACTION ON FURAN DERIVATIVES THE PATERNO
83
The Paterno`-Bu¨ chi reaction is a photocycloaddition reaction of a n,p* carbonyl compound to an alkene in the ground state from either the S1 or the T1 state. The reaction can occur through the initial C O bond formation or through a previous formation of the C C bond. A frontier orbitals approach can be used to explain the formation of oxetanes. We can observe the HSOMO-LUMO interaction in which the half-occupied p* carbonyl orbital interacts with the unoccupied p* molecular orbital of an electron-deficient alkene, and a C,O-biradical is formed. The LSOMO-HOMO interaction in which the half-occupied n orbital of the carbonyl O atom interacts with the p orbital of an electron-rich alkene, and a C,C-biradical is formed [13, 14]. More recently a theoretical study of the Paterno`-Bu¨ chi reaction was written. This study showed that there are two conical intersection points located near the C C and C–O bonded biradical regions of the ground state. These two conical intersections support a mechanism by which the decay from the excited state is accompanied by a geometric rotation of the terminal group, in the case of C O attack, and by an orbital rotation at the oxygen center, in the case of C C attack. Furthermore, for C O attack, the triplet surface must cross the singlet to reach a diradicaloid minimum. For C C attack, the triplet biradical minimum is located at the same geometry as the conical intersection between the two singlet states, and the efficiency of the intersystem crossing will be determined by the nature of the spin-orbit coupling. Thus, for the triplet, the reaction path can be predicted by the most stable biradical rule [15]. A conformational analysis of the biradicals was also published [16]. The biradical intermediate in the reaction between benzophenone and electron-rich alkene has been determined by using laser flash photolysis. An absorption with lmax 535 nm has been observed [17, 18]. The formation of exciplex is used to explain the reaction behavior of simple alkene. However, evidence of monoelectron-transfer processes are reported for electron-rich alkenes [19].
` -BU ¨ CHI REACTION ON FURAN II. THE PATERNO DERIVATIVES A. Furan The first report on the photochemical reactivity of furan toward carbonyl compounds appeared in 1963. Schenck et al. [20] reported that the irradiation of benzophenone in furan gave the corresponding adduct in 94% yield (Scheme 3.3). The structure assignment was confirmed by Gagnaire and Payo-Subiza [21]. On the basis of the NMR analysis of both the products of the same reaction and the reaction between 2,5-dimethylfuran and benzophenone. Two years later, furan
84
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
Ph + Ph2CO
Ph
hv
O O
O
Scheme 3.3
and 2-methylfuran were found to react with propanal and benzaldehyde. NMR spectral data supported the formation of the same type of products (Scheme 3.4) [22]. H + C2H5CHO
hv
O H O
88%
O
H
H + C2H5CHO
O
Ph
hv
O
H3C
80%
O
O
+ PhCHO
C2H5
hv
O H3C
O
C2H5
C2H5 O
+ O
CH3
Scheme 3.4
The structure of the products was confirmed also by NOE analysis [23]. The yields were very high. In this case, the authors did not specify the exo or endo stereochemistry at C-6 on the dioxabicyclo[3.2.0]heptene skeleton. This problem was solved some years later assigning the exo configuration to this carbon [24]. When 2-methylfuran was used, a mixture of regioisomeric products was obtained, but the authors did not report the regioisomeric ratio. In contrast with this result, other authors reported a complete regioselectivity in the reaction between substituted furans and benzophenone: in these reactions, the coupling occurred on the most hindered side of the molecule (Scheme 3.5) [25, 26]. The Japanese group supposed that the reaction involved the formation of the biradical 6, due to the attack of the n,p* triplet oxygen on the site of higher free valence or higher electron density of the furan ring (Fig. 3.1). Afterward, the same group tested the reaction on a large group of aliphatic and aromatic aldehydes and ketones [27]. They found that ketones reacted, giving lower yields than the corresponding aldehydes; in particular acetone, butanone, and acetophenone
` -BU ¨ CHI REACTION ON FURAN DERIVATIVES THE PATERNO
Ph
85
Ph
hv O CH3 Ph2CO
O 98% O CH 3 Ph H3C Ph
CH3 hv O
O O
Ph2CO
98% Ph
O
hv CH2OH Ph CO 2
Ph
O 80% O CH OH 2
Scheme 3.5
Ph
H O O 6
Figure 3.1. Biradical intermediate in the Paterno`-Bu¨ chi reaction between furan and benzaldehyde.
gave yields in the range of 0.9–1.7%. Furthermore, the oxetane yields increased with the number of carbon atoms. With acetaldehyde, they obtained 15% yield, while cyclohexylcarbaldehyde gave 27% yield, and benzaldehyde 35%. The kinetics of the reaction is in agreement with a mechanism involving the formation of the biradical 6 following Scheme 3.6, where B is benzophenone, F is furan, (B-F)* is the biradical 6, and Ox is the oxetane [28]. Subsequently, more complex carbonyl compounds were used to give the corresponding
Scheme 3.6
86
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
adducts. n-Butyl glyoxylate gave the adduct with furan in 77.3% yield, while diethyl ketomalonate gave the corresponding product in 30% yield [29]. Good regioselectivity was observed using silyl and stannyl furan derivatives. The reaction in this case occurred on the less hindered side of the molecule (Scheme 3.7) [30]. On the contrary, 2-furylmethanol and the corresponding silyl ether gave low regioselectivity [31, 32]. Ph
H + PhCHO
O
hv Me3Si
SiMe3
O
O
O
2.5
O
+ PhCHO Si(i-Pr)3
:
O
>20
+ PhCHO O
SnBu3
O O
O
SnBu3
Si(i-Pr)3
1 Ph
H
O
+ O
H 2.5
+ BuO2CCHO
: Ph
hv Bu3Sn
O
+
H
H
Ph
H
O O
SiMe3
1
Ph
hv i-Pr3Si
O
+
H
H
Ph
H
SnBu3
:
1
H
CO2Bu
hv
H
O
O + Bu3Sn
O
O
H
> 20
:
CO2Bu
SnBu3
1
Scheme 3.7
The high exo stereoselectivity of the reaction has been extensively studied. The formation of the product occurs on a triplet 1,4-biradical, which must be converted into the singlet biradical to give the product. Three mechanisms operate for the interaction between singlet and triplet states of 1,n-biradicals: electron-nuclear hyperfine coupling (HFC), spin-lattice relaxation (SLR), and spin-orbit coupling (SOC). HFC is an important control factor for biradicals with long carbon chains between the radical centers. SOC plays a dominant role in biradicals with shorter distances between the radical centers, whereas SLR seems to contribute only marginally. SOC strongly depends on the geometry of the triplet biradicals; in particular, SOC decreases with increasing distance
` -BU ¨ CHI REACTION ON FURAN DERIVATIVES THE PATERNO
87
R O H
O 7
Scheme 3.8
between the two spin-bearing atoms. The best geometry for SOC requires that the axes of the p orbitals at the radical centers are perpendicular to each other. To explain the pronounced exo stereoselectivity, a secondary orbital effect can be postulated: an interaction between the rather flexible a-oxy radical center and the allyloxy ring localized radical in 7, likely plays a major role (Scheme 3.8) [33, 34]. When the reaction was carried out in benzene, new products were obtained. Ogata et al. [35] reported that two 2:1 adducts are the product of the reaction (Scheme 3.9). When 2-methylfuran was irradiated in the presence of benzophenone
Ph H H Ph
Ph + Ph2CO O
hv benzene
H
O
O
Ph H
H
27%
+ Ph2CO H3C
O
Ph O O
+
O
O
CH3
H
19%
O
HH
Ph + Ph
14%
+ Ph2CO O
CH3
Ph O
Ph H
O
CH3
3%
Ph H H Ph
Ph
H3C
Ph
O
CH3
hv benzene
Ph O
O H3C
O
CH3
10%
Scheme 3.9
O
18%
hv benzene
Ph H H Ph
Ph Ph
Ph
O
+ Ph
O
O
H H Ph
Ph
H
88
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
Ph H H Ph
Ph O
Ph O
H
O
R
8
Figure 3.2. Double cyloadduct in the reaction between furan and benzophenone.
the 1:1 adduct was obtained with two 2:1 adducts (Scheme 3.9). On the contrary, 2,5-dimethylfuran gave only one 2:1 adduct (Scheme 3.9); benzaldehyde showed minor selectivity, and this behavior was attributed to the higher triplet energy of benzaldehyde (71 kcal mol1) in comparison with that of benzophenone (68.7 kcal mol1). The more energetic excited benzaldehyde would not entirely discriminate between the two double bonds. The formation of the major 2:1 adduct in the reaction of Ogata et al. was immediately questioned by Leitich [36]. He considered that structure 8 was more acceptable (Fig. 3.2). Evanega and Whipple [37] made the reaction obtaining the 1:1 adduct, the adduct 8 in 29% yield, and the other 2:1 adduct reported in Scheme 3.9 in 18% yield. Finally, the structure of the adduct 8 was confirmed through X-ray analysis [38]. Furan reacts with chloral to give unexpectedly the corresponding 2-furyl carbinols [39]. Furthermore, 2-cyanofuran did not react, while 2-furfural diacetate and furfural ethylene acetal showed low reactivity. In particular, furfural diacetate reacted with benzophenone in benzene, giving the monoadduct with high regioselectivity (only the adduct on the most hindered side of the molecule was obtained) but in low yield (21%), and in the presence of significant amounts of diadducts (two diadducts were obtained in 21% and 18% yields, respectively). Furthermore, furfural ethylene acetal reacted with benzophenone in benzene showing high regioselectivity because the ketone attacks the furan derivative only on the side bearing the substituent, but in low yield (22%) and in the presence of diadducts in 28% and 15% yields [39]. The cycloaddition reaction can be performed on esters. In this case, the adducts can be obtained in a few cases. In most of the examples, they underwent a cycloreversion reaction to give the ring opening products (Scheme 3.10) [40]. More recently, this result has been questioned and a 95:5 mixture of stereoisomeric adducts was identified as the product [34]. Coupling products can be obtained carrying out the reaction between an amide and furan. Also in this case the cycloadduct cannot be isolated, but the subsequent decomposition products can be isolated (Scheme 3.11) [41]. Furan quenches the fluorescence of the substrate, while a small new emission at 500 nm appears. This evidence is in agreement with a mechanism involving a reaction in the excited singlet state via the formation of an exciplex.
` -BU ¨ CHI REACTION ON FURAN DERIVATIVES THE PATERNO
H
O OCH3
OCH3
Ph OCH3
hv
+
O
Ph
hv ∆
O
O
89
O
H
31%
CHO
13%
CHO O
O OCH3
hv
+ O
OCH3 36%
Scheme 3.10 O O
O
N
N O
hv
+
O
O OCHO H H OCHO N
+
O 41%
N O
49%
Scheme 3.11
The Paterno`-Bu¨ chi reaction on furan has been studied from a theoretical point of view [42]. The HOMO of furan has an energy of 0.32 eV, and the atomic coefficients of this orbital (Fig. 3.3) are in agreement with an attack of the carbonyl oxygen on the a carbon. The structures of the possible adducts deriving from the attack of benzaldehyde in the a position of furan are depicted in Figure 3.4. The exo isomer is more stable (0.5 kcal mol1) than the other isomer, in agreement with the experimental results. The reaction between 2-methylfuran and benzaldehyde gave only the photoadduct deriving from the attack on the less hindered side of the molecule (Scheme 3.12) [42]. The HOMO of 2-methylfuran showed the atomic
90
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
0.21 0.32 O 0.00
Figure 3.3. HOMO’s atomic coefficients in furan.
6 5
7
4
1
2 3 exo adduct endo adduct
Figure 3.4. Possible cycloadducts in the reaction between furan and benzaldehyde.
H Ph hv O
CH3
PhCHO
O H3C
O
Scheme 3.12
coefficients reported in Figure 3.5. These data are in agreement with an attack of the carbonyl oxygen on the a carbon. The structures of the possible adducts deriving from the attack of benzaldehyde in the a position of 2-methylfuran are depicted in the Figures 3.6. The formation of the exo isomers of both 1,2 and 4,5 adducts is favored. Furthermore, the 4,5-adduct is favored for 0.14 kcal mol1. This low energy difference does not allow us to explain the observed regioselectivity. To explain this behavior, the energy of the possible biradical intermediates was considered (Fig. 3.7). −0.19 0.25 0.28 −0.29 O CH3
Figure 3.5. HOMO’s atomic coefficients in 2-methylfuran.
` -BU ¨ CHI REACTION ON FURAN DERIVATIVES THE PATERNO
91
1,2-exo
1,2-endo
Figure 3.6. Possible cycloadducts in the reaction between 2-methylfuran and benzaldehyde. H .
. Ph
H .
O
O O
CH3
. Ph
H3C
O
Figure 3.7. Biradical intermediates in the reaction between 2-methylfuran and benzaldehyde.
The second biradical intermediate is favored for 1.18 kcal mol1, thus allowing us to explain the observed regioselectivity. The HOMO of 2-furylmethanol showed the atomic coefficients (Figure 3.8). These data are in agreement with an attack of the carbonyl oxygen on the a carbon. The structures of the possible adducts deriving from the attack of benzaldehyde in the a position of 2-furylmethanol are depicted in the Figure 3.9. The exo stereoisomer is favored for all the regioisomers. On the basis of these data, the attack on the most hindered side of the molecule should be favored for 2 kcal mol1. However, considering the results obtained studying the reactivity of 2-methylfuran, we considered the relative stability of the possible biradical intermediates (Fig. 3.10). The calculations of these biradicals showed that the first is more stable than the other for 0.8 kcal mol1. This difference could account for the preferential formation of the product deriving from the attack on the most hindered side of the molecule. Furthermore, it can account for the formation of two regioisomers. In conclusion, the regiochemical behavior (the attack of the oxygen on an a-carbon or on a b-carbon of the furan ring) of the Paterno`-Bu¨ chi reaction on furan, 2-methylfuran, and 2-furylmethanol can be explained considering the −0.19 0.23 0.28 −0.29 O CH OH 2
Figure 3.8. HOMO’s atomic coefficients in 2-furylmethanol.
92
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
1,2-endo
1,2-exo
Figure 3.9. Possible cycloadducts in the reaction between 2-furylmethanol and benzaldehyde. Ph
H
H
O O CH2OH HOH2C
Ph O
O
Figure 3.10. Biradical intermediates in the reaction between 2-furylmethanol and benzaldehyde.
atomic coefficients of the HOMOs. Furthermore, the stereochemical behavior can be explained, on the basis of the Scharf’s hypothesis, considering the formation of the most stable stereoisomer. The origin of this behavior can be found in the spin control formulated by Griesbeck. Finally, the regiochemical behavior on monosubstituted furan can be explained on the basis of the relative stability of the possible biradical intermediates, where the most stable one is favored.
B. Benzofuran Schenck and co-workers [43] reported also the reaction of some aromatic carbonyl compounds (benzophenone, benzaldehyde) with benzofuran (Scheme 3.13). They showed that when high triplet energy compounds were used, dimers of Ph +
Ph2CO
O
hv O
O
75%
Scheme 3.13
Ph
` -BU ¨ CHI REACTION ON FURAN DERIVATIVES THE PATERNO
93
benzofuran were obtained, while oxetanes were the products of the reaction when low triplet energy carbonyls were used. As in the case of furan, they obtained only one product, showing that the reaction has a high regioselectivity. Reinvestigation of the reaction of benzofuran with acetophenone or propiophenone showed that, also in this case, oxetanes were obtained [44].
C.
2,3-Dihydrofuran
The irradiation of 2,3-dihydrofuran with benzophenone in benzene gave adduct 9 (Scheme 3.14). It should be noted that in this case a completely different regioselectivity was obtained in the coupling reaction [45]. The same behavior was observed irradiating 2,3-dihydrofuran with aliphatic aldehydes and ketones. By using acetone, the adducts were obtained in 52% yield in a 200:1 isomeric ratio (Scheme 3.15) [46]. When acetaldehyde was used, the adduct was obtained in 63% yield as a mixture of stereoisomers. When benzaldehyde was used as carbonyl compounds, the adducts were obtained with an overall yield of 98% as a > 98:2 regioisomeric mixture. The major isomer is 88:12 endo/exo mixture. The reaction showed a good regioselectivity and stereoselectivity (Scheme 3.16) [46–48]. This stereoselectivity is partially lost when benzaldehyde reacts with 2methyl-2,3-dihydrofuran [47]. Benzaldehyde reacts in its triplet state. This way, a triplet biradical is formed as an intermediate. For the formation of the products, intersystem crossing into O + Ph2CO O
Ph
hv benzene
O
Ph 9
Scheme 3.14 O hv
+ O
O
O
+ O
O 200
:
1
H H + O
O
H O
hv
O H
+ O
H H 4 :
Scheme 3.15
O 3
H
94
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
H
H O
PhCHO hv 98%
O
O Ph
O
+
88
Ph
O
H :
H
12
H
H
O
O O
PhCHO CH3 hv 68%
O
Ph + 65
Ph
O
CH3 :
CH3
35
Scheme 3.16
the singlet manifold is necessary. The most important factor influencing an intersystem crossing for flexibile triplet biradicals is spin-orbit coupling. The angle between p orbitals at the radical centers is approximately 90 for maximun spin orbit coupling. For the pronounced endo selectivity in the reaction between aromatic aldehydes and 2,3-dihydrofuran, we can consider the two biradical conformers 10 and 11 to be responsible, with the alkyloxy substitutent localized in a pseudoequatorial position and 10 being more populated because of fewer steric interactions (Scheme 3.17). When a methyl group is present, the increasing gauche endo
exo
Ph
O
O
H H
H O 11
O 10 Ph
H
H
O H
CH3
O Ph
O 12
O 13
exo
endo
Scheme 3.17
CH3
Ph
` -BU ¨ CHI REACTION ON FURAN DERIVATIVES THE PATERNO
−0.36 −0.29
95
0.10 0.09 O 0.32
Figure 3.11. HOMO’s atomic coefficients in 2,3-dihydrofuran.
4 3
2
5
1
6
endo -adduct
exo-adduct
Figure 3.12. Possible cycloadducts in the reaction between 2,3-dihydrofuran and benzaldehyde.
interactions with the b-alkyloxy substituent lead to a certain concentration of 12 and 13, again with 12 being preferred because of fewer steric interactions. The a- and b-Naphthaldehydes, on the contrary, gave high exo selectivity. To understand this behavior we have to consider that they gave cycloadducts also in the presence of triplet quenchers, while the authors observed fluorescence quenching in the presence of dihydrofuran; in this case, singlet excited states are responsible for the high exo selectivity [33, 49]. The same theoretical approach was used to explain the photochemical behavior of 2,3-dihydrofuran. The HOMO of 2,3-dihydrofuran showed at 0.32 eV is reported in Figure 3.11. The atomic coefficients of the HOMO are in agreement with an attack of the LUMO (the carbonyl oxygen) on the b-carbon. On the basis of this result, the thermodynamic stability of the possible adducts was tested. The structures of these compounds are reported in Figure 3.12. The endo isomer is more stable for 2 kcal mol1 than the other isomer, in agreement with experimental results. When 2,3-dihydrofuran derivatives react with a,b-unsaturated carbonyl compounds, 2þ2 cycloaddition between the olefins occurs (Scheme 3.18) [50]. O
H
O H H H
O hv
+
O
O O
O
H
H
H
Scheme 3.18
O
H
96
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
D. Synthetic Applications Oxetanes can be converted into the corresponding 3-furylcarbinol through reaction with TsOH (Scheme 3.19) [29]. The formation of 14 is in agreement with H H
H CO2C4H9 O
TsOH O
O
OH CO2C4H9 O+ H
CO2C4H9 O
73%
14
Scheme 3.19
the high negative entropy of the activation found (Sz ¼ 141.6 J K1) [51]. If the irradiation was performed in the presence of an acid, 3-furylcarbinol can be obtained in a one-step procedure in better yields [52]. This procedure has been used in the synthesis of perillaketone, a naturally occurring 3-substituted furan [52]. Irradiation of furan with 4-methyl-pentanal in the presence of MsOH gave a mixture of 15 (18%) and 16 (5%). Oxidation of 15 with Collins reagent gave perillaketone (Scheme 3.20). It is noteworthy that this is the first report on a probable Norrish type II on the carbonyl compound. In fact, this type of reaction can be responsible for the formation of 16. The excited aldehyde can undergo 1,5-hydrogen shift to give acetaldehyde, and this aldehyde, through the cycloaddition reaction, gives the 3-furylcarbinol 16. OH + O
CHO
hv MsOH
OH
+ O
15 (18%)
O
16 (5%)
Collins O
O
66%
Scheme 3.20
On the contrary, Lewis acids catalyzed a different behavior. Treatment with BF3 Et2O gave only 3-substituted furan in THF and 89% of 2-substituted furan in acetonitrile (Scheme 3.21) [51]. This behavior is in agreement with the formation of a carbocation 17, which can be stabilized by splitting off a proton to
` -BU ¨ CHI REACTION ON FURAN DERIVATIVES THE PATERNO
97
OH H CO2C4H9
CO2C4H9 +
BF3.Et2O
O
O
O
CO2C4H9
O
OH H
H
CO2C4H9
CO2C4H9 + OA
+ O
O
+
OA
CO2C4H9
O
17
OA
Scheme 3.21
form 3-substituted furan or can undergo a 1,2-rearrangement to a more stable carbocation, which splits off a proton to form 2-substitued furan (Scheme 3.21). The oxetane derivatives can also be treated with KMnO4, and the resulting cis diol was treated with acetone in the presence of an acid. This catalysis induces a ring-opening reaction, giving two products (Scheme 3.22) [53]. The major product was obtained through a spontaneous epimerisation of intermediate 18 in these reaction conditions. This product was 3-deoxy-1,2-O-isopropylidene-b-DL-streptose. The same epimerisation reaction occurs on the trans diol obtained through treatment of the oxetane with mCPBA (Scheme 3.23) [53]. CH3
H
1) KMnO4
O O
2) Acetone, H +
H
CH3
HO H
CH3
HO H OHC
O HO
O
HO H
H CH3
HO H
HO
O
OHC
OH CHO
CH3
HO H
OH OH
O
O O O O
O 15%
OHC
OH OH
H3C OHC
OHC O
O H3C
O OH OH
O
H3C
O 33%
18
Scheme 3.22
98
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
CO2Bu
H O + H
Bu
O
O
hv
O O
O CH2OAc
H
H
LiAlH4 Ac2O
CH2OAc
HO H
MCPBA O O
O m-ClC 6H4CO2
H
O
1) MeO−
O
H
OMe
2) IR120 O H
O
OH
OMe
Scheme 3.23
The possible chemical modifications on the cycloadduct deriving from the reaction between an aldehyde and a furan was extensively studied by Schreiber et al. [54], and some results are reported in Scheme 3.24. Hydrolytic ring opening, reductions, hetero Diels-Alder, the reaction with mCPBA, and hydroboration-oxidation were the object of this work. The reaction with mCPBA was used by Schreiber and co-workers in the synthesis of asteltoxin 19, a potent inhibitor of E. coli BF1-ATPase activity used as a fluorescent probe for mitocondrial F1- and bacterial BF1-ATPase (Fig. 3.13) [55]. The synthetic sequence implies the photochemical coupling of 3,4-dimethylfuran with a functionalized aldehyde to give the corresponding adduct in 63% yield. The subsequent reaction with MCPBA gave the protected trans diol that was converted into 20 through acid hydrolysis. Then the aldehyde was converted into the corresponding hydrazone, and this substrate was treated with EtMgBr to give 21 with complete stereochemical control obtained through chelation. The subsequent conversion of the benzyl ether into a seleno derivative and the elimination of the selenoxide gave 22. Compound 22 could be treated with ozone to give the corresponding aldehyde, which was then converted into asteltoxin (Scheme 3.25) [56]. In fact, the aldehydic product could be alkylated with a vinyl anion, and the subsequent double [2,3] sigmatropic rearrangement at room temperature furnished 23 as a 3:1 mixture of epimers. Camphorsulfonic acid induced the transposition of the substrate into a bis(tetrahydrofuran) derivative. Pummerer
` -BU ¨ CHI REACTION ON FURAN DERIVATIVES THE PATERNO
Scheme 3.24
OH
OH
O
O
O
H
O 19 OCH3
Figure 3.13. Asteltoxin.
99
100
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
OBz
OBz HO
O MCPBA
hv
+ BzO
CHO
O
O
HO
OBz
O 20
BzO
HO
OBz
O
2) Acetone, CSA
o-NO2C6H4Se
O O
H
CHNNMe2 1) EtMgBr
HO
HO
O
HO Me2NNH2
CHO
HCl HO
O
m-ClC 6H4CO2
H
O
1) Li, NH3 SeCN
O H
O NO2
2)
O H
HO
O
H2O2 O
O H 22
21 OHC HO
O
1) O3 2) Me2S
O
O H
Scheme 3.25
rearrangement and hydrolysis gave aldehyde 24. Aldol condensation gave 25, and elimination gave asteltoxin (Scheme 3.26) [57]. The same research group reported a formal synthesis of avenaciolide, an antifungal metabolite (Scheme 3.27) [58]. In this case, the oxetane (obtained in multigram quantities in high yields and with complete stereochemical control) was treated with hydrogen to give the saturated compound. The key step in this synthetic procedure is a reaction with ozone followed by a basecatalyzed epimerization with potassium carbonate and cyclization in acidic medium. The reaction of tributylstannylfuran with butyl glyoxylate [30] was used in the synthesis of a ginkgolide B-kadsurenone hybrid of two inhibitors of a platelet-activating factor (Scheme 3.28). In the obtained oxetane, the stannyl substituent was used to give a coupling Stille reaction; the subsequent hydrolysis gave the corresponding carbonyl derivative, which was reduced with NaBH4 in the presence of CeCl3: this reaction allowed for the selective reduction of the aryl
` -BU ¨ CHI REACTION ON FURAN DERIVATIVES THE PATERNO
HO
OHC HO
SOPh
HO
O
O
BuLi O
O
PhSO
H
O
101
CSA
O H 23
OH
OH
O
O
1) CF3CO2COCH3, lutidine, Ac 2O
H
2) HgCl2, CaCO3 SOPh
O O OH
OH
O
O
OMe LDA
H
CHO 24
OH
OH
O
O
1) TsCl OH
H
O
19 2) DMAP, Et 3N
O 25 OMe
Scheme 3.26
ketone in the presence of the aldehyde. The use of distannoxane transesterification catalyst gave a 1:1 mixture of the hybrid. Further elaborations of the stannyl substituents in the adduct can be obtained via reaction with RCOCl in the presence of Pd(0) to give the corresponding ketone, or via reaction with BuLi and reaction of the corresponding anion with PhCHO to give the corresponding alcohol [59]. Furthermore, the cycloadduct obtained from the reaction between furan and an aldehyde can be treated with an excess of Schlosser’s base (BuLi, t-BuOK) to give the corresponding
102
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
C8H17
H +
C8H17
H
O
hv O
O 1) HCl 96% H 2) H2C=CHMgBr 80%
H2
O
O
C8H17
H
Rh, Al 2O3
H
O
97% H
HO
C8H17 OH
H H
HO
H OH
1) Acetone, TsOH, 85%
K2CO3
H
HO
O
O O H
OMe
HCl
O
C8H17
H MCPBA
MeO
H MeOH
H OH H
BF3.Et2O
O
C8H17 O
O
O3 Me2S
O H
C8H17
C8H17
OHC
2) PCC, 91%
OH
O O
H
O O H
OHC
OH
OHC
H
OH
H
C8H17 O
80'%
C8H17
O
O O H
O
C8H17
O
31%
Scheme 3.27
anion, which can react with carbonyl compounds or alkyl halides (Scheme 3.29) [59]. A possible further application of the Paterno`-Bu¨ chi reaction between carbonyl compounds and furan could be related to the use of the oxetane ring. The most important target in this field was oxetanocin (26), a nucleoside isolated from Bacillus megaterium NK 84-0218 showing anti-HIV activity (Fig. 3.14) [60, 61]. The treatment of the oxetane 27 with N-iodosuccinimide in the presence of methallyl alcohol gave the corresponding iodoacetals 28 (Scheme 3.30) [62]. The subsequent treatment of 28 with iodonium di-symcollidine perchlorate (IDCP) [63] gave 29 in low yields, as a relatively unstable mixture of diastereoisomers. Alternatively, treatment of 30 with dimethyl dioxirane gave the corresponding epoxide, and treatment with methallyl alcohol gave the hydroxy acetal 31. The treatment of 31 with IDCP gave 32 in moderate yields. Oxetanocin was obtained by carrying out the reaction between 2-methylfuran and benzoyloxyacetaldehyde. The corresponding adduct was treated with ozone, and the product was reduced with NaBH4. The obtained alcohols were protected.
` -BU ¨ CHI REACTION ON FURAN DERIVATIVES THE PATERNO
103 Br
CO2Bu
H
O
SnBu3
hv
+ BuO2CCHO
35%
O O
O Bu3Sn
Pd(PPh3)4, THF 55% OH H CO2Bu
H
H
1) HCl
MeO OMe
OMe
HO SCNBu2SnOSnBu2OH Toluene, 50°
OH
O
2) NaBH4, CeCl3 90%
MeO
O
CO2Bu
H
OMe OMe
O
MeO
O O
MeO
H
+
H
68% HO
O O
H O H
MeO
OMe
Scheme 3.28
Product 33 was treated with N-benzoyl-disilyladenine and SnCl4 to give 26 (Scheme 3.31) [64].
E. Asymmetric Reactions The reaction of glyoxylates with furan can also be performed using chiral glyoxylate. In particular, the use of R-(-)-menthol, chiral 2-octanol, and chiral 2,2dimethyl-3-butanol as chiral auxiliaries gave the corresponding oxetanes in high yields. These compounds can be converted into the corresponding
104
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
O O
BuLi, t-BuOK, THF
O
t-Bu
t-BuCHO
O
H
OH
R
BuLi, t-BuOK, THF O O
H
63−72%
H
R
H
R
H
R
H
O
MeI
O
H
H 41% R
H
R
H
BuLi, t-BuOK, THF O O
O Me3Sn
Me3SnCl
H
O
H
50−65%
Scheme 3.29
3-substituted furans. These furans showed low enantiomeric excess (Scheme 3.32) [65]. On the basis of the experimental data, the configuration of the oxetane was 1R, 5S, 6S. The low optical purity was explained considering that two conformers of the glyoxylate can approach furan (Fig. 3.15). The use of chiral phenylglyoxylate gave better results. The use of chiral alcohols gave diastereoisomeric excess in the range of 4–80% (Scheme 3.33) [66– 68]. The authors observed also an important variability of the diastereoisomeric excess in function of the temperature, with the presence of an inversion temperature. These results were explained assuming that the diastereoselection is produced on two levels: (1) the preferred formation of the diastereoisomeric 1,4biradical intermediate 34 and (2) the preferred retrocleavage of the energetically NH2 N
N
N
N HOH2C
O CH2OH 26
Figure 3.14. Oxetanocin.
` -BU ¨ CHI REACTION ON FURAN DERIVATIVES THE PATERNO
H
H
O
O
O
O
NIS
O
I
OCH3 O
O
IDCP
105
O H
Ph
27 H
Ph
28 H
H
O
O
O
O
O
O PhCO2H2C
H
PhCO2H2C
H
I
29
O O
O
O PhCO2H2C
I
H
Ph
I
30
H
OH
31
OAc O
O
1) Ac 2O
O
2) IDCP PhCO H C 2 2
OAc 32
Scheme 3.30
unfavored diastereoisomeric intermediate 35 to the starting materials (Fig. 3.16) [69]. The diastereoisomeric excess in the high-temperature region (T > Tinv) is dominantly controlled by steric effects of the chiral auxiliaries, whereas in the low-temperature region (T < Tinv), the nature of the olefin has a dominating influence. When the reaction is carried out on 2-methylfuran, a 2:1 regioisomeric H O
CH3
PhCO2CH2CHO N(TMS)2
45−50%
N
N OAc
N
N OCOPh
O CH2OAc H
PhCO2H2C
CH3
H
PhCO2H2C
H
O
O
hv
SnCl4
33
Scheme 3.31
26
1) O3 2) Me2S 3) NaBH4 4) Ac 2O, Py
106
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
O H
O
O H
+
O
O
hv O
O
O
O
H+
O
OH O
Scheme 3.32 L O
M
S M
O
S H
H
O
O
L O
O S-trans
S-cis
Figure 3.15. Possible conformers of glyoxylate esters. O Ph
O
R*
+
O
hv
R*O2C
O
O
+
O
O Ph
Ph
O
O R*O2C
O O
O
Ph
H H
CO2R*
H H major
Scheme 3.33
CO2R*
Ph
` -BU ¨ CHI REACTION ON FURAN DERIVATIVES THE PATERNO
O
O R*O2C
O
O Ph
Ph 34
107
CO2R*
35
Figure 3.16. Biradical intermediates in the reaction between furan and phenylglyoxylates. O Ph + CH3
O
hv
O O Ph
O
O Ph
O
Ph
Ph
O
O
Ph
O
+ CH3
O
O 2 de 96%
H3C :
1 de 96%
Scheme 3.34
mixture was obtained with a very high diastereoisomeric excess (Scheme 3.34) [70]. The reactivity of some chiral phenylglyoxylates was studied [71]. Three phenylglyoxylate esters were used (Fig. 3.17). All the reactions gave, in good yields, the corresponding oxetanes (Scheme 3.35). Considering the stereochemical behavior of the reaction, while the reaction with 8-phenylmenthol glyoxylate gave a high diastereoisomeric excess, the reaction with the glyoxylate ester of (S)-1-methyl-1-propanol gave only 15% de and the reaction with the glyoxylate ester of (S)-2-methyl-1-butanol resulted in no diastereoselectivity. In the reaction of the first glyoxylate ester with furan, we could obtain the observed diastereoisomer (10 R,1R,5S,6S) (Fig. 3.18A) and the other one (10 R,1S,5R,6R). These two stereoisomers showed a different energy for 0.07 kcal mol1, with the most stable one being the (10 R,1R,5S,6S) isomer. In the reaction of the
108
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
O
O
O
O O
O
O O O
Figure 3.17. Phenylglyoxylate [71]. O O O
Ph
O hv
O
O
O
H
O
O O O
hv
Ph
O O
O
O
H
O
O
O hv
Ph
O
O
O
O O
Ph
H
O
Scheme 3.35
second ester, we could obtain the observed diastereoisomers (10 S,1S,5R,6R) (Fig. 3.18B) and (10 S,1R,5S,6S) (Fig. 3.18C). These two stereoisomers showed a different energy for 0.14 kcal mol1, with the most stable one being the (10 S,1S,5R,6R) isomer. In the reaction of the third ester with furan, we could obtain the observed diastereoisomer (10 R,20 S,50 R,1R,5S,6S) (Fig. 3.18D) and
` -BU ¨ CHI REACTION ON FURAN DERIVATIVES THE PATERNO
A
109
B
C D
Figure 3.18. Possible cycloadducts in the reaction between furan (or 2-methylfuran) and chiral phenyl glyoxylates.
the other one (10 R,20 S,50 R1S,5R,6R). We calculated that these two stereoisomers showed a different energy for 3.05 kcal mol1, with the most stable one being the (10 R,20 S,50 R,1R,5S,6S) isomer. This high difference in the energy of the possible diastereoisomers could explain the observed diastereoisomeric excess; furthermore, the calculated differences in the energy of the possible diastereoisomers that can be obtained in the reaction of the other esters can explain the low diastereoisomeric excess observed when the first was used as substrate. However, the lack of diastereoselectivity when the second one was used as substrate
110
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
O
O Ph
O
Ph
O H
O O
O
O
Ph
Ph
H O
O O
O
H
O O
O O
H
O
O Ph
O
Ph
Ph H
O O
O O
O
H
Ph
Figure 3.19. Biradical intermediates in the reaction between chiral phenylglyoxylates and furan.
in the same reaction cannot be explained by these results, considering the differences in the energy of the possible diastereoismers are higher than those obtained using the first ester as substrate. The best results can be obtained considering the energy of the triplet biradical intermediates (Fig. 3.19). Calculations on these biradical intermediates showed that the first (the precursor of the observed product) was more stable than the other by 0.73 kcal mol1. Furthermore, the first and the second possible biradical intermediate in the reaction of the ester of (S)-2-methyl-1-butanol and differed by only 0.02 kcal mol1, in agreement with the observed no stereoselectivity of the reaction. Finally, the first biradical intermediate in the reaction of 8-phenylmenthol ester proved to be more stable than the other one by 21.9 kcal mol1. This result is also in agreement with the observed high diastereoisomeric excess. Isatine derivatives gave the corresponding cycloadducts with high stereoselectivity when irradiated in the presence of furan and benzofuran [72]. The reaction of furan with acyl cyanides yields the corresponding oxetanes, but both diastereoisomeric endo- and exo-oxetanes are formed (Scheme 3.36). When chiral acyl cyanides are used, low asymmetric induction is observed [73]. Furan also reacts with chiral ketones. In this case, an a-cleavage reaction before the 2þ2 cycloaddition modifies the expected products (Scheme 3.37). When ()menthone was used as a substrate, a chiral product was obtained as a 2:1 diastereoisomeric mixture; the most abundant product has the (1R, 3R) configuration [74]. When the reaction was performed on carbohydrate 36, a complex reaction mixture was obtained (Scheme 3.38) [74, 75].
` -BU ¨ CHI REACTION ON FURAN DERIVATIVES THE PATERNO
OAc
OAc NC
Ph
OAc CN
+ Ph
O
O
+
−55° 71%
O
CN
Ph
O
hv
O
O 6.2 : de 5%
111
1 de 13%
Scheme 3.36
O CHO hv AcO
AcO
O
O
O
HO
O H+ AcO
AcO
O
hv
H+ CHO
O
O O
OH
O
Scheme 3.37
Attempts to obtain stereoselective Paterno`-Bu¨ chi reactions were performed carrying out the reaction between 3,4-dimethylfuran and R-isopropylidene diastereoisomer. The coupling products were obtained with an overall yield of 35% as a 1.2:1 mixture of diastereoisomers. Furthermore, the compound 37 was obtained with 54% ee (Scheme 3.39) [76, 77]. This behavior suggests the operation of a mechanism that is insensitive to the substitution pattern of chiral aldehydes. Reaction between an excited aldehyde (singlet or triplet state) and furan
112
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
O
O
O O
O hv
O
O
O O
O
O
O
O
O
O
O
O
O
O
O 9.1%
O 36
16%
O
O
O
O O
O
O O O
O
O O
O O
55.3%
19.6%
Scheme 3.38
O
H O + OHC O
O
hv
O O
H
O
H
O
O O
+ O
H
37
Scheme 3.39
proceeds with initial carbon-oxygen bond formation to produce either of the two biradical species. The stereocenter adjacent to the carbonyl is now in a 1,4-relationship to the newly formed stereocenter at the acetal carbon and is expected to exert little influence as a stereocontrol device [76]. The extensive racemization observed probably reflects the photolability of the aldehydes toward racemization under the conditions of the reaction [77]. Nevertheless, compound 37 can be used in a chiral synthesis of the bicyclic part of asteltoxin, confirming the assigned absolute configuration [77]. In 1990, Griesbeck and Stadtmu¨ ller [46] found that the reaction of benzaldehyde with homoallylic alcohols did not show diastereoselectivity. Ten years later Adam and co-workers showed that allylic alcohols reacted with benzophenone to give the corresponding adducts with high regioselectivity and diastereoselectivity (Scheme 3.40) [78–80].
` -BU ¨ CHI REACTION ON FURAN DERIVATIVES THE PATERNO
OH
OH
OH
O
hv
O
+
Ph Ph
Ph2CO
90
Ph Ph
:
10
OH
OH Bu-t
hv Ph2CO
113
O
OH
Bu-t
O +
Ph Ph > 95
Bu-t
Ph Ph
:
5
Scheme 3.40
The diastereoselectivity dropped drastically in presence of protic methanol and totally disappeared for the corresponding silyl ethers. These data are in agreement with the presence of a hydroxy directing effect in the Paterno`-Bu¨ chi reaction. Threo stereoisomer can be favored through the formation of an hydrogen bond between triplet excited benzophenone and the substrate in the exciplex, while the formation of the erythro stereoisomer would be less favored due to allylic strain (Scheme 3.41). The formation of hydrogen bond to direct the Paterno`-Bu¨ chi reaction has been considered by other researchers. Diastereoselective cycloaddition has Ph2C O H HO R
H H OH H
Ph2C O H HO H
Ph Ph
O
R
H
H OH R
threo
R
HH R OH HH R
R O Ph2C O H
R
HH O Ph2C O H
erythro
Scheme 3.41
Ph Ph
O
H OH
114
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
been obtained using chiral enamide [81, 82], or in the reaction of allylic alcohols with naphthalene rings [83]. When unsymmetrical carbonyl partners, such as acetophenone or benzaldehyde, were used, high diastereoselectivity was observed to give the corresponding cis isomer. The regioselectivity was high with acetophenone but lower with benzaldehyde [79]. Cis diastereoselectivity can be explained by using the Griesbeck rule on the possible triplet biradicals formed in the reaction. Steric interactions are minimized when the biradical assumes the optimal conformation, and this conformation accounts for the formation of the observed stereoisomer [84]. When chiral allylic alcohols were used as substrates in the reaction, cis diastereoisomers were formed. Furthermore, also in this case, a pronounced threo diastereoselectivity was observed, in agreement with a less pronounced hydroxy-directing effect when acetophenone and benzaldehyde were used [79, 84]. Chiral allyl ether gave the corresponding adduct with high diastereoselectivity [85]. The reaction of allylic alcohols with carbonyl compounds was tested on a particular type of allylic alcohol, such as 2-furylmethanol derivatives. While the reaction of 2-furylmethanol with benzophenone showed low regioselectivity, the presence of larger substituents on the carbon bearing the alcoholic function allows a high regioselectivity (Scheme 3.42) [86]. Furthermore, when 2-furylethanol (38) was used as substrate, a 1:1 mixture of stereoisomers was obtained; Ph
O
CH2OH
+ Ph2CO
hv
O O
+
:
Ph Ph O
+ Ph2CO
hv
O O
OH 38
OH
CH3 1:1 diasteroisomeric mixture Ph Ph Ph + Ph2CO
O OH 39
O O
CH2OH 3
CH3
Ph
Ph
Ph
hv
O O
OH Ph
one diastereoisomer
Scheme 3.42
1
CH2OH
` -BU ¨ CHI REACTION ON FURAN DERIVATIVES THE PATERNO
115
whereas when 1-(2-furyl)-benzylic alcohol (39) was the substrate, only one diastereoisomer was obtained (Scheme 3.42) [86]. NOE experiments are in agreement with a (1RS, 10 RS, 5RS) configuration. The high diastereoselectivity observed was confirmed using optically active 39. The regioselectivity of the reaction was explained assuming that the reaction involves the formation of an exciplex with the following coupling to a zwitterionic intermediate. The polar nature of the intermediate is in agreement with the observed r value in a Hammett free energy correlation [86]. When 5-methyl-2-furyl derivatives were used as substrates, a different regioselectivity was observed. Compound 40 gave a 1:1 mixture of regioisomers, when irradiated in the presence of benzophenone, and a single regioisomer in the presence of benzaldehyde (Scheme 3.43) [87]. In agreement with the results obtained with 2-furyl derivatives, the products deriving from the attack on the side bearing the alcoholic function were obtained as a single diastereoisomer, while those deriving from the attack on the side bearing the methyl group were obtained as a mixture of diastereoisomers. Also in this case, the regioselectivity of the reaction can be explained on the basis of the different stability of the possible zwitterionic intermediates [87]. Ph Ph CH3
O
+ Ph2CO
hv
O H3C
O
OH
Ph O OH
+ PhCHO
hv
Ph
O H3C
O OH
Ph
Ph H3C
+ OH
Ph
40
Ph
Ph
Ph
Ph
O H3C
O OH
40
Scheme 3.43
The reaction of 2-furylmethanol derivatives with aliphatic aldehydes and ketones gave the corresponding adducts with high regioselectivity but no diastereoselectivity (Scheme 3.44) [88]. The observed diastereoselectivity in the reaction with aromatic carbonyl compounds clearly shows that it increases in relation to the nature of the substituents on the carbon bearing the alcoholic function, as described by Adam. However, while Adam considers the allylic strain with a methyl group as the driving force for the diastereoselectivity; in this case, the methyl group is not present. Therefore, allylic strain cannot be used to explain diastereoselectivity.
116
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
H CH3 Ph + CH3CHO
O
hv
O O
OH
OH Ph
Scheme 3.44
We supposed, on the basis of theoretical calculations, that 2-furylmethanols showed a preferential conformation able to induce, through the formation of a hydrogen bond between the alcoholic function and the carbonyl compound, the high diastereoselectivity observed (Scheme 3.45) [88]. The lack of diastereoselectivity in the case of aliphatic carbonyl compounds represents a problem. The results cannot be explained on the basis of 1,3-allylic strain as reported by Adam. In fact, the same ‘‘allylic’’ substrates were used giving high regioselectivity with aromatic carbonyl compounds. Furthermore, aliphatic and aromatic ketones could give the same hydrogen bond interaction with the hydroxy group in the furan derivatives. R' R'
O
pro-S
H O H
O R R
Scheme 3.45
` -BU ¨ CHI REACTION ON III. THE PATERNO PENTAATOMIC HETEROCYCLES DIFFERENT FROM FURAN A. Thiophene The reactivity of pentatomic aromatic heterocycles that are not furan toward carbonyl compounds to give the corresponding oxetanes was the object of other review articles [89]. These compounds show lower reactivity than furan. The reason for this behavior is not clear. It could be related to the different aromaticity of these compounds in comparison with that of furan or, as reported below, to the quenching properties of the heterocycles.
` -BU ¨ CHI REACTION ON PENTAATOMIC HETEROCYCLES THE PATERNO
Ph
Ph
Ph
Ph S
hv Ph2CO BF3
117
OH
O + S 10%
S 90%
Scheme 3.46
Thiophene does not react with benzophenone. It reacts only when irradiated in the presence of BF3 (Scheme 3.46) [90]. In this reaction, BF3 is able to catalyze the ring opening of the oxetane. Benzophenone BF3 complex, excited by light, leads to an exciplex whose excitation energy is lower than the lowest triplet energy level of thiophene, which under the circumstances cannot act as a quencher. On the contrary, 2,5-dimethylthiophene reacts with benzophenone at 10 to give the corresponding cycloadduct with high regioselectivity in 62% yield (Scheme 3.47) [26, 91]. Ph Ph
S
hv Ph2CO
O S
Scheme 3.47
The reaction product can be obtained also by using 1-naphthaldehyde (50%), 2-, 3-, and 4-benzoylpyridine (62%, 58%, and 60%, respectively), and 2-benzoylthiophene (50%), while 2-naphthaldehyde, benzaldehyde, and acetophenone do not react [92]. 2,3-Dimethylthiophene also gave the corresponding oxetane when irradiated in the presence of benzophenone (60%) [93]. On the contrary, 2,3-dimethyl- and 2,3,5-trimethylthiophene do not react.
B. Pyrrole Pyrrole, like thiophene, does not react with benzophenone to give the corresponding oxetane. However, pyrrole reacts with aliphatic aldehydes and ketones to the corresponding 3-pyrryl carbinols. The alcohols derive from the cleavage of the corresponding oxetanes (Scheme 3.48) [94]. The yields increase when N-methylpyrrole is used as substrate, while the reactivity is depressed in the presence of substituents on the pyrrole ring. Pyrrole can give the corresponding oxetane when irradiated in the presence of benzophenone only when an electron-attracting group, such as benzoyl, is bound
118
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
H3C
CH3 OH
hv N H
CH3COCH3 Vycor
6%
N H H3C
CH3 OH
hv N CH3
CH3COCH3 Vycor
92%
N CH3
H3C
CH3 OH
hv H3C
N CH3
CH3 CH3COCH3 Vycor
H3C
CH3 N CH3
56%
Scheme 3.48
to the nitrogen atom (Scheme 3.49) [95, 96]. When N-phenylpyrrole was used as substrate, the corresponding 2-pyrryl carbinol was isolated [95].
HH Ph hv N
Ph2CO
Ph
O
Ph H
hv N Ph
Ph Ph
N Ph
12%
N H COPh
COPh
Ph2CO
Ph
O
OH
Scheme 3.49
C. Selenophene Selenophene does not react with benzophenone [26]. On the contrary 2-methylselenophene gives the corresponding cycloadduct. Also in this case the reaction occurs on the most hindered side of the molecule (Scheme 3.50) [97].
` -BU ¨ CHI REACTION ON PENTAATOMIC HETEROCYCLES THE PATERNO
119
Ph Ph hv Se
O 34%
CH3 Ph2CO
Se
CH3
Scheme 3.50
D. Imidazole Imidazole, N-methylimidazole, and 1,2-dimethylimidazole react with aliphatic aldehydes and ketones to give the corresponding 4-imidazolyl carbinols in good yields (Scheme 3.51) [94, 98]. The carbinols are obtained through the formation of the corresponding oxetanes. However, the reaction gave only 7.7% yield of the corresponding alcohol when treated with benzophenone [98]. Furthermore, the reaction with acetophenone gave only 1.2% yield of the product [98].
N
hv
HO CH3 H3C N
CH3COCH3
N H
N N
hv
N H
HO CH3 H3C N
CH3COCH3
N CH3
CH3
N
HO CH3 H3C N
hv CH3 CH3COCH3 N CH3
N
CH3
CH3
Scheme 3.51
There is no agreement between the work of Jones et al. [94] and that of Matsuura et al. [98] on the behavior of N-methylimidazole. While Jones’s group reported that the corresponding imidazolyl carbinol was obtained in excellent yields, Matsuura’s group reported that N-methylimidazole gave the corresponding carbinol in 9% yield. They also reported that 2-methylimidazole gave a
120
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
N N H
hv
HO CH3 H3C N
CH3 CH3COCH3
HO CH3 H3C N CH3
N H
+ H3C
N OH H 14.5%
H3C
8%
CH3
Scheme 3.52
mixture of two products with an overall yield of 22.5% (Scheme 3.52) [98]. 1,2Dimethylimidazole reacts with benzophenone, giving in low yield a reaction on the methyl in 2 position [99, 100]. N-Benzylimidazole, when irradiated in the presence of benzophenone in t-BuOH, gave a reaction on the methylene group [99, 100]. N-Acetylimidazole allows us to isolate the corresponding oxetane when irradiated in the presence of benzophenone (Scheme 3.53). The same result was obtained by using N-benzoylimidazole and N,N0 -carbonyldiimidazole [99, 100]. The oxetane can also be obtained in the reaction of N-methyl-2,4,5triphenylimidazole with aromatic ketones. It is noteworthy that, in this case, the reaction does not work in the presence of acetophenone and benzaldehyde [101]. Ph N N
Ph hv
N O
Ph2CO
COCH3
50% N COCH3
Scheme 3.53
E. Thiazole, Isoxazole, Isothiazole 2,4-Dimethylthiazole gives the corresponding oxetane when irradiated in the presence of benzophenone, while the same reaction fails when acetophenone is used as reagent (Scheme 3.54) [102]; 3,5- and 4,5-dimethylisoxazole give the corresponding oxetane in good yields. 4-Methylisothiazole give a reaction on the methyl substituent (Scheme 3.54) [102].
F. Indole Indole does not react with benzophenone under irradiation conditions. On the contrary, a benzoyl derivative reacts with it, giving the corresponding
` -BU ¨ CHI REACTION ON PENTAATOMIC HETEROCYCLES THE PATERNO
H3C hv Ph2CO
CH3
S
N
H3C
hv Ph2CO
O
40%
N
Ph N
CH3
O
Ph
H3C H3C
25% CH3
S
Ph hv Ph2CO
O
N O
Ph
CH3 H3C
Ph CH 3
Ph
N
O CH3
O
65%
N
H3C
121
O
Scheme 3.54
oxetane (Scheme 3.55) [102]. It does not react with acetophenone, benzaldehyde, acetone, and propionaldehyde. When methyl pyruvate is used as reagent, the corresponding 3-indolyl carbinol is isolated (Scheme 3.55) [102].
Ph
Ph O
hv
83%
Ph2CO
N
N O
O
Cl
Cl
H3C
OH
hv N
24%
CH3COCO2CH3
N O
O
Cl
CO2CH3
Cl
Scheme 3.55
122
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
G. Azaindole The same reaction has been described on N-acetyl derivative of 7-azaindole. Although the reaction represents a method able to obtain a new class of compound, the low yields of the product (4%) prevent from a synthetic use of this reaction [103].
H. Possible Explanation of the Observed Reactivity The reported results represent all the available data on the Paterno`-Bu¨ chi reaction on pentaatomic heterocycles that are not furan. There are very few data; in particular, (1) most of the unsubstituted compounds tested did not react and (2) only a few substituted derivatives showed a significative reactivity toward excited carbonyl compounds. This behavior may be due to different reasons. First, the different aromaticity of the compounds could play an important role to define the reactivity of the compounds. Furan is the lowest aromatic pentaatomic heterocyclic compound known, while the other compounds show higher aromaticity. However, this type of explanation cannot justify why thiophene does not react while simple dimethylthienyl derivatives react and why some dimethylthienyl derivatives react while some others do not show any reactivity. A different explanation was identified in the quenching properties of these heterocycles. Thiophene and monomethyl derivatives are efficient quenchers of triplet benzophenone. The Stern-Volmer plot showed a linear relationship [104, 105]. On the contrary, 2,5-dimethylthiophene (a compound able to give the cycloaddition reaction) is not a good quencher of benzophene [106]. NBenzoylpyrrole also does not act as a quencher of the triplet benzophenone [106]. On the contrary, pyrrole and selenophene are quenchers of the excited benzophenone [106]. In this case, the Stern-Volmer plot is not linear. This situation is commonly encountered when the quencher employed quenches two excited states. It seems reasonable that pyrrole acts as quencher of both triplet benzophenone and the exciplex between triplet benzophenone and pyrrole [106]. On the basis of these investigations, the common five-membered heterocycles may be classified in two categories in regard to their quenching properties: those with electron-donating groups, which give Stern-Volmer plots in the shape of straight lines, and those substituted with halogens and electron attracting groups, such as N-benzoylpyrrole, which give parabola-shaped curves. Those in the first category give oxetanes when they are bad quenchers. Since these compounds do not presumably form exciplexes, they should go from starting materials to products through a biradical intermediate. On the other hand, those in the second category most likely give oxetanes when they are good quenchers, as may be
REFERENCES
123
shown for 2,5-dibromothiophene, and furthermore they go from starting materials to products through an exciplex [107].
IV.
CONCLUSIONS
After reviewing these studies we think it could be useful to discuss some open aspects in the Paterno`-Bu¨ chi reaction on furan derivatives. The first problem is encountered in the regioselectivity of the reaction. By using substituted furans, in most of the cases the reaction occurs on the most hindered side of the molecule. This behavior could be explained assuming, on the basis of a theoretical description of the mechanism [15], the formation of the most stable biradical; in some cases, this assumption has been verified. However, the described regioselectivity was observed with both 2- and 3-methyl substituted furans, whereas a methyl group plays a role in the biradical stabilization only when present in position 2 on the furan. Furthermore, silyl- and stannyl-substituted furans showed inverse regioselectivity, in spite of the possible capto dative stabilizing effect of these substituents. Another open question is related to the diastereoselectivity of the reaction. We have shown that Adam’s rule cannot be used to explain the observed stereoselectivity in the reaction with 2-furylmethanol derivatives. If Adam’s hypothesis does not work, the diastereoselectivity of the Paterno`-Bu¨ chi reaction with allylic alcohols remains to be explained.
REFERENCES 1. E. Paterno` and G. Chieffi, Gazz. Chim. Ital. 1909, 39, 431. 2. G. Bu¨ chi, C. G. Inman, and E. S. Lipinsky, J. Am. Chem. Soc. 1954, 76, 4327. 3. D. R. Arnold, Adv. Photochem. 1968, 6, 301. 4. G. Jones, Org. Photochem. 1981, 5, 1. 5. H. A. J. Carless, In Synthetic Organic Photochemistry, W. M. Horspool, Ed., Plenum, New York, 1984, p. 425. 6. J. A. Porco and S. L. Schreiber, In Comprehensive Organic Synthesis, Vol. 5, B. M. Trost, I. Fleming, and L. A. Paquette, Eds., Plenum, New York, 1991, p. 151. 7. A. G. Griesbeck, In Handbook of Photochemistry and Photobiology, W. A. Horspool, and P.-S. Song, Eds., CRC Press, Boca Raton, FL, 1994, p. 550. 8. D. G. I. Kingston, Pharmac. Ther. 1991, 52, 1. 9. J. Huang, R. Yokoyama, C. Yang, and Y. Fukuyama, Tetrahedron Lett. 2000, 41, 6111.
124
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
10. Y. Wang, G. W. J. Fleet, R. Storer, P. L. Myers, C. J. Wallis, O. Doherty, D. J. Watkin, K. Vogt, D. R. Witty, F. X. Wilson, and J. M. Peach, Tetrahedron Asymmetry 1990, 1, 527. 11. C. O. Yang, W. Kurz, E. M. Eugui, M. J. McRoberts, J. P. H. Verheyden, L. J. Kurz, and K. A. M. Walker, Tetrahedron Lett. 1992, 33, 41. 12. Y. Kawahata, S. Takatsuto, N. Ikekawa, M. Murata, and S. Omura, Chem. Pharm. Bull. 1986, 34, 3102. 13. J. Kopecky´ , Organic Photochemistry, VCH, New York, 1992. 14. N. J. Turro, J. C. Dalton, K. Dawes, G. Farrington, R. Hautala, D. Morton, M. Niemczyk, and N. Schore, Acc. Chem. Res. 1972, 5, 92. 15. I. J. Palmer, I. N. Ragazos, F. Bernardi, M. Olivucci, and M. A. Robb, J. Am. Chem. Soc. 1994, 116, 2121. 16. A. G. Kutateladze, J. Am. Chem. Soc. 2001, 123, 9279. 17. S. C. Freilich and K. S. Peters, J. Am. Chem. Soc. 1981, 103, 6225. 18. S. C. Freilich and K. S. Peters, J. Am. Chem. Soc. 1985, 107, 3819. 19. J. Mattay, Angew. Chem., Int. Ed. Engl. 1987, 26, 825. 20. G. O. Schenck, W. Hartman, and R. Steinmetz, Chem. Ber. 1963, 96, 498. 21. D. Gagnaire and E. Payo-Subiza, Bull. Soc. Chim. Fr. 1963, 2623. 22. S. Toki, K. Shima, and H. Sakurai, Bull Chem. Soc. Jpn. 1965, 38, 760. 23. T. Nakano, C. Rivas, and C. Perez, J. Chem. Soc. Perkin Trans. 1, 1973, 2322. 24. E. B. Whipple and G. R. Evanega, Tetrahedron 1968, 24, 1299. 25. C. Rivas and E. Payo, J. Org. Chem. 1967, 32, 2918. 26. C. Rivas, R. A. Bolivar, and M. Cucarella, J. Heterocyclic Chem. 1982, 19, 529. 27. K. Shima and H. Sakurai, Bull. Chem. Soc. Jpn. 1966, 39, 1806. 28. S. Toki and H. Sakurai, Bull. Soc. Chem. Jpn. 1967, 40, 2885. 29. A. Zamojski and T. Koz´ luk, J. Org. Chem. 1977, 42, 1089. 30. S. L. Schreiber, D. Desmaele, and J. A. Porco, Tetrahedron Lett. 1988, 29, 6689. 31. H. A. J. Carless and A. F. Halfhide, J. Chem. Soc. Perkin Trans. 1, 1992, 1081. 32. M. Abe, E. Torii, and M. Nojima, J. Org. Chem. 2000, 65, 3426. 33. A. G. Griesbeck, H. Mauder, and S. Stadtmu¨ ller, Acc. Chem. Res. 1994, 27, 70. 34. A. G. Griesbeck, S. Buhr, M. Fiege, H. Schmickler, and J. Lex, J. Org. Chem. 1998, 63, 3847. 35. M. Ogata, H. Watanabe, and H. Kano, Tetrahedron Lett. 1967, 533. 36. J. Leitich, Tetrahedron Lett. 1967, 1937. 37. G. R. Evanega and E. B. Whipple, Tetrahedron Lett. 1967, 2163. 38. S. Toki and H. Sakurai, Tetrahedron Lett. 1967, 4119. tibranyi, Coll. Czech. Chem. Commun. 1984, 49, 71. 39. S. Sekretar, J. Ruda¯ , L. S 40. T. S. Cantrell, A. Allen, and H. Ziffer, J. Org. Chem. 1989, 54, 140. 41. Y. Kubo, M. Suto, S. Tojo, and T. Araki, J. Chem. Soc. Perkin Trans. 1 1986, 771.
REFERENCES
125
42. M. D’Auria, L. Emanuele, and R. Racioppi, J. Photochem. Photobiol., A: Chem. 2004, 163, 103. 43. C. H. Krauch, W. Metzner, and G. O. Schenck, Chem. Ber. 1966, 99, 1723. 44. S. Farid, S. E. Hartman, and C. D. DeBoer, J. Am. Chem. Soc. 1975, 97, 808. 45. H. A. Carless and D. J. Haywood, J. Chem. Soc., Chem. Commun. 1980, 1067. 46. A. G. Griesbeck and S. Stadtmu¨ ller, J. Am. Chem. Soc. 1990, 112, 1281. 47. A. G. Griesbeck and S. Stadtmu¨ ller, J. Am. Chem. Soc. 1991, 113, 6923. 48. A. G. Griesbeck and S. Stadtmu¨ ller, Chem. Ber. 1990, 123, 357. 49. A. G. Griesbeck, H. Mauder, K. Peters, E.-M. Peters, and H. G. von Schnering, Chem. Ber. 1991, 124, 407. 50. A. B. Smith III, G. A. Sulikowski, M. M. Sulikowski, and K. Fujimoto, J. Am. Chem. Soc. 1992, 114, 2567. 51. S. Jarosz and A. Zamojski, J. Org. Chem. 1979, 44, 3720. 52. T. Kitamura, Y. Kawakami, T. Imagawa, and M. Kawanisi, Synth. Commun. 1977, 7, 521. 53. T. Kozluk and A. Zamojski, Tetrahedron, 1983, 39, 805. 54. S. L. Schreiber, A. H. Hoveyda, and H.-J. Wu, J. Am. Chem. Soc. 1983, 105, 660. 55. M. Satre, Biochem. Biophys. Res. Commun. 1981, 100, 267. 56. S. L. Schreiber and K. Satake, J. Am. Chem. Soc. 1983, 105, 6723. 57. S. L. Schreiber and K. Satake, J. Am. Chem. Soc. 1984, 106, 4186. 58. S. L. Schreiber and A. H. Hoveyda, J. Am. Chem. Soc. 1984, 106, 7200. 59. S. L. Schreiber and J. A. Porco, J. Org. Chem. 1989, 54, 4721. 60. N. Shimada, S. Hasegawa, T. Harada, T. Tonisawa, and T. Fujii, J. Antibiot. 1986, 39, 1623. 61. H. Hoshino, N. Shimizu, N. Shimada, T. Takida, and T. Takeuchi, J. Antibiot. 1987, 60, 1077. 62. R. Hambalek and G. Just, Tetrahedron Lett. 1990, 31, 4693. 63. D. R. Mootoo, V. Date, and B. Fraser-Reid, J. Am. Chem. Soc. 1988, 110, 2662. 64. R. Hambalek and G. Just, Tetrahedron Lett. 1990, 31, 5445. 65. S. Jarosz and A. Zamojski, Tetrahedron, 1982, 38, 1447. 66. R. Pelzer, P. Ju¨ tten, and H.-D. Scharf, Chem. Ber. 1989, 122, 487. 67. R. Pelzer, H.-D. Scharf, H. Buschmann, and J. Runsink, Chem. Ber. 1989, 122, 1187. 68. S. Hu and D. C. Neckers, J. Chem. Soc. Perkin Trans. 2, 1999, 1771. 69. H. Buschmann and H.-D. Scharf, N. Hoffmann, M. W. Plath, and J. Runsink, J. Am. Chem. Soc. 1989, 111, 5367. 70. H. Buschmann, H.-D. Scharf, N. Hoffmann, and P. Esser, Angew. Chem., Int. Ed. Engl. 1991, 30, 477. 71. M. D’Auria, L. Emanuele, and R. Racioppi, Photochem. Photobiol. Sci. 2003, 2, 904. 72. Y. Zhang, J. Xue, Y. Gao, H.-K. Fun, and J.-H. Xu, J. Chem. Soc., Perkin Trans. 1, 2002, 345.
126
1,2-CYCLOADDITION REACTION OF CARBONYL COMPOUNDS
agar and H.-D. Scharf, Chem. Ber. 1991, 124, 967. 73. C. Z 74. S. Jarosz and A. Zamojski, Tetrahedron 1982, 38, 1453. 75. J. M. J. Tronchet and B. Baehler, J. Carbohydrates Nucleosides Nucleotides 1974, 1, 449. 76. S. L. Schreiber, Science 1985, 227, 857. 77. S. L. Schreiber and K. Satake Tetrahedron Lett. 1986, 27, 2575. 78. W. Adam, K. Peters, E. M. Peters, and V. R. Stegmann, J. Am. Chem. Soc. 2000, 122, 2958. 79. W. Adam and V. R. Stegmann, Synthesis 2001, 1203. 80. W. Adam and M. Prein, Angew. Chem., Int. Ed. Engl. 1996, 35, 477. 81. T. Bach, J. Schro¨ der, and K. Harms, Tetrahedron Lett. 1999, 40, 9003. 82. T. Bach, H. Bergmann, and K. Harms, J. Am Chem. Soc. 1999, 121, 10650. 83. A. Yokohama and K. Mizuno, Org. Lett. 2000, 2, 3457. 84. A. G. Griesbeck and S. Bondock, J. Am. Chem. Soc. 2001, 123, 6191. 85. T. Bach, Liebigs Ann. 1997, 1627. 86. M. D’Auria, R. Racioppi, and G. Romaniello, Eur. J. Org. Chem. 2000, 3265. 87. M. D’Auria and R. Racioppi, Arkivoc, 2000, 1, 145. 88. M. D’Auria, L. Emanuele, G. Poggi, R. Racioppi, and G. Romaniello, Tetrahedron 2002, 58, 5045. 89. C. Rivas and F. Vargas, in CRC Handbook of Organic Photochemistry and Photobiology, W. M. Horspool and P.-S. Song, Eds., CRC Press, Boca Raton, FL 1995, p. 536. 90. F. Vargas, C. Rivas, M. Navarro, and Y. Alvarado, J. Photochem. Photobiol., A Chem. 1996, 93, 169. 91. C. Rivas, M. Velez, and O. Crescente, J. Chem. Soc. Chem. Commun. 1970, 1474. 92. C. Rivas, and R. A. Bolivar, J. Heterocyclic Chem. 1973, 10, 967. 93. C. Rivas, D. Pacheco, F. Vargas, and J. Ascanio J. Heterocyclic Chem. 1981, 18, 1065. 94. G. Jones, H. M. Gilow, and J. Low, J. Org. Chem. 1979, 44, 2949. 95. C. Rivas, M. Velez, M. Cucarella, R. A. Bolivar, and S. E. Flores, Acta Cient. Venezolana 1971, 22, 145. 96. C. Rivas and R. A. Bolivar, J. Heterocyclic Chem. 1976, 13, 1037. 97. C. Rivas, D. Pacheco, and F. Vargas, Acta Sud Am. Quim. 1982, 2, 1. 98. T. Matsuura, A. Banba, and K. Ogura, Tetrahedron 1971, 27, 1211. 99. T. Nakano, C. Rivas, C. Perez, and J. M. Larrauri, J. Heterocyclic Chem. 1976, 13, 173. 100. T. Nakano, W. Rodriguez, S. Z. de Roche, J. M. Larrauri, C. Rivas, and C. Perez, J. Heterocyclic Chem. 1980, 17, 1777. 101. Y. Ito, M. Ji-Ben, S. Suzuki, Y. Kusunaga, T. Matsuura, and K. Fukuyama, Tetrahedron Lett. 1985, 26, 2093. 102. D. R. Julian and G. D. Tringham, J. Chem. Soc. Chem. Commun. 1973, 13.
REFERENCES
127
103. T. Nakano and M. Santana, J. Heterocyclic Chem. 1976, 13, 585. 104. R. A. Bolivar and C. Rivas, J. Photochem. 1981, 17, 91. 105. R. A. Bolivar and C. Rivas, J. Photochem. 1982, 19, 95. 106. R. A. Bolivar, R. Machado, L. Montero, F. Vargas, and C. Rivas, J. Photochem. 1983, 22, 91. 107. F. Vargas and C. Rivas, Int. J. Photoenergy 2000, 2, 97.
THE INVENTION OF DYLUX1 INSTANT-ACCESS IMAGING MATERIALS AND THE DEVELOPMENT OF HABI CHEMISTRY—A PERSONAL HISTORY Rolf Dessauer (Retired) Greenville, DE
CONTENTS I. II. III. IV. V.
Introduction My Background Dessauer at DuPont We Must Have Done Something Right! The Beginning A. Lawrence A. Cescon—An Inventor B. A New Phototropic Substance C. George R. Coraor—An Inspiring Leader VI. 1961—Permanent Color A. What Do You Do with This Stuff? B. What Is Needed? Advances in Photochemistry, Volume 28 Edited by Douglas C. Neckers, William S. Jenks, and Thomas Wolff # 2005 John Wiley & Sons, Inc.
129
130
THE INVENTION OF DYLUX
VII.
VIII.
IX.
X. XI.
XII.
XIII.
XIV.
C. Kay Looney—A Creative Spectroscopist D. Radiation Physics Laboratory and Al MacLachlan E. Chuck Yembrick, Peter Strilko, and Optimization F. Patents 1962—Improvements in Technology A. A New Opportunity—Foot Imaging B. Optical Printing C. Proofing? 1964—A Higher Speed System A. HABIs Really Are Photopolymerization Initiators B. Management Change 1965—Other Markets A. Microfilm B. Cathode Ray Tube Imaging 1966—We Become a Venture 1967—We Start a Marketing Effort: How Do You Get to the Market, Any Market? A. Harold Wilbur—The Electrician B. Point of Sales Opportunity C. The UVI Movie D. Another Trip to the West Coast E. The Name Dylux1 F. More Marketing G. Large Wall-Screen Displays 1. Project Apollo 2. Boeing H. Is There a Business Here? I. Opto-Magnetic Printing J. How to Get Rich? 1968—More Marketing Activities A. W. H. Brady Co. and the Kalograph1 B. John de Campi—An Enthusiastic Marketeer C. Mattel 1969—William S. Wartell: A Dynamic Decision Maker A. Proofing, for Certain B. Spring Boston SPSE Meeting C. What Is So Good about Dylux1 Proofpaper? D. Europe 1970—Expansion of Opportunities A. Military Applications B. What Is/Was Wrong with Dylux1 Proofpaper? C. Transfer of the Dylux1 Venture
131
XV. 1971—After the Venture A. Photodecoration XVI. 1972—What Else? XVII. 1973—Universal Product Code Opportunity XVIII. 1974–Philip Botsolas: A Breath of Fresh Air XIX. Progress A. Dylux1 DFF Filter B. Europe C. Medical Imaging Systems–‘‘aca’’ Automatic Clinical Analyzer XX. 1976—Still More Opportunities A. GTE-Sylvania Home Office Copier B. The End of Photodecoration C. Dylux1-4C Overlay Films D. Phenidone XXI. 1977—And Still More Opportunities A. New HABIs: TCTM-HABI B. Photomarker1 Corp. C. Miscellaneous Activities D. J. C. Penney—Inventory Control E. Tokyo SPSE Meeting F. X-Raylux: Instructional X-Ray Prints XXII. 1978—Transitions A. Another Dual Response System B. Transfer to the Photo Products Department XXIII. 1979—Black Dylux1 535 Proofpaper XXIV. 1981—Add-On Toning XXV. 1983—A New Wind Blows XXVI. 1984—China XXVII. 1984–2004 A. Xeroprinting B. Photoimaging, Ltd. C. Competition D. Commercial Products E. Watermarks F. Thermal Dylux1 Proofpaper G. HABIs in Photopolymer Products H. The HABI Literature I. Could We Have Done Better? J. HABI Statistics Acknowledgments Rewards and Awards
132
THE INVENTION OF DYLUX
I. INTRODUCTION Many publications in the literature describe the technical aspects of hexaarylbiimidzole (HABI) chemistry. The following is an attempt to put these and the reported patent literature into a historical perspective and to add my personal reflections about the times, the scientists, and the emotions that were involved. Although several publications preceded the DuPont work, I believe that it was work in industry that elucidated and commercialized products based on these unique compounds. I have attempted to put events in chronologic order and have relied on Chemical Abstracts and information derived from issued U.S. patents to support the technical details. This is not entirely satisfactory because the dates when patents are allowed may not actually relate to the date the work was done. I used the issue dates of the patents for chronologic order, although the work, of course, had to precede these dates. I have mentioned the names of the individuals who did the actual research and marketing, although there were many people who contributed to this work, who were not recognized in patents or whose names are now forgotten; to them, my apologies. To report this work accurately, it must be pointed out that there were supporters and detractors. In some cases, their deliberations were unknown to me. In summary, however, the DuPont Company supported a very exciting effort and was rewarded with financial returns. The Photo Products Department changed from a producer of silver films to an innovator and powerhouse in the prepress proofing and electronic products area. This work started around 1958, when the DuPont Company was organized very differently from the way it is organized now. To understand better some of the problems that occurred in the development of this chemistry, it is necessary to explain this structure and comment on some of the individuals who were involved in this work.
II. MY BACKGROUND I was born in Germany in 1926, and as a result of the rise of the Nazi Party, my family was forced to emigrate to the United States. My older brother arrived in America in November 1937, my father in December 1938, and my mother and I in June 1939. My father had been a physician in Germany and spent nearly four years in the United States taking language and medical state board examinations before resuming his career. I completed my schooling in New York City, enlisted in the U.S. Army in 1944, and in 1947 enrolled at the University of Chicago,
DESSAUER AT DUPONT
133
where I received BS and MS degrees in chemistry. In 1949, I transferred to the University of Wisconsin, where I received my Ph.D. in organic chemistry in 1952. In September 1952, I began my career as a chemist at DuPont, where I continued to work until 1991. I married Peggy in 1965, divorced in 1967, married Nicky in 1968, was widowed in 1981, and married Angela in 1991. I have two stepsons, Nicky’s children, neither of whom became chemists.
III. DESSAUER AT DUPONT At the time I joined DuPont, the company had a number of operating departments, which covered certain business areas, sometimes with considerable overlap and resulting conflicts. The largest of these then was the Textile Fibers Department, which conducted research on novel fibers, such as nylon, Dacron1 and Orlon1 fibers; manufactured these in a number of domestic plants; and sold them through sales offices throughout the United States. Other departments were the Organic Chemicals Department, which was in the dyes, petroleum additives, fluorinated, and elastomeric products businesses; the Pigments Department, which sold colored and white pigments; the Fabrics and Finishes Department, which sold coating materials and some coated products; the Photo Products sold a number of silver-based photographic products, including X-ray, graphic arts, and black and white motion picture films; the Film Department handled a series of film products, such as cellophane and polyester films; and the Polychemicals Department manufactured a number of polymeric products, such as nylon, Delrin1, and polyolefins. Other departments handled agricultural products (Grasseli), certain specialty chemicals (Electrochemicals Department), explosives (Explosives Department), etc. To assist in the research and develop new businesses, there were staff departments, such as the Chemical Department, later to become Central Research, the Engineering Department, the Development Department, the Treasurers Department, etc. The operating departments each had some similarity in structure; there were separate research, manufacturing, and marketing/sales divisions, each headed by directors, who reported to a general manager. These then reported to the company’s executive committee and president and chairman. At the upper levels of the company there were frequent shifts in personnel from one department to another in order to develop management with a broader background. On some occasions, this resulted in favorable results; but frequently it occurred that the leadership was lacking in experience in the areas that it was to manage. In my experience, I found very few individuals who had the breadth of intellect and energy to learn a business from the ground up, resulting in lack of positive leadership.
134
THE INVENTION OF DYLUX
DuPont had grown enormously after World War II and realized that the development of the company depended on the entry into new fields. The main research center, the DuPont Experimental Station, was enlarged significantly in the early 1950s. Each operating department had representation there, and transmittal of technology from the various research groups of the operating and staff departments was encouraged. Several thousand scientists were stationed there, and it was a great time to do research. At that time, the DuPont Company was still very much in the hands of the DuPont family, and then was more oriented toward growth than day-to-day changes in the value of the company’s stock. Hence there was less pressure then to think in terms of short-term objectives. In the mid-1960s, the venture concept, with a certain autonomy for groups inside an operating department was pursued. Typical of these were Corfam, DuPont’s leather replacement offering, and UVI, the ultraviolet imaging program; there were some 29 others. The ventures varied in size from a few to many employees. In the 1980s, there were major acquisitions; and later there were joint ventures with major companies, such as DuPont Merck and DXI (DuPont Xerox Imaging). In the 1990s, Apex programs, designed to identify major opportunities in new business areas were considered. In time, this changed; and the current structure, which had evolved as the company grew and later contracted, is better organized to take advantage of short-term new opportunities. When I joined the Organic Chemicals Department (Orchem), it had a very large staff of scientists working at Jackson Laboratory in Deepwater, New Jersey. It operated several highly profitable businesses, such as petroleum additives, for which the principal product was the gasoline additive tetraethyl lead; Freon1 propellants; elastomeric products, such as neoprene; and the dyes and chemicals business, which sold a broad range of colored materials to the textile and paper industries. During the immediate post–World War II period, DuPont developed a number of profitable novel fibers. The Textile Fibers Department supported research at Orchem to identify and manufacture novel dyes, which were required to impart color to these new materials. In some cases, the new dyes enlarged the business opportunities for these fibers, so that they became widely successful. In time, the textile dye field dominated the thinking of the department, until the market identified major opportunities in the coloration of paper products as well as consideration of related materials, such as fluorescent whitening agents, UV screening agents, and colorants for the graphic arts and other industries. In writing this story, years after much happened, I am struck by the fact that my colleagues and I developed some excellent chemistry in the Organic Chemicals Department, which resulted in myriad novel products that generated significant earnings for the DuPont Company, but that there was also much opposition—especially from some of the management of the Photo Products
THE BEGINNING
135
Department. Was it departmental envy or a not-invented-here attitude? Or was it that novel technologies always have to fight harder to get support? Were there personal animosities? For example, Dr. Robert Upson, the director of research for Photo Products Department, believed chemists should invent what their managers tell them, but Dr. M. L. Ernsberger, Director of Research of the Organic Chemicals Department (and I), favored a pragmatic approach—if a new and better opportunity arises—pursue it!
IV. WE MUST HAVE DONE SOMETHING RIGHT! My colleagues and I developed a novel imaging system, which resulted in products for the prepress proofing industry that were successful for over 35 years. They were ‘‘big earners’’ because the research was done with an eye on establishing a strong patent position, and no effective competition came along. The photo-oxidants that were developed became enormously successful as photopolymerization initiators, which allowed DuPont to manufacture a series of novel products of importance to the printing and electronics industries. Now, some 40 years after this work was done, these biimidazole derivatives still appear to be initiators of choice and are found to play a role in current patents.
V. THE BEGINNING It all began in 1958. I was assigned by the management of the Organic Chemicals Department of DuPont to examine chemistry that would lead to new products, which could use the facilities in the Chambers Works, which had been dedicated to the manufacture of dyes and related aromatic chemicals. At that time, automobiles were designed to have larger glass areas; and one possible application was to design glass that could automatically darken when sunlight became more intense. Such photochromic materials could also see application in the architectural field. DuPont manufactured Butacite1, a plastic material used to provide stability and protection to automotive window glass, and it was hoped that incorporation of a photochromic material in the automotive glass interliners would provide a ready market. I had earlier found that certain classes of metalized hydroxyazo dyes could be dissolved in Butacite1, and when sandwiched between glasses, exhibited extraordinary stability. This work led to a good relationship with the Butacite research group, who endorsed our work with photochromic materials.
136
THE INVENTION OF DYLUX
It may be useful to think of what we now accept as routine in 2004 that we had not seen, experienced, or anticipated in 1958. Here is a brief list: Personal computers Office and home copiers Digital music Compact disks Color television Video tape and recorders Audio tape recorders Photoresists GPS The Pill Cable TV Faximile Low-cost telephony e-mail and the Internet DNA Infrared lasers
In the summer of 1958, I was assigned to a group at the DuPont Experimental Station under Dr. W. R. Remington, with the objective of working with a very talented summer employee, John Harriman, who was then still an undergraduate at the University of Wisconsin. One of my colleagues, Dr. C. Wheaton Vaughn urged me to examine a reference to N-pyridylsydnone (N-PS) (Scheme 4.1). H N N
O O
N
Scheme 4.1 N-3-pyridylsydnone
which had recently appeared in the Journal of the American Chemical Society (J. M. Tien and I. M. Hunsberger, J. Am. Chem. Soc. 1955, 77, 6604). We prepared samples of this compound and examined it under a variety of conditions.
THE BEGINNING
137
Harriman and I had considerable encouragement from Dr. W. D. Phillips, a talented and highly regarded physical chemist then at the Chemical Department of DuPont. In the end, we concluded that N-PS was photochromic in the crystalline state but not in solution or in Butacite films; and by the end of the summer, we decided that other materials would be needed. Also in our group was Dr. C. F. Wahlig, a physicist formerly with the Photo Products Department, who had been transferred to help Harriman and me. Another chemist, Dr. Theodore Mill, was assigned this program; and after he left DuPont, Dr. L. A. Cescon, who had recently joined DuPont was asked to work in this area.
A. Lawrence A. Cescon—An Inventor Cescon obtained his Ph.D. at The Pennsylvania State University. He was born in Allentown, Pennsylvania, the son of Italian immigrants. He was a very methodical person, who studied the literature, endlessly purified every material he ever used in syntheses or evaluations, and seemed at times very slow. But he was very thorough. Cescon remained on this program only until 1965, and finished his career at the company’s Explosives Department, where he also made significant contributions in the field of emulsion-containing explosive compositions. During the 5 years of our collaboration, Cescon and I obtained many patents and continued to stimulate each other. As I had several years of experience in the dye field from earlier work, Cescon was content to accept many of my suggestions, and I in turn was stimulated by his perception of organic chemistry. I had found it very difficult to understand why he and I could not collaborate in developing better imaging systems, but his later work on waterborne explosives (L. A. Cescon and R. W. Trebilcock, U.S. Pat. 4,619,721, Oct. 28, 1986; L. A. Cescon, R. W. Trebilcock, and R. H. Moffett, U.S. Pat. 4,714,503, Dec. 22, 1987) was very successful, too. He contracted cancer and retired from DuPont in 1985 and moved to Italy, where his second wife had been a schoolteacher. He became interested in the effect of nutrition on cancer and was invited to give talks about his studies at many institutions. He taught chemistry at Italian universities. We remained friends and visited each other for many years, until he died in 1995.
B.
A New Phototropic Substance
One of the many benefits of working for the DuPont Company in the 1960s was the availability of a company-organized abstract publication that covered the technical and patent activities in areas of interest to the operating departments.
138
THE INVENTION OF DYLUX
Hence it was possible to be somewhat more abreast of the literature than if one had to rely on Chemical Abstracts, which then lagged considerably behind. It must be recalled that it was far more difficult during that period to obtain copies of articles that appeared in journals, because the electrophotographic copiers were not with us yet, and copies of a referenced article had to be microphotograped and reprinted on photographic paper in our excellent Lavoisier Library. Of course, there were then far fewer journals that had to be followed. Cescon saw a reference to work by T. Hayashi and K. Maeda, in which the Japanese scientists described a new phototropic substance, based on 2,4,5-triphenylimidazole (T. Hayashi, K. Maeda, and S. Shida, J. Chem. Phys. 1960, 32, 1568; T. Hayashi, and K. Maeda, Bull. Chem. Soc. Jpn. 1960, 33, 565). Cescon and I discussed this work, which required lophine or triarylimidazole as a starting material for this photochromic compound. As an experienced dye chemist, who had worked with aromatic heterocyclic compounds, I realized that lophine and its analogs were easy to prepare. As a matter of fact, we had found that they had been considered as fluorescent whitening agents earlier. At that time, I had laboratories at the Experimental Station and at Jackson Laboratory, where I was still involved with research on dyes and UV absorbers. Working for me were three very talented technicians, Robert Jenkins, John Willister, and Walter Balon. The latter was a self-taught chemist who possessed great synthetic skills; the former two were also very talented and enthusiastic to try new things. I asked Jenkins to make a sample of triphenylimidazole for Cescon, and delivered it to the Experimental Station, where Larry then studied the oxidation of this compound. About that time, the Organic Chemicals Department reassigned Dr. Remington; Dr. George Coraor, who came from the Explosives Department, took his place. I had been friends with Remington and was sorry to see him leave the program, but was encouraged by Coraor’s enthusiasm for new chemistry.
C. George R. Coraor—An Inspiring Leader The enthusiastic leader of Cescon’s and my endeavors was Dr. George R. Coraor, then in his mid-thirties. He had begun his career in DuPont’s Explosives Department in 1950 and had several promotions. Coraor was an immensely curious man and shared with me the idea that one should seek alternate routes toward success if and when management stood in our way. It appeared that much of our collaboration involved trying to bypass his overseers, who were more concerned with avoiding failure than encouraging success. Coraor was short and stout and did not fit the perception of DuPont management, who usually promoted tall, thin men. He was just plain smart and more supportive of his subordinates
139
THE BEGINNING
than anyone I had ever met. In time, his independence cost him the leadership of our group, which in turn almost killed our program. George asked me if I could spend more time on the program; as I enjoyed working with him, this was a pleasure for me. Cescon repeated Hayashi’s work and concluded that the compound synthesized by the Japanese workers, though not identified, was probably the dimer of the triphenylimidazolyl radical. We quickly did the requisite analytical work to confirm this. The radical formed by irradiation of the dimer was purple, and we decided that we should make related triarylimidazoles, a relatively simple task, considering that the Jackson Laboratory store room, under the dedicated supervision of Mrs. Cy Bleistine and Herb Dayton, had a large repertoire of aromatic aldehydes, benzils, and other necessary ingredients. The synthetic approach used is outlined in Scheme 4.2.
H CHO
O
OH O
CHO
O NH4 Ac
N
N H
N
N
2
Scheme 4.2
Cescon said that he would study the oxidation of these materials and, of course, purified everything that Jenkins and Balon made. Again, an historical note: In the 1960s, the Aldrich catalog was not as all-encompassing as it was to become later. DuPont’s Jackson Laboratory had a huge store room, where organic chemicals were stored according to the Beilstein classification system. Hence one could easily obtain samples of complex compounds with minimum effort. The structure of the dimer was believed to be as shown in Scheme 4.3. Before long we had prepared several biimidazoles, which gave rise to a wide range of colored radicals, including blues and greens. Coraor asked for more manpower, and Dr. Ernest Silversmith was assigned to study the physical chemistry of these materials. He in turn also synthesized some novel triarylimidazoles. I was promoted to senior research chemist and assigned to study applications of these
140
THE INVENTION OF DYLUX
N
N
N
N
UV
N
N
DIMER
N
N
RADICALS
Scheme 4.3
biimidazoles and, in general, provide some guidance to the group, which grew to five, while under Coraor. My days as a dye chemist were just about over! Our work was summarized in two patents: 2,4,5-Triphenylimidazole dimers. E. I. du Pont de Nemours & Co. (by Lawrence A. Cescon). Fr. 1,351,818 (Cl. C 07d, G l 02b, G 02c), Feb. 7, 1964; U.S. Appl. Mar. 21, 1962; 32 pp. Triphenylimidazoles were oxidized with K3Fe-(CN)6 in EtOH contg. KOH to give phototropic dimers. Thus, 1.4 parts o-ClC6H4CHO was added to a soln. of 2.1 parts benzil in 50 ml. HOAc contg. 6 parts NH4OAc, and the mixt. refluxed 2 hrs. and poured into 200 parts cold H2O to give 2-(o-chlorophenyl)-4,5-diphenylimidazole (I), m. 196–7 (EtOH). A soln. is prepd. from 1.1 parts I and 100 parts EtOH contg. 12 parts KOH, 450 parts 1% K3Fe(CN)6 added over 1.5 hrs., the mixt. filtered, and the ppt. solvated with EtOH (2 moles EtOH per 3 moles diimidazole) to give solvated I dimer, m. 95–110 , softens at 90 , lavender at 170 , red brown at 190 , red at 220 . Similarly prepd. were the following II and solvated II dimers: (X, Y, m.p. and softening point of solvated dimer given): Cl, Cl, 174.5 75 90 ; MeO, H, 207.5 8.5 , 160 ; MeO, MeO, 164 5 ,—Br, H 205.5-6.5 , 106 ; and F, H, 205.5–206 , 139 40 . 2-(Naphthyl)4,5diphenylimidazole (III) m. 289.5–90 ; solvated III dimer softened 153 . Dimers of IV were prepd. (R, R1 and softening point of solvated dimer given): 2-EtOC6H4, Ph, 138 ; 2,3-(MeO)2C6H3, Ph, 95 ; 9-phenanthryl, Ph, 165 ; 2,4,6-Me3C6H2, Ph, —; 2-MeOC6H4, 4-MeOC6H4, —; 1-C10H7, 4-MeOC6H4,—. Also prepd. was the solvated dimer of 2-(o-methoxyphenyl)-4-(p-methoxyphenyl)-5-phenylimidazole. Biimidazoles. E. I. du Pont de Nemours & Co. Belg. 635,804, Dec. 2, 1963, Appl. Aug. 2, 1963; 32 pp. I are phototropic. A mixt. of benzil 2.1, AcOH 50, NH4OAc 6, and 2 ClC6H4CHO 1.4 was refluxed for 2 hrs. to give 2-(2-chlorophenyl)-4,5-diphenylimidazole (II), 3.1 parts m. 196-7 . A 1 % aq. soln. of K3Fe(CN)6 450 was added at the rate of 5 parts/min. to II 1.1 in EtOH 100 contg. KOH 12, the ppt.1 part filtered off, washed, and vacuum-dried (0.1 mm.) at 50-6 for 8 hrs. to give I
141
THE BEGINNING
N
N
R1 R2
2
R3 I 1
2
3
(R ¼ Cl, R ¼ R ¼ H) (III), solvated with 2 mols. EtOH/3 mols. III, as pale yellow crystals, softening at 90 , forming a gel at 95–110 , becoming red-brown at 190 and red at 220 . A colorless 0.5% benzene soln. of III exposed to sunlight became purple immediately and the color disappeared in 1 min. after removal of the illumination. Comparison in a spectrophotometer with I (R1 ¼ R2 ¼ R3 ¼ H) (IV) showed that the color of III disappeared 16 times as rapidly as that of IV. Similarly, other I were
R1
R2
R3
color in sunlight
Cl OMe OMe CH:CH:CH:CH Br F
H H H
Cl H OMe H H H
purple blue greenish blue orange purple purple
H H
rapidity of decolorization compared with IV 35 17 2 2 16 13
prepd. (see Table). Also prepd. were 2,2,4-dichlorophenyl) (m. 174.5-75 ), 2-o-anisyl- (m. 207.5–8.5 ), 2-(2,4-dimethoxyphenyl)- (m. 164–5 ), 2-(1-naphthyl)- (m. 289.5–90 ), 2-o-bromophenyl)- (m. 205.5–6.5 ) and 2-(o-fluorophenyl)-4,5-diphenylimidazole (m. 205.5–206 )
We soon realized that others also found the reference to Hayashi and Maeda’s work intriguing, and a number of publications showed considerable activity in this new field. We scouted Chemical Abstracts for any work by any of the chemical or photographic companies, but only one publication from a major company, General Electric by White and Sonnenberg came to our attention. Was this only a scientific curiosity? Didn’t anyone else want to make photochromic materials commercially? Here is a sampling of references from 1960 to 1972.
142
THE INVENTION OF DYLUX
T. Hayashi and K. Maeda, Preparation of a new photochromic substance, Bull. Chem. Soc. Jpn. 1960, 33, 566. T. Hayashi, K. Maeda, S. Shida, and K. Nakada, A new phototropic substance and its ESR. J. Chem. Phys. 1960. 32, 1568. T. Hayashi and K. Maeda, Mechanism of chemiluminescence of 2,4,5-triphenylimidazole. Bull. Chem. Soc. Jpn. 1962, 35, 2057. H. Baumgaertel and H. Zimmermann, Synthesis of 1,10 -biimidazoles and dissociation into radicals. Z. Naturforsch. 1963, 18b, 406. T. Hayashi and K. Maeda, Storage of light energy by a solution of photochromatic-1,10 biorg(2,4,5-triphenylimidazyl) at low temperatures. Bull. Chem. Soc. Jpn. 1963, 36, 1052. J. Sonnenberg and D. M. White, Chemiluminescent and thermochemiluminescent lophine hydroperoxide. J. Am. Chem. Soc. 1964, 86, 568. E. H. White and M. J. C. Harding, The chemiluminescence of lophine and its derivatives.J. Am. Chem. Soc. 1964, 86, 5686. T. Hayashi, K. Maeda, and M. Morinaga, Mechanism of the photochromism and thermochromism of 2,20 ,4,40 ,5,50 -hexaphenyl-1,10 -biimidazole. Bull. Chem. Soc. Jpn. 1964, 37, 1563. T. Hayashi, K. Maeda, and M. Takeuchi, Kinetic study of the photochromism of 2,20 ,4,40 ,5,50 -hexaphenyl-1,10 -biimidazolyl with electron spin resonance. Bull. Chem. Soc. Jpn. 1964, 37, 1717. T. Hayashi, M. Kuyama, E. Takizawa, and M. Hata, The mechanism of dehydrogenation of 2,4,5-triphenyl-4,5-dihydroimidazole over solid catalysts. Bull. Chem. Soc. Jpn. 1964, 37, 1702. T. Hayashi and K. Maeda, Mechanism of the piezochromism of hexaphenylbiimidazole. Bull. Chem. Soc. Jpn. 1965, 38, 685. T. Hayashi, K. Maeda, and T. Kanaji, The mechanism of the photochromism and thermochromism of 2,20 ,4,40 ,5,50 -hexa-phenyl-1,10 -biimidazolyl in a solid state. Bull. Chem. Soc. Jpn. 1965, 38, 857. T. Hayashi and K. Maeda, Infrared absorption spectra of photochromic and piezochromic systems of the dimers of triarylimidazolyls. Bull. Chem. Soc. Jpn 1965, 38, 2202. H. Baumgaertel and H. Zimmermann, Triarylimidazolyls and triarylimidazole dyes. Chem. Ber. 1966, 99, 843. D. M. White and J. Sonnenberg, Oxidation of triarylimidazoles. Structures of the photochromic and piezochromic dimers of triarylimidazyl radicals. J. Am. Chem. Soc. 1966, 88, 3825. M. A. J. Wilks and M. R. Willis, Kinetics of the photochromic decay reaction of solutions of 2,20 ,4,40 ,5,50 -hexaphenyl-biimidazolyl. Nature 1966, 212, 500.
THE BEGINNING
143
T. Hayashi and K. Maeda, Mechanism of reversible photochemical color change of bis(triphenylimidazolyl) at low temperature. Bull. Chem. Soc. Jpn. 1967, 40, 2990. M. A. J Wilks and M. R. Willis, Kinetics of the photochromic decay reaction of 2,20 ,4,40 ,5,50 -hexaphenylbiimidazolyl. J. Chem. Soc. B 1968, 1526. M. A. J. Wilks and M. R. Willis, Interpretation of the high-resolution electron spin resonance spectrum of the 2,4,5-triphenylimidazolyl radical. J. Phys. Chem. 1968, 72, 4717. T. Hayashi and K. Maeda, Radical mechanism of the photochromism of bis (2,4,5-triphenylimidazolyl). Nippon Kagaku Zasshi 1969, 90, 325. A. L. Prokhoda and V. A. Krongauz, Photochromism of organic compounds. I. Kinetics and mechanism of photochromic reactions of 2,20 ,4,40 ,5,50 hexaphenyl-1,20 -biimidazole. Khim. Vys. Energ. 1969, 3, 495. B. S. Tanaseichuk, K. V. Stanovkina, A. N. Sunin, and L. G. Rezenova, Nitrogen-containing heterocyclic free radicals. VII. Use of diphenylpicrylhydrazine for studying the dissociation rate of dimers of triphenylimidazolyl radicals. Zh. Org. Khim. 1969, 5, 2054. H. Tanino, T. Kondo, K. Okada, and T. Goto, Chemiluminescence of organic compounds. I. Structures of three isomeric dimers of 2,4,5-triphenylimidazolyl. Bull. Chem. Soc. Jpn. 1972, 45, 1474. Soon after our initial experiments, we realized that we had something unusual, with potentially important opportunities. We were able to obtain some foreign patents on the biimidazoles and were concerned that these might tip off others in this field, for there was considerable activity, as can be seen from a listing of biimidazole work as reported in Chemical Abstracts. By 1964, we filed our first application on the leucodye oxidation. By then, we had a sufficient head start on this chemistry that we became less concerned about competitive activities. We had no equipment initially to do any photochemistry. We obtained a Cary recording spectrophotometer after a long wait, and our light sources inevitably were photographic flashguns, which I brought from home to work, and we used flashes of light, albeit of uncertain spectral distribution, to produce photochromic colors. In time, we were able to buy more powerful flashguns. Our initial attempts at making photochromic plastics for windows were not successful. In general, the radicals may have recombined, but they also underwent fatigue, i.e., they gave rise to compounds which no longer were photochromic. The Polychemicals Department, the business unit that marketed Butacite1 assigned Dr. Jean Paris to the program and, in time, Dr. Carleton Sperati. Both were enthusiastic supporters of the program and tried to provide us with media in which the biimidazoles, soon to be nicknamed HABIs were to be incorporated.
144
THE INVENTION OF DYLUX
Professor Dr. George Hammond, who then consulted for DuPont’s Film Department heard about our work and agreed to consult for us, too. He was the first photochemist of stature whom we sought as adviser. In time, Professors Ron Breslow and Nick Turro of Columbia University were also brought in. Hammond was then starting the series of books Advances in Photochemistry and suggested it would be good for me to write a literature survey about photochromism for the first volume of the series. I asked Paris if he would cooperate on this, and he agreed. I did a thorough literature survey, and Jean tried to put the theoretical aspects together. After about 6 months of this I began to question whether the type of photochromism that we required for windows was ever going to be possible with dye-like molecules. My reservations were based on the fatigue of the HABIs, the weak colors that were formed, and the likelihood that these colors would fade, rather than reverse themselves. I also suspected that the fact that no one had really succeeded in commercializing any photochromic materials was because the fatigue reaction was too considerable an obstacle. I said to some of my friends that the academic scientists who reported in this field were content with a few dozen reversals, which might be achieved by low quantum yields of formation, and so there was an ample reservoir of dimers, provided one did not push too hard. It became fairly obvious in our program that the development of automotive or architectural photochromic windows would be a lengthy task. Although we were a research group, we knew that sooner, rather than later, someone would ask the question: How soon can we make some money from all that? We thought that print advertising with photochromic inks might be an opportunity. We could spread solutions of HABIs onto paper, and with flashes of light, or even outdoor sunlight, produce blue or green coloration. We talked to Phil Wheeler, a printing expert in DuPont’s Philadelphia Printing Plant, and he gave us several ink vehicles to which we added HABIs. Initially we painted these onto paper and did produce photochromic images. These worked quite well in sunlight; but before long, we realized that people would not likely read magazines outdoors, and we either had to be able to generate color rapidly with indoor light, or forget this application. We were not yet familiar with sensitizing dyes that might have been useful in allowing our inks to work with indoor light! We analyzed the fatigue product of the photolysis of HABIs, and found these to be the parent triarylimidazoles (Scheme 4.4). At that time we became aware of photochromic sunglasses; Lord and Taylor’s (a New York City department store) offered some for around $16, and we bought a pair. But were not impressed with their performance. We thought that they employed some materials then available from American Cyanamid, which reported that they had photochromic materials for sale. We continued to dissolve HABIs in ever more purified solvents obtained from Paris and Sperati, and finally managed to fill some glass cells that actually lasted about 2 weeks,
145
1961—PERMANENT COLOR
N
N N
DIMER
N
N
.
N
IMIDAZOLYL RADICAL
N
NH
IMIDAZOLE
Scheme 4.4
when the photochrome was the dimer of 2-(p-methoxyphenyl)-4,5-diphenylimidazole. I took one such cell through the Holland Tunnel on the way to New York City, and it did discolor when in the tunnel and revert to the blue color on the sunny New York side. The solvent was a cyclobutane derivative especially purified for us. We analyzed the fatigue product of the photolysis of HABIs and found these to be the parent triarylimidazoles (Scheme 4.4). Could we have formulated a mixture in which air-oxidation of the imidazole could regenerate the dimer? Surprisingly, that may work. In order to establish the possible stability of the biimidazoles in a ‘‘window’’ environment, we poured a benzene solution of the dimer of 2-(p-methoxyphenyl)-4,5-diphenylimidazole into a hollow glass brick, and put in a rubber stopper to ‘‘seal’’ the system. We positioned the glass brick in a window, and watched it turn blue during the day and less blue when the sun set. It worked for several months; but by then, we had turned our attention to photo-oxidation.
VI.
1961—PERMANENT COLOR
Cescon and I had wondered if we could make polymeric HABIs. One proposed way was to make a biimidazole in which terephthaldehyde was used as starting material, and Jenkins provided us with this material. When Cescon oxidized this under his standard conditions, an intense bright blue color resulted (Scheme 4.5). I was impressed by it, but it would not reverse. Cescon said perhaps we should make permanent color from HABIs photochemically and not try to
146 N
H N
N N
N H
N
N
N
H N
N
Scheme 4.5
Biimidazole from terephthaldehyde
N H
N
N N
Blue-colored oxidation product
N N
147
1961—PERMANENT COLOR
achieve reversibility. I told him that perhaps we could oxidize leucodyes instead, remembering some work done earlier at DuPont, in which a series of novel leucodyes were made for a program related to spirit duplication, Until the development of electrophotography, a facile means of making multiple copies of a document was to type on a waxy stencil material, which permitted ink to penetrate through the typed and ruptured areas—this was called mimeographing. Another, less expensive process was called spirit duplication. Here one typed on a waxy sheet of paper, and the typing ruptured a coating, which was applied over a layer containing triarylmethane dyes, usually Crystal Violet. The rupture of the wax layer permitted transfer of the dye to a sheet of high holdout paper temporarily wetted with ethanol (spirits of methanol). With some of these, as many as 100 copies could be obtained. The process could also be repeated with other colors, and so a multicolored copy sheet could be obtained at relatively low cost. It is safe to say that most examinations in schools and colleges up to the 1950s were written to questions, which were printed out on Crystal Violet image papers. Major manufacturers of these products were the Ditto Corp. and Columbia Carbon and Ribbon. DuPont’s Organic Chemicals Department manufactured large quantities of Crystal Violet dye—many hundreds of thousands of pounds per year—it was the largest volume dye at one time. Someone in Orchem research management had conceived an idea for a better product. The weakness of spirit duplication systems was that the wax sheets inevitably leaked color onto the hands of the people who handled them. Secretaries hated them, as the intense colors in the wax matrix transferred to clothing, causing severe staining. Why not have a leucodye, i.e., a colorless precursor of a dye present in the wax layer, and moisten the receiver sheet with an alcoholic solution of an oxidant? For a variety of reasons, the preferred oxidant was chloranil (tetrachlorobenzophenone), an inexpensive, colorless, and effective material (Scheme 4.6). Unfortunately, this compound also gave rise to hydrochloric acid
(C 2 H5 )2 N
CH3
(C 2 H5 )2 N
N(C 2 H5 )2
CH3
CH3
C 2 H5
CH3 CHLORANIL
N
CH2
C 2 H5
CH3 N(C 2 H5 )2
(C 2 H5 )2 N DYE
LEUCODYE TLA-454
Scheme 4.6
148
THE INVENTION OF DYLUX
and thus caused corrosion of the metal parts of the duplicating machinery. Were other oxidants tried? I do not know. Just the same, the program was terminated; but Walter Balon, who had developed the leucodye tris(p-diethylamino-o-tolyl)methane, retained samples in his laboratory for several years. This dye had the code name TLA-454, which simply indicated that it was the 454th dye sample that had been submitted to the Technical Laboratory at the Chambers Works, which was in charge of evaluating materials that were considered for commercialization. I remembered Balon’s TLA-454 and thought it would be worth trying as a leucodye in combination with the radicals generated by photolysis of triarylimidazoles. To our great satisfaction, a solution containing HABI, and TLA-454, p-toluenesulfonic acid in acetone turned an intense blue in sunlight. We made fresh solution as above, dipped filter paper in it, let it dry, and produced an intense color with a flash of light. This ultimately lead to a number of patents, typical of which was U.S. Patent 3,585,038. Our goal was to cover all possible aspects of HABI-related photo-oxidation. During the period from 1964 to 1972, many leucodye oxidation patents issued to Cescon and his co-workers. We called this technology UVI–Ultraviolet Imaging. Photosensitive composition. Cescon, Lawrence. (du Pont de Nemours, E. I., and Co.) U.S. 3,445,233 (Cl. 6-90; G 03c), 20 May 1969, Appl. 16 Apr 1965; 9 pp. A simple, rapid, and dry process for the production of colored visible images is described. A light-sensitive compn. consisting of essentially colorless oxidizable N-contg. org. color-generator (I), a photooxidant (II), such as CBr4, a tetraarylhydrazine, etc., and a light-sensitive aromatic diazonium compd. (III) such as p-anilinobenzenediazonium sulfate, is irradiated with light of wavelength W1 to form the image. Deactivation achieved when irradiated with light of wavelength W2. A neg. image is formed when the compn. is irradiated in reverse order. Three types of I are: (a) leuco dyes, preferably aminotriarylmethane; (b) org. amines; (c) aromatic diamines with a coupling agent. II when mixed with I and irradiated with W1, of 200–420 nm oxidizes I to said-color permanently. Initiator and acceptor types of mechanisms are shown. III is added to compn. in order that the photosensitive compn. produced by I and II may be deactivated by exposure to light of wavelength W2, 200–550 nm, thus producing a reducing agent that is a 1,2- or 1,4-aminohydroxy- or -dihydroxybenzene or -naphthalene. Preferred ratios of II to I range from 1:1 to 2:1, and of III to II 0.2:01.0–10:1. Hexaarylbiimidazole oxidation systems. Cescon Lawrence A.; Dessauer, Rolf (du Pont de Nemours, E. I., and Co.) U.S. 3,585,038 (Cl. 96-90; G 03c), 15 Jun 1971, Appl. Apr 29, 1964–13 Nov 1967; 8 pp. Systems useful in photog. thermog., pattern layout, etc., are composed of a 2,20 ,4,40 ,5,50 -hexaarylbiimidazole (I) component and an oxidizable component. Methods for activating the I compd. are heat, pressure, light, electron beam which form a free radical that then oxidizes the oxidizable component. The aryl group for I may be Ph, phenyl, naphthyl, furyl, or thienyl groups having an ortho substituent of F, Cl, Br, lower alkyl, or lower alkoxy group. The oxidizable components are selected from the following classes of
1961—PERMANENT COLOR
149
compds.: p-arylenedi-tertiary-amines, p-phenylene diamines, hydrazones, o,o0 -disubstituted phenols or org. sulfhydryl compds. The 2 components are mixed by dissolving in a solvent or by mech. blending and can be used in a polymeric binder, coated on a substrate or on a film. Leuco dye photocopy system. Cescon, Lawrence A. (du Pont de Nemours, E. I., and Co.) Ger. l,772,534 (Cl. G 03c), 02 Mar 1972, Appl. P 17 72 534.7-51, 31 May 1968, 8 pp. A leuco dye photocopy system is based on photooxidn. of an aminotriarylmethane leuco dye by 1-2 moles of a hexaarylbisimidazole to the dye in uv (250–370 nm) radiation (U.K. 1,047,569; CA 66:50700m) and the inactivation of the oxidant at longer wavelengths by a redox combination of 1,6- or 1,8-pyrenequinone and (or) 9,10-phenanthrenequinone with 1–40 moles of a Cl–4 alkyl nitrilotriacetate (or -propionate) as reductant. When irradiated at 380–500 nm the quinone is reduced to a hydroquinone which then reacts with the photooxidant to form a colorless compd., thus inactivating the light-sensitive system. Either wavelength range may be used for the imagewise (1st) exposure. Thus, a paper support coated with an Me2CO soln. contg. polypropylene glycol, cellulose acetate butyrate, pyrenequinone, 2,20 -bis(o-chlorophenyl)-4,40 ,5,50 -tetraphenylbisimidazole, tris(4diethylamino-o-tolyl)-methane, tri-Me-3,30 ,300 -nitrilotripropionate, and p-MeC6H4SO3H, was imagewise exposed to radiation at 250–390 nm and then re-exposed to radiation >390 nm to give a clear sharp image on a white background. Further irradn. showed no visible effect on the background or the image. Storage-stable, light-sensitive compositions comprising an aminotriarylmethane and an organic photooxidant. Cescon, Lawrence A. (du Pont de Nemours, E. I., and Co.) U.S. 3,598,592 (Cl. 96-85; G 03c), 10 Aug 1971, Appl. 29 Apr 1964–07 Nov 1967; 7 pp. Mixts. of aminotriaryl-methane compds., such as tris-(4-diethylaminoo-tolyl)methane (I), and photooxidants, such as Cl2BrCCBrCl2, form light-sensitive compns. that develop color rapidly when exposed to light of certain wavelengths, but will not undergo color formation under ordinary darkroom conditions. Thus, 2.6 in.2 of Whatman No. 1 filter paper was treated with a soln. contg. 0.0125 M I and 0.025 M Cl2BrCCBrCl2 in 50/50 by vol. DMF-C6H6 and dried at room temp. to give per in.2 of dried paper 0.5 mg I and 0.8 mg Cl2BrCCBrCl2. After exposure of a portion of the sheet to a 275 W sun lamp, the paper was stored in a darkroom for 2 weeks. Upon removal and examn., the paper showed no color change in the unexposed areas. Hexaarylbisimidazole oxidation system. Cescon, Lawrence A.; Dessauer, Rolf (du Pont de Nemours, E. I., and Co.) Ger. Offen. 2,028,124 (Cl. CO9k, C 07), 16 Dec 1971, Appl. 08 Jun 1970; 42 pp. A free radical imaging system contains a bisimidazole I (R ¼ H, Cl ) and an oxidizable compd. with a formal oxidn. potential of 1.35 V vs. SCE. The oxidizable compd. is preferably p-(Me2N)2C6H4, PhNMe2, 2,6-(Me3C)2C6H30H, 2,6-(MeO)2C6H30H, hydroquinone, 3-methyl-2benzothiazolone hydrazone, or 2-thenoylacetonitrile. Flash photolysis of 0.01 104 II p-(Me2N)2C6H4.2HCl and 2 104 M I (R ¼ Cl) caused the formation of Wurster’s blue with a rate const. of 7 107 l./mole sec. The p-(Me2N)2C6H4 2HCI had a formal oxidn. potential of 0.96 V. Paper soaked in p-(Me2N)2C6H4 0.272, PhOH 0.094, and I (R ¼ Cl) 0.66) g, and dried turned blue on exposure to sunlight.
150
THE INVENTION OF DYLUX
N
N
R 2
I Hexaarylbiimidazole oxidation systems. Cescon Lawrence A.; Dessauer, Rolf (du Pont de Nemours, E. I., and Co.) Brit. 1,271,177 (Cl. G 03c, C 07cd, C 09b), 19 Apr 1972 Appl. 26,161, 70, 29 May 1970; 13 pp. Color copies are obtained with an imaging system comprised of a heat-, pressure-, light-, or electron beam-sensitive hexaarylbiimidazole and an oxidizable compd. having an oxidn. product that reacts with an admixed coupler to form a dye image. Upon activation the biimidazole forms the free radical triarylimidazolyl, which oxidizes the color producing compd. such as N-acylhydrazone oxidizable to diazonium compds., in combination with a coupler to form the dye. Thus, to 50 ml of a soln. contg. 20 parts by vol. of DMF and 8 parts MeOH are added 0.212 g of N,N-dimethyl-p-phenylene diamine, 0.094 g of PhOH, and 0.66 g of 2,20 -bis(o-chlorophenyl)-4,40 ,5,50 -tetraphenylbiimidazole. Unsized paper is impregnated with this soln. and dried by ir heating, the operation being carried out in dim light. Exposure of this treated paper to a sunlamp turns the paper blue.
Cescon and I repeated the experiment for George Coraor, who was duly impressed by it, as were others in our group. At that time, George reported to Dr. Alfred C. Haven, who initially was somewhat slow in sharing our enthusiasm for photo-oxidation, as we called our process, because he had sold his management on the fact that we were making photochromic materials that were to be reversible. This was one of many instances in my career with DuPont where I found that new applications, derived from programs that were expected to deliver another specified application, were not enthusiastically received by management. It was simply that to get a program started it had to be sold to too many people, who would then object to deviations from a selected path. In time, however, the idea of generating dyes with light was augmented with a program to generate dyes electrolytically. Still, Coraor said he saw that we had something more marketable than photochromic compounds, although we had a limited idea as to what to do with the light-induced color change. In any event, photo-oxidation did not lead to reversible, reusable materials, and conceptually, this would lead to a larger market than a product that was reversible. If only one could identify a market. A few weeks later, we were to have a research review, a session at which scientists
1961—PERMANENT COLOR
151
presented the results of their work to the entire research management of Orchem. This then included Ernsberger, the director of research; Dr. G. M. Whitman, the assistant director of research; Haven; and a few others. A few days before the review, on a ski trip, I met Tim Chilton, who was in the company’s Development Department. Tim’s father was a highly respected engineer, who was well known throughout the company. Tim had just completed a survey on office copying for the company and appeared knowledgeable in the reprographics field. I had told him what we had come up with, and he asked if he could come to our research review to learn more. I went to his home for dinner the night before, and he said he could probably sell this chemistry to our management. When the review began, Ernsberger asked Tim who he was and what he was doing there. Tim presented himself very positively, described his background and allowed that we might be on the verge of a great discovery, if it were properly pursued. He said that he knew of nothing like it, which may have been of ignorance, but it sounded as though he knew what he was talking about. After that, everyone wanted to expose filter papers to form colors with a flash of light; our color repertory was somewhat enhanced with leuco dyes that gave green, blue black and violet colors. Enthusiasm really ran high, and this was the only review in my career at which management people stayed around well after quitting time, suggesting what we should do. It looked like photooxidation had rescued HABI chemistry just in time! Even though our experimental work was crude, Ernsberger said that we should show what we had to the Photo Products Department, who after all were in charge of the company’s photographic business and were considered to have expertise in the field. Coraor and I went to see their director of research, Dr. Frank Signaigo, who in turn had invited a few of his subordinates to our show. They were not very impressed, explaining to us that the color-forming process requiring contact with a photographic flashgun was photographically slow, and that until we had good black-and-white images, we had little of interest to them. By that time we had made stencils and even obtained some photographic negatives, which we used to modulate the light from the flashgun. The photographic negatives inevitably fried when exposed with a flashgun, until we learned that the intense heat absorbed by the silver coating could be deflected by employing a cobalt blue or similar filter. We then made more respectable images of the photographic negatives, albeit, still onto filter paper.
A. What Do You Do with This Stuff? After our first research review, where Cescon and I demonstrated photo-oxidation with HABIs, Chilton was assigned to the program. He and I speculated for many hours about possible applications. Obviously, if we could apply this technology
152
THE INVENTION OF DYLUX
to replace spirit duplication or mimeographing, there would be a ready fit. Coraor asked several engineers to design a glass drum that could be used in connection with a modified duplicator, but it never worked out. We thought of toys, but that seemed too frivolous for a company like DuPont. Somewhere in the early 1960s I had had an opportunity to partake in a very interesting visit to Meredith Publishing Company in Des Moines, Iowa. Meredith at that time invited scientists from major companies to visit them for a week to learn about all aspects of putting the magazine Better Homes and Gardens together. Of course, they also published books. Their idea was that researchers in industry probably knew little about the printing industry and might actually have materials that could affect their business. So a group of us from the Pigments, Textile Fibers, Organic Chemicals, Photo Products, and other departments were invited for a strenuous week of learning all about printing. It was fascinating and provided me with some comprehension of the opportunities in that field. I also recalled those years earlier, before I graduated from Flushing High School, we had to select pictures from proof prints submitted by the school’s photographer. These proofs were made on a paper that would in time fade out, as the photographer wanted us to buy real prints—not use these proofs. So it occurred to me that we might be able to adapt our blue-and-white, still-nottoo-permanent images to this yearbook business, perhaps to make proofs. Also, I saw proofs made at Meredith. So on one of our occasional visits to the Photo Products Department’s R&D management, I suggested that we might find an application in this area. We were told that the quality of our images was not adequate, their permanence had not been established, and that Photo Products, which was primarily in the silver-photographic business, was not interested in getting into an application that would compete with ferrocyanide paper, a product that cost pennies per square foot. Of course, years later, proofing was recognized as a major opportunity. Another thought then was that we would cooperate with the Polaroid Company, who had found a way of making single black-and-white prints and negatives by an instant process, and we thought that one could make attractive copies by contact printing from the latter onto our paper. Years later, we did discuss this with Polaroid, but they were not very interested. I guess that they felt that if one wanted two or more prints, one was to take two or more pictures.
B. What Is Needed? We realized, of course, that these images built up background color all too quickly, as there still was an ultraviolet light component in daylight or even roomlight. As we knew that radicals could be destroyed with hydroquinone, we ‘‘fixed’’ our images by dipping them into an ethereal solution of hydroquinone; more
153
1961—PERMANENT COLOR
conventional solvents were avoided to eliminate bleeding of the photogenerated dye loosely held in the filter paper structure. Naturally, we were not in this for the science of it all, and we had to conceive of an application that fit into Orchem’s business activities. It seemed to me that as the department sold huge quantities of basic dyes for spirit duplication, we might resurrect the scheme in which chloranil corroded the equipment, and use ultraviolet light to effect the color formation. Also, we obtained mimeograph stencil materials, which provided us with good masters. In addition, Photo Products had developed Cronapaque1, an opaque pressure-coalescible film, used to duplicate letterpress plates; and this looked like another useful stencil material. Our immediate need was to develop a more elegant way of stabilizing the images that were made by photo-oxidation of leucodyes. One obvious way was to truncate the absorption spectrum of HABIs, so that the activation occurred only in the ultraviolet region—at wavelengths <380 nm. Despite a significant synthetic effort, this was not accomplished. The spectra of a number of HABIs are shown in Figure 4.1. Another route was to consider thermal generation of inhibitors. Dr. Phil Manos, a talented chemist in Coraor’s group, who was working on another program, prepared several ether derivatives (as in Scheme 4.2) of hydroquinones, e.g., of di-t-butylhydroquinone, which thermally gave rise to reducing agents. This was
o-Cl HABI o-Methyl HABI TCTM HABI TCDM HABI
l/mo l/cm
8000
6000
4000
2000
0 300
350
400 Nano meters
450
Figure 4.1. The spectra of several HABIs.
500
154
THE INVENTION OF DYLUX
O
O C(CH 3 )3
(CH3 )3 C O
O
Scheme 4.7
described in U.S. Patent 3,390,995. Manos was typical of chemists in our building; when someone knew something that could help another group, he did it! Light-sensitive composition consisting of organic color-generator, photooxidant and organic thermally activatable reducing agent progenitor. Manos, Philip (DuPont de Nemours E. I., and Co.) U.S. Patent 3,390,995 (Cl. 96-48), 02 Jul 1968, Appl. 29 Apr 1964; 10 pp. A photosensitive compn. contains in intimate assocn., essentially nonhygroscopic interreactant pro genitors of intensely colored org. color bodies. The progenitors are activated by radiation of wavelength 2000– 4200 A. which can be applied in a graphic pattern. They are permanently deactivated when heated to 80-160 . The compn. consists of (1) an essentially colorless, oxidizable N-contg., org. color-generator which, when contained in the photosensitive compn., is stable to oxidn. by atm. O under normal room and storage conditions but which can be oxidized to an intensely colored species; (2) a photo-oxidant which, when mixed with the oxidizable color generator and irradiated with 2000– 4200 A. radiation, will oxidize the color generator to an intensely colored species as an essential part of the compn.; (3) an org. progenitor of a reducing agent which, when mixed with components (1) and (2) but prior to the heat treatment it does not function as a reducing agent. It is chem. changed by the heat treatment to produce a reducing agent, which prevents the photo-oxidn. A preferred compn. contains an amino triarylmethane with an orthosubstituent in at least 2 of the aryl groups as the org. color generator, a hexaarylbiimidazole as photo oxidant, and an acetal of hydroquinone as the org. progenitor of reducing agent. Thus, photosensitive paper is prepd. by dipping unsized paper in a 4:1 (by vol.) methanol-N,N-dimethylformamide soln. contg. (by wt.) 0.4% tris(4-diethylamino-o-tolyl)methane-3HCl and 0.4% 2,20 -bis(o-chlorophenyl)-4,40 ,5,50 -tetraphenylbiimidazole followed by drying under an ir lamp. The paper is then dipped into a 0.5% benzene soln. of a progenitor of a reducing agent and again dried. The paper is folded so that part of it is exposed for 10 sec. to a 275-w. sun lamp at a distance of 10 in. whereupon an intense blue color forms. The whole paper is then heated for 5 sec. between the plates of a hydraulic press at 125 and the unirradiated portion of the paper exposed to the sun lamp for 10 sec. In the absence of an agent other than the color-generator and the photooxidant no deactivation occurs with heating. When
1961—PERMANENT COLOR
155
a reducing agent deactivates the photosensitive compn. prior to heating, no image is formed on light exposure. With acetals of hydroquinone, substituted hydroquinones, and other phenolic compds., however, the photosensitive compn. gives sharp images on uv exposure and is readily deactivated by moderate heating to preserve the initial image during subsequent light exposure.
By this time, Cescon had decided that to make good images we needed to dissolve the imaging chemistry in binders; and since we had experience with filter papers, which did not degrade the HABIs, we went to Eastman’s cellulosic esters as vehicles for our chemistry. With Manos’s chemicals, we could now make coatings on paper, that could be exposed with UV light and stabilized by heating. Cescon discovered that the reaction could be speeded up when we included plasticizers in the formulations; and before long, we were making attractive coatings on small pieces of paper. We obviously needed someone who knew something about paper. I had met Dr. Albert Deutsch many years earlier, when he was a lab technician at Sloan Kettering Institute for a friend of mine from graduate school days. After he had obtained a Ph.D., he was hired for the company’s Paper Laboratory and seemed knowledgeable in areas in which Cescon, Silversmith, and I were ignorant. Al did help some, but we made too little progress to suit our management; and he was then replaced with another chemist, David Springs. Slowly, we learned about coating rods, drying, storage, and the lot. Al apparently retained his enthusiasm for HABI chemistry and obtained several patents, that incorporated HABI-chemistry, while working for other companies. A. S. Deutsch, R. Dennis, and R. Gumbinner, Ultraviolet Curing Printing Inks Having Improved Shelf Life, U.S. Pat. 4,129,486, Dec. 12, 1978. A. S. Deutsch, L. D. David, and D. B. West, Chemical Imaging of a Lithographic Printing Plate, U.S. Pat. 6,691,618, Feb. 17, 2004.
C. Kay Looney—A Creative Spectroscopist Earlier in my career as a dye chemist, I worked with Dr. Monroe Sadler of Central Research, who introduced me to Dr. Catherine Looney, a spectroscopist, who had obtained her doctorate working on the spectra of triphenylmethane dyes. Looney was very helpful to me with reflectance measurements before she retired to have two sons. It was the beginning of a very long association, which proved invaluable to our program. Soon after we started the program on photo-oxidation of leucodyes, I mentioned to Coraor that I knew of a retired spectroscopist, who might be available to consult with us. Coraor thought this to be a great idea, since we might obtain
156
THE INVENTION OF DYLUX
someone with experience in modifying dye structures to give us better blackforming leucodyes and also to bring a new discipline, spectroscopy, to a group that was trying to develop novel photochemistry. In our usual excitement for coming up with a good solution to a problem, we arranged for Kay’s contract to be written hastily and had ignored how to deal with inventions that she might make or participate in. Several years later, this resulted in a hilarious episode. A patent, in which she was involved, was to be filed, and as no compensation was ascribed to the contract, which had to be filed urgently, a hurried secretary sent Kay a document with a blank space for her perception of the value of the patent. When it became known that Kay could have inserted a large sum of money into the blank, a senior manager was asked to retrieve the patent from Kay at her home, posthaste. Based on synthetic work that Cescon, Jenkins, Balon and I did, Looney calculated some structural changes for triphenylmethane dyes, which resulted in two patents. L. A. Cescon, R. Dessauer, and C. E. Looney, Selectively Substituted Methane Leuco Dyes, U.S. Pat. 3,423,427, Jan. 21, 1969. L. A. Cescon, R. Dessauer, and C. E. Looney, Triphenyl Methane Derivatives, U.S. Pat. 3,440,379, Apr. 22, 1969. More important, Looney suggested an alternate way of stabilizing our images. She proposed that we embed the imaging chemistry in a thermoplastic binder, which could be thermally softened to permit migration of leucodyes and radicals, but when cold, lock the chemicals in place. Cescon quickly reduced this to practice, and we were able to perform a remarkable trick—image a coating with light and obtain no color change until heat was subsequently applied. The stability of imidazolyl radicals made all this possible. This concept was the basis of U.S. Patent 3,615,481. Leuco dye/hexaarylbiimidazole thermally activated imaging process. Looney, Catharine E. (DuPont de Nemours, E. I., and Co.) U.S. Patent 3,615,481 (Cl. 96/48; G 03c), 26 1971 Appl. 19 May 1969; 13 pp. A thermog. imaging system consists of a leuco dye and 2,20 -bis(o-chlorophenyl)-4,40 ,5,50 -tetrakis(m-methoxyphenyl)biimidazole (I) (mostly 1,20 -isomer in a polymer with a glass transition temp. above room temp. but below the decompn. temp. of the 1st 2 components). The system is heated to fluidize it before imaging. Thus, 0.9 ml 0.1M [2,4-Me(Et2N)C6H3]3CH, 0.6 ml 0.3M p-MeC6H4SO3H, and 1.8 ml. 0.05M I, all in Me2CO, were added to 750 mg poly(me methacrylate) with a glass transition temp. 100 in excess Me2CO. The mixt. was concd. to 15 ml, poured on a 5 thick Mylar film, and the residual Me2CO evapd. off give a substantially colorless film. After heating to 105 , the film was exposed to a Xe flash lamp through a Ag halide emulsion negative to give a blue pos. image. At room temp., the system was inert to the Xe flash.
1961—PERMANENT COLOR
157
Still, we felt that systems involving thermal development or activation were not what we really wanted to develop. A fortunate series of events solved that problem.
D.
Radiation Physics Laboratory and Al MacLachlan
I belonged to the Wilmington Ski Club, which at that time had about 300 members, mostly single people, who found this a great way to meet others. We ran weekly car trips to the Vermont ski areas to Mount Snow, Killington, Bromley, and similar places. One weekend in 1961 we scheduled a bus trip, but as bad luck would have it, we had to cancel it because of relatively warm weather in the Catskill area we had intended to visit. Dr. Rudy Pariser, a DuPont chemist, neighbor, and friend decided to use the weekend instead to throw a party for some of the disappointed skiers. He invited me as well as a nonskier, Dr. Richard G. Bennett. I had not met him before, but we started chatting—naturally about work. Dick said that he was a research associate at the Radiation Physics Laboratory (RPL), close to the Building 336 where my group was located. Dick had not heard of what we were doing and thought it might be a good idea if I came to give a talk about our work to his group. It seemed the thing to do, and on the following Monday I gave a presentation there, with photochromic and photoxidant solutions and flashgun, etc. The effect was remarkable. The entire audience was excited about this work, and Dick and one of his colleagues, Dr. Alexander MacLachlan, said they could and would help us with our research. It turned out that RPL hade been set up in the early 1950s at the suggestion of Victor F. Hanson, an engineer in the DuPont Engineering Department, who was well connected with the top management, especially DuPont president Crawford Greenewalt. Hanson thought that it would be useful for DuPont to have a research institute devoted to radiation chemistry, although he considered that to be the study of atomic radiation, rather than solar radiation. The lab was very well equipped and had measurement devices the rest of us could only dream of. The dozen or so scientists there were mostly of very high caliber, having been recruited for something really special. Their problem was that the effect of nuclear and atomic radiation on organic materials did not yield any particularly useful results. MacLachlan especially seemed very excited about working with us, and I invited him to visit with Coraor, which he did promptly. At that time, RPL received funding from operating departments, and MacLachlan suggested that if Orchem would arrange funding of some programs at RPL, he could request being assigned to them, and his background in the study of kinetics would help us greatly. Coraor was enthusiastic about this, because we desperately needed quantitative measurements of the type that our group was
158
THE INVENTION OF DYLUX
incapable of doing. Joint programs were rapidly arranged; and for 2 years, a great relationship between Orchem and RPL existed. Hansen, despite being the laboratory’s director, was a very down to earth person, who really felt that his role was to help all the people in his lab, which included building equipment. In time, he built equipment for us, which allowed us to measure rates of recombination of HABI radicals. Dr. Roland Riehm, another scientist we supported, examined alternate photo-oxidants, but found nothing that worked better than HABI. He had been interested in horseradish peroxidase, and studied this in combination with our leucodyes. We hired Dr. Reid Kellogg, a spectroscopist, and assigned him to RPL, where he conducted kinetic studies. MacLachlan studied the oxidation of triphenylmethane leucodyes, which we were synthesizing in an attempt to develop different colors, as well as biimidazole photolysis. RPL had flash photolysis equipment. RPL had contacts with scientists from all over the world. Professor George Porter, who later won a Nobel Prize in this area, was one of their guests, and we were able to consult with him. So was Professor T. Fo¨ rster, with whom we discussed our work. It was a great relationship and a great time, scientifically. MacLachlan soon afterward demonstrated a composition in which a thermal system was employed to produce stable images. Dry silver-free photographic process. MacLachlan, Alexander (DuPont de Nemours E. I., and Co.) U.S. 3,383,212 (Cl. 96-48), 14 May 1968, Appl. 29 Apr 1964; 10 pp. A direct neg. stencil color image with heat- and light-stable backgrounds is obtained by irradiating at 2000–5500 A. a coating compn. contg. an oxidn. activatable org. color former, a thermally activatable org. color former, a thermally activatable active oxidant, and a redox couple, then heating it at 90– 150 . If the heating step is 1st, a pos. image is obtained. Unlike other dry photochem. processes these images are stable to further activation. The org. color former can be a leuco dye an org. amine as in U.S. 3,042,517, a 2-benzothiazoline hydrazone, or an aromatic diamine and a compd. capable of producing an indoaniline or azomethine dye. The thermally activatable oxidant can be a diaryl or diacyl peroxide, hydroperoxide, azo compd., N-chlorinated compd., or a biimidazole. The oxidant of the redox couple can be a quinone, substituted lH-benzonaphthalen-1one, phenazine, acridine, phenoxazine, quinoline, phenanthroline, or an anil, and the reductant can be an ether, alc., ester, acetal, aldehyde, amide, a compd. contg. an allylic or benzylic H, Ph3SnH, dioctyl phosphite, or Ph3SiH. Thus, a piece of filter paper impregnated with a soln. of 890 mg. [4,2-H2N(Me)C6H3]3CH.3HCl, 210 mg. 9,10-phenanthraquinone, and 240 mg. Bz2O2 in 50 ml. EtOH, and 50 ml. Carbowax 600 is air dried, irradiated by 2 flashes of a Hico-Lite Electronic Flash Model K, and heated 10 sec. between metal plates of a hydraulic press at 125 to give a neg. blue image. No color develops in the exposed areas by subsequent heating. Stable coatings of these materials in which the color former and the oxidant are physically sepd. can be prepd. by known thermography methods or by microencapsulation techniques.
1961—PERMANENT COLOR
159
One day in 1962, MacLachlan called me and said he had an idea, which he had demonstrated, but he did not know how stable the solution he had prepared was—so hurry. I rushed to his lab, and he showed me that he had dipped a filter paper in a solution that contained a biimidazole, a triphenylmethane leucodye, phenanthrenequinone, and Carbowax. He exposed it with a flashgun through a UV-blocking filter and a stencil and was able to bleach out the yellow color, to give a near white pattern. When he subsequently exposed the paper through a cobalt blue glass, the previously unexposed areas turned blue. It looked like he had conceived and demonstrated a photofix reaction! We repeated the experiment for Coraor, who immediately saw its significance. We had found a way of incorporating the entire imaging/fixing chemistry into a single coating, which obviated the need for a chemical after-treatment, or thermal fixing. We definitely had something new! MacLachlan continued to investigate the invention, and Dr. Charles Yembrick and Dr. Peter Strilko, who had joined our group recently, were assigned to optimize MacLachlan’s system. They did a superb job; and before long we had developed coatings, which functioned reliably. Of course, it was fortunate that we had access to leucodyes that were not oxidized by phenanthrenequinone, and that the spectra of the various components had minimum overlap in the critical spectral regions required for color formation and photodeactivation. U.S. Patent 3,390,996 Photosensitive composition comprising an organic nitrogen-containing color-generator, a photo-oxidant and a redox couple: A. MacLachlan, (1968) (DuPont de Nemours E. I., and Co.) (Cl. 96-48), A composition which forms color when irradiated with light of one wavelength and becomes relatively insensitive to that light when irradiated with light of a different wavelength. The composition comprises (a) an organic nitrogen containing colorgenerator, such as a leuco dye. (b) a photooxidant, such as a HABI, which upon being irradiated oxidizes the color-generator to its colored form, (c) a redox couple of (1) a reductant, and (2) an oxidant which when activated by light reacts with the photo-oxidant to deactivate it. The composition can be coated on a substrate such as plastic, paper or metal. Leuco triarylmethane/hexaarylbiimidazole color forming system containing a deactivator. Strilko, Peter S. (DuPont de Nemours, E. I., and Co.) U.S. Patent 3,579,342 (Cl.96-90); (G 03c), 18 May 1971, Appl. 27 Jun 1968; 11 pp. A photosensitive compn. is claimed, which contains a photoactivatable color forming system of a leuco triarylmethane and a hexaarylbiimidazole a and a system for photodeactivating the color-forming system comprising a mixt. contg. selected quinones and an aliphatic ether. Thus Me2CO 46, cellulose acetate butyrate 6.0, p-nonylphenolethylene oxide adduct 4.0, tris(4-diethylaminoethyltolyl)methane 0.30, p-toluenesulfonic acid monohydrate 0.344, (o-chlorophenyl)-4,5-bis(m-methoxyphenyl)imidazolyl dimer 0.624, 1,6- and 1,8-pyrenequinones (1 to 1 mixt.) 0.0371, and 9,10-phenanthrenequinone 0.0664 part were made up into a soln. and coated on
160
THE INVENTION OF DYLUX
paper. This coating has marked deactivatic capability to resist the development of unwanted color under room light, daylight and sunlight.
A number of chemists had joined our group and were given specific assignments to optimize the total system. In this period, DuPont had an active hiring program, and during interviews that preceded the hiring of more scientists, we inevitably showed them novel chemistry that had a lot of intellectual appeal. Almost everybody who saw what we had wanted to work on it. A number of new employees thus came to this program. George Coraor, who wanted people to spend time to obtain background in this area, was a patient man, too patient for his immediate supervision, who wanted products as soon as possible. Coraor was interested in establishing a patent position that was sufficiently comprehensive so no one else could wipe out our eventual commercial offerings. This, of course, took time and interfered with product development! Several who joined the program were contributors; others did not have the requisite background and in spite of enthusiasms were of little value. Dr. Robert L. Cohen was one of the contributors; he had the job of selecting a HABI that was most effective with the commercially available filters that would be useful with our dual-response imaging system. He found the best spectral match was with CDM-HABI. The question was whether we could effectively shift the absorption spectrum of HABI by synthetic means or by addition of sensitizing dyes. Both approaches showed promise. Ultimately, we found the synthetic approach was more effective; but many years later, when we sought to form images with near-infrared radiation, we had to depend on sensitizing dyes to shift the absorption. Photoactivatible hexaarylbiimidazole-coumarin compositions. James, Daniel S.; Witterholt, Vincent G. (du Pont de Nemours, E. I., and Co.) U.S. 3,533,797 (Cl. 96-90, G 03g) 13 Oct 1970. Appl. 13 Mar 1967; 12 pp. Photoactivatible compns. consisting of a hexaarylbiimidazole, which absorbs radiation in the uv at short wavelengths, and a substituted coumarin, which absorbs in the uv at relatively long wavelengths are disclosed. These compns. can also be combined with a leuco dye or a leuco dye and fixing component. The substituted coumarins can transfer light energy absorbed at longer wavelengths to the biimidazole. This enhances the spectral sensitivity of the biimidazole at wavelengths where they normally absorb poorly. These compns. are useful for light-activated colored image formatlon, sun screens, and dry, non-Ag photog. processes. For developing colored images by exposure to light, biimidazole coumarin-leuco dye are used in a molar ratio of 1/0.1–2/0.1–10. Visible light-sensitive phototropic compositions comprising a hexaarylbiimidazole and a carbocyanine dye. Cohen, Robert L. (du Pont) U.S. 3,554,753 (Cl. 96 90; C 03c), 22 Jan 1971, Appl. 20 Jul 1967, 6 pp. Compns. contg. 2,20 ,4,40 ,5,50 -hexaarylbiimidazoles, tris(N,N-diethyl-amino-o-tolyl)methane, and
161
HABI N
Cl
(C 2 H5)2N
LEUCODYE SALT
N(C 2 H5 )2
N(C 2 H5 )2
HX
HYDROGEN DONOR
O
DYE
+
Figure 4.2. UV-light causes the biimidazole to form radicals that oxidize the leucodye.
UV LIGHT
N
O
N H
N
Cl
162 N
N
Cl
(C2H5)2N
LEUCODYE SALT
N(C2H5)2
HX
N(C2H5)2
HYDROGEN DONOR
OH
VISIBLE LIGHT
O
N H
N
Cl
Figure 4.3. Visible light converts the phenanthrenequinone into a reduction product, which intercepts and reduces the radical before it can oxidize the leucodye salt.
HABI
UV LIGHT
HO
O
163
1961—PERMANENT COLOR
oxa- oxathia-, oxaselena-, thia-, thiaselena-, or selena-carbocyanines are used in conjunction with leuco dyes to form color images upon exposure to visible light. Visible light-activated phototropic compositions comprising a hexaarylbiimidazole and a hydroxyphthalein dye. Walker, Peter (du Pont de Nemours, E. I., and Co.) U.S. 3,563,750 (Cl. 96-90, G 03c), 16 Feb 1971, Appl. 20 Jul 1967; 5 pp. The title compns. useful for imaging were manufd. by mixing 2,20 -bis(ochlorophenyl)-4,40 ,5,50 -tetraphenyl-1,10 -biimidazole (I) with a hydroxyphthalein dye, and optionally, tris [4-(N,N-diethylamino)-o-tolyl]methane (II). For example, an acetone soln. contg. 0.03 moles each of I and p-toluenesulfonic acid, 1.5 moles II, 0.04 moles Erythrosine B, and 6.8% polyethylene ether was spotted on a com. filter paper, the acetone evapd., and the paper irradiated with a flash > 470 mm;, giving an intense light intensity for the irradiated compn. compared to no intensity for a control without sensitizer. Visible light-activated phototropic compositions comprising a hexa-arylbiimidazole and a bis(alkylamino)acridine dye. Cohen, Robert L. (du Pont de Nemours, E. I., and Co.) U.S. 3,563,751 (Cl. 96-90; G 030), 16 Feb 1971, Appl. 20 Jul 1967, 5 pp. Acridine dyes of general formula I, where alkyl; R2,R4 ¼ H or
R1
R1 N
R6
R5 N
R3 N
R4
A−
R8
R7
I
alkyl; R5 ¼ H, alkyl or C6-10 aryl, R7,R8 ¼ H, alkyl or halogen and A ¼ inert saltforming anion, sensitized (activate) the imidazole to visible light in the 310–520 nm range. Listed are 12 representative acridines, 40 imidazoles, 28 amino triarylmethanes (leuco).
Dr. Dan James was given the task of developing more soluble quinone derivatives but had little success in this area. MacLachlan’s selection of phenanthrenequinone was a stroke of genius, although we found the addition of small amounts of pyrenequinone was beneficial in improving the room light performance of our coatings. Two of the ingredients, which MacLachlan employed, the leucodye TLA-454 and phenanthrenequinone were used during the entire product-life of DYLUX 503.
164
THE INVENTION OF DYLUX
E. Chuck Yembrick, Peter Strilko, and Optimization Yembrick was a superb experimenter, who conducted hundreds of experiments to establish the effect of component variations. Peter Strilko understood the criticality of the light sources and contributed significantly in showing us that certain Blacklite blue lamps from GTE Sylvania offered advantages as exposure sources—after all photographic flashguns were not practical. Strilko and I also identified a supplier of cobalt blue glass for church glass windows, the Kokomo Glass Co, of Kokomo, Indiana, as a valuable source of inexpensive UV pass filters that were magnitudes less expensive than the Corning 754 filters, which we had used in all our flashgun experiments. Eventually, they were incorporated in commercial DYLUX exposure equipment. Silversmith left us after a few years to pursue a career as a college professor. Cescon and I pursued HABI chemistry, and I developed syntheses for pyrenequinone. Bob Jenkins and Red Ireland in Jackson Lab supported our synthetic efforts. A new hire, Dr. Plutarchus Papannou, quantified the photofix chemistry. I was to concentrate more and more on applications, and establishing contacts with other laboratories that could help us. Coraor’s principal mission was to persuade management to let us focus on photo-oxidation chemistry and to obtain funding to support the programs in our laboratory and at Radiation Physics Lab. I had been promoted to Senior Research Chemist and was asked to supervise the chemists on our program. We all had jobs to do and were enthusiastic. Other members of Coraor’s group, who were involved in other research areas also contributed to our program, especially Dr. S. V. R. Mastrangelo, who became interested in the electrochromic properties of our leuco dyes.
F. Patents As we had little background in the overall area, and there was not too much literature available to us, I decided that we would make a determined effort to at least build up a patent file, to learn what others in this area had done or were doing. I then dated a friend in Washington, who was studying Hindi and Urdu at the Foreign Service Language School and found it useful to combine visits to the patent office with my social life. So I spent many days searching patents, which then, unlike now, was a tedious job, involving physically going through different classes of patents, which were neatly stacked in pigeon holes in the search rooms. In time, Coraor persuaded his bosses to assign two chemists to help us in this area, and Drs. Harold Jackson, and Roy E. Starn, a former roommate of mine at the University of Wisconsin, were given this assignment. They
1962—IMPROVEMENTS IN TECHNOLOGY
165
did a superb job, and we established a patent strategy that protected this chemistry for a very long time. Before Jackson and Starn joined the program, I was asked to assist the patent group at Orchem with their work, as I was best read in this area, thanks to my many visits to the U.S. Patent Office. We had filed an extensive application, which sought to cover all HABIs that we had (and could have) made and showed improvements over the oxidation product of lophine, which was not claimed in any patent but was reported in the literature. We sought to get composition of matter claims. Dr. Herb Walker, an elderly chemist who had previously worked on rubber chemistry, was assigned as patent chemist and was close to retirement when the first office action, a rejection came. The examiner told us that similar imidazoles had been reported to give off light when heated and that what we had showed no novelty over the prior art! Walker allowed me to reply directly, and I wrote that the phenomenon that we had invented was photochromism, a color change induced by light, as in certain organisms like salamanders, while what he cited was chemiluminescence, a phenomenon attributed to fireflies. Since fireflies are different from salamanders, a patent should be granted. The examiner had a sense of humor. He allowed my response but suggested that a patent could be granted only if we could chemically identify the HABI dimer. At that time, this could be done only by X-ray crystallography, which was a long and tedious job. However, we were pleased with his response as we had a priority date and could delay the ultimate issue for a considerable length of time. As it turned out, this application, filed in the early 1960s, became a patent (L. A. Cescon, U.S. Pat. 3,784,557, Jan. 8, 1974) in 1974, 5 years after products based on this chemistry hit the market. It had issued many years earlier as Belg. Pat. 635,804 (see above). The crystallographic examination required the growth of single crystals, which took time, and was finally accomplished by one of DuPont’s foremost crystallographers, Dr. Gunther Teufer. He identified the bonding between rings as 1,20 .
VII. 1962—IMPROVEMENTS IN TECHNOLOGY During the period of 1962 to 1964 we made significant improvements in the technology. Thanks to Yembrick’s work, coatings achieved a high level of stability. Bob Cohen’s work had resulted in the replacement of o-Cl-HABI with the preferred biimidazole, CDM-HABI, with a spectrum that maximized the ‘‘imaging’’ and ‘‘deactivation’’ peaks of the dual response spectrum (structures depicted in Scheme 4.8).
166
THE INVENTION OF DYLUX
OC H3 OC H3
N Cl
N
N
Cl N
N
Cl
N
Cl N
N
OC H3 OC H3
o-ClHABI
CDM-lHABI
Scheme 4.8
Cescon had studied binders and plasticizers. The chemistry improved, but the photosensitivity of the system remained relatively slow. We, like all photochemists, were striving for higher speed materials.
A. A New Opportunity—Foot Imaging In 1962, soon after we had demonstrated the photofix concept, we still had no good idea what to do with our light-sensitive papers. Our management had little experience in the area, and we were constrained in going outside the company for help, for fear of compromising our patent position. Where could we get help? In the 1950s, DuPont invested a major effort in a product Corfam1, a poromeric substance that was to be the synthetic replacement for shoe leather. This program was driven by the Fabrics and Finishes Department, which had considerable experience in the coated materials business, and high hopes were held out for the opportunities derived from these new markets. Corfam was a good product, and many people who owned Corfam shoes were delighted with them, especially golfers, who were frequently walking on wet grass. Many people, however, became dissatisfied with Corfam because it stretched little, and traditionally, shoes are bought with the expectation that they widen over a period of time; Corfam shoes did not. An extensive study on how to overcome this problem was undertaken, and it was decided that perhaps better techniques of fitting shoes were desired. Shoe fitting was based on old arbitrary techniques. Before they were outlawed, X-ray machines in shoe stores gave the customer some indication of how well the shoes fit. The Brannock gauge attempted to make some measurements based on the
1962—IMPROVEMENTS IN TECHNOLOGY
167
length and width of the foot, but it was pointed out that shoe salesmen pretended to be scientific, but actually peeked into the shoe to establish size rather than rely on the measurements. DuPont was going to put science into this area! A suggestion was made to measure the feet of visitors to DuPont’s pavilion at the 1964–1965 World’s Fair in New York, and a facile way of recording this information was sought. When the group charged with this program heard about our light-sensitive paper, they visited our laboratory and asked us to help them. We showed them that we could make silhouettes of feet by a simple exposure of the foot on a piece of our paper, using only two blacklite fluorescent lamps. It seemed like a great opportunity for getting broad exposure for our paper, and we were delighted finally to have support from someone in the company who wanted what we could deliver. In a short time, an exposure device was built, and the process was demonstrated at the highest level of the company. The right foot of the then CEO, Mr. Crawford Greenewalt, was duly exposed. Before long we had a large collection of white silhouettes of feet on a blue background. It was remarkable how the outlines of people’s feet differ. Several of the company’s footwear consultants were ecstatic about the proposed program, and considered the approach as a major step toward better shoe fitting. We learned a great deal about shoes and feet in those days. Bill Cryer, the marketing manager, who had responsibility for coordinating this program, suddenly took ill and died shortly thereafter. His successor did not share his enthusiasm and withdrew his support. The concept did not die entirely—in 1966 we were approached by an independent inventor to get into the foot measuring business again; but by then, this was no longer of much interest to our management. In 1963, our management decided we needed some more guidance toward commercialization, and Dr. Ross Fassick, a supervisor from another group was assigned to get us better focused. He stayed with us for a few months, before being promoted to another task, ultimately winding up in a high management position.
B. Optical Printing Most everybody remembers where he or she was on November 22, 1963, the day John F. Kennedy was assassinated. I was sitting in my office at the Experimental Station, waiting for a meeting with a group of engineers, who had been hired to consult with DuPont to identify products for the expected information explosion. David Nettleton and Fred Palmer had worked at RCA, had formed Data Communications Inc. (DCI), and had been by contacted the DuPont Development Department. We showed them our photosensitive coatings, and they immediately suggested that there was an opportunity for our technology in optical printing.
168
THE INVENTION OF DYLUX
At that time, high-speed printing was done either with IBM’s Selectric typewriters or Teletype machines. UV lasers were around the corner, as were computers; and the need for high-speed, quiet printing was obvious. DCI offered to propose to build an optical printer, which could expose a beam of light via lenticular plate optics, to print silently as many as 20 characters per second. Our management, which had been frustrated by Photo Products Department’s lack of enthusiasm for our technology felt that this could be an Orchem type of application and agreed to fund this development. In time, DCI, did indeed produce an optical printer. However, by mid-1965, Orchem decided that they would not want to market a printer and instead sought partners in this development. They established contacts with companies such as AT&T, Western Union, and Litton Industries. Chuck Yembrick and John Clemens, a former Freon1 salesman went to establish these, but were very limited in what information they could disclose to our would-be partners. The result was that no one was very interested in working with DuPont, and the program was severely curtailed.
C. Proofing? In December 1963 I had an opportunity to visit DuPont’s Rochester plant, which was primarily oriented toward silver-based products. I gave a seminar about our technology, and met Larry Friar, who soon afterward told me that we should use our technology for proofing. Friar, I recall, gave me a negative of a girl sitting in a bathtub, and we made a nice print on UVI-paper—although it was blue and yellow. Friar said, ‘‘Tell your bosses you have the ideal proofing medium!’’ I went back with that message but we could not persuade anyone in the Photo Products management. Inevitably, the question was asked as to whether there could be a market for a proof paper, and if so, why did Eastman Kodak not have a product to meet that need. I read that the inventors of desktop computers at Xerox’s PARC laboratory were asked a similar question: If there were a need for these, why had IBM not put it on the market. Innovators do sometimes have hurdles to cross.
VIII.
1964—A HIGHER SPEED SYSTEM
One afternoon in 1964, Cescon, Bob Cohen, and I had a bull session where we discussed how we could improve the imaging speed of our materials. Two ways were obvious: a chain reaction to form color more effectively or a route to a higher deactivation speed.
1964—A HIGHER SPEED SYSTEM
169
Our deactivation, or fixing speed, was purposely low, so that the materials could be handled in ambient light. Could we speed it up? During our discussion I suggested that as we needed plasticizer to permit color formation to occur at all, perhaps we needed to consider the rapid photochemical removal of the plasticizer as a route toward a higher speed. All of a sudden, an obvious approach appeared: Use a monomer as plasticizer, and polymerization to ‘‘fix’’ the system. Cescon and Cohen immediately began to purify monomers, because they were excited to demonstrate new technology. They proposed a system in which our preferred leucodye and HABI were incorporated in a binder–monomer system and in which a conventional anthraquinone photoinitiator would effect polymerization in the visible portion of the spectrum. It worked! Over the next weeks, they optimized the system, showing that visible light initiated polymerization and UV-exposure would yield color. Cescon, as always, was thorough. So he tested the system by omitting the anthraquinone photoinitiator and found that it still worked! Then he and Cohen demonstrated that the system was intensity sensitive; at a high intensity of light, polymerization and color formation ensued, but at low-intensity irradiation, only polymerization occurred. We took the system as demonstrated to our colleagues in Radiation Physics Laboratory, and they were duly impressed. MacLachlan immediately postulated a mechanism, by which the ion-radical formed on photo-oxidation of the triphenylmethane leucodye could, at low concentrations, act as a photopolymerization co-initiator, initiating polymerization only. At highintensity exposure levels, sufficient ion radical was formed to form color (second order reaction) and effect polymerization as well. The polymerization locked the imaging components in place, so that subsequent intense exposures did not result in further color formation. The images were fixed. Thus, by controlling intensity of exposure, we could form positive or negative-mode images. We termed this technology the RF-system, indicating that rigidification-fixing could take place. Imaging and fixing radiation-sensitive compositions by sequential irradiation. Cescon, Lawrence A.; Cohen, Robert L.; Dessauer, Rolf (DuPont de Nemours, E. I., and Co.) U.S. 3,615,454 (Cl. 96/35.1; G 03c), 26 Oct 1971, Appl. 26 Jun 1968; 22 pp. A photopolymn. photog. system contains an image precursor, such as a leuco dye, in a fluid system activated by imaging radiation, and immobilized, but not chem. inactivated by the fixing irradn. Preferably, a single compd., e.g. 2,20 ,4,40 ,5,50 -hexaphenylbiimidazole (I), acts as both the oxidant for the leuco dye and the polymn. initiator. A chain transfer agent such as N-phenylglycine, may also be added to the system. Thus, I 0.035, [2,4-Me(Et2N)C6H3]3CH 0.013, p-MeC6H4SO3H 0.010, pentaerythritol triacrylate (II) 0.030, pyrenequinone 0.0013, tri-Me nitrilotripropionate 0.30, and poly(vinylpyrrolidone) 0.25 part was coated 1-mil thick on a Mylar film and covered with a 3-mil thick polypropylene film. The system was exposed to a Xe lamp through a stencil and a 240–420 nm filter
170
THE INVENTION OF DYLUX
to give a deep blue image, which was fixed by a Xe flash through a filter transmitting >350 nm. When II was replaced by pentaerythritol tripropionate, the image could not be fixed.
Many years later, Kay Looney and I presented a paper at an SPSE Meeting in which we discussed this imaging system in detail. Modulation of biimidazole-leucodye photochemical reaction by photopolymerization. Dessauer, Rolf; Looney, Catharine (Chem., Dyes Pigm. Dep., E. I. DuPont de Nemours and Co., Inc., Wilmington, DE 19898 USA). Photogr. Sci. Eng. 1979, 23 (5), 287–9 (Eng). A photoimaging system is reported in which the diffusion controlled color-formation reaction is modulated by the polymn. of the essential plasticizer. Polymn. of the plasticizer by light-induced reactions allows pos. or neg. image formation depending on the sequence of exposures; an intense exposure leads to simultaneous color-formation and deplasticization, while low-intensity exposure results only in polymn. without color formation.
We perceived the mechanism of the reaction as shown in Scheme 4.9. The mechanism, as postulated by MacLachlan, involves an ion radical formed by oxidation of the leuco dye (by a triarylimidazolyl radical), which can act as polymerization initiator (at low concentrations) or will disproportionate into dye and a colorless species and also act as polymerization initiator at high concentrations.
UV
HABI DIMER
IMIDAZOLYL RADICALS CH3
(C 2 H5 )2 N
[(CH2 =C(C H3 )-CO 2 CH 2 ]3 CCH2 OH
CH3
N(C 2 H5 )2
H CH 3
MONOMER
LEUCODYE N(C 2 H5 )2 ION RADICAL
LOW ION RADICAL CONCENTRATION
POLYMER
HIGH ION RADICAL CONCENTRATION
POLYMER and DYE
Scheme 4.9
1964—A HIGHER SPEED SYSTEM
171
A. HABIs Really Are Photopolymerization Initiators Almost from the beginning of our work on HABI chemistry, we had wondered whether they could act as photopolymerization initiators. We then knew next to nothing about photopolymerization but had some relationship with our research colleagues at Photo Products Department, and asked them to help us. Dr. Marion Burg agreed to test a purified sample of o-Cl-HABI, which we provided her, and she reported that the results were negative. Some time later, Dr. V. C. Chambers, who headed a group, that worked on a photopolymer based copying system visited us and requested information about how we could make black images using our leucodye chemistry. The Photo Products office copy system was based on ingenious chemistry, in which an unpolymerized coating, containing a triphenylmethane dye, was thermally transferred onto a piece of plain paper. We supplied Chambers with some neutral shade forming leucodyes and HABIs, and his technician observed that the polymerization reaction, which was an imagewise reflex-exposure, had higher speed as a result of the presence of our chemicals. Photopolymerizable compositions and elements. E. I. DuPont de Nemours & Co. (by Glen A. Thommes, Peter Walker, and Vaughan C. Chambers, Jr.). Fr. 1,481,819 (Cl. C08f, G 03c), May 19, 1967; U.S. Appl. June 3, 1965, and March 4, 1966; 14 pp. Photopolymerizable compns. for use in the production of color images contain 1 nongaseous ethylenically unsatd. monomer capable of undergoing free-radical addn. polymn. with chain propagation; 1 free radical-producing electron donor, e.g. a leuco dye, a tertiary amine a mixt. of a leuco dye and a photoreducible substance, or a mixt. of a leuco dye and triethanolamine; a 2,4,5-triphenylimidazole dimer; and optionally an energy-transfer substance to improve the sensitivity and (or) an additive to eliminate or reduce the induction period. The compn. is dispersed in a thermoplastic binder and contains various dyes, pigments,
Cl N
N
Ph
Ph
2
I
thermographic compns., and chromogens, and is coated on any appropriate support. Thus, a standard soln. was prepd. from a 10% cellulose acetate butyrate (20.5% Ac, 26% butyryl, 2.5% OH, and viscosity 9.0–13.5 poises) soln. in Me2CO
172
THE INVENTION OF DYLUX
110.0, a 10% cellulose acetate (39.4% Ac 55% HOAc, viscosity 130–82 poises) soln. in Me2CO 32,5, pentaerythritol triacrylate 30.0, poly(ethylene oxide) (mol. wt. 400) (in 14 ml. MeOH) 2.0 g., and Me2CO to make 200.0 g. The standard soln. 5.0, the I dimer 2-(o-chlorophenyl)-4,5-diphenylimidazole 80, and CBr4, 100-mg. were mixed, coated as a 0.15-mm. layer on a 0.03-mm. poly(ethylene terephthalate) support precoated with a layer of vinylidene chloride-Me acrylate-itaconic acid terpolymer, and dried for 30 min. at ambient temp. to give a 0.03-mm. film. The laminated product was contacted with an original, exposed to a photoflood lamp at a distance of 40.6 cm. for 30 sec., delaminated, developed, and passed with a receptor paper through 2 rollers heated at 125 to transfer the under exposed zones from the photopolymer layer to the receptor sheet thereby forming a well-contrasted and defined image on the paper.
The reason Chambers and we had success and Burg did not was that our system included a tertiary amine, which acted as co-initiator. HABIs themselves indeed do not initiate polymerization but do so most effectively when paired with a co-initiator. Even though the project was terminated because office copy arena became exclusively an opportunity for electrophotography, DuPont applied for a patent, which was the first of many that identified HABIs as photopolymerization initiators. Cescon, Cohen, and Dessauer’s patent, in true Orchem style, issued much later, because we never were in a hurry to see our patents issue—the later the patent issued the better our products would be protected! In time, HABIs were to become important components of photopolymer systems; as of December 2004, there were over 1000 U.S. patents, in which biimidazoles, lophine dimers, HABIs, and 2-o-chlorophenyl-imidazole derivatives were mentioned! A word about patents. At Du Pont, the names of inventors are listed alphabetically, even though the lead inventor might have a name that was at the bottom of the alphabet. There were exceptions. Sometimes, one of the inventors would request the attorneys to list a lead inventor first.
B. Management Change George Coraor was the ideal research leader. He was pleasant, never showed anger or disappointment at the slow pace of research, tried to be helpful to all his subordinates, and fought endlessly with his management to provide us with more and more support. He and his bosses simply were not made for each other. They were not interested as much in technology as he was, nor did they have the long-range vision that precluded going to the market before one has technology locked up. By December 1964 things had come to a crisis, and Dr. Robert Terss, who had headed the Orchem patent group, was appointed division head to oversee our UV-Imaging Program. I had shared a lab with Terss some years earlier,
1965—OTHER MARKETS
173
when we were both dye chemists, and had liked him both as a chemist and as a friend. He interpreted his mission as one of changing our program from a research activity to a product development effort. He agreed with me that the discovery of the HABI-initiated photopolymerization was significant, and that we needed to undertake whatever research was required to establish a patent position, but he felt that his bosses were determined to produce a salable commodity, whatever that could be. He allowed Cescon, and Cohen to continue under my supervision, but slowly transferred the other members of our group to Jackson Laboratory. In time, Cescon, Cohen, and I were reassigned to Coraor, with the expectation that we might find something useful to do with the newly discovered photopolymerization technology. Our group there was even expanded, by the addition of two new chemists, Elva Mae Nicholson and Mark Krakov. We were interested in gaining a better understanding of the chemistry involved.
IX. 1965—OTHER MARKETS A. Microfilm In the mid-1960s, microfilm seemed like a rapidly growing field. Many documents were microfilmed, journals appeared in microfilm form, and both microfilm and microfiche were areas of growth. One limitation of microfilm was that there was a time-lapse between the photography, using silver film, and the use thereof, because it was not practical to develop one frame of silver microfilm at a time. I thought that the use of our materials, which could form images without processing, might find an interesting market here. Some time earlier, Ray Firmani, the Jackson Laboratory photographer, and I had built a box in which an intense light illuminated a document, and a 35-mm Leica lens was positioned to capture the image on a piece of UVI film. A 1-minute exposure to visible light emitted by the lamps deactivated the UVI film, and a flash of UV light subsequently would effect color formation in the areas that had not been completely deactivated. A blue-and-yellow image was formed. I thus photographed a dollar bill, which I could then project with a 35-mm slide projector. I thought that that was really cool. Almost everyone I showed this to wondered whether we could get into difficulty because of the counterfeiting laws. I finally called up the U.S. Secret Service and asked them if there was a law against what I did, and they said not to worry. Later on, when we developed the rigidification system, we formed even better images, faster. Still, the interest level in this concept could not be advanced. Another area where I thought we might do something unique was in microfilm duplication. Here there were other products on the market, such as the
174
THE INVENTION OF DYLUX
vesicular films (e.g., Kalvar) or diazofilms. Still, there seemed some advantage in negative working materials, which did not require processing. A few years later, there actually was an effort by DuPont to explore this market, but not much came of it. At several trade shows that I visited, I saw demonstrations by a small English company, CAPS, of a UV projector that was used to enlarge images of engineering drawings from aperture cards onto diazo paper. The unit was relatively large and generally attracted a lot of attention. The images formed on diazo paper were in a positive mode; a negative-mode image would have been more desirable. I tried to interact with CAPS, but it was difficult because there seemed so little interest in this area by my management. In May 1965, I visited the National Microfilm show in Washington, D.C., with John Roberts, an engineer who followed the field for the Development Department. I finally asked Mr. Terry Wilson, who worked the CAPS exhibit, if he had a moment to learn of an experimental product being developed by DuPont. He asked me to speak to his associate, Mr. Gerald Frankel, who was enthusiastic and asked me to come to their suite in the Hilton Hotel. He ordered three steak dinners and listened attentively to what I told him. He invited me to come to their showroom in Manhattan that weekend, before his return to England, to test materials and suggested I take someone from our management along, to convey his high level of interest. My success in enthusing CAPS was irrelevant. Dumont had no plans to use his weekend on a business trip and did not think we had much business going into microfilm copying, blowback, or whatever. He said we could go to CAPS if we wanted to, but nothing would likely come of it. Caris and I did go to Manhattan on Saturday morning and had a great time working with the CAPS equipment. Frankel was disappointed with our management, and I suggested that as a company president he could call our general manager, Bill Kay, and tell him he wanted to do business with him. Caris was a little worried, and suggested that Frankel should call Dumont. The result was a meeting the following week, in which Wilson came to Wilmington and reviewed their business for some of the DuPont research people, and they asked him questions about the future of microfilm, which no one could have answered. Some months later we did actually buy some equipment from CAPS; it worked quite well, but Dumont decided that microfilm was in the province of the Photo Products Department, and if they did not want to work in that area, there was nothing we could do about it. In September 1965 I went to Greece on a vacation and spent considerable time with my colleague Arch Papannou, who had taken a leave of absence from DuPont. I remember two events that preceded the trip. In 1965, DuPont decided to make a television commercial, in which were to be shown research
1965—OTHER MARKETS
175
activities at the DuPont Experimental Station. Dr. Henry Rothrock, who was in charge of selection the scientists to participate, offered me a chance to be on national television, and I jumped at it. I was photographed showing how we could form color with a flash of light. We were to record a brief commentary, but the tiny tape recorder available was not likely to do a good job. So it came as no surprise that all of us had to make a hurried trip to New York to redo the voice recording. The program was to be shown in November 1965 as part of a commercial on the DuPont Cavalcade of America program; unfortunately, this coincided with the first great blackout in U.S. history; and my parents, who had invited a number of friends to see their son on TV, were disappointed—the program could not air in New York. It was seen on the West Coast, and one selfstyled inventor sent long letters to DuPont suggesting applications for our photosensitive materials. Our legal people urged me to ignore these. In time, the inventor started to complain to the upper-most DuPont management. I finally called him up one evening, and had a nice chat with him. He said he understood the situation, but just wanted to have a little fun! After the trip to the sound studio, I drove to Parlin, and tried to persuade Vaughn Chambers to delay the issuance of his photopolymerization patent, but was unsuccessful. I still do not know why people rush to get patents, when a later issue ensures them of longer life for any products based on the patented technology. In time, many products were based on technology that was described in that patent. Fortunately, our delayed issue of the omnibus HABI patent protected all this until 1991! When I returned from my holiday, I learned from Coraor that a DuPont subsidiary, Holotron, which was involved in holography, asked me to investigate the use of our UVI technology in this area. I asked Kay Looney if she would like to work with me in this area, and we did a nice job, proposing the use of our photopolymer system in this application area. For some unknown reason, several years after this work was reported, a patent issued naming the inventor as Eugene Haugh, who had never worked on this. Mysterious! Image-producing layer of nearly uniform thickness for holography. Haugh, Eugene E. (DuPont de Nemours, E. I., and Co.) Ger. Offen. 2,041,072 (Cl. G 03rb), 18 Mar 1971, US Appl. 25 Aug 1969; 4, pp. Light resistant holograms are prepd. by coating a transparent carrier layer with a photopolymerizable layer comprising an acrylatc monomer, actinic radiation-activated free radical initiator, and solvents and exposing to light to polymerize the acrylic monomer. A photopolymerizable soln. comprising cellulose acetate butyrate, triethylene glycol diacrylate, 2-(o-chlorophenyl)-4,5-bis(m-methoxyphenyl) imidazole dimer, (3,4dimethoxyphenyl)-bis(2-methyl-4-dietbylaminophenyl)methane, N-phenylglycine, ethylene bis(2-hydroxy ethylacetate), CHCl3 and CH2ClCH2Cl is coated on a poly(ethylene terephthalate) carrier as a 0.10 mm layer and dried. The light sensitive
176
THE INVENTION OF DYLUX
film is placed in a vacuum copier where the emulsion side of the carrier is contacted with a holographic master copy and exposed 15 sec to uv radiation to form a holographic copy of the original. The copy is light resistant.
B. Cathode Ray Tube Imaging We had made contacts with other departments in the company to explore the use of our technology and were surprised one day, when a note came from someone in the Treasurer’s Department, which was charged with maintaining stockholder relations and the like, that we ought to consider the use of our UVI solutions to address stuffed envelopes. Whoever it was, he felt that we had a unique self-contained imaging system, and that if one could use it to sensitize nonflat surfaces and print on them, this would be a unique capability. Of course, as Orchem was already in the Freon propellant business, it was not far-fetched to develop a spray system. Imagery was another matter. UV lasers were proposed but not really available. Someone suggested that UV-emitting cathode ray tubes were available; and engineer Charley Ford was dispatched to attempt to image a piece of UVI paper with a specialty tube, manufactured by the Litton Electron Tube Division in California. When a blue pattern was formed, the few of us still working on UVI realized that we had been saved once again. In the 1960s, General Robert Sarnoff, the president of RCA, had proposed that newspapers would be delivered via television. Sarnoff had been the power behind color television, and he was widely respected. When our management realized that we could use our photosensitive paper in combination with cathode ray tubes, their enthusiasm for our technology again rose. I do not believe that it was fully comprehended that the Litton tube that was used was a very expensive item; it had quartz fiber optics and UV-emitting phosphors and was driven so hard that its lifetime was unlikely to be very long. Still, someone thought that these limitations could and would be overcome. We again had identified an application that was not in the photographic field to be freely explored by the Organic Chemicals Department. To expedite our work, a tube was ordered post-haste. It might have been an ominous sign, but after the tube had arrived at Jackson Laboratory on a Friday afternoon, it was discovered on Monday that a janitor had thrown out the package in which it was still sitting. So another tube was ordered. During the early part of 1966, there was renewed enthusiasm for our technology, and the manpower of our program was again increased from its low point of five. Only Peter Strilko, Chuck Yembrick, Bob Cohen, and I had remained from the original crew. My problem was that I was concerned that the energy to image our materials with a CRT was marginal and kept on suggesting that perhaps our photopolymer technology was a better match for the proposed
1966—WE BECOME A VENTURE
177
CRT imaging. However, the buzzword was electronic imaging, and it was not good to be opposing something that was deemed to be the future, even if it was someone else’s. I kept on plugging away at microfilm and proofing. These applications, to me, seemed real and I could conceive of having possible novel products.
X. 1966—WE BECOME A VENTURE In the mid-1960s, there was a lot of money in the DuPont Company, resulting from the divestiture of General Motors stock, which was mandated by the U.S. government. Some of this was doled out to support ventures, which were sponsored by the various operating departments. In August 1966, the UVI-group became a venture. Leon Dumont, who had been manager of the program, became venture manager. Bob Terss became the head of its research division, engineer Ed Lightcap became the manufacturing division head, and Irene D. May was named head of the marketing division. Regrettably, none of these men had any experience in the imaging field. The idea behind ‘‘ventures’’ was great. Let a group exist with some independence in a nurturing environment and start out as a ‘‘future business.’’ It would have the advantage of being a highly motivated small business, with the backing of a larger business. In actual practice this was not the case. We could not go outside to hire people as we needed them. When a business downturn occurred, the ventures were the first to feel the reduction of support. So actually, we had all the disadvantages of being small. I was assigned to the marketing division and was joined by Walt Godfrey, whose marketing experience was in the field of Freon propellants. Godfrey had no experience in the field, and I was somewhat handicapped in working up a great deal of enthusiasm for the impossible quest of making a product that could be imaged with cathode ray tubes. Within a few weeks after we became a venture, a consultant from the Photo Products Department, Mr. Ted Harding, an elderly gentleman who had been assistant plant manager of the Parlin Photo Products plant, was assigned to us. I was given the task of educating Ted about our technology. He was able to comprehend my concern about the lack of sensitivity of our materials for the proposed CRT-imaging applications; and when I asked him to try to persuade our management that we were going in the wrong direction, he pointed out to me that this would be unwise—had our leaders not sold the venture to the upper echelons of DuPont on the premise that there was a huge opportunity facing us in this business area which had such huge potential. He admonished me with ‘‘You don’t want to make your bosses look like liars!’’
178
THE INVENTION OF DYLUX
Ted Harding came on a number of trips with me to visit Defense Department sites and was always affable, but never really helped steer our program into a meaningful direction. What was needed was a dynamic leader, who understood image technology or how to sell a totally novel material. Or both!
XI. 1967—WE START A MARKETING EFFORT: HOW DO YOU GET TO THE MARKET, ANY MARKET? In the 1960s, DuPont’s Photo Products Department sold through distributors, but had a network of technical representatives (TRs), who would call on accounts that would then buy from the distributors. Technical people, who were to listen to customer needs and direct their research efforts accordingly, occasionally accompanied these TRs. DuPont felt that it gave the TRs some more credibility if scientists accompanied them. Unfortunately, the research management in many departments gave little credence to what information the scientists brought back from the market. Soon after we became a venture in 1966, it was decided that the UVI group would occasionally be permitted to accompany TRs to customers. At that time, our venture manager, Leon F. Dumont, was quite generous in setting aside expense account money for these trips; and inevitably, the TRs were happy to see us, because we could take them and their customers out for lunch—something that was not common practice with the somewhat tightfisted Photo Products Department.
A. Harold Wilbur—The Electrician We needed equipment to expose our materials. Initially, we bought the most powerful portable flashguns that were available on the market—these were manufactured by the HiCo Company. These flashguns were useful to show that our materials gave images instantly; but what was needed was a flat-bed exposure unit that could make good contact prints of lithographic negatives. I found such available in camera stores and brought one to our lab. These were designed to work with silver photographic films and had relatively weak light sources. By 1967 an electrician, Harold Wilbur, was assigned to our group, He modified these small contact printers, so that they could expose an 8 by 10-inch negative onto a piece of our photosensitive paper. He inserted a number of blacklite blue fluorescent lamps and suitable ballasting, so that a print could be obtained in ca. 20 seconds. We could thus make contact prints of continuous tone and lithographic negative in very little time. Wilbur made dozens of these, and they were used by our group, and later by sales personnel to demonstrate the
WE START A MARKETING EFFORT: HOW DO YOU GET TO THE MARKET, ANY MARKET?
179
utility of our materials. They were the most powerful sales tool that we could conceive. When we found that the manufacturers of exposure equipment for the graphic arts, companies like Douthitt and Newark, were slow in designing equipment for us, we asked Wilbur to get involved in building commercial equipment, and he did a superb job in building such equipment in his own home workshop. In time, this proved to be quite profitable to him and for us, as it helped move the sale of Dylux proof paper. Without this early equipment, the venture might well have succumbed, as commercially available equipment for exposing lightsensitive materials did not image our materials well, as the light sources contained too much visible light, which acted to deactivate the coatings. We had also encouraged people to employ frames holding banks of blacklite blue lamps.
B. Point of Sales Opportunity One novel development which was pursued by us involved Litton. During the presentations given by Yembrick and Clemens to various companies that might have been interested in optical printing, a contact was established with Litton’s Monroe Division, which sold cash registers. They had conceived a point-of-sales business opportunity, where a magnetically readable coated paper, optically printed alphanumerics were superimposed in a sandwich structure. Litton requested that we coat our photosensitive materials on a very thin paper base and then laminated this to a magnetic iron oxide layer previously deposited on paper. They appeared very serious about this opportunity and advised our management of their need for our unique materials. Their program was ultimately tested in a shoe department of a Sears Department store in Kentucky and performed remarkably well. Several years later, in 1969, it was found that impact printing with typewriters did not adversely affect the performance of the magnetic layer, and the need for an optically imagable product was found unnecessary. End of an application!
C. The UVI Movie During the early days of the UVI-Marketing Division, I suggested that we prepare brochures, describing our technology; this would make them more than a laboratory curiosity. May was supportive of that, and as I was the only one in our small group who had any understanding of the technology, writing the brochure, getting the needed data, etc. fell into my hands. I also suggested that it might be
180
THE INVENTION OF DYLUX
desirable to make a brief movie describing our technology. Again, in 1966 this was not the easy task that it would have been if videotape had been around. Instead, the movie had to be made with Professional Kodachrome1 film, and the company’s professional movie photographer was enlisted to work with me. Fran Ryan was a charming and enthusiastic man, who had made many movies of various DuPont products and was obviously challenged by having to make a movie about a light-sensitive product. He immediately persuaded me that the movie had to be made at night, because the drain on the electrical circuits at the DuPont Experimental Station was too much – after all we had to have a number of flood lights to illuminate whatever we photographed. So, we usually started ‘‘shooting’’ around midnight and quit around 4 A.M. Fran was not above shooting the same scene over and over. He convinced me that a movie should be free of technical flaws, and after I had seen it a few hundred times, I realized that he was right. He suggested I write a script, and that we should shoot anything that came to my mind. Ultimately we wound up with around 12 minutes of 16-mm movie film, and I think that each minute represented a day of shooting. Ryan put the film together, and suggested to May that we get a sound track. May was reluctant to spend more money on this and pointed out that I should give a commentary as we showed the movie. We ran a few trials, and found that inevitably someone would interrupt the show and ask a question - after that I could never catch up. So, I had to write a proper script around what we had photographed, and obtained the services of an announcer of DuPont’s Cavalcade of America on radio, and later TV, to read my lines. Ryan even selected a musical accompaniment for all that. I was quite proud of all this—it was fun. About 2 years after we made the movie, I had occasion to show this to Douglas Shearer, Technical Director of Metro-Goldwyn-Mayer, then one of the major movie studios. I was surprised to learn that MGM was interested in novel imaging materials that could be employed in films for drive-in movie theaters, where the intense heat of the projection apparatus adversely affected the silverbased films. He thought my film was a nice piece of work and said MGM could have made a great film about our materials using their special-effects department. But, then, he was the owner of 20 Oscars for his past work. All things considered, the movie was a success. It allowed us to send it to would-be users, without having to make trips to show our materials. To my great surprise, even some large companies lacked 16-mm sound movie projectors; and in time, we simply learned to take them along on airplanes. By 1970 we had learned enough about our materials and some new potentially attractive applications that Ryan and I were encouraged to make another movie. This time we were even able to hire a model to demonstrate our photosensitive papers and films. Just as we had begun to edit the film, the Dylux Venture was transferred to the research division of Photo Products Department, who had no desire to support this effort.
WE START A MARKETING EFFORT: HOW DO YOU GET TO THE MARKET, ANY MARKET?
181
Other than writing brochures and making movies, I tried to stimulate anyone who wanted the materials, which by then were becoming somewhat known in the industry. By May 1967, Dumont decided we should make some presentations about our technology to several potential customers His first choice was IBM, even though that company had several groups working on similar nonsilver technologies. Our agenda was for Dumont or May to give an introduction, for me to give a technical presentation, to show the movie, and to have an engineer to follow this with a talk about the use of our technology in a CRT-directed imaging system. Initially, John Knupp was the engineer, some time later, E. W. (Ted) James took on that task. Our first presentation at IBM’s Armonk headquarters was received politely, with a request to repeat it at an IBM Laboratory on the West Coast.
D. Another Trip to the West Coast In 1967, the principal centers of innovative research were on Route 128, surrounding Boston, and in California. Transcontinental travel was then not what it became later, and many of us looked at going to California as though we were going to a different country. Dumont suggested we spend a week in California and visit potential customers for our electronic-imaging program. High on his list was North American Aviation, which had a group working on computer displays: Larry Hendricks, a scientist at NAA had developed some novel CRTs and was anxious to work with us. We developed quite a relationship with Hendricks, but ultimately little came of this work. Litton Advanced Data Systems had conceived a scheme of using our technology to print tickets for entertainment events and was hoping that DuPont would support this with materials and money. This too was to fail. Our visit to IBM at San Jose was not much of a success; the head scientist there presented us with a set of questions, which were unanswerable by us at that time. Some weeks later, we made another trip to the West Coast, this time to visit Boeing, Tektronix, and again NAA. This time we were more successful. Boeing had already developed a large-wall screen display system, which depended on forming images on Kalvar1 film, and considered the use of our UVI materials a route to a better system. More about this later. We arranged for a presentation at Tektronix. They were politely interested, but nothing came out of this, other than a nice day in Oregon. NAA again gave us some encouragement.
E. The Name Dylux1 One afternoon, Ernie May, Ted James, and I finished early with our customer visits, and sat at a motel bar in Laguna Beach, California. I suggested that we
182
THE INVENTION OF DYLUX
use the time to think up some names for our proposed product while we were drinking mai-tais. There were a few rules for naming products in those days. The name had to be novel, it should not directly describe the product or its features, and it should be easy to pronounce. Years earlier DuPont had a paint line called Dulux1 and I came up with the name Dylux1, based on the idea of making dyes with light. We all seemed to like it. May suggested we have a contest in the venture group for a name; but in the meantime, we did a search for the term Dylux among trade names. When nothing showed up, he generously suggested that I had worked so hard on this that I should be permitted to give it a name. So, the trade name Dylux was used to identify DuPont’s instant-access imaging materials. Only in Germany was there some opposition to this name: Osram, a manufacturer of electric lamps had a product Bilux, which they claimed sounded similar. In general, DuPont attached some descriptor to identify the trade name, and for Dylux it was to be instant-access imaging materials. By the year 2000 the name Dylux had become so identified with proofing products, that some paper base, manufactured for ink-jet proofing systems, was also given this trade name.
F.
More Marketing
In August 1967, we made still another trip to the West Coast, this time to attend an electronics show in San Francisco’s Cow Palace and to visit Boeing, NAA, and a few other companies. We had had an inquiry from a small company, Marks Systems, who had become recognized as a manufacturer of developing equipment for Eastman Kodak’s Bimat1 system. I visited Mark Systems alone and was impressed with their enthusiasm for our technology. They wanted to get into the microfilm field and had the equipment-building capability that DuPont lacked. Of course, because of Photo Products Department’s lack of enthusiasm for our materials, microfilm was not an acceptable application—but what the heck! By this time, the venture had grown a lot in size. The Marketing Group was expanded, but primarily to examine the market areas for our technology from the sanctity of our offices rather than by going to trade shows, customers, etc. John Caris, who decided he was going to be the venture planner, began to assert himself more and more. He was bright and could talk convincingly on any subject he chose; but unfortunately, he had little comprehension of the capability of our materials or the effort required to improve them. To Caris, the challenge was the size of the opportunity and not its attainability. The bigger, the opportunity, the better. The longer the opportunity required to achieve fruition, the better, because it would give us more time to get there. What seemed of no importance at all was where we were at the start of that opportunity.
WE START A MARKETING EFFORT: HOW DO YOU GET TO THE MARKET, ANY MARKET?
183
G. Large Wall-Screen Displays 1. Project Apollo During our 1967 visit to General Electric, where we described our materials, I met Mr. Max Kerr, a senior scientist who was involved in GE’s contribution to Project Apollo. He was keenly interested in obtaining materials, that had instant accessibility for large displays. Max invited me to come to Florida and see his facility; and in May 1967, I took him up on this, combining it with a visit to the microfilm show in Miami. Kerr and I worked together for another year; and in March 1968, I delivered some of our higherspeed photopolymer materials, which were imaged successfully on their unit. Nevertheless, nothing ever developed beyond that.
2. Boeing Perhaps the only electronic-imaging application that we could apply our technology to was in the field of large wall-screen displays. It must be remembered that in the 1960s, television sets seldom exceeded 32 inches, and so were relatively small in comparison to what we accepted 40 years later. Hence, the monitors were too small to be viewed simultaneously by more than four or five people. The military wanted to have displays that could be seen simultaneously by dozens of people; and in the absence of large TV monitors, a projection system was required. Ideally, such a system would also allow permanent picture storage, which in the absence of then nonexistent videotape was only possible by conventional photography. Boeing had developed a system in which signals could be sent to a remote location, Wright-Patterson A.F.B., where it was imaged with a UV-emitting CRT, which exposed a piece of vesicular film, that could then be thermally developed and projected. Thus, with only a few seconds of delay, a permanent record was obtained that could be seen by a large group simultaneously. When the Boeing group that was charged with information technology heard about our instantly-accessible materials, they realized that a far more elegant system could be designed. This group, led by Mr. Mac Gardiner developed a system in which the key components were a UV-emitting CRT, with quartz fiber optics and a dichroic coating. The UVI-film was contact imaged on the face of the tube and projected onto a large area, with a light source, which due to filtering, lacked UV radiation. Thus the image could be projected as it was formed. By suitable optical arrangements, Boeing designed a system, which actually projected three different colored patterns onto a single screen (Fig. 4.4). Those of us who saw the system in operation observed a command post of a ship, with the suitable display. We understood that this program cost many millions of dollars and that the DuPont film was an essential component. By that time, we were able to coat a composition containing a HABI, leucodye salt, binders, and plasticizers on a 6-inch wide 4-mil polyester film with our laboratory coating equipment. Boeing wanted to
184
THE INVENTION OF DYLUX
Figure 4.4. The light-sensitive film coming from a supply reel is being imaged on the background exposure station, where nonvariable information (e.g., a map or a tabulation) is printed from photographic negatives. The exposed film is then transported to the faceplate of the CRT, where it is exposed through a fiber-optics faceplate containing a dichroic mirror, to UV light produced by the UV-emitting phosphors. Simultaneously, the film is projected with a 500-W tungsten source through a UV-absorbing filter from the faceplate via a lens to a large screen on the left of the diagram.
pay for the film, but May decided to give it to them free of charge. It would have been our only sale of a product for the electronic imaging market.
H. Is There a Business Here? Caris was fascinated by the concept of large-screen displays. He felt that this was a major business opportunity, as it merged computerized data management with a military and possible commercial need. Large airport displays, which to this day (2004) are mostly opto-mechanical, were conceived to be a major nonmilitary application, as was information management for business managers. It would eliminate the need for hand-drawn charts and or mechanical displays, by
WE START A MARKETING EFFORT: HOW DO YOU GET TO THE MARKET, ANY MARKET?
185
which data or figures would be moved by hand or mechanical means. Caris persuaded Orchem management that our unique materials and Boeing’s information-handling capability would make for a successful joint venture and a considerable amount of travel between upper management from DuPont to Seattle ensued, until it was decided that this application was far too speculative to provide any real financial incentive. By the end of 1968 our marketing group was expanded significantly, mostlyhowever, to help the DuPont Company’s Development Department track the different ventures on a comparative analytic scheme, which consumed great amounts of data, to allow evaluation of the prospects of success in a slew of applications. Each of the 30 ventures required inputs on technology, competition, prospects of developing technology to overcome competitive products, etc. Not only existing competition but prospective competition was evaluated. Thus when the 3M Company commercialized their Dry Silver technology, this was seen as a threat toward future Dylux applications; and more than half of the marketing group suddenly was assigned to evaluating 3M’s prospects, even though none of them had ever seen the 3M offering. The system of analysis was so complex that a full-time mathematician, Lane Bailey, was assigned to establish mathematical models for the input of our technology on the corporate Venture Analyzer. The principal result of all this was that it almost entirely paralyzed our effort to develop a real understanding of markets. Dylux was further handicapped in that there were so many possible applications that had to be analyzed. Of course, proofing, which later on became the most dominant application, was excluded from the analysis, because Caris felt that market was not big enough to warrant study!
I.
Opto-Magnetic Printing
The idea of using a photosensitive paper in connection with a high-speed printer seems of questionable merit after the development of electrophotography and ink-jet printing; but in 1967 it still was on the list of opportunities for our venture. Was the approach employed by DCI, using a lenticular plate printer, exposing the characters one at a time, flawed? Could a scheme by which one could print one line of text at a time with a single flash offer a better approach? Our engineering consultant, E. W. James, established contact with a young engineer, Richard Sinnott, who had worked on printers for Hewlett-Packard and who came up with some novel ideas after leaving that company. Sinnott proposed a scheme by which we would form characters by alignment of magnetic particles on a transparent support. Our venture supported him quite generously, and there were some initial successes; but in the end, the progress was not sufficient. Later,
186
THE INVENTION OF DYLUX
when we were looking for a simple way to print UPC-bar codes, we reconsidered this approach.
J. How to Get Rich? Our relationship with Mark Systems continued through 1967 and early 1968. Bernard Markus, the president and chief stockholder became a friend, and frequently asked me to modify materials for their tests. In time he said that Mark Systems would go public and that it would be possible for residents of Delaware and Pennsylvania to purchase stocks when the initial public offering became available. I was somewhat concerned with the morality of being involved with a company that would have an IPO and announce that they were working with DuPont. Markus and one of his investors had already visited Wilmington and discussed his company with Dr. Ed Gee, then the General Manager of Photo Products Department at a meeting, which I attended. Markus several times invited me to buy stock at $8/share, and I vacillated. I went on a trip to Rome Air Force Base in Upstate New York, mostly to discuss a flawed evaluation of our material but also to enthuse them in our possible reconnaissance film-duplication program. The mission was successful. When I returned to work the following morning, I found that all my colleagues in the marketing group were exuberant, for their shares of Marks Systems had already gone to $28. I then called Markus, and he sold me 50 of his shares for $400. A few months later when I needed some money for an engagement ring, I sold half of the shares, which had gone over $65, and retained the rest. Within a few months, the unsold shares had lost nearly all their value. Our relationship with Mark Systems continued, but in the end nothing came of it.
XII. 1968—MORE MARKETING ACTIVITIES A.
W. H. Brady Co. and the Kalograph1
In 1968 I was invited to visit the W. H. Brady Co., in Milwaukee, Wisconsin. This company had developed a very impressive line of labels and adhesive backed products, which were used widely for designating safety areas, mark containers, etc. They made specialty labels, such as were used to decorate football players’ helmets. One of their chemists had an idea for making a labeling system that involved a photochemical approach. I gave my usual presentation and was received with great enthusiasm. At that time, William Brady, the son of the founder, ran the company, and he was intrigued by what I showed him. Brady thought that a simple means of generating high-quality labels for a variety
1968—MORE MARKETING ACTIVITIES
187
of purposes would give them a product to compete with the raised-letter (by pressure) label makers than marketed by Dymo Corp. In time the Brady Co. wanted to buy Dylux solutions and coat them on an adhesive-backed film. They had designed a small optical printer, which involved different fonts and logos produced by a photographic technique, through which one would expose the photosensitive-coated material with a flash of light. We sent them some samples, but warned Brady that the decision to sell such materials had not been made. DuPont was reluctant to sell the chemical mix to anyone for fear that they would compete with us and make their own, competitive, proof paper. In the end, we did make and sell solutions to Brady, and they did indeed produce a compact label maker, the Kalograph, measuring about 10 by 10 by 9 inches. They coated the Dylux solution on films with different colored backing and sold it in cartridges. The labels were attractive; and to improve the light stability, Brady manufactured and sold a solution containing wet-fix chemistry. This was supplied in little applicators, similar to those containing shoe polish. The pricing of the machine was a challenge. At the initial price of $300 it was too expensive for many prospective users. The Brady machine unfortunately never developed a market. It was not clear whether this was related to the marketing effort, pricing, or other issues. On the technical side, however, it was a real pleasure to work with the scientists there, especially their research director, Dr. ‘‘Red’’ Strouse. He had an enthusiasm for Philadelphia-produced Tastycake1, and on my trips west, I generally brought him a good supply of these. On the return from Milwaukee, I always came home with a large supply of excellent sausage products, sold by the highly regarded Usinger’s Famous Sausage Store.
B. John de Campi—An Enthusiastic Marketeer In 1968, John de Campi, a very able young man whose background included an MBA from the Harvard Business School and a BS from the University of Rochester, joined our group. He too felt that the use of our technology in the printing industry represented a great opportunity and did a superb job of demonstrating the advantages of a dry instant-access material. John visited a number of printers and studied the operation of making printing plates and all the necessary prepress operations. He analyzed the cost of materials vs. the space required to use them, and the labor savings incurred by using a dry, instantly accessible proof paper. He calculated that a $0.07/square foot price was acceptable to the user of single-sided proof, and a higher price for the double side-coated version. John and I traveled around to visit printing companies, usually in the company of TRs; but what we could not do was to convince the Photo Products
188
THE INVENTION OF DYLUX
Management that they should market this product, and the Organic Chemicals Department simply lacked a distribution system to get a photosensitive product onto the market.
C. Mattel On several of our trips to California, my engineering colleague Ted James and I wondered whether there would not be an opportunity to develop toys based on photosensitive papers. We had once invited a number of children to play with Dylux paper and light sources and saw that there was considerable interest there. We observed that children found it fascinating to arrange objects over the paper, expose this assemblage to light, and then give it a flash with UV light to develop a picture. I contacted the Plastics Department, which sold materials to Mattel, and asked them to arrange for a visit. This was duly set up. Several days later I received a phone call from a Jack Ryan, who introduced himself as Director of Research at Mattel, and said he had heard about my impending visit. Ryan and I had quite a chat for nearly an hour. I was quite impressed that someone would make a transcontinental call for so long a period, for in those days, such calls were very expensive. At the end, Ryan said I should come out in a few days, not in a few weeks—he was excited at the prospect of working with Dylux paper. I said that we had scheduled our visit to coincide with an electronics show to be held in Los Angeles in a few weeks, He asked where I was staying and told me that my choice of hotels was poor—he would get me something better. Actually, because of the electronics show, Los Angeles hotels were fully booked, and he could not find a better one. His secretary called the following day, and suggested that I might enjoy staying at Ryan’s estate in Bel Aire. I said that my wife and a colleague were coming, and she said that that would not be a problem, there was plenty of room. I had some misgivings about staying at a potential customer’s home, but May said it would be okay, so I called back an announced that I would be delighted to accept the invitation. My wife, Nicki, and I flew to Los Angeles and drove to a rather impressive home on Nimes Road. The estate was in the process of being converted into a French castle, and large blocks of stones—from France—were strewn all over the property. I was asked by the estate manager to visit Ryan’s personal secretary, an attractive blond woman, who I later learned had been a Miss Sweden. She offered drinks and asked me to fill out two questionnaires, which resembled IQ and personality tests. When Ryan came in, I asked him if it was standard procedure to evaluate his guests in detail, and he said oh no, his secretary had mixed me up with someone who was applying for a position at Mattel. He checked my scores and concluded that I was smart enough, but probably too nice to work at Mattel! He treated Nicki and me quite royally; his French
1969—WILLIAM S. WARTELL: A DYNAMIC DECISION MAKER
189
chef prepared dinner, and then Ryan us to the Mark Taper Forum to see a Tennessee Williams play. We stayed with Ryan for a week; his home was certainly the most spectacular place I had been to up to then. The technical presentation at Mattel went well, and they agreed to develop a toy that would utilize Dylux paper. Mattel had earlier developed the Barbie doll, in whose development Ryan had a major involvement, as well as Creepy Crawlers1 and Hot Wheels1, which depended on using chemistry and physics to create best-selling toys. I was to visit Mattel many more times, and often stayed at Ryan’s home. A year later I suggested to our venture manager that he and I should visit Mattel, and that visit turned into a disaster. Dumont asked whether Mattel could predict how much paper they would require over the next 1, 3 and 5 years. Ryan said that with toys you either have a hit or a bomb, and that it would be difficult to predict sales for year 2 or beyond. Dumont concluded that DuPont would not want to get into a business with so little long-term potential. End of an opportunity!
XIII. 1969—WILLIAM S. WARTELL: A DYNAMIC DECISION MAKER By the end of 1968, May decided to leave the company, and the Marketing Group was entrusted to Walt Godfrey and Alexander Roe. It was soon quite obvious that we were not going anywhere. As a matter of fact, ever since the Marketing Group was established in mid-1966, it seemed to be directed toward evaluating competitive products rather than to establish a market for what we could deliver. All of a sudden, in February, a dramatic event occurred. I was on a field trip to visit potential users, when I learned on my return that a new division head of the Marketing Group had been appointed. His name was Bill Wartell, and he came from Jackson Laboratory, where he had been assigned after a brief sojourn at Pfanstiehl Corp., who had bought some parts of the DuPont Metals business, which involved some relatively rare metals, such as Columbium. Bill had been involved in the marketing of these, but did not like working at Pfanstiehl and returned to DuPont. I learned that Wartell had given a rousing speech to our somewhat demoralized marketing group, and had commented on several dissenters, who were not satisfied with the way things were going. It was quite obvious to me that the remarks were about me, because I had been very outspoken about our ‘‘marketing effort,’’ which seemed more directed toward ‘‘feeding’’ a computer program than exploring to opportunities for our technology. After all, I tried to marshal our limited resources toward proofing and not some long range, far-fetched, hard-to-achieve products.
190
THE INVENTION OF DYLUX
I went to Wartell’s office, and he said he was glad to meet me and that he had missed me the day before. I said that I had heard his remarks, which seemed to be about me, and I did not appreciate that, as I would have liked to hear about them face to face, rather than behind my back. Wartell was somewhat surprised, but sensed that we had to resolve our differences, and suggested we meet for dinner that very evening at the DuPont Country Club. After martinis, he asked me what I thought of the Venture’s operation, and I told him how frustrated we all were by the time-wasting Venture Analysis exercise, when we should have been out in the field moving our products. Our meeting ended only when the restaurant closed. Wartell then shook my hand, thanked me for being so honest and said that he would take appropriate action. The following morning he visited each of the people who were involved in Venture Analysis and reassigned them to other programs. He and I became great admirers of each other, and I am firmly convinced that he was one of the best things that ever happened to our program. He shared with me the feeling that proofing might be the real starting point for commercialization of our technology and realized that some sort of breakthrough was required—after all we were going to encroach on Photo Products Department’s territory. He arranged for de Campi and me to attend Photo Products Sales meetings, where we were allowed to talk about ‘‘experimental’’ materials, but were urged to be very conservative in our assessments of the future of our technology. I found this to be difficult, because we inevitably aroused a great deal of enthusiasm by the experienced field sales people, who were delighted at the prospect of having a new material that was easy to demonstrate and would provide them with a real excuse to visit potential customers. Gradually, we were breaking through.
A. Proofing, for Certain In April 1969 I was on a trip to St. Louis, to visit an Air Force Research Laboratory, where we wanted to show the capabilities of our research. The TR, who met me at my hotel, said that they had called to cancel the visit and suggested we visited a few printers instead. One account, which he said was a yellow box account, i.e., they used Eastman Kodak products exclusively, might be a hard sell. We met the plant manager in the lobby. He was fascinated by my demonstration of a 20-second proof. He said that if DuPont did not put this product on the market soon, it would only confirm his view that they had little comprehension of customer’s needs. Then he took us to lunch! Later in the afternoon, we visited another printer. The plant manager’s office was on the third floor, and I had to carry my small, but heavy contact printer up several steep flights of stairs. After we arrived, the manager apologized that he
1969—WILLIAM S. WARTELL: A DYNAMIC DECISION MAKER
191
had little time for us, and urged us to return the following day. I asked if he could spare a minute, and show me where to plug in an appliance. I made a 20-second proof for him, packed up, and said he could call when he had more time. His response: ‘‘Sonny, I suddenly have all afternoon for you!’’ As a matter of fact, he wanted to borrow my printer and some paper, so that he could give a demonstration at a Printers Club meeting that night. We left these for him and returned the following morning. He said that it would be nice if DuPont sold this product only in St. Louis—so that the local printers would have some advantages over the rest of the country. He was ready to buy. I got on the phone and called Bill Wartell. I told him that what we had was ready to sell, and while some improvements might be needed, these could be made once we had established a business. Wartell believed me. That afternoon he told Dumont that we should figure out a way of producing Dylux proof paper, and laid the foundation for a manufacturing plan.
B. Spring Boston SPSE Meeting In the spring of 1969, the Society of Photographic Engineers and Scientists (SPSE) held a meeting titled Unconventional Imaging Materials in Boston. We decided that this was a good opportunity formally to introduce Dylux to the imaging community, and I was selected to present a paper. Dumont had asked me to prepare this early in the year, and it was modified greatly by everyone. Even though many of the patents that covered the basic principles of our technology were already filed, Dumont was anxious to avoid including much chemistry in the talk. During the conference, there was also an opportunity to display samples of imaged materials and actually expose samples to demonstrate our materials. Here we had a great advantage over other processes, which could not compete, as they required processing steps, which could not be performed in the exhibit area. There were sections dealing with silver-based systems, organic colorforming systems, electrophotographic systems, and photopolymerization. A number of other presenters covered then-novel imaging chemistry. The session on organic color-forming systems contained these papers; ‘‘Recent Advances in Organic-based Imaging Systems,’’ by Dr. William R. Lawton of Bard Laboratories (Amherst, New Hampshire); ‘‘Dylux Instant Access Photoimaging Products’’ by Dr. Rolf Dessauer of DuPont; ‘‘A Novel Heat Developable, Heat Fixable Photographic Material—Warren’s 1264,’’ by Dr. Ben Millard of S. D. Warren Co., Westbrook, Maine; and ‘‘Dry, Heat-Fixable Nitrone Photoimaging System,’’ by Dr. Sheldon I. Schlesinger of the American Can Co., Princeton, New Jersey. With the exception of Dylux, none of the
192
THE INVENTION OF DYLUX
systems discussed survived for very long. Would they have been a commercial success if Dylux had not been invented? We shall never know. By that time, many people in the audience had been aware of DuPont’s interest in this area, as several patents had issued and a fairly extensive marketing effort had been under way. I felt quite proud that I represented the work of so many able colleagues and also sensed that the other presenters were a bit envious of all the resources that DuPont had been able to throw behind this development. What surprised me most was that there was no similar presentation from some of the then major companies in the imaging field: Eastman Kodak, Agfa-Gevaert, 3M-Company, or Polaroid Corp. These companies and others, of course, were represented in the audience, and some of their scientists interacted with my colleagues and me during the display session that preceded and followed the presentation. The audience was quite impressed with what we offered, and our technology was well received. The speaker following me, Dr. Millard of S. D. Warren talked about their Fotoproof1 technology and commented negatively on ‘‘the plethora of blues’’ of Dylux chemistry. Many in the audience were surprised to find one speaker making negative remarks about a preceding speaker. I had enjoyed giving my presentation, and while I explained the ‘‘dual response imaging system’’ I suddenly realized that our technology might lend itself to microfilm blowback. We had never considered that deactivation might proceed more rapidly if we exposed the coating to both visible and UV light, because this would allow formation of radicals, which could be reduced almost immediately by the reduction products of the quinones, as they were formed. Immediately after the session I got together with some of my colleagues and proposed an exposure system in which an image was projected with visible light onto a piece of Dylux 503 paper, which was simultaneously exposed to UV light. We demonstrated this with a 35-mm slide projector and some hand-held blacklite blue exposure sources. The effect was quite dramatic. When we returned to Wilmington, we immediately considered patenting this process of ‘‘coirradiation,’’ and approached several possible development partners with the concept. Canon Inc. was interested in this area, and visited Wilmington a number of times to develop products. While this approach to positive blowback offered some advantages, we made little effort to modify the coating recipe so as to maximize performance; and after a few months, we turned our attention to other opportunities. The process was described in U.S. Patent 3,661,461 (R. Dessauer, Co-Irradiation System for Producing Positive Images, May 9, 1972). We even extended it to a multicolor projection system (R. Dessauer, Co-Irradiation Method for Producing Positive Images Utilizing Phototropic Spiropyrans or Indenone Oxide or Dual Response Photosensitive Composition, U.S. Pat. 3,704,127, Nov. 27, 1972).
1969—WILLIAM S. WARTELL: A DYNAMIC DECISION MAKER
193
C. What Is So Good about Dylux1 Proofpaper? Printing plates were generated from lithographic negatives, which were photocomposed, and onto which additional information, either alphanumeric, or pictorial, was ‘‘stripped.’’ The quality of this composite determined the quality of the plate. If there was a typographic error in the text, or if the position of the ‘‘stripped’’ material was incorrect, the resulting plate was useless and had to be remade. Other flaws in the litho negative, such as scratches and pinholes, would similarly result in useless plates. Clearly, the printer was looking for a reliable, fast way of producing a single print from the lithographic negative or the plate. The latter, of course, was costly because it necessitated using a printing machine, adjusting ink, etc. Hence prepress proofing, exposure of silverphotographic materials for this would require processing, with attendant time delays. A 20-second exposure in ambient light without a processing step clearly offered advantages. What were the important properties that printers saw in this technology?
1.
2.
3.
4. 5.
6.
The ability to add on information to an exposed sheet by a subsequent exposure permitted sequential positioning of alphanumeric as well as pictorial information. The ability to form images of different densities by double burns allowed the printer to simulate different colors in the actual print. Here one might expose a portion of the proof to a density of 1.2, while another might be exposed to a density of 0.8, by reducing the exposure time. This would give a clearly distinct difference and was found useful in indicating that different printing inks would be used in different areas. Two-side coated papers, which could be imaged without a processing step allowed the second negative to be lined up against the image on the reverse side—the paper was sufficiently translucent to allow the cyan color to peek through it. This allowed the printer to easily produce signatures (large sheets printed with four or a multiple of four pages that when folded becomes a section of the book). The foldability of the paper was important, to prepare signatures easily. As the Dylux paper coatings possessed a fairly low gamma of around 0.7, it was possible to produce attractive prints from continuous-tone negatives. The dry process eliminated the problems of shrinkage or expansion inherent in wet processes, and dimensional stability of the proof was considered very important.
194
THE INVENTION OF DYLUX
7.
The cyan color, which we thought would be less attractive than black actually offered an advantage—the user knew he or she was dealing with a proof and could mark it up with a black or colored pencil. 8. Negative images were formed in a single step. The photo deactivation was only necessary if the imaged proof was to be handled in daylight. 9. When the proof was handled in room light it eventually ‘‘fixed’’ itself. The exposure process could be carried out in subdued or yellow light, handling of the unexposed paper was not much of a problem, and exposure could be controlled readily. 10. Either negative or positive images could be formed with the same material. 11. The paper was priced to reflect value-in-use, and customers soon realized that the labor-saving achieved by obviating the need for processing allowed greater productivity. By this time, we had found that the addition of matting agents, such as silica, would improve the ‘‘tooth’’ of the paper—i.e., it would accept ink or pencil writing. We soon realized that we had lucked out, and had put together a new product that filled a market need. Had all these requirements been presented to us before we started our research, we would probably not have succeeded with so complex a task. A triumph for pragmatism! From this point, technology advances were primarily in coating as reported in the cited references. Preparation of papers coated on both sides with photosensitive composition. Connair, Michael John (DuPont de Nemours, E. I, and Co.) U.S. 3,674,534 (Cl. 117-34; B 44d), 04 Jul 1972, filed, 30 Jan 1970; 4 pp. Blister-free photosensitive paper is obtained by coating opaque paper (< 10 mil thick) on both sides with a photosensitive compn. contg. for example a leuco dye, a photooxidant, a film-forming binder such as poly(vinyl alc.), and a volatile org. solvent boiling between 55 and 85 . The paper is coated 1st on 1 side with this compn. in an amt. suffcient to provide a dry film coating of 6–15 lb/3000 ft2. Of paper, dried by evapg. the solvent at a rate to prevent blister formation, and coated on the other side with the same amt. of photosensitive compn., but in multiple stages of coating and drying such that in each stage 10–60% of the total wt. is applied.
Were there any limitations? The resolution should have been excellent because of the molecular nature of the image; but in actual practice it suffered somewhat because addition of a matting agent resulted in some scatter, and the UV-reflectance of the paper base added further light scatter, reducing image acuity. The coatings initially gave off some odors, as a result of the formation of acrylonitrile from the hydrogen donor, and an alternate hydrogen donor,
1969—WILLIAM S. WARTELL: A DYNAMIC DECISION MAKER
195
Figure 4.5 The dual response of Dylux1 material. (Top) Nagative image formation by imagewise exposure to UV light. (Bottom) Positive image formation with initial imagewise exposure to visible light followed by flooding exposure with UV light.
triethanolamine triacrylate was substituted for it. Shelf life of 2 years was considered to be acceptable, and even if the coatings underwent some loss of sensitivity on prolonged storage, this never turned out to be a problem. The photofix system was optimized by careful selection of components, with the principal contributors being Dr. Howard Gerlach and Dr. Cyrus Henry.
Dual-response photosensitive compositions containing an acyl ester of triethanolamine. Gerlach. Howard G., Jr.; Looney, Catharine E. (DuPont de Nemours, E. I., and Co.) U.S. 3,658,543 (Cl. 96 90; G 03c) 5 Apr 1972, Appl. 99,512, 18 Dec 1970, 15 pp. Improved storage stability and reduced sensitivity to human handling of the photoimageable/photo deactivatable compns. described in U.S. 3,390,994 are obtained by replacing the nitriloalkanoic acid ester reductant with an acyl ester of triethanolamine, such as N(CH2CH2OAc)3. Thus, a suitable coating compn. contains Me2CO 54 ml, Me2CHOH 6 ml., 2,20 -bis(o-chlorophenyl)4,40 ,5,50 -tetrakis(m-methoxyphenyl) biimida-zole 0.4180, tris(l-diethylamino-otolyl)methane 0.0900, p-MeC6H4SO3H.H2O 0.400, 9,10-phenanthrenequinone 0.054, cellulose acetate butyrate 6, poly-ethylene oxide adduct of o-phenylphenol 3 g., and l ml of N(CH2CH2OAc)3.
196
THE INVENTION OF DYLUX
Dual-response photosensitive compositions containing an alkylbenzenesulfonic acid and an arene sulfonamide. Henry Cyrus P., Jr.; Jeffrey, John R. (du Pont de Nemours, E. I., and Co.) U.S. 3,658,542 (Cl. 96/90; G 03c), 25 Apr 1972 Appl. 99 511, 18 Dec 1970, 13 pp. The use of arene sulfonamide plasticizers in the photoimageable/photo-deactivatable compns. of U.S. 3,390,994 and 3,445,234 imparts in them outstanding storage stability with freedom from crystn. and phase sepn. Thus, a coating formulation, which is free from crystn. and phase sepn. on storage, contains Me2CO 54, Me2CHOH 6, N(CH2CH2OAc)3 0.4, N(CH2CH2CO2Me3 0.6 ml, 2,20 -bis(o-chlorophenyl)-4,40 ,5,50 -tetrakis(m-methoxyphenyl)biimidazole 0.418, tris(4-diethylamino-o-tolyl)methane 0.0900, dodecylbenzenesulfonic acid 0.70, 9,10-phenanthrenequinone 0.0486, cellulose acetate butyrate 6, polyethylene oxide adduct of o-phenyl phenol 1.14, pyrenequinone (a 1:1 mixt. of 1,6- and 1,8isomers) 0.003, and p-MeC6H4SO2NHEt 1.3 g.
In assessing the value of Dylux as a proofing medium vs. competitive wet processes, de Campi had accurately predicted higher value in use for a material which could yield a proof in a fraction of the time of that previously required, as well as that the dry process eliminated space required for wet processing and drying associated with other processes. The most serious limitation was to be the need to install dedicated equipment in order to form proofs with good image density. Printers had light sources rich in UV light, but these also emitted significant amounts of visible light, thus partially ‘‘fixing’’ the papers as they were imaging, resulting in unacceptably low visual contrast. An inexpensive light-filtering system was clearly required to modify such exposure sources! At several SPSE meetings I had met a German chemist, Dr. D. Schultz, who had a high position at Agfa-Gevaert in the United States. He had followed my attempts to bring this technology to the market, ever since I presented the first technical paper at an SPSE meeting in Boston. Once he asked whether Agfa-Gavaert could market this material, if DuPont had no interest. I urged him to send me a letter to this effect. When the letter finally came, we passed it on to Photo Products management. They then decided that perhaps they should consider marketing Dylux after all. In the fall of 1969, Dylux 503 proof paper was introduced to the industry at the National Association of Photolithographers In Chicago. Reception was very favorable, but the market grew only after exposure equipment became available. As this occurred, there was steady growth, and sales increased steadily every year, well into the 1990s, when the digital revolution finally took its toll of the conventional proofing business. In time, two-sided cyan-forming proof paper became the biggest Dylux product; but there also were neutral shade proof papers, Dylux 515 and later Dylux 535, as well as the registration master film, Dylux 608. Dylux 503 paper itself was superseded by Dylux 503A and Dylux 603B, which contained different photo-oxidants. As a matter of fact, the term blue-line became an acceptable
1969—WILLIAM S. WARTELL: A DYNAMIC DECISION MAKER
197
synonym for proof paper, no doubt, because most printers were familiar with Dylux 503. In time it was realized that Dylux proof paper offered a ‘‘pull-through’’ for other Du Pont graphic arts products. Our TR’s could visit customers to demonstrate Dylux, which everyone wanted to learn about, and additionally demonstrate some lesser known products, with occasional sales of the latter. Dylux 503 proof paper was the first Du Pont product for the prepress proofing market, and was followed by many others.
D.
Europe
Wartell always thought big. So he arranged for me to spend three months in Europe to determine if there was a product opportunity there for Dylux paper. I was delighted to have an opportunity to see more of Europe and my wife, Nicki, was happy to come along for part of the trip. It started ominously. We drove to New York so that we could leave a car at my mother-in-law’s home, and went out to dinner—only to find out that the house had been burglarized. Fortunately, the thieves did not take any of the Dylux samples, printer, flashguns, filters, and literature, which was part of our luggage. We took a day flight and arrived in London; we stayed at the Hotel Brittania, where we ran into a group of DuPont colleagues, who had stopped over in London from a Fluorine Chemistry Symposium in Moscow. They celebrated their return to the West with a wild party, which the hotel closed down. The following day I was met by several of the English tech reps and was taken to potential customers. Everyone was polite, but told me that they were less interested in a negative then a positive working system. Still, people were impressed. After a few days I was to go to Frankfurt, and give a talk to the R&D group in DuPont’s Neu Isenburg plant. Here was an opportunity to show off my linguistic skills, and the group was delighted to have a German-speaking American to listen to. I started off in German, but suddenly was asked by the American director of the operation to speak English. Like many other transferees, he was unfamiliar with the local language. I was to be directed in my efforts by Mr. Pete Piateski, who had been assigned to Germany to bring some American methods to the operation. Unfortunately, Pete, whom I had known for some time, also spoke no German, and did not especially like his German colleagues. As a result, my marketing efforts were bounced around, and finally a Herr Joachim Rau directed my efforts. He, and a Herr Rudi Kuempel, who was a marketing manager sent me to many places, and inevitably, I was told that the Germans wanted a black-and-white proof, a positive proof, a better quality paper, a lower price, etc. There was little positive response. None of desired properties was beyond our technology, and so the
198
THE INVENTION OF DYLUX
question was whether I could persuade my American bosses to undertake research in these directions. Piateski and I visited Paris together and had several interesting days with the Photo Products group there. We visited some would-be customers; and in general, there was admiration for the cleverness of our chemistry, but no one was ready to buy. Piateski also sent me to Sweden, with similar results. In Germany, the most memorable visits were in Kiel, to Hell GMBH, where Dr. Hell, the founder of that great company shook my hand and congratulated me on the technology, and to Daimler Benz, where the manager, to whom we showed our possible products, was so intrigued that he started exposing papers and testing them on the spot. Until his boss came in and told him that he was not to do that sort of thing—he had assistants for that. Finally, in October 1969, there was the annual European graphic arts show in Milan, Italy. I was given a space in the DuPont booth, and important customers were brought in for me to show them one of the new DuPont technologies. Nearly everybody had nice words to say about Dylux. The Italian distributor of DuPont Photographic materials took an especial interest in Dylux and me, took me to several of his better customers, and a large number of excellent lunches and dinners. Toward the end of the show, there was a ‘‘family dinner’’ for all the DuPont people; Phil Wingate, the General Manager, attended it also, and everyone was complimentary about my contributions to the company’s European business. I felt that I had made a lot of friends—even among people who did not get along with each other. Case in point: Piateski did not even attend the show, saying that unlike Germans, he did not believe in travel when there is work to be done in the office! I returned to America in mid-October, with full enthusiasm for making the changes in technology that would ensure some market opportunities in Europe, but these fell on deaf ears. There were too many domestic crises, and besides that, if the Europeans could not use what we wanted to give them, they would have to do without. A year later, the German DuPont organization gave Dylux a real try. Under Piateski’s guidance, they assigned Herr Dieter Kuckei, an up-and-coming marketing man to explore uses for Dylux technology broadly. He would have had an easier time if we had made some of the modifications in our technology that I had suggested. Kuckei sought to move Dylux into nongraphic arts fields, as well. At one German graphic arts show they had a man, dressed as a sailor, make tattoo-like images on his arm, using a Dylux-spray solution, a series of negatives, and a flashgun. By the time Kuckei was making the needed contacts, the Dylux Venture had been transferred to Photo Products Department, and I was no longer part of it. I could help him only by by-passing the Wilmington management, and this did not necessarily help either his or my careers.
1970—EXPANSION OF OPPORTUNITIES
199
XIV. 1970—EXPANSION OF OPPORTUNITIES Our management decided that we should have available a publication that would describe the basic characteristics of Dylux technology. I wanted to make this a scientific publication, but this approach was rejected on the basis that imaging scientists really did not care what kind of chemistry was involved, as long as the materials worked! Naturally, as a chemist, I disagreed. Du Pont Dylux1 Instant Access Photosensitive Products. Dessauer R. Image Technol. 1970, 12, 27. The photosensitive coatings contain a mixt. of org. compds. Colored images are obtained by exposure to uv radiation without chem. processing, and light is used to stabilize the nonimaged areas. By control of the sequence of exposures, either pos. or neg. images can be obtained.
A.
Military Applications
One of the major competitors for our technology, from the standpoint of obtaining military funding was Photohorizons, a small research company located in Cleveland, Ohio. Their director of research, Dr. Eugene Wainer, had been honored at several SPSE meetings for his contributions to free-radical photochemistry, and he was very successful in obtaining funding from the U.S. Air Force. The Photohorizons technology was somewhat similar to ours in that instantly accessible images could be generated by exposure to light, but it lacked a good fixing method and involved the use of relatively toxic materials such as carbon tetrabromide, in combination with aromatic amines. Photohorizons claimed ever greater imaging speeds, and indicated that they could extend their activation spectrum to longer and longer wavelengths. They claimed black and near black images. Fixing was to be accomplished by application of heat to drive off the unreacted halocarbons. As they strove to higher and higher speeds, the preimage shelf life decreased, and their materials had to be delivered under refrigeration. Somehow, they succeeded in convincing the airforce that they could in time solve their shortcomings. Thanks to the small office, that DuPont then maintained in Washington, staffed by Gus Davis, Fred Gerretson, and David Peek, we were able to make contacts with the military as well as government agencies. They in turn encouraged contacts with the U.S. Navy Photographic Center, and Beale AFB, in Marysville, California. The Navy laboratory, headed by Mr. Glen Bull was especially sympathetic to our offerings and made considerable efforts to help us. In this my colleague, Garrett Forsythe, who had been a naval aviator in the Korean War and could communicate with military people far better than I, helped us. He was especially successful with the group at Beale, persuading them that
200
THE INVENTION OF DYLUX
instant printout of aerial reconnaissance film by the photo-interpreters would be a great advantage over then existing processes. Garrett was able to get our research group to produce a blue-black film that could be contact printed in a device which he built himself one weekend, when he had become frustrated with the lack of speed with which we could manufacture instruments. We learned how important personal contacts could be and were helped a lot by an able Photo Products Manager, Ron Stewart, who worked on the West Coast. The Navy Photographic Center in Washington, D.C., looked for a nonsilver color duplicating film. This would be difficult, but might have been developed with our RF-technology, as we had shown that we could initiate photopolymerization over a wide range of wavelengths. At that time, even a two-color duplication film sounded interesting, and Bull gave generously of his time to help us. We identified other military applications. One, which looked exciting, was film annotation. At that time, aircraft maps would be on color filmstrips, which would move as the plane traveled. Pilots were looking for ways to add information, such as special sites, to such maps, and applying a Dylux coating over the mapstrips, and then imaging this offered a route to accomplish this. We showed that we could contact print or even project an image, which could be stripped off when it was no longer desired. Inevitably, we ran into the Air Force’s support of Photohorizons. They obviously were reluctant to withdraw support from them, and back a different, newer system. So, Colonel Hoy, who ran that office, suggested we interact with Photohorizons to see if we could help them to get into a more practical system. Dumont, Dr. Ray Panos, and I went to Cleveland in October 1969; and after a somewhat awkward meeting, where Photohorizons’ marketing manager, a Mr. Cameron, decided that I should not participate in the meeting, because I was so committed to Dylux technology. Gene Wainer said he was confident that I would not do anything dishonest, and I could participate. Our conclusion was that their system could not ever be coated in our coating facility at Towanda, as it involved toxic halocarbons, and, while we were all impressed with their achievements, we could not help Photohorizons.
B.
What Is/Was Wrong with Dylux1 Proofpaper?
The Dylux 503 formulation contained a HABI, the leucodye TLA-454, a mixture of quinones, triethanolamine triacetate, several plasticizers, an organic acid, cellulose acetate butyrate binders, as well as antiblocking agents (fluorinated derivatives) as well as a silica derivative to provide tooth to the coating. The selection was made to provide maximum performance at minimum mill cost. The paper substrate required had high holdout, so as to permit two-sided coatings as well as to minimize the wicking of chemistry into the base.
1970—EXPANSION OF OPPORTUNITIES
201
Dylux 503 proofpaper had several weaknesses, which probably limited its penetration into additional markets. The formulation, paper base, and coating method were optimized for maximum profitability and not for maximum performance. Proof papers are essentially a temporary medium and are not expected to have a very long period of life after usage. Resolution of images as formed was adequate for the application but did not approach what could have been achieved with a molecular imaging system. The reflectance of the paper base resulted in light scatter, as did the introduction of a matting agent, which allowed the coating to be written on. Years later, we found that the incorporation of small amounts of phenidone enhanced image quality resulting from the scattering of light. The cyan image was not the most attractive color, but it employed a leucodye that could be readily manufactured, possessed excellent dark stability, and gave an intense image on exposure. Attractive black-and-white images were made, but the mill cost, i.e., cost/square foot was appreciably higher. Inevitably, there was a desire to make a higher-speed product. While this was a worthwhile goal, it was not important for proofing, where most of the time is spent evacuating the contact frames in which the materials were exposed. Making a singleexposure proof might take 80 seconds, with 60 seconds for evacuation of the contact frame. A significantly faster product would have made the ambient light handling significantly more difficult. Dylux 503 is essentially a negative working product, and for the American market, this was very desirable. I demonstrated a proofing product in which the deactivation speed was appreciably higher and that could be used in diazo machines for deactivation, followed by a UV exposure to form color. It moved through diazo machines at speeds in excess of 60 ft/min. However, our marketing groups never pursued this very actively. Another failure of the cyan Dylux images was that they could not be replicated on the same medium. Thus we could not copy Dylux 503 onto itself. We did demonstrate that it was possible to take black Dylux images, made by using a purple-forming and a yellow-forming leucodye, and expose these in a positive mode, and replicate them. We coated such materials on a transparent polyester base, and made copies therefrom. In addition, the moisture sensitivity of cellulose butyrate binders tended to influence the image density, yielding better image density when the humidity was high. We realized all of the above when we introduced Dylux proofpaper to the market, and some improvements were certainly made during the course of the product life. It is important that the advantages of simplicity, convenience, and relatively low cost, were sufficient incentives for the proofing community to use ever-expanding amounts of Dylux 503, making it a very successful product. The most significant changes over the course of the product’s life were the substitution of different hydrogen donors and plasticizers as well as different hexaarylbiimidazole derivatives.
202
THE INVENTION OF DYLUX
C. Transfer of the Dylux1 Venture The Dylux and HABI technology were developed to provide novel chemistry that could be manufactured at the Chambers Works. The manufacture of a coated product and marketing required different skills than were available in the Organic Chemicals Department. Still, the Organic Chemicals Department persisted with the development of the technology from 1961 to 1970, and even after that, as they deemed the chemical manufacture of these products to be an attractive business. Sometime in late 1970, the company decided that the Dylux group would be transferred to the Photo Products Department; and Dumont called me, and several of the more senior people in the group to advise us of the transfer ‘‘lock, stock, and barrel.’’ We all hoped that we would profit from this and that we could then draw on the expertise of people who had been in the imaging field long before most of us. I was on a field trip with Jim Clark, who was given the assignment to probe military contacts, when I called my secretary from the Denver airport. She said that I should call Dumont at home as soon as possible. When I did he said that he was sorry to tell me that the venture transfer was not as had been planned, and that only about half of the group would move to the Photo Products Department. Among the ones not going would be Dessauer, de Campi, Terss, and a few others who had been involved with the program for considerable time. I was shattered. Clark and I continued our trip to Beale AFB; then to Ogden, Utah; and finally to Drake AFB. I had no idea what would happen to me, and there was no one to talk to. Soon after the Dylux Venture had been transferred to the Photo Products Department, there was one more visit with Mr. Bull and Colonel Hoy in 1971, but Dylux and the military never came together again. The Photo Products Department was not then interested in military projects and/or funding. A lot of effort to build up friendships and understand needs in areas beyond our horizon was lost. I had given all my energies and enthusiasms to the development of a technology, which I understood better than anyone else. What weird decision making was going on in Wilmington? That was in December 1970. By 1972, I concluded that it was fortunate that I did not move to Photo Products; by 1986, I concluded it was the best thing that ever happened to me and Dylux.
XV. 1971—AFTER THE VENTURE Even though Coraor was no longer associated with the program, he was enough of a scientist years later to dedicate himself to publishing some of the HABI
1971—AFTER THE VENTURE
203
chemistry in a series of articles in the Journal of Organic Chemistry and encouraging the publication of a number of related papers. Had he not done this work, only the patent estate would have been a document of this research! Flash photolysis of a substituted hexaarylbiimidazole and reactions of the imidazolyl radical. Coraor, G. R., Riem R. H.; MacLachlan, A.; Urban, E. J. (Exp. Stn., E. I. DuPont de Nemours, and Co., Wilmington, Del.). J. Org. Chem. 1971, 36 (16), 2272–5. (Eng). The rate of reaction of 2-(o-chlorophenyl)-4,5-diphenylimidazolyl radicals (L with additives was studied in various solvents. Evidence based on measured rate consts., including kinetic D isotope effects, prove that the rate detg. step in the reaction L þ aromatic amine is an electron change reaction at the amino N, while in the reaction L þ hydroquinone the rate-detg. step is H abstraction. Reactions of triarylimidazolyl free radicals. Coraor G. R.; Cescon, L. A.; Dessauer, R.; Deutsch, A. S., Jackson, Harold Leonard; MacLachlan, A.; Marcali, K.; Potrafke, E. M.; Read, R. E.; et al (Exp. Stn., E. I. DuPont de Nemours and Co., Wilmington, Del.). J. Org. Chem. 1971, 36 (16), 2267–72 (Eng). Triarylimidazolyl free radicals oxidize electron-rich substances by rapid electron abstraction from tertiary amines, iodide ion, and metal ions and H atom abstraction from phenols, mercaptans, primary and secondary amines, and activated C-H compds. The rate consts. for electron abstraction from tertiary amines were related to sþ values via oxidn. potentials which were detd. by cyclic voltametry. Properties of triarylimidazolyl radicals and their dimers. Coraor, G R; Cescon, L.A; Dessauer, R.; Silversmith. E. F.; Urban, E. J. (Exp. Stn., E. I. DuPont de Nemours and Co., Wilmington, Del. J. Org. Chem., 1971, 36 (18), 2262–7 (Eng) The electronic spectra of 2,4,5-triarylimid-azolyl radicals (I) are strongly influenced by substitution on Ph rings. The rates of disappearance of I in benzene at 27 vary over 100 fold with substitution of Ph rings. Any ortho substituent in the aryl group increases the rate const. relative to position isomers, a fact consistent with radical destabilization by ortho substituents through steric disruption of ring coplanarity. Substituent effects on the reactivity of triarylimidazolyl free radicals toward tris(2-methyl-4-diethylaminophenyl)methane. Cohen, Robert L. (Exp. Stn., E. I. DuPont de Nemours and Co., Wilmington, Del.). J. Org. Chem. 1971, 36 (16), 2280–4 (Eng). The effect of aryl substituents on the reactivity of substituted triarylimidazolyl free radicals, photolytically generated from their corresponding dimers, in an electron exchange reaction with an aminotriphenylmethanc substrate was studied by flash photolysis, The reaction rate was retarded by electron-donating substituents and enhanced by electron-withdrawing groups on the imidazolyl radical. Bulky substituents ortho to the imidazole ring, also increased the reactivity. Biimidazole-sensitized photooxidation of leuco triphenylmethane dyes. MacLachlan, Alexander; Riem, R. H. (Exp. Stan., E. I. DuPont de Nemours and Co., Wilmington, Del.). J. Org. Chem. 1971, 36(1G), 22/5-80 (Eng). Oxidn. of tris(2-methyl-4-diethylamino-phenyl)methane by photogenerated 2-(o-chlorophenyl)-4,5-diphenylimidazolyl radical (L ) was studied by flash photolysis. An electron-exchange reaction involving L occurs at an unprotonated amino N of the
204
THE INVENTION OF DYLUX
leuco dye and is responsible for the first oxidn. step. Subsequent reactions do not involve the L radical and depend only on the structure of the leuco dye and environmental effects. The influence of pH on both the course and rate of the dye-forming reaction is discussed.
In January 1971, those who were not transferred to the Photo Products Department faced an uncertain future. Bob Terss went back to being head of the patent group at Orchem. I was assigned to an exploratory division headed by Bill Remington, where our work on photochromism started 10 years earlier. A few people, like Max Deinzer and Bob Engel were to lose their jobs altogether. The people who went to Photo Products Department continued to report to Lee Dumont, who remained the Venture Manager—for less than 6 months, when the whole group was split up, and research on Dylux chemistry was completely discontinued, except for a limited effort at Towanda, the manufacturing site for DuPont’s nonsilver imaging products. Among these was Tom Sheets, an engineer who had joined the Dylux Venture a few years earlier. Sheets was a remarkably dedicated individual. The survival of Dylux in the rough waters of the Photo Products Department depended very much on his remarkable mix of talents. He shepherded Dylux through a large number of manufacturing crises until he retired on 1985. Before he retired, he managed to shift his enthusiasm and talents to David Logrando, another engineer who deserves a huge amount of credit for his quiet work. I seriously doubt that Dylux would have been a commercial success without Tom and Dave. Most of the transferees were not happy. Many had been trained as organic chemists and were suddenly shifted to totally different tasks. Peter Strilko became involved in the marketing of Riston1 Photoresists. Howard Gerlach wound up working in the medical products field. Dr. Wlliam Hardam, retained his supervisory position, and some years later was briefly involved in a Dylux program. I was puzzled for the longest time why I was not allowed to continue working in the Dylux Venture. Ultimately I concluded that the animosity, which had developed between Robert Upson, the Photo Products Department’s authoritarian director of research and me was the cause. Upson believed that research was directed from the top down and those who strayed and made unprogrammed inventions were not to be encouraged. And what was worse—I had tried to get around the ‘‘system’’ and sought support from marketing people, who did not report to him!
A. Photodecoration Remington suggested that I examine some work done at Central Research Department on liquid crystals. Two scientists, Walter Mahler and M. Panar, had discovered that certain half-acid esters of cholesterol gave rise to solid compositions, which possessed the iridescent qualities of conventional cholesteric esters
1971—AFTER THE VENTURE
205
(W. Mahler and M. Panar, J. Am. Chem. Soc. 1972, 94, 7195). This sounded interesting, and offered me an opportunity to do some exploratory marketing as well as research. After all, I had worked on steroids as a graduate student and at least had no difficulty drawing their complex structures. Sometime in the spring of 1971 I attended a development conference for General Motors; these conferences were designed to create interactions between two companies and to identify sites for cooperation. I sat next to a Bob Ellefson, who told me that he had been a designer of leather patterns for Corfam1, the leather-substitute that DuPont had developed with considerable effort and recently terminated. I showed Bob some samples of our solid liquid crystal materials and he immediately suggested we have lunch and see where he could help me with my marketing efforts. Bob was highly regarded in the footwear industry and said that he could find some people who would consider our new materials, which we had dubbed Irlux in the fashion shoe business. As Bob and I talked, I naturally told him about Dylux chemistry, and he thought that here too, there were opportunities in ‘‘photodecoration.’’ Ellefson took me to see a leading shoe designer, Nancy Quinn, who worked with Genesco, one of the major manufacturers in the garment industry. Both she, as well as the Chairman of Genesco were enthusiastic about the concept of finding a totally novel way of decorating leather. Ellefson and I also made contacts with George Newman Co., of Boston, which had bought out the Corfam inventory and was interested in our approach of photodecoration. In one brief marketing trip we had uncovered a potentially large use for our novel chemistry in an industry that was quite stagnant. Ellefson had many friends at the Fabrics and Finishes Department, the organization that had developed Corfam. Among these was Jack Lowe, who headed the furniture finishes group stationed at Greensboro, North Carolina. Ellefson quickly arranged for a visit there; and before long, we had an active program to apply photosensitive lacquers to plywood and other low-cost wood compositions. Ellefson indicated that the ability to put intricate designs on case goods or other furniture offered the manufacture a flexibility, that would be commercially attractive, and contacts with Bassett and Thomasville established the validity of this. At Greensboro, Wayne Deveise, a talented chemist was assigned to work with me; and we devised a system in which a blocking coat, which separated the wood from our photochemistry, was applied to the surface, followed by a photosensitive layer containing leucodyes and biimidazoles in a furniture lacquer binder. The structure was then imaged, and finally a fix coat, containing a phenol, UV absorber, and a metalized hydroxyazo dye, was applied. Harold Wilbur, who had built so much equipment for us in the early days of the Dylux Venture also built an exposure unit which he and I delivered by truck to Greensboro. When the managers of the Equipment Division of Photo Products heard about that, we were not applauded for our entrepreneurial spirit but rather
206
THE INVENTION OF DYLUX
chastised for having raised a potential conflict of interest between an employee of Du Pont, his private business and the company. As furniture was generally coated with many, many layers of lacquers, this approach was quite acceptable. Ellefson supplied the lab with high-quality photographic negatives, and we made significant progress in relatively little time. We tested the light-fastness of our imaged surfaces and found that they exceeded that of conventional furniture. Everything was going along famously. Within a few months, we had identified new applications for Dylux chemistry, which definitely were outside the province of Photo Products Department and would expand the markets for the chemicals, which the Organic Chemicals Department planned to produce. It was unbelievable train of events. The Dyes Marketing Division agreed to support this work, and Ellefson and I were assigned to Robert Moyer, a sales type, who was sympathetic to a new, unique product line. By 1972, it was quite apparent that there were a number of opportunities for photochemical color formation outside the graphic arts area, which was Photo Products Departments business area. We now had a reason to manufacture the Dylux chemicals for that department, but anything else that we could do, would simply expand this business. This time we tried to keep the organization small. Remington gave technical direction, Moyer was to define business opportunities, and I was in charge of doing the technical work. By that time, it was agreed that we would have ample technician support, and with Looney to provide me with a sounding board for new chemistry, we had a very efficient force. Ed McBride took care of some of the other technical problems. We had a small empire, with a few very able former technicians, including. Mike Dycio, John Brawders and C. Gerald Chetkowski whom I was able to rescue from the Photo Products Department. And, of course, Aurelio Baccino to make novel leucodyes. A patent on the concept of photodecoration was applied for. U.S. Patent 3,847,608: Photodecorating Sheet Material with Matched Colored Designs Dessauer; Rolf, Greenville, DE and Ellefson; John R., West Chester, PA, E. I. du Pont de Nemours and Company, Wilmington, DE Issued/Filed Dates: Nov. 12, 1974 filed Aug. 8, 1972 Application Number: US1972000278778 IPC Class: G03C 005/04; G03C 005/00; Class: 430/352; 002/243.1; 430/351; 430/ 394; 430/538; A process for color decorating sheet material with a design, which material will be used to form an article by assembling multiple pieces of the designed sheet material, whereby the shapes of the pieces are laid out on sheet material in closely-spaced arrangement and the pieces are then cut from the sheet material, is improved by A. either before or after cutting out the pieces, coating the sheet material with a photosensitive composition capable of generating color subsequent to exposure to activating light, B. exposing the coated shapes to activating light through a light modulating means which registers at least a portion of the desired design on the shaped piece, the piece design being one which matches with an adjacent piece in the assembled article, and C. color stabilizing the design
1972—WHAT ELSE?
207
by fixing the coating after registration of a latent design or after color has been generated. Such a process permits minimum waste of sheet material and maximum convenience in accurately matching piece designs in an assembled article.
XVI. 1972—WHAT ELSE? Kay Looney and three colleagues had sought to quantify the imaging system in a computer simulation. This work, begun years earlier, was published in 1972. Long Exposure density reversal in dye-based imaging systems. Gordon Michael D.; Looney, Catharine E.; (Org. Chem. Dep., E. I. DuPont de Nemours and Co. Inc., Wilmington, Del.). Photogr. Sci. Eng. 1972, 16 (2)151–6 (Engl.) Sensitometric curves of imaging systems based on formation of organic dyes in viscous media show long exposure ‘‘d. reversal.’’ In such cases the concn. of the product dye increases monotonically, reaches a max., and then decreases with continuing exposure and finally stabilizes. The conditions are (1) a high rate of absorption of photons, (2) a large excess of photoinitiator; and a highly viscous medium. The obsd. ‘‘d. reversal does not arise from dye instability, but is a natural consequence of reaction kinetics the viscous system. Computer simulation of the chemical dynamics in a complex imaging system. Looney, Catharine E.; Gordon Michael D.; Laird, J. P.; James, E. W. (Org. Chem. Dep., E. I. DuPont de Nemours and Co. Inc., Wilmington, Del.). Photogr. Sci. Eng. 1972, 16(6), 433–42 (Eng). A computer simulation has been successfully applied to Dylux films and papers under a wide variety of imaging conditions and formulation variables. The math. models of both uv-activated color formation and visible light deactivation involve multistep reaction sequences. Fine adjustment of the estd. rate consts. is accomplished by optimizing the fit between computer and actual exptl. results. Once so established, the model permits computer generation of practical sensitometric curves as functions of intensity, wavelength (mono- and polychromatic) for various sample thicknesses and concns. of actives. The program includes light absorption characteristics of any transient inter mediates as well as those of sensitizers, actinic species, screeners, and products. Specific examples of the effects of the independent variables on the sensitometric properties of Dylux are discussed
Our work with the Greensboro F&F group went well, primarily thanks to Wayne Deveise and several of his colleagues, who were delighted to have some more interesting chemistry to develop. Our major technical effort was directed at finding a color combination that would yield a neutral shade image, rather than the cyan, which resulted from the photooxidation of TLA-454. We found that the mixture of the leucodye LB-8 and an orange-forming dye gave an attractive neutral shade. The leuco-orange dye itself had an interesting past. In the 1950s, the DuPont Company’s Central Research Department, under the leadership of T. E. Cairns made a very dedicated effort to explore the chemistry
208
THE INVENTION OF DYLUX
of tetracyanoethylene. This program was very extensive, and one of the proposed applications was the development of a new class of dyes, which could be formed readily by the condensation of TCNE with aromatic amines. Many of the dyes made gave attractive yellow, orange and red shades. Some of the brighter lights of CRD’s scientists were involved, including Richard Heckert, who became CEO of the company in 1975. He was the inventor of a class of dyes described in U.S. Patent 2,803,640 (R. E. Heckert, Aug. 20, 1957). My job at the time was to assist in the application of these dyes. At that time, most dye applications involved the coloration of textiles from aqueous systems; and the cyano groups would generally hydrolyze, destroying the color. Solventbased dye application systems were not considered to offer a significant market opportunity. Unfortunately, I was unable to convince the scientists of the Chemical Department to develop hydrolytically stable compounds, and in the end, after considerable effort, the program was terminated. For the photodecoration program, where coatings were applied from solvent systems, the TCNE-derived dyes were totally acceptable. Hence, the only question was whether the leucodyes could be readily oxidized by HABIs. The answer was yes. NC CH(CN) 2
CN
[OX]
(C 2 H5 )2 N
(C 2 H5 )2 N CN
CN
The tricyanonvinyl dyes were reddish in shade; orange dyes could be obtained by replacing one of the b-cyano groups with a carboalkoxy moiety. We found that Orchem manufactured a dye intermediate that could be hydrocyanated to give the desired precursor; and with some help from the Special Services Department of the Central Research Department, especially Gene Vanover, we were able to produce sufficient materials to supply Greensboro. The leucodye LB-8 (bis(p-N-diethylamino-o-tolyl)-(3,4-dimethoxyphenyl)methane) had already been considered as a possible replacement for TLA-454 during the Dylux Venture days; and thus process information, which allowed manufacture of large amounts, was in place. Technical problems, such as producing quality negatives that could accommodate the relatively low g-images were handled by Ellefson; and before long, F&F decided to work primarily with Thomasville, a division of Armstrong Cork that was deemed to have the best scientific approach to new technology in its field. Thomasville installed equipment to photodecorate tabletops with attractive inlay patterns. One technical problem that we had not considered was that furniture manufacture involves considerable dust, especially sawdust. This could, of course, significantly interfere with the imaging process when it came between negatives and photosensitive surfaces. In time, this problem was overcome by carefully
1973—UNIVERSAL PRODUCT CODE OPPORTUNITY
209
control of the imaging locations. Our effort to establish the photodecoration concept with George Newman Company went slowly. They did not have the resources to move the program. One of their marketing managers, Jay Scott, left them and set himself up as an independent consultant; he offered to guide us to other collaborators. Before long, it was found that a major opportunity presented itself in upgrading leather splits. Hides are ‘‘split’’ to give a top-grain and a ‘‘split’’ portion—the latter is used to manufacture low-cost shoes or similar. We were told that if we could employ photodecoration to produce patterns on splits that simulated the patterns of topgrain hides, there would be a great opportunity for our chemistry. The rationale was that fashion demanded attractive leather patterns, which we could produce with photographic quality. The fashion industry could rapidly switch from one pattern to another, and the gravure rolls to print on leather were expensive and took considerable time to produce. Imaging a photosensitive layer, which could be imaged from more accessible photographic negatives would offer an attractive alternative. We soon realized that the photodecoration of leather was a big challenge. The tanning process introduced a large number of chemicals into hides, which already contained many reactive species, including phenolic substances that could interact with our chemistry. We thus had to learn to put blocking layers onto the hides and, subsequently, embed our chemistry in polyurethane binders, which would provide the required flexibility to these coatings. After imaging, a fix layer, which would contain phenolic radical traps and UV absorbers, was applied. We considered alternate routes, such as applying urethane layers on release papers, with the concept of imaging these offline and transferring the images onto the block-coating covered hide. This transfer was to be thermal and gave us an opportunity to introduce our thermal fix chemistry (Manos patent) into the coating. The development of this chemistry was generally undertaken in Massachusetts, which at that time was still the center of America’s footwear industry. We found several companies, such as Permuthane Corp., which produced the polyurethane binders, were eager to cooperate with us. The program went international when we made contact with a major leather tannery in Bolton, England, and one in Toronto, Canada. Our program offered our collaborators a chance to apply some new technology to an industry that was leaving the United States and was not very technically oriented.
XVII.
1973—UNIVERSAL PRODUCT CODE OPPORTUNITY
The Universal Product Code (UPC) was conceived in the early 1970s, and I was asked to follow its development as part of an effort to expand Dylux technology.
210
THE INVENTION OF DYLUX
At that time, DuPont’s Film Department had excellent contacts in the packaging field, and several of their marketing people considered that the availability of a novel, dry-imaging system could have application in this area. Specifically, UPC codes are either printed on the package or container (e.g., cereal boxes, dog food) or are generated in the store (meat, fish, and produce labels). The former labels represented an opportunity only in special applications; e.g., the UPC symbols as initially proposed were to be on the bottom of the container, and in some cases (e.g., bottles) printing on irregularly shaped surfaces by impact techniques was considered impossible. We, therefore, considered optical printing on the bottoms of glass bottles and other nonflat containers. This offered us some opportunities for novel label concepts. For instance, to accommodate dark bottles, or clear bottles containing dark liquids, we proposed a pigment-containing formulation. Universal product code marking composition containing a photosensitive dye former, a pigment and a binder. Dessauer, Rolf (du Pont de Nemours, E. I., and Co.) U.S. Patent 4,029,506 (Cl. 96-48R; G03C5/24), 14 Jun 1977, Appl. 516,483, 21 Oct 1974; 6 pp. Photosensitive color marking compns. providing upon exposure patterns that are obsd. by diffuse reflectance scanning means are given. Thus, a compn. obtained by mixing a binder comprised of cellulose acetate butyrate 45, pigment TiO2 (rutile form) 15 g, and a solvent 600 mL with a photosensitive component comprised of Santicizer 3 plasticizer 9.3, dodecylbenzenesulfonic acid 10, leuco dye tris(N,N-diethylamino-o-tolyl)methane 1.8, 2,20 -bis(o-chlorophenyl)4,40 ,5,50 -tetrakis(m-methoxyphenyl)-1,20 -biimidazole photooxidant 8.36, pyrenequinone (1:1 mixt. of 1,6- and 1,8-isomers) 0.07, 9,10-phenanthrene-quinone 1.144, polyethylene adduct of o-phenylphenol 8.6 g and triethanolamine triacetate 11 mL was coated on a Mylar film rendering it opaque, imagewise exposed for 1 min to 2.75 mW/cm2 UV light peaked at 365 nm, and the blue color formed photofixed by exposure to UV-free visible light. The diffuse reflectance of the exposed area was 1.10, that of the nonimaged area 0.08, giving a contrast sufficient for UPC marking.
The major opportunity for our technology was to be in store-printed labels. The most expensive items in supermarkets are found in the meat and fish counters. Generally, meat is cut and packaged in the store, and a label must be printed then and there. The store-printed labels require a somewhat different code than the package-printed labels, as they reflect the price of the item. That, however, is not important here. At the time we considered this application it was anticipated that some 30,000 supermarkets in the United States and Canada would go with UPC labels, and the average supermarket would require some 2 MM labels/year. Hence, the requirement was to be 60 MMM labels/year; at 1.5 square inch/label, this would amount to 600 MM square foot of label material/year. UPC patterns
1973—UNIVERSAL PRODUCT CODE OPPORTUNITY
211
were to be scanned with a HeNe laser, for which our cyan images were an ideal match. These numbers were based on market studies made around 1975. Fred Clarke, a marketing manager in the Film Department, Ted James, and I were invited to present this technology to Hobart in Dayton, Ohio, then the largest weighingscale manufacturer in the United States. We were received with considerable enthusiasm, except that the Hobart people told us that our numbers were too conservative. They pointed out that in time, meat and dairy labels would need to contain information beyond the bar code, such as dates, ingredients, etc., and hence more label surface would be required. This would, of course increase the volume of label material considerably. We agreed to cooperate in this area. The Hobart people were to design an electro-optical printer, and DuPont was to consider developing a stable, photosensitive label that contained a thermally activated adhesive. The competition then was a thermal transfer printing system that involved transferring a black pattern onto the label material. Hobart felt that these labels tended to smudge and the disposal of the waste tape was considered undesirable. Hence, an alloptical system was preferred. Another advantage of an optical system was that one might be able to contact expose certain information or patterns through a negative or stencil, thus minimizing the exposure energy required only to that of the actual bar code. Hobart told us that weighing scales then cost ca. $10–12,000 and that a printer could be sold for a similar price. They anticipated that stores would require at least four printers/store; one for meat, one for dairy, one for fruit and vegetable products, and one as a backup. Thus a printer population of 120,000 at maximum market penetration could be anticipated. Obviously, that represented a nice business opportunity for them. The DuPont Film Department then assigned a chemist and an engineer, located at the Spruance Plant, in Richmond, Virginia, to work in this area with help from me. We made fairly good progress on our side of the program. Hobart, however, ran into considerable problems. They had no in-house electro-optical expertise and seemed unwilling to go outside for it. They had difficulty finding a proper source of UV light. We then contacted GTE-Sylvania and identified several sources with sufficient power and permanence. Some of the technical problems that required attention were the stability of the exposed image under conditions of fluorescent irradiation, any toxicology, etc. By that time, we had begun to assemble considerable background on Dylux proof paper, and the toxicology was not considered a major challenge. The preimage stability of a Dylux material (ca. 2 years) was considered adequate. The primary issues then were photosensitivity of the coatings and image stability. Of course, photosensitivity was an issue that required consideration of the light source and its modulation. By that time, we had developed related
212
THE INVENTION OF DYLUX
technologies that might have been shared with Hobart, but were not. This included lenticular plate imaging, the use of higher speed HABI/leucodye combinations, and the consideration of thermally activatable coatings. We had collaborated with W. H. Brady Co., in Milwaukee to develop a photosensitive label for their ‘‘Kalograph’’ printer. We had filed a patent application for a photosensitive coating on a transparent support that could be imaged through the base, thus providing a smooth, clear label, etc. Marking transfer sheets. Dessauer, Rolf (du Pont de Nemours, E. I., and Co.) U.S. Patent 4,207,102 (Cl. 430-252 G03C5/04), 10 Jun 1980, Appl. 516,483, 21 Oct 1974–6 pp. Cont.-in-part of U.S. Ser. No. 617,540, abandoned. A thermally transferable imaging sheet is comprised of a carrier film coated with an imaging layer comprised of a pigment, a binder, and a color-forming light sensitive component. The imaging sheet is exposed to form an image, the imaging layer heated, contacted with a support more adherent to the imaging layer than the carrier film to transfer the image, and, optionally, the carrier film is sepd. from the imaging layer. Thus, a photosensitive compn. prepd. from cellulose acetate butyrate 10.28, TiO2 1.37 Me2CO 80.14, N-ethyl-p-toluenesulfonamide 2.225 dodecylbenzenesulfonic acid 0.856, tris(N,N-diethylamino-o-tolyl)methane 0.154, 2,20 -bis(o-chloro-phenyl)4,40 ,5,50 -tetrakis(m-methoxyphenyl)-1,20 -biimidazole 0.7153, triethanolamine triacetate 1.882, pyrenequinone (1:1 mixt. of 1,6 and 1,8-isomers) 0.0061, 9,10-phenanthrenequinone 0.10, and PhC6H4O(CH2CH2O)2.23H 1.951 g was sprayed on a yellow polyimide film, dried to give a 0.45 mil layer, UV (365 nm) exposed in contact with a photog. neg. of a UPC symbol at 2.75 W/cm2, and adhered to a black plastic bottle cap by heating at 70 . The greenish-blue and yellow UPC marking obscured the blackness of the bottle cap and appeared unchanged after several days exposure to daylight.
Although the numbers continued to look good, the program was terminated, when the Film Department was absorbed into the Plastics Department, who decided that they were not interested in this technology. As DuPont was then unwilling to offer the chemicals to a third party to coat, the program ended in 1976.
XVIII. 1974–PHILIP BOTSOLAS: A BREATH OF FRESH AIR One of my heroes in all this was Philip Botsolas, who first appeared on my scene in 1974. Botsolas had co-owned a company that he sold at a sizable profit and thus was independently wealthy. This allowed him to be himself, instead of being endlessly concerned about the consequences of his actions. Botsolas had been in the Sales Group of the Photo Products Department and was manager of the Clifton, New Jersey office, when he was reassigned to Wilmington and was given the task of looking into Dylux proof paper sales. Why was this unique product not doing better?
PROGRESS
213
When I learned that someone with sales experience would have this task, I immediately called him up—still in Clifton—and expressed my delight in finding someone who could help me in my efforts to expand the Dylux business. We immediately took a liking to each other. Botsolas said he had heard about me and my enthusiasm for this technology and said he was delighted to work with an enthusiastic chemist. Nevertheless, when we first met, he was somewhat skeptical about Dylux; he had heard about it only from his Photo Products colleagues, who never really understood it and the potentials of this then still novel technology. He knew that the department had not given it much technical support. I had once suggested that Dylux, being a self-contained product, did not need development and was not getting any from the Company. Soon, Botsolas knew what I meant by that. In time, he agreed to convene a meeting in which I could present some of my ideas for improvements and next-generation products. Among the attendees were Marvin Yates, the New Product Manager and Glen Thommes, then an assistant director of research. After I had given my presentation, Thommes countered that much of what I had talked about was tried and had failed during the Dylux Venture days and that he saw little sense in another round of rediscovery. I was livid, because that was not true, and told Thommes he really did not know what he was talking about. One of the challenges in developing applications with collaborative research organizations is that one must offer them some incentives in exchange for their efforts. These might be an exclusive position, or some payments or joint patents. The task is infinitely more difficult when one himself is uncertain that the product will be manufactured, is safe, has good shelf life, etc. It is even more difficult, when the product is not a product, but a research development. Hence our early efforts were always met with a certain amount of skepticism, and it should have been obvious to Thommes that we were in a different position now than during the venture days. Yates allowed that a new product manager spends more time putting out fires, i.e., satisfying customers, than conceiving new products. Two weeks later, Thommes became my director of research, but he never referred to our meeting; as a matter of fact, he became very supportive of my efforts!
XIX. PROGRESS A. Dylux1 DFF Filter One of the limitations of Dylux is that it requires relatively pure UV light to form images effectively. Many customers had graphic arts light sources, e.g., xenon arcs, which were relatively rich in visible light. Interposing a glass UVonly pass filter would overcome this problem, but this was not a practical solution for the various exposure units. Clearly, a flexible filter would be the answer,
214
THE INVENTION OF DYLUX
that could be interposed between light source and the lithographic negative would be the answer—if we could develop one. It turned out that the orange-forming dye produced in our photodecoration work, had relatively low absorption in the UV and good absorption in that portion of the visible spectrum in which 9,10-phenanthrenequinone absorbed. We coated samples of this dye onto a Mylar base and were encouraged that this dye could be the basis of a flexible filter. Kay Looney calculated the required concentration and purity of the dye in the coating; and with some enthusiastic support from Botsolas, we were able in the short period of several months produce a commercial product that was successful in significantly expanding the market for Dylux proof paper, as customers could make good proofs in existing equipment. After that, Botsolas became a real admirer of my little group of abandoned photochemists. The availability of the DFF served a need beyond improving the image quality of Dylux proof paper; it was used in enhancing the performance of other imaging materials, and is referenced in a number of patents issued to other companies. U.S. Patent 4,167,490 September 11, 1979 Flexible ultraviolet radiation transmitting filters Inventors: Looney; Catharine E. (Wilmington, DE) E. I. DuPont de Nemours and Company (Wilmington, DE) Filed: December 22, 1975 Abstract: Flexible filters capable of transmitting radiation in the spectral range between 315 and 380 mm and absorbing radiation strongly in the spectral range between 400 to 550 mm maximally absorbing in the spectral range between 440 to 500 mm, consisting essentially of a thin flexible film substantially transparent to radiation in the spectral range between 315 and 380 nm having homogeneously added thereto a dye in a concentration of 15 to 60 mg/dm2 of the formula I where R1 is lower alkyl of 1 to 5 carbon atoms; R2 and R3 , which may be the same or different, are selected from the group consisting of R1 , 2-cyanoethyl, 2-hydroxyethyl and formula II or from a saturated 5- or 6-membered ring; R4 is taken from the class consisting of lower alkyl of 1 to 5 carbon atoms, aryl of 6 carbon atoms or alkaryl of 7 to 9 carbon atoms, and arakyl. The filters are useful in exposing dual response photosensivtive materials with ultraviolet radiation sources which contain appreciable quantities of visible radiation.
R1O
C C(CN)
C(CN)
NR2R3
O I
CH2CH2O C R4 O II
PROGRESS
215
B. Europe During the summer of 1975, I was invited to give a lecture at the ETH in Zu¨ rich, where Professor Berg had a distinguished department of photographic science. The invitation came via my friend Al MacLachlan, who at that time was director of DuPont’s Neu Isenburg Laboratory in Germany. This was a prestige job, which several other chemists held on the way to promotion to bigger and positions in the company. Al, as usual, was enthusiastic about his assignment and learned sufficient German to be able to conduct business in that language. Prof. Berg of ETH had asked for a speaker from DuPont, and Al suggested that I might like to give a lecture there. Of course, I would have to pay my own expenses—Photo Products Department had no money for someone who was not in the department. So, my wife, stepson, and I flew across on Loftleidir, then the least expensive way to fly to Europe. Al invited me to give the talk first at his laboratory, and my wife and I then traveled to Zu¨ rich for the presentation there. I was able to give my talk in German, which pleased the group in Neu Isenburg, as well as the ETH attendees, except for one American who attended, and was dismayed that he could not understand me. Subsequent to this talk I gave another presentation at the Agfa-Gevaert Laboratory in Mortsel, Belgium, where I spent an entire day and was received with utmost interest and hospitality. I was impressed that my presentation was given in a conference room equipped with a bar and excellent food and drink. I did not return to my hotel until after 8 P.M. The management wondered if there were any areas of technology that DuPont had ignored and that they might follow up on. I pointed out that our RF-technology was virtually abandoned, and that I would be delighted if they requested information and samples, with some request for licensing this technology. This letter was duly received and caused the anticipated furor among the Photo Products Management, especially the director of research, Robert Upson. Who had given me permission to visti Agfa-Gevaert? Still, it created respect for our technology. Some samples were indeed transmitted, but no serious effort was made by DuPont to work with AgfaGevaert.
C. Medical Imaging Systems–‘‘aca’’ Automatic Clinical Analyzer DuPont developed a novel medical analytical instrument in the late 1950s. One of the drivers of this program was Dr. Donald R. Johnson, a contemporary of mine at the University of Wisconsin who came to Wilmington in 1953, shortly after I did. Don was an analytic chemist and for several years lived in an apartment above mine. He was aware of my work with UVI/Dylux and found an interesting application for it.
216
THE INVENTION OF DYLUX
The automatic clinical analyzer (aca) was a rather large instrument, which automatically performed spectrophotometric tests on body fluids and was readily accepted in the field in the 1970s. Another Wisconsin contemporary of mine, Dr. Robert C. Doban, led the effort to introduce it to the market. Johnson and some of his colleagues wanted to develop a foolproof method of transferring information from the sample container, a transparent plastic bag, to the report sheet and proposed attaching a matted polyester card on which the patient’s information, such as name and date, was written by hand. The report slip was a sheet of photosensitive, i.e., Dylux paper, on which the analysis results were printed by a conventional impact printer. The information from the polyester card was contact printed on the same paper with a flash of UV light. As the machine was designed to use Dylux paper, this was indeed a captive market. In time the paper changed from a cyan to a blue-black shade, to facilitate copying with electrophotographic copiers. The photosensitive layer was coated onto a relatively thin chart paper and then finished into long rolls for the aca. For many years we sold Dylux 504 paper for the aca. This application was unique and accepted by an industry that was determined to avoid any errors in data transfer, as would be the case if alternate data transcription were used. It appeared to me that other medical applications could have benefited from this approach; and I frequently, without success, tried to promote extending the market. No one was particularly interested—the aca people had their hands full with solving their internal problems, and no one was available to promote the paper as a unique product. Sometime in the mid-1970s, when I was looking for some support inside of the company for improving the Dylux application technology, I began to interact with Dr. Bruce Booth, a physicist, who had been involved with holography and a number of other imaging technologies and who had become associated with the medical instrument business. Booth understood the need for improving what we had, and supported some effort to upgrade the report slip chemistry, so that we could make multiple copies with a single flash. He was also of help in supporting my efforts with medical identification systems. In 1975, I received an invitation to visit Georgetown University Medical School in Washington, D.C., where a professor of clinical chemistry, Dr. Martin Rubin, was researching better patient identification systems. Rubin had proposed inserting a bar-code-bearing lithographic negative in a patient’s wrist bracelet, and using this as a master to generate human/machine readable labels at the patient’s bedside. Dylux was just what he needed! He also proposed providing photosensitive containers and forms that could be used with the same wrist bracelet. Rubin, an enthusiastic, politically oriented gentleman and I immediately became friends, and I provided him with many samples, which were enthusiastically evaluated by him and some of his sponsors at the National Institutes of Health. Still, what was needed was a corporate provider of this technology to
1976—STILL MORE OPPORTUNITIES
217
the industry. Strangely enough, Don Johnson showed only modest enthusiasm for this, but I found another supporter in Mr. Mike Lanham, an enthusiastic marketing type then in DuPont’s Film Department. Lanham and I tried to persuade companies like Becton-Dickinson, who supplied the medical field, to examine this market, but met with mixed success. In time, we made some technical improvements, such as providing opaque coatings that could be coated on transparent glass containers and multilayer-labels that could be simultaneously imaged; but there was no real way of getting a foothold here.
XX. 1976—STILL MORE OPPORTUNITIES A. GTE-Sylvania Home Office Copier Lanham made contacts with GTE-Sylvania, realizing that a combination of unique light-sources and photosensitive paper might be an approach to reach the market. We had a need for unique light sources for the UPC-label printer, and we obtained very enthusiastic support from GTE-Sylvania. At that time, their flashbulb business was being eroded by faster photographic-films and built-in electronic flash units in newer cameras. They were looking for new flashbulb applications. Several of their engineers decided that a very low cost copier could be designed, in which flashguns would in a single exposure capture an image on a photosensitive paper. Where was such a paper? This seemed like an intriguing challenge. When they learned about the Dylux technology, it seemed that there might be a way of tweaking it to achieve singleexposure copies, and I naturally enjoyed the opportunity to do something new and different. Of course, the problem was that conventional Dylux required two distinct exposures. Reflex copying, in which the light would pass through the photosensitive coating, and a second pass to provide enough light to fix the paper could be considered an option. Films coated with Dylux-type formulations, with suitable changes, actually yielded, on flashbulb exposure, images that with a subsequent UV exposure produced a positive-mode image. The image quality was enhanced when the film was moistened slightly. We found that the images could be developed in sunlight, when a sheet of the Dylux DFF filter was placed over the fixed coating. In time, it was realized that the approach would never yield the quality of copies attained with electrophotographic copiers and was abandoned. All that remains is the GTE-Sylvania patent. U.S. Patent 4,153,365 Portable copier using flash lamp article English, George J.; Fohl, Timothy, GTE Sylvania Incorporated (Stamford, CT) Issued/Filed: May 8, 1979/April 5, 1978. Application Number: US1972000278778 IPC Class G03B
218
THE INVENTION OF DYLUX
027/04 U.S. Class: 355/99; 347/224; 362/11; 362/13; 396/317: A low-cost, portable copier for exposing a photosensitized sheet through a stencil placed thereon. The sheet and stencil, in the form of a ‘‘sandwich’’, are positioned on the copier’s cover member which is then pivoted to a closed position against the copier’s housing. A flash lamp article, including a tapered casing having a percussively-ignitable flash lamp therein, is inserted within the housing and activated by a resilient cantilever member located therein. In the patent, it is reported that the imaging medium is ‘‘preferably a a photosensitized material sold by the E. E. du Pont de Nemours Company under the trade designation ‘‘Dylux’’. This material is made by the application of organic coatings containing complex organic compounds onto suitable papers and films. It is unique in that ultraviolet light is used to form permanent colored images while visible light may be used to stabilize the unimaged areas and prevent further color formation. It is also possible to produce both positive and negative images in a variety of colors. Understandably, all of the above results are obtained without the need for chemical processing. When using paper of the above variety, it is preferred to utilize an ultraviolet filter 24 (FIG. 2) to filter out all light except that within the ultraviolet frequency range. Filter 24 is located within first end portion 27 of chamber 25 and is thereby immediately adjacent label 19 when cover 13 is closed.’’
B. The End of Photodecoration By 1975 we were optimistic about the opportunities for photodecoration. However, in 1976, the Fabric and Finishes Department concluded that the manufacture of coating chemicals for the furniture industry would not fit into their future plans, and in spite of our success, closed their 20-man laboratory in Greensboro, North Carolina. They would not allow us to work with any of the other suppliers that could have offered us a facility in the area; and after paying Thomasville for the equipment that they had bought for this program, furniture photodecoration became a dead issue. The Dye Sales Division then got cold feet about photodecoration, and our program to modify leather was also terminated. I was back with Remington full time. In retrospect, we should have done better, but our management was too timid to assert itself.
C. Dylux1-4C Overlay Films Botsolas and I went on a trip to Boston to demonstrate the filter to a number of customers, and he asked me there why we had never made any colored Dylux overlay films. At that time, 3M Co. sold Color-Keys1, a series of colored films, which after exposure were immersed in an aqueous solution to remove unexposed areas, thus giving a negative-working colored film. These were overlaid to demonstrate the ultimate color rendition of a printed page. This system was
1976—STILL MORE OPPORTUNITIES
219
far less complicated, though results not as realistic, as DuPont’s Cromalin1, which required lamination, exposure, toning in several sequences to produce a surprint. One reason we had not made a full-color Dylux was that we had never succeeded in making good red- and yellow-forming leucodyes; those that we had were oxidized too easily by phenanthrenequinone. I concluded that we might not need a photofix in a colored overlay system, and called Al MacLachlan who had by now been promoted to Director of Sales for Graphic Arts Products. He agreed that a four-color overlay, which worked without fixing, might indeed be a desirable product for making ‘‘quick proofs’’; we assumed that a wet fix might ultimately be acceptable. I told Botsolas that it could probably be demonstrated quickly, and he encouraged me to go ahead with some experimental work. He then wondered why the neutral-shade Dylux 515 was olive rather than black, and I said that I could probably solve that problem, too. By that time, W. R. Remington and I were once again separated; this time my new leader was Dr. Albert Bauer, with whom I had worked at Orchem some 25 years earlier. Then, Robert Terss, Al, and I were a three-man team, which was, to identify new classes of dyes for acid-modified Dacron polyester fiber. After that work, he was transferred to a position in the Freon business. He now was transferred back to the Orchem Research Division to impart some of the wisdom that he had accumulated in that field to another business area. Bauer and I wrote several proposals for work to be undertaken by me on behalf of the Photo Products Department and submitted these to Botsolas. We set a low price tag of $25,000 on each proposal. We rationalized that this was a ‘‘bargain’’ for the Photo Products Department, but then, the Organic Chemicals Department would gain potential new business manufacturing the chemicals. When Botsolas received these, he went to see his general manager, Dr. Phil Wingate, and suggested that this work should be supported. Wingate told Botsolas to arrange for his research director, Upson, to evaluate these proposals; and in time, a meeting was set up for me to present my approaches to an assemblage of Photo Products research managers, including Upson. For several months after that there was no response. Wingate often stopped at the Greenville, Delaware, post office, where he and I had mail boxes. I finally decided that I would take samples of my tentative color-overlay film in my briefcase and be ready to show it to Wingate if by chance we met. This opportunity soon came, and Wingate was quite impressed with what I had accomplished. He asked me how soon we could have a product, and I replied that we were still looking for support from Photo Products to undertake the needed research. I called Botsolas, and he said he would see Wingate the following day. He did and pointed out that if his job was to improve Dylux business, he should get some funding to do it. Later that day, Wingate left a note for Botsolas that the $50,000 would come from Upson’s research
220
THE INVENTION OF DYLUX
budget, post haste. I learned later that Upson was very upset about this turn of events. We made good progress on both programs, although not without a set of new problems. In February 1976, Orchem had decided to vacate its laboratory building at the Experimental Station and concentrate all its efforts at Jackson Laboratory. This meant relocating into a facility that had no laboratories equipped to do photochemical research, losing all the skilled technicians, and spending weeks packing and unpacking. I was in an awkward position—we had been authorized to do work that required facilities and support that was no longer available. After a 1 month hiatus, the Orchem management finally decided to build a laboratory for me to do photochemical research. Chuck Kelly, an elderly chemist who had worked in the patent group for many years was assigned to work with me, and a technician was assigned to us. It took several more weeks to unpack all the materials that we had brought from the Experimental Station. Unfortunately, when the move was completed, there was a paucity of room to store equipment and chemicals; they were left outdoors and covered with a tarpaulin. Naturally, much of the metal equipment rusted. It was soon evident that our facilities were inadequate to the task. Our technician supplemented his income by selling fresh fish in the laboratory during the first work hours of most days, and was seldom available for experimental work before 10 A.M. When I complained about this to Al Bauer, my then Division Head, he said he was aware of this, and said that the fish was very good and reasonably priced, and I should take advantage of this. Unbelievable! In photochemical research it is necessary to make coatings of some uniformity, to have materials available for meaningful evaluations. A Talboys coater was available at the Photo Products laboratory building at the Experimental Station, and Botsolas invited me to visit Dr. Harry Knop, the laboratory director there to find out if some space could be made available there, so that I could work in proximity to the coater. Knop, who took direction from above and never was prepared to venture any opinion, said there was no space around, although we actually saw that about 10 laboratories in the building were unoccupied. Obviously, Upson had gotten the message across that we were to be given minimum help. He suggested I prepare my coating solutions at Jackson Laboratory and occasional coating time would be made available for me to use the Talboys coater. While our laboratory was being constructed and equipped, Kelly and I made occasional trips to the Towanda R&D laboratory, where Tom Sheets and some of his colleagues found space and technician time for me. Regrettably, the trip to Towanda inevitably involved driving nearly 500 miles and required overnight stays. Finally, the Experimental Station management, which was still angry about the Organic Chemicals Department’s early departure, allowed me to have a
221
1976—STILL MORE OPPORTUNITIES
CH 3
(C 2 H5 )2 N
H CH 3
H N
CH 3
(C 2 H5 )2 N
O OC H3
H LEUCO GREEN
CH 3
N(C 2 H5 )2
(CH2 )n CH 3
LEUCO PURPLE
N(C 2 H5 )2
Scheme 4.10
lab—but I could use it only on alternate days, and they would not supply a technician. So someone from Chambers Works had to come over. Still we made some progress. I found some old leuco dyes from the TCNE period, and Dr. Owen Webster of Central Research kindly reduced them to yield yellowforming leucodyes; so we could make almost all shades. Our black was a mixture of LG-1 (green) and LP-1 (purple), and looked very good (Scheme 4.10). After we moved into Building 334 at the station, we had about 2 weeks before Upson decided we should have a research review—after all, we had now been funded for about 3 months, though we had only about 4 weeks lab work, because of the lack of facilities. The review went well, and Upson was complimentary. We continued on our program, the ‘‘black’’ Dylux was beginning to look better; it required some attention to the leucodye ratios, and purification of ingredients. We had no fix chemistry built into Dylux-4C system, and so the question was what could we do to stabilize the background. Here we are dealing with a coating on polyester film, and the thought occurred that we could quickly pass the film through a fix solution. Not totally elegant, but it avoided having to get rid of the yellow color of a quinone photofix agent. We tried a number of phenols; one of the Dylux sales managers, L. K. Jordan, who was enthusiastic about the quick-proof concept. brought me a small X-ray film processor, made by Hope Co. A 1-minute pass through a solution of a hydroquinone, dissolved in a water-propanol mixture stabilized the image nicely, but many hydroquinones oxidized in the film, and after a few days, imparted a grayish background.
D. Phenidone An unexpected assist! Ray Firmani, the photographer for Jackson Laboratory, on learning of my problems, responded to a question about whether there were some photographic reducing agents that did not discolor. He asked me if I had ever considered phenidone. I had not even heard of phenidone (3-Nphenylpyrazolidinone) (Scheme 4.11), but Ray obtained an ounce from a camera shop, and we fixed and had no background problems thereafter. No one had ever
222
THE INVENTION OF DYLUX
used phenidone in any nonsilver imaging system, but it was well known to photographic scientists as part of a superdeveloper combination. The results were so outstanding that we filed and were granted a patent on the use of phenidone as a radical trap. Firmani, who had worked at Jackson Laboratory for almost 30 years, finally received his first patent. We also found that incorporation of phenidone in the different recipes allowed us to build an induction period into the films, so that the materials could be handled quite well in ambient light. Only after about 30 minutes in a proofing chamber did background color begin to develop. Photoimaging systems with cyclic hydrazides. Dessauer, Rolf; Firmani, Raymond Alexander (du Pont de Nemours, E. I., and Co.) Eur. Pat. Appl. 19,219 (Cl. G03C1/00), 26 Nov 1980, US Appl. 38,056, 11 May 1979; 40 pp. Photoimaging system useful in applications such as light-actuated color image formation on various supports (fabric, paper, film) comprises 1 leuco dye, a photooxidant (from the group of 2,4,5-triarylbiimidazolyl dimer, benzophenone, p-amino-phenyl ketone, polynuclear quinone, or thioxanthenone), and 1 cyclic phenylhydrazide. Thus, 4-mil thick poly(ethylene terephthalate) support having a subbing layer and an antistatic back layer was coated with a compn. contg. cellulose acetate butyrate (binder) 8.63 2,20 -bis(o-chlorophenoxy)-4,40 ,5,50 -tetrakis-(m-methoxyphenyl)biimidazole (initiator) 1.99, 2-methyl-3-[(10 ,30 ,30 -trimethyl-indolin-2-yl)vinyl]indole (leuco orange dye) 0.17, p-dibutylaminophenyltricyanoethane (leuco magenta dye) 0.13 p-toluenesulfonic acid 0.22; nonylphenoxypoly(ethyleneoxy)ethanol (plasticizer) 3.02, 1-phenylpyrazolidin-3one 0.04, iso-PrOH 8.6, and CH3Cl 77.2%, dried, and imagewise exposed (2 W Berkey ASCOR exposure unit, 30 40 vacuum printer) for 2-4 min to give an image having green region spectrum d. ¼ 1.1 and blue region d. ¼ 1. 5. Obtained image exhibited good color stability in the dark and improved room light stability. It also issued as U.S. Patent 4,247,618 (1/27/1981).
In later years, I found that phenidone was truly a remarkable material. It worked at low concentrations, and some analogs were similarly effective. Attempts to encourage someone to develop some chemistry by which we could form phenidone photochemically or thermally failed; it still seems like a good idea. The phenidone patent was allowed in 1979, and apparently kindled some ideas at Fuji Photo Film Ltd., who in the late 1980s began to show an interest in HABI chemistry. As a matter of fact, they were interested in developing a Dylux-like proofing product, to be described later. N
N H
Scheme 4.11
O
1977—AND STILL MORE OPPORTUNITIES
223
XXI. 1977—AND STILL MORE OPPORTUNITIES One of the continued problems with all chemical researchers is the lack of adequate technician help. Throughout much of my career I was subject to the rule of one, or even half a technician per chemist, which was applied irrespective of how productive a scientist was. There were, of course, exceptions, when I had somewhat more aggressive management, who realized we could make more progress if we could conduct more experimental work. I often wonder how industry would advance if each manager just had one person under him or her!
A. New HABIs: TCTM-HABI In 1975, while I was in Europe and gave lectures at Neu Isenburg, ETH in Zu¨ rich and Moertsel, Belgium, one of my technicians, Aurelio Baccino, was available to do some synthetic work on new materials. I had always wondered why we had not developed some lower-cost HABIs—at that time, the only commercial materials were o-Cl-HABI and CDM-HABI. I suggested to Aurelio, among other things, that it might be of interest to prepare HABIs from the low-cost starting materials o-chlorobenzaldehyde and veratraldehyde (3,4dimethoxybenzaldehyde). Baccino prepared the benzil via the classical benzoin condensation, followed by oxidation, and then formed the triarylimidazole and, in turn, the biimidazole depicted in Scheme 4.12. I had assumed that we obtained a statistic distribution of the various possible isomers until I learned that the preparation of unsymmetrical benzils was described in details in Organic Reactions (W. S. Ide and J. S. Buck, Org. React. 1948, 4, 269) the preparation of 2-chloro-30 ,40 -dimethoxybenzoin being specifically reported.
OC H3 OC H3 Cl N
N Cl
2
Scheme 4.12
224
THE INVENTION OF DYLUX
We found that all samples prepared were effective in initiating photopolymerization, and on returning from Europe, I submitted samples of several of these new HABIs to Photo Products Department for evaluation, but no response was received. Of course, it is difficult to develop much enthusiasm for incorporating yet another new material into existing coating recipes, as this would entail a long round of testing, examination, etc. Two years later, Dr. Jim Morgan of Photo Products’ Parlin Plant contacted me. He had heard that I had made some new HABIs and would be happy to evaluate them in Dylux; he had recently been assigned to find ways of lowering the manufacturing cost of our proof paper. In September 1977, I had asked my then technician, Steve Ryan, to test these HABIs for stability and asked him to incorporate them in a Dylux type formulation. Ryan had found that TCTM-HABI, derived from 2-chloro-30 ,40 -dimethoxybenzoin had indeed worked well—too well as a matter of fact, because it could not be easily fixed. I submitted samples to Morgan, but when he was slow in evaluating them, I asked Ryan to re-examine his work and concluded that the poor fixing could be corrected by using half the quantity of the photo-oxidant. This was a significant finding, as we had shown that the mill cost of the new HABI was about a third of that of the currently used CDM-HABI. And in turn, we needed only half as much in the coating recipe. I communicated these results to Dr. Vaughn Chambers, who was then director of marketing of graphic arts products, and he was impressed. In addition, I pointed out that if we could obtain patent protection on this and related new HABIs, we could extend the Dylux patent estate till the late 1990s. Chambers asked what was needed, and I responded that we had to synthesize a number of analogs, so as to make for a better patent. Chambers urged me to go ahead, but Robert Upson decided that one of his chemists should do the synthetic work. At that time, Orchem still thought that the manufacture of HABIs was to be a long-range business, and I had little difficulty in getting some help in manufacturing the new dimer. The very talented chemist, Dr. Lawrence Q. Green, who had developed all the process information for o-Cl-HABI and CDM-HABI, took it upon himself to scale up my process or, perhaps more realistically, to modify the processes for the then commercial HABIs. Everything went nicely. Green made first 1 pound, then 5 pounds, and all the tests confirmed our initial success. We felt confident that making a large batch of several thousand pounds would not be a big deal. At that time, our coatings were made in Towanda with methylene chloride as solvent, but our custom coating was done with butanone and isopropyl alcohol. The samples of TCTM-HABI had good solubility in this solvent pair. When Green precipitated the product from the oxidation of the 2,4-bis(2chlorophenyl)-5-(3,4-dimethoxyphenyl)imidazole, he made a startling discovery. Some of it went into solution in the ketone solvent, but most did not. What did
1977—AND STILL MORE OPPORTUNITIES
225
not, did go into methylene chloride. I found that both the soluble and the insoluble material functioned as an effective photooxidizing agent, working well in the now preferred Dylux recipe. But what was going on? Green, who was working frantically to see if he could convert the less soluble material into a ketonesoluble HABI had scheduled a safari in Africa, and time ran out on us. We were stuck with over 1000 pounds of a biimidazole that could not be used in the available coating facilities, which were not designed for solvent recovery of a halogenated solvent. Our colleague in Towanda, Tom Sheets, came to the rescue. He studied the order of addition of the multicomponent recipe and found that the ketoneinsoluble HABI was soluble in butanone when acid was added to the solvent. Somehow, we tricked the HABI into going into solution, and by the time Green had returned from Africa, things had resolved themselves. Orchem, which had charged Photo Products a fairly high price for the CDM-HABI based on the expensive intermediate, m-methoxybenzaldehyde, decided that money would be made by charging Photo Products the same price for TCTM-HABI. As the recipe required but half as much of the new HABI, there were still substantial savings to be made, and both departments profited. But Photo Products decided that this was not fair. An unfortunate oversight was that the m-methoxybenzaldehyde had to be custom manufactured, and orders for many thousands of pounds were placed before it was realized that it would no longer be needed. So CDM-HABI was manufactured for another year, mostly because no one had the courage to throw it out and make the better Dylux proof paper with the new HABI. In spite of the fact that this resulted in excess of $1,000,000 savings on production costs of Dylux proof paper, almost 4 years were required before CDM-HABI was replaced. The performance of proof paper that was formulated with TCTM-HABI was significantly better than that containing CDM-HABI. The visual difference was striking. Still, the patent office kept on rejecting our application. We had some evidence that TCTM-HABI had much greater radical stability and attributed its better photo-oxidizing ability to that. I had prepared films containing several different HABIs in cellulose acetate butyrate binder. We exposed these films to UV light and could observe remarkably long free radical life. Dr. Paul Krusic of the Central Research Department measured the ESR signal of exposed films and found significant differences. In spite of these data the U.S. Patent office rejected our claims. I suggested to Mr. Wes Wardel, the attorney handling this case that we visit the patent office and speak to the examiner. He was reluctant to do so, until we received a final rejection. I never figured out why a third person had to defend you when you are capable to do so yourself. We drove to the patent office and met Mr. Jack Brammer, a tall, physically handicapped man, who was very courteous and asked me why I should receive a
226
THE INVENTION OF DYLUX
patent. I showed him comparisons of Dylyux 503 proof paper made with the old and new HABIs, and he said that I had very convincing evidence and that a patent was going to be allowed. Brammer said that his name was on the patent, as well as mine, and he wanted to be certain that the evidence was solid enough that it should not be overturned. He asked me if I could explain why the new HABI was better, and I did. He said he was grateful for some technical insights, and that companies like DuPont should make more of an effort to submit technical information to the examiners, who after all also wanted to be technically updated. I said that there was an SPSE meeting in Washington a few weeks later, and perhaps he would like to attend. He said that he appreciated hearing about this from me. Then he said we should go to the cafeteria and have some coffee— but no one should pay for someone else’s! Some time later, I met him at the SPSE meeting, and he told me that he enjoyed the opportunity to be better informed. Unfortunately, some time afterward he died, and we never met again. He was a real gentleman, who simply wanted to do a good job. Two patents issued, covering different aspects of the chemistry. U.S. Patent 4,252,887 (February 24, 1981) Dimers derived from unsymmetrical 2,4,5-triphenylimidazole compounds as photoinitiators Dessauer; Rolf E. I. (DuPont de Nemours and Company (Wilmington, DE) Filed August 14, 1979, IPD G03C 001/68, 430/281.1; 430/283.1; 430/287.1; 430/288.1; 430/915; 430/ 917; 430/920; 522/16; 522/26; 522/79; 522/83; 522/89 Photoimaging compositions comprising (A) 2,4,5-triphenylimidazolyl dimer having selected substituents on the 2,4 and 5 phenyl rings and an extinction coefficient determined in methylene chloride at 105 to 103 mol/liter at 350 nm of at least 4000 liters/mol-cm and at 400 nm of at least 250 liters/mol-cm; and at least one of (B1) leuco dye or (B2) addition polymerizable ethylenically unsaturated monomeric compound. The new imaging compositions are useful in preparing dual response photoimaging products such as proofing papers, printout paper, overlay films and photopolymerizable elements. Improved imaging speed is achieved at equal concentration levels when compared with conventional 2,4,5-triphenylimidazolyl dimers. U.S. Patent 4,311,783 (January 19, 1982) Dimers derived from unsymmetrical 2,4,5-triphenylimidazole compounds as photoinitiators Dessauer; Rolf E. I. (DuPont de Nemours and Company (Wilmington, DE) Filed 7/17/1980. IPD G03C 001/72; G03C 001/5, U.S. Classes 430/270.1; 430/292; 430/293; 430/334; 430/ 346; 430/358; 430/905; 430/915; 430/916; 430/917; 430/920; 522/26; 522/63 Photoimaging compositions comprising (A) 2,4,5-triphenylimidazolyl dimer having selected substituents on the 2,4 and 5 phenyl rings and an extinction coefficient determined in methylene chloride at 105 to 103 mol/liter at 350 nm of at least 4000 liters/mol-cm and at 400 nm of at least 250 liters/mol-cm; and at least one of (B1) leuco dye or (B2) addition polymerizable ethylenically unsaturated monomeric compound. The new imaging compositions are useful in preparing dual response photoimaging products such as proofing papers, printout paper, overlay
1977—AND STILL MORE OPPORTUNITIES
Cl H N
Cl
Cl +
N
227
H N
TCDM-HABI
N
H3 CO OC H3
Scheme 4.13
films and photopolymerizable elements. Improved imaging speed is achieved at equal concentration levels when compared with conventional 2,4,5triphenylimidazolyl dimers.
A few years later, Sheets suggested that co-oxidation of the biimidazole derived from 2-chloro-30 ,40 -dimethoxybenzoin and o-chlorobenzaldehyde and 2-(o-chlorophenyl)-4,5-diphenylimidazole would give a still less expensive photo-oxidant (Scheme 4.13) with less visible sensitivity, thus making for better photofixing. U.S. Patent 4,622,286 (November 11, 1986 ) Sheets; Thomas M. E. I. DuPont de Nemours and Company (Wilmington, DE) Photoimaging composition containing admixture of leuco dye and 2,4,5-triphenylimidazolyl dimer. Photoimaging composition comprising an admixture of leuco dye, and at least one 2,4,5-triphenylimidazolyl dimer prepared by an oxidative coupling reaction, a reaction product, 2,20 ,5-tris-(o-chlorophenyl)-4-(3,4-dimethoxyphenyl)-40 ,50 -diphenylbiimidazole, being present in an amount of 0.01 to 90.0% by weight based on the weight of solids in the composition. The composition when coated on a support is useful for prepress proofing.
B. Photomarker1 Corp. By late 1966, news of the UVI-venture had spread throughout the industry. I received a phone call from Leon Stern of New York City, who told me he had need for a large quantity of a low-quality photosensitive paper to be used in the manufacture of cutting patterns for garments. He visited me in November 1966, and I had to tell him that we were far from commercializing anything, and as our venture was directed toward ‘‘electronic imaging,’’ we could do little as garment patterns did not fit into our program. After he explained to me what was needed I was really disappointed, because had we considered this application seriously, we might have generated a financial base from which we could develop newer and better technologies.
228
THE INVENTION OF DYLUX
Specifically, diazo paper was used at that time to generate patterns from which garments would be cut in the clothing industry. These patterns were called markers, and Stern’s company was called Photomarker. In addition to coating an inexpensive diazopaper, this company also sold exposure machines, all based on diazo-chemistry. Why go to UVI? Diazo papers required moist development, and this caused considerable dimensional stability problems with the 6-footwide, 18þ-foot-long markers. A dry process would minimize this problem. In 1976, the Photo Products Department was approached by one of Stern’s colleagues, and the program was reconsidered. By this time, it was agreed that this was a business that did not fit into their business plan, and as it might have consumed large amounts of chemicals, it was deemed as an opportunity, which Orchem could pursue. I was told that we could sell a dry powder mix, which had to be designed so that no effective proofing product could be manufactured from it, and I took up the challenge. Photomarker was to identify a coater, who could handle this material, and my task was to make a formulation that would give an image with acceptable contrast. To keep the cost low, I decided we needed to employ the lower-cost o-ClHABI, and to prevent the final product from competing with Dylux proof papers, I added a trace of the red dye Rhodamine 5G. This actually improved visual contrast. At this time, electronic calculators were just hitting the market, and simplified the development of the requisite formulations of calculations with slide rules. Had desktop computers been available, we could have worked even faster. After a search for coating facilities, we found that the Fletcher Paper Company in Alpena, MI, had facilities for solvent coating and also manufactured a paper that was sufficiently neutral to be acceptable as a base for the proposed product. In the short spell of 9 months we developed technology that would permit manufacture of a new product. We demonstrated the concept at a Garment Manufacturers’ show at Atlantic City and elicited a positive response. By this time, I had developed a good relationship with Mr. Jack Yentis, who was president of Photomarker; Stern was chairman. Photomarker had a factory in the south Bronx, and I frequently visited it to undertake coatings and evaluations. The program went well, but hit a serious roadblock. Interest rates! By the mid-1970s, interest rates assumed double digits, and at the time Photomarker wanted to commercialize, interest rates reached nearly 20%. They did not have the resources to go on, and we again were stymied in the development of our technology. Soon thereafter, the relationship between Stern and Yentis soured, and Photomarker disappeared from our horizon.
C. Miscellaneous Activities By spring of 1977, we were running low on the funded programs and considering the progress we had made, it seemed certain that Photo Products would
1977—AND STILL MORE OPPORTUNITIES
229
continue the programs. We did not count on Upson’s desire to exclude me from this work. Photo Products assigned three chemists, to work under Bill Hardam, a former Dylux Venture supervisor, to see where they could take the Dylux-4C chemistry, which I had developed, over a 3-month spell. I was never invited to any meetings, only a written report and samples were requested. After 3 months, the Photo Products Department decided that the four-color Dylux program would not lead to a product that could compete with the Cromacheck1 peel-apart program, which they were pursuing with more enthusiasm, and that was the end of colored Dylux. In retrospect, this was another one of Upson’s devious activities. The fourcolor Dylux, which the marketing people liked, offered a quick, and inexpensive four-color overlay system, which could have competed effectively with 3M’s color overlay systems. The sensitometry, which we had developed, showed remarkable consistency; only the black required a bit longer exposure to bring the color to full density. This was established by Dr. E. E. Grubb, of Photo Products, who had been asked to establish the performance of these films. The cyan was an excellent color match, and the red and yellows could be tuned close to the lithographic standards. Another company, with a more open-minded research management would no doubt have made a product out of all that.
D. J. C. Penney—Inventory Control In April 1977, I received a phone call from a Richard Koenig, an engineer, who was working on an inventory control system at J. C. Penney. He asked me if we could sensitize a portion of a form, so that it could be employed in an inventorycontrol system that they were developing. The goal was a method and means for automatically inventorying items having coded tags or labels using a characterrecognition device, such as a wand scanner, in combination with a system for specially marking the labels. A flash device was to be actuated upon the successful completion of a reading to irradiate the label with UV light, to produce a visible and permanent mark. The question was whether a preprinted form could be overcoated with Dylux or whether a Dylux-coated paper could be overprinted. This was easy to test at the DuPont printing plant, where a friend, Mr. Russell K. Johnson was the technical manager. As in so many other instances where it was easier to run the experiment than to analyze it to death, we found that it could be done—either way. I communicated the results to Koenig, and he invited me to visit a potential equipment maker on the West Coast. Johnson was an incredibly enthusiastic man, who welcomed any scientist to his plant and helped in steering research into useful channels. By that time, I visited the modern printing plant almost
230
THE INVENTION OF DYLUX
on a weekly basis and at lunch soaked up as much of Johnson’s extensive knowledge of the printing industry. The program continued and Penney invited DuPont’s Photo Products Distribution Manager to visit the system as it was being tested. He was very impressed, and thought it might help the company broadly. Unfortunately, his enthusiasm could not be converted into a new product opportunity for DuPont. A patent subsequently issued. U.S. Patent 4,180,204 Automatic inventorying system R. W. Koenig; P. R. Girard, and A. J. Neuhauser of J. C. Penney Corp. (December 25, 1979) A method and means for automatically inventorying items having coded tags or labels using a character-recognition device, such as a wand scanner, in combination with a system for specially marking the labels when they have been successfully read and recorded by the scanner, which system comprises the use of a photosensitive dye-forming material on the label and an ultraviolet light-producing flash device on the scanner. The flash device is actuated upon the successful completion of a recognition reading and irradiates the label with ultraviolet light which produces a visible and permanent mark on the label.
E. Tokyo SPSE Meeting Professor E. Inoue, of Tokyo Institute of Technology (TIT), one of Japan’s leading photochemists had interacted with us when he consulted for Canon Inc., with whom we were trying to build a relationship in the 1960s. He was a frequent visitor to my home and asked me if I would like to give an invited paper at the Tokyo SPSE meeting in 1977. It was an interesting opportunity for me to see another part of the world, but I found that Orchem would not pay for the trip, and the Photo Products R&D management was also not a likely source of support. I paid for the trip myself and was embarrassed that DuPont had no other representative at what certainly was a great technical meeting. I gave a paper about Dylux, which I authored jointly with Dr. Ed Abramson, who had done sensitometric studies on Dylux. It was well received. Later that day, the Fuji Photofilm Co. had a reception for the foreign visitors at a beautiful Japanese restaurant. One of their chemists asked if I would like to visit their research facility and perhaps give my talk there again—only more slowly. The talk was scheduled for the afternoon of a day on which I had been asked to give a talk at TIT in the morning. Around noon, a limousine, with uniformed chauffeur, arrived to drive me to the Fuji Lab, where I was received most graciously. I was invited to lunch, and this was followed by a tour of the laboratory, during which I was shown a number of new products, which no one in Wilmington was aware of. My talk went well, and then the director of laboratory invited me to his office for tea. He asked me to sign a document that he said was
1977—AND STILL MORE OPPORTUNITIES
231
for tax purposes and was required because of my honorarium. I countered that I could not accept money from Fuji for my talk, and he said that he had heard from Professor Inoue, that I had to pay for the trip myself, and that the money would only cover part of that. Well, we argued about it for nearly an hour, and he then asked me if they could send a present to my hotel. I said that I saw no reason why that would create a problem. I expected a fairly complex camera or similar, but instead received a small 110 camera. My wife used it for years, and in spite of its modest cost, gave us a lot of excellent pictures that would not have been taken otherwise. Also at the Tokyo meeting I met a Chinese chemist, who said that if I ever came to China, he would like to take me around. I sent him some Dylux paper, and in 1984 actually visited the Beijing Institute of Photographic Sciences. Several companies were hosting the foreign visitors. Canon invited me to a fine dinner, and I was impressed that the seven or eight Canon people there included employees at many levels—from research director to chemist. That was not a common practice in the ‘‘democratic’’ American industrial system. Ricoh invited several us for a plant tour and provided a large bus, equipped with bars, TV, and attractive women guides for the two Americans who showed up. I am certain that more had signed up, but did not come. I was embarrassed, but the Japanese took it in excellent stride and showed their usual superb courtesy. They also asked if they could visit me in Wilmington, and a week after I returned to the United States, I hosted three Ricoh visitors. We had some interesting discussions, to which I invited Phil Botsolas. After a few months of negotiation about what we might do together, the relationship cooled. However, Ricoh must have invested a fair bit of time exploring Dylux chemistry, because after that, nearly 90 Japanese patent applications as well as one U.S. patent were issued. Photosensitive and heat-sensitive preparations recording materials containing them. Kunikane, Ioto; Yasumori, Akiyoshi; Taniguchi Kiyoshi; Yamamuro, Tetu: (Ricoh Co., Ltd.) Ger. Offen. 3,012,954 (Cl. G03C1/72), Oct 1980, Japan. Appl. 79/39,348, 03 Apr 1979; 50 pp. Photo- and heat-sensitive prepns. consist of a leuco pigment, a H-donor, a photoreducing agent (which upon exposure to visible light forms a reducing substance through the simultaneous action of a H-donor), a photooxidizing agent (which upon exposure to UV light activates the leuco pigment to form a color and loses its own oxidizing ability through reaction with the reducing agent, Co complex (which reacts with the reducing agent and thus suppresses the reaction of the color-forming system). Recording elements consist of a support with a layer contg. such photo and heat-sensitive prepns. Thus, a soln. was prepd. from (NH3)6(CF3CO2)3 200, the reducing agents p-benzoquinone 60 and 9,10-phenanthrenequinone 10, poly(ethylene glycol) 500, the oxidizing agent 2,20 -bis(o-chlorophenyl)-4,40 ,5,50 -tetraphenylbiimidazole 132, the leuco pigment tris(4-diethylaminoo-tolyl)methane 6.0, p-toluenesulfonic acid monohydrate 40 mg, Me2CO 9 and
232
THE INVENTION OF DYLUX
iso-PrOH 1 mL. A filter paper was impregnated with this soln. and dried at 30 . A portion of the paper was exposed repeatedly to visible light of 50,000 Ix from a UV-lamp through a filter (UV 39) and the exposure times were 4, 6.5, 12.5, 30, and 40s. The sample was then passed through a pair of heated rollers (130 ) and subsequently exposed 5s to 30-W pure UV-radiation at 4 cm. The regions exposed 12.5 sec. with visible light form no color even with 5-s exposure to UV radiation, whereas the regions not exposed to visible light form blue color of d. 0.8. For comparison, a filter paper impregnated with a prepn. identical to the above but without the Co complex. This paper underwent the same treatment as above except the heat-treatment at 130 ; it required 30 s exposure to visible light before the image was fixed. (U.S. 4,332,284 (6/1/1982).
I spent a week in Japan subsequent to the meeting. During this time, Bill Simkins, who ran the Tokyo Sales office, and Jack Rodowski, the Director of Sales for Asia, asked me to visit several Japanese companies, including Dai Nippon Printing, and more important, paid for my hotel stay and frequently took me to the American Club, where the chef prepared excellent black forest cake. It was another instance where my excellent relations with the Sales Division of Photo Products helped them and me, much to the irritation of Robert Upson.
F.
X-Raylux: Instructional X-Ray Prints
Another opportunity for Dylux paper was as a possible medium for instructional X-ray prints. One of DuPont’s consultants in Germany, a Professor Ackermann, was quite intrigued with the idea of having an inexpensive paper print, which could be marked up, to show particular aspects of the radiogram. He referred to this as X-Raylux, an instructional X-ray picture, unlike the diagnostic Xray film, which of course would be an expensive silver-based product. This aroused some interest in the X-ray section of Photo Products, and I was asked to look into how we might modify the proof paper to make it fit better into that application. One way to heighten the contrast was to add an orange dye to the recipe, and the now black-and-orange prints has visually more contrast and appeared to meet some of Ackermann’s proposed requirements. What was needed, of course, was a simple printer, which would allow the exposure of the film radiogram onto Dylux paper. As X-ray film was fairly opaque, some work to make a good exposure was needed. I asked Wilbur to consider building an exposure unit for the instructional X-ray print market, and as usual, he delivered an operating unit within a few weeks. We gave it to Botsolas, who was enthusiastic about its performance, and the speed with which the unit came into being. Of course, circumventing the system is not how to become popular in large companies. We had bypassed the
1978—TRANSITIONS
233
equipment division of Photo Products Department, who had not deemed this worthwhile, and we had a potential conflict of a DuPont employee being set up in business to provide equipment, even if he did it on his own time. There was so much opposition to the use of Wilbur’s machine that Botsolas had to apologize to him to get him involved with that. We never field tested the device. Wilbur was so angry, he ran a truck over it!
XXII.
1978—TRANSITIONS
A. Another Dual-Response System Dr. Jose L. Pazos in DuPont’s Central Research Department worked on a photopolymerizable compositions containing an o-nitroaromatic compound as photoinhibitor. He and the late Dr. George Nacci invented several interesting approaches to improving the hexaarylbiimidazoles system when applied to photopolymer technology. Some of his work could have been applied to color-forming chemistry, but was not. U.S. Patent 4,198,242 Photopolymerizable composition containing an o-nitroaromatic compound as photoinhibitor. Pazos; Jose F. E. I. Du Pont de Nemours and Company (Wilmington, DE) (Filed: June 8, 1978) IPC G03C 002/ 6 Pat.Cl. 430/286.1; 430/301; 430/910; 430/917; 522/26; 522/28; 522/63; 522/ 65; 522/79; 522/109; 522/121 A photopolymerizable coating composition comprising (1) a nongaseous, ethylenically unsaturated, polymerizable compound, (2) a specified nitroaromatic compound, and (3) an organic, radiation-sensitive, freeradical generating system is useful for making a positive or negative polymeric image on a substrate
B. Transfer to the Photo Products Department By late 1978, there were ominous signs that DuPont would leave the dyes business. This had been rumored for many years. When DuPont built a modern dyemanufacturing plant in Puerto Rico in the late 1960s, it was felt that this was a commitment that the dye business would be a continuing activity. Unfortunately, the plant was a disaster. It suffered from its distance to suppliers and consumers, and it was a considerable distance from the technical support that was required to keep things going. As a results of these and other factors, the plant hemorrhaged earnings from the dye side of Orchem’s business. On top of this, in the late 1970s, the company decided to rearrange the dyes and chemicals business to be part of a new conglomerate, the Chemicals, Dyes and Pigments Department (CDP). It was sort of an internal takeover, and the ones who were
234
THE INVENTION OF DYLUX
managing it were not interested in rescuing the dye business, which was already in ill-repute as a result of previous mismanagement. My immediate supervision tried to place me in Photo Products Department again, and since Robert Upson had retired, it seemed that it might happen now. One afternoon in October I received a call to visit Dr. Buck Arrington, the new Research Director, who had taken over from Upson. I had interacted with him briefly on a number of occasions, and expected a pleasant meeting. Nothing of the sort. Arrington closed the door to his office after I sat down, and asked me if I had any idea of what his opinion of me was. I said that I hoped that it was good. He countered that as far as he was concerned I was ‘‘a royal pain in the—,’’ that from what he had seen I had done little of value for the company, and from what he had heard, I was very difficult to manage. I was somewhat shocked, particularly by his language. By that time I had worked for DuPont 26 years, had received a number of promotions, over 20 patents, obtained bonuses and had always had positive performance reviews. I asked him if he had ever heard of Dylux, TCTM-HABI, proofing, etc. Arrington said that they had always been done by others and that I was claiming credit for things that I had not done. If his goading me was an attempt to provoke me to hit him, he came close to achieving his goal. I did not because I thought that he might be joking. He then told me that he would reluctantly take a chance of transferring me to his department, but that I would work under the close supervision of Dr. Peter Walker, then a manager, and that I would be treated like everyone else—half a technician, no travel, nothing without permission, etc. I was in no position to bargain and went home, very upset. That evening I had dinner with Dr. Suzukawa, Director of Research of Canon Inc., who was visiting the United States and who had asked me to make some contacts for him with the DuPont group, which was working on magnetic media. I came close to asking him for a job. I spoke with Walker the following day, and expressed my concerns. Walker, an able and gentle man, told me that we would have no difficulty in getting along, and that I should work on improving the socalled black Dylux, which was still olive, as my work several years earlier had been ignored. I must say that after this initial interaction, my relationship with Arrington was always quite good, but I came to realize that I was working in an organization where the authoritarian style of the research director, whether Upson or Arrington, made for some very unhappy researchers. Years later, after Arrington retired, things took a great change for the better.
XXIII. 1979—BLACK DYLUX1 535 PROOFPAPER I found it rather stupid that the half a technician per chemist would be so rigidly enforced. First of all, most technicians then had more service than their chemists
1979—BLACK DYLUX1 535 PROOFPAPER
235
and always had longer vacations. Second, technicians were always drafted for tasks like inspections, cleanups, etc., and so the chemist time was poorly utilized. Still, I made progress, particularly since I now had available TCTM-HABI, which more effectively oxidized leucodyes than the hitherto used CDM-HABI. I found that the mixture of leuco yellow and leuco purple dyes invented by the very talented Dr. John F. Neumer during the Dylux Venture days were still effective as precursors for a black shade; what was critical was their purity and balance. Leuco compounds oxidizable to dyes. Neumer, John Fred (du Pont de Nemours, E. I., and Co.) Ger. Offen. 2,220,383 (Cl. C 07c, C O9b, C 03c), 09 Nov 1972, US Appl 139,281, 30 Apr 1971; 47 pp. Colorless indanones (I) were prepd. which, when oxidized, gave strong yellow dyes (II) or, in combination with triphenylmethane deriv. (III) [38615-38-2] black dyes. In structures I and II, R ¼ H or Me and Rl ¼ H, CONHC6H4NO2, CONHPh, or CONHC10H7-1. I and I - III mixts. together with a light-activated oxidizing agent, e.g. a hexaarylbiimidazole, were used in photoimaging systems. Thus, 1,3-indandione was reduced, the 3-hydroxy1-indanone treated with p-Et2NC6H4CHO to give II (R ¼ Rl ¼ H)(IV), and IV
O
O H
H H R
R
H OR ′
H OR ′ NEt2
NEt2
I
II NHCO(CH 2 )6 CH 3 CH 3
OC H3 CH 3
(C 2 H5 )2 N
N(C 2 H5 )2
IV
reduced to the transleuco dye (I, R ¼ Rl ¼ H) [38615-39-3]. The carbamate derivs. of I (Rl ¼ H) were prepd. by reaction of the hydroxy compds. with arom. isocyanates. III was prepd. by the condensation of 3,4-Me(HCO)C6H3NEt2, o-MeOC6H4NH2, and urea in the presence of H2SO4, acylation of the intermediate 2,4Me(Et2N)C6H3CH(NHCONH2)C6H3(NH2)OMe-4,3 with Me(CH2)6COCl and reaction of the octanamide with m-MeC6H4NEt2, in HCO2H.
236
THE INVENTION OF DYLUX
Within a few months, I was able to visit James River Graphics, who were then doing custom coating of Dylux proof paper and make large coating runs in association with Tom Sheets of Towanda, and Mr. Cary Giles, who had been a big supporter of our technology from its inception. The marketing manager then was Tom Haller, an enthusiastic supporter of my work. He and I went on a few field trips, and he arranged for me to present a paper about recent developments in Dylux chemistry at the 1980 SPSE meeting at the Homestead, in Hot Springs, Virginia. When the conference chairman learned of my proposed paper, he invited me to be sessions chairman, a small, but nevertheless significant honor. Naturally, it was more complicated then that. Haller had to contact Dr. A. B. Cohen, who was Walker’s boss, who in turn had to contact Arrington. Walker then told me to prepare a paper, which I did. We submitted it to Cohen, who urged me to expand it to cover not only black Dylux but also competitive technologies. Walker felt that this would weaken the presentation, but that we should give some lip service to competitive materials. At that time, it was common practice to submit papers separately to each layer of management. Harry Knop, the laboratory director, had been away on a vacation trip to China, and Arrington had been away someplace else. I asked Walker what to do. We scheduled a meeting with Knop the first day he returned; and while the latter had difficulty remaining awake, he was alert enough to make all sorts of criticisms, such as by whose authority I had called the new product Dylux 535, why I included the names of chemicals (giving competitors unnecessary information), etc. I made the needed changes, and waited for my audience with Arrington and Cohen, to which Walker and Knop accompanied me. Cohen had more important things to do and, anyway, had wondered why we were giving the paper at all. After all, he knew nothing of the paper! Of course, it was Cohen who had suggested to Walker that I give the paper. Walker remained silent, not wanting to stir things up further. When Arrington looked at my slides, he said they were ugly—no one working for a photographic company should have such miserable exhibits. The slides were ugly because Knop would not allow us to use the Experimental Station’s slide-making facilities, and we had to draw them ourselves. Arrington suggested that I go to the General Services Department, where an attractive young woman helped beautify my slides. Some years later I dated her briefly; I do have to thank Arrington for many pleasant evenings. Arrington stated that a technical paper should contain sufficient information that a scientist skilled in the art should be able to repeat it. For that, he had to know what chemicals were used. I was glad to hear that. Knop did not have the honesty to tell Arrington that he asked me to delete these. Arrington then wondered whether I had to go to the meeting at all, after all, what would DuPont gain from that? Walker said I was already on the program, and it would look bad if I withdrew. Arrington sort of gave me weak endorsement to go. On the way
1981—ADD-ON TONING
237
back to the Experimental Station, there was a stony silence between Walker, Knop, and myself. Walker suggested that I take an early flight to the Homestead, so that I could make it back and forth in a day. As I already had decided to drive and spend the time of the full meeting there with my wife, I did not care for the one-day trip; besides that, my paper was to be given around 8:30, making it impossible to get there in time. I paid for the stay at the Homestead out of my own pocket but had a good time. At that meeting there were many young Japanese scientists and many older American scientists. My wife wondered why that was so. I said that the Japanese send people to learn at meetings—the Americans use it as a reward for good behavior. In my case, perhaps an exception.
XXIV. 1981—ADD-ON TONING After the work on Dylux 535 was completed, Walker asked me to look into addon toning (AOT) for electronic applications. It was another extension of Dylux chemistry, where the photo-oxidation of a dihydropyridine derivative caused a change in the tackiness of a surface. AOT had been developed at the Neu Isenburg Laboratory by several talented scientists, including Drs. Walter Abele, Mario Grossa, and Dieter Tiegler. It was to be a negative-proofing system, in which a surface would be sequentially tackified and toned. The basis of the chemistry was the photo-oxidation of dihydropyridines (DHP) and similar materials. It was believed that the oxidation resulted in formation of an aromatic pyridine derivative, which added adhesiveness to the coating surface. A number of dihydropyridine derivatives were prepared from readily available aldehydes and acetoacetic esters. Typical of these was derivative I. H
CH 3 CO 2 C 2 H5
C 2 H5 O 2 C H3 C
N H
CH 3
I
To produce a near-white surface, a new HABI, o-EtO-HABI was developed. Conceptually, it was beautiful work. Practically, it failed to overcome a problem— the ability to superimpose one toner (color) over another so that, for example, green shades, which depended on sequential yellow and cyan layers, could not be readily formed. Grossa, in 1979 gave a paper about this technology at the Washington, D.C., SPSE meeting and was approached by a manager at RCA who wondered if DuPont would consider this technology for the manufacture of color TV tubes. He pointed out that the color CRT manufacturing process
238
THE INVENTION OF DYLUX
was complex, requiring multiple coatings, exposures, and washings to position the different phosphors and shadow mask. The add-on toning would be potentially simpler and less costly. Surprisingly, few people know how color television tubes were manufactured. The different color phosphors must be carefully positioned, and this is accomplished by coating the glass sequentially with a slurry of different phosphors, exposing these through a mask to harden some areas, and then removing the unexposed areas, before applying the subsequent phosphor slurry. Three cycles (R, B, G) of this were necessary, in addition to application of the black shadow mask. Obviously, AOT would offer advantages. Walker and Cohen asked me to work in this area, as well as examining the feasibility of using toning to position metallic particles so as to produce, on fusing, circuits. It was an interesting program, and I was able also to count on help from Dr. V. C. C. Chu, a respected chemist who worked in our building and who had also been involved in the AOT for negative proofing. Competitive technologies existed, such as a system investigated by Hitachi, which relied on photolysis of zinc salts of diazonium compounds. I tried to persuade Walker that we should get together with RCA or another tube manufacturer before long, as I felt we needed to learn more about tube manufacture. We finally made contacts with the RCA Laboratory in Princeton, New Jersey, and I was asked to work with Phyl Branin, a chemist who had worked in the field for many years. I was able to supply her with samples of coated glass slides, and she evaluated different coating variations and we made good progress. The conventional wisdom was that the system functioned because the generation of the pyridine gave rise to an oily species, which provided the adhesive component of the coating. On one occasion my technician, inverted the ratio of dihydropyidine and HABI, and found that the unexposed areas toned, but not the exposed areas. Clearly something was counter to our comprehension of the system. I repeated the experiment several times, but was never able to postulate a satisfactory mechanism for this. The program was then turned over to another chemist who was located in Parlin, New Jersey, much closer to the RCA Laboratory. She soon persuaded her bosses that there was no market for this technology and was promptly promoted! RCA eventually became part of GE, the Sarnoff Laboratory where Branin worked was closed, and the CRT program was returned to Neu Isenburg, where Grossa et al. managed to make progress.
XXV. 1983—A NEW WIND BLOWS Arrington retired in the early 1980s, and his replacement, Dr. Don Rodgers, was a far cry from the authoritarian martinet who preceded him. In 1983, Dr. Thomas
1984—CHINA
239
D. Smith, who was at the same level as A. B. Cohen assumed responsibility for the direction of the Experimental Station Laboratory and encouraged a far broader attitude toward research than had been hitherto the case. I was asked to look at higher-speed HABI systems, as well as thermal imaging technologies. We found that purification of HABIs and removal of impurities from photosensitive layers did indeed speed up the chemistry; Dr. R. David Mitchell, whose laboratory was adjacent to mine worked on a novel amplification system, and things seemed at long last turn into a positive direction. A new high-speed photopolymerization system. Mitchell, R. David; Nebe, William J.; Hardam, William M. (Photosyst. Electron. Prod. Dep., E. I. DuPont de Nemours and Co., Inc., Wilmington, DE 19898 USA). J. Imaging Sci. 1986, 30(5) 15–17 (Eng). In a novel high-speed photopolymer imaging system an imagewise exposure to long wavelength radiation forms a photopolymn. sensitizer. A second overall exposure to shorter wavelength, which is not absorbed by the sensitizer precursor polymerizes the image areas where sensitizer was formed. This system has high speed not only because the high intensity overall exposure generates the image polymer, but also because the sensitizer promotes formation of addnl. photopolymn. sensitizer. This allows an imaging step in which only a very low radiation dose is required to generate enough sensitizer for the second overall exposure. Imaging energies as low as 0.1 mJ/cm2 were achieved; the mechanism and system requirements are discussed.
We also investigated the effect of using highly purified HABIs in photopolymerization, and saw improvements; the U.S. Patent Office, however, did not reward us with a patent.
XXVI. 1984—CHINA In 1984, I visited China on a vacation trip. I recalled that in 1977 I had met a Chinese scientist, a Dr. Liang, who asked for samples of Dylux to be sent to his lab at the Beijing Center of Photochemistry of the Chinese Academy of Sciences. He said if ever I were to come to China, he would be happy to have me visit his laboratory. In September I wrote to him, and within a fortnight I had an invitation. I was impressed at the speed at which things like that happen. On the morning of my visit, an old limousine, with a uniformed driver and Dr. Liang met me at my hotel, just as the other tour members were assembling. They were impressed as I was driven to the lab. There I saw a message there on a blackboard, in Chinese, except for the words Dylux and Dr. Rolf Dessauer. A professor, who had studied in the United States told me he would be my interpreter. About 40 people attended my seminar, and after about 5 minutes there
240
THE INVENTION OF DYLUX
was a strange noise emanating from the entire audience. I asked my interpreter what was going on. He said it was a signal that they could understand me, and he sat down, somewhat dejected. I continued slowly, and there were lots of questions. It was an interesting experience. Afterward I was given a tour of the laboratory, and then we had lunch. Someone had brought in a semiprepared meal—was it a Chinese takeout?—and cooked it over a Bunsen burner. I found everyone very friendly and interested in everything I had said. I thought that I should bring my host a gift, and had bought a book about the National Parks in the United States. It seemed a fairly neutral present. He and the chauffeur returned me to my hotel, and on the way back, I asked if we could all celebrate my visit at the hotel bar. I was told that that would be fine and that the chauffeur should join us, as he too was a member of the institute. I admired this sort of democracy—I could not imagine that the chauffeur of the CEO of the DuPont Company ever was invited to join his boss for a drink. However, the chauffeur could not drink an alcoholic beverage. I had brought along a bottle of expensive Scotch whiskey in case my host could accept it. Having established that he liked gin and tonics, I asked if I could present him with the whiskey. He was delighted. He then invited me to his house for dinner— something that no one on my tour had ever experienced. Unfortunately, I could not escape the tour for that. But it would probably have been the most interesting moment on my trip. Instead, the following morning, a package of excellent Chinese tea was delivered to my hotel. A few weeks after my return from China, I received a phone call from the CIA. They asked me if I could tell them about my visit to China. I said I would be glad to, and fortunately, had just read an excellent article about China in The New Yorker magazine by Orville Schell, so I was impressively knowledgeable.
XXVII. 1984–2004 A limitation of the proofing systems available in the 1980s was the difficulty of making multiple proofs. If a second or third proof was required, the process of making a proof had to be repeated. Marketing studies indicated that there was indeed a need for a proofing system that could deliver as many as 10 copies, on conventional paper, as used by the printing industry. Dr. Thomas D. Smith, Assistant Director of Research at the Photo Products Department assumed responsibility for identifying a new technology, which would address this need. A group at the Experimental Station, headed by Dr. John Bierlein and later Dr. Alan S. Dubin, was organized and I was assigned to it. What technologies were available?
1984–2004
A.
241
Xeroprinting
An ambitious concept, to use a combination of photopolymerization and electrophotography was considered, especially after some earlier work by a former DuPont chemist, Jim Riesenfeld, had shown that Lydel printing plates might be used as electrophotographic masters. I was able to follow up on Riesenfeld’s work, and found that photopolymerizable coatings on a metallic substrate, here aluminized Mylar1, could be charged and toned; the exposed areas were more dielectric than the unexposed areas, and toners could be designed to work with this system. This approach, identified as xeroprinting, would permit a single exposure to generate an electrophotographic plate, which could be relied upon to produce multiple copies as desired. Xeroprinting with a photopolymer master. Riesenfeld, James; Bindloss, William; Blanchet, Graciela; Dessauer Rolf; Dubin, Alan Sander (du Pont de Nemours E. I., and Co.) Eur. Pat. Appl. EP 243,934 (Cl. G03G5/026), 04 Nov 1987, US Appl. 858,172, 01 May 1986; 24 pp. A xeroprinting process is described which involves imagewise exposing a photopolymer master to selectively polymerize and thereby increase the resistivity in the exposed areas of the layer, charging by corona discharge to form an electrostatic latent image in the exposed areas, developing the image by applying an oppositely charged electrostatic toner, and transforming the toner image to another surface. Photopolymer systems for use in the process contain polymerize binders, ethylenically unsatd. monomers, hexaarylbiimidazole initiators, and chain transfer agents. Thus, an Al coated PET support was overcoated with a soln. contg. poly(memethacrylate), ethoxylated trimethylolpropane triacrylate 2,20 ,4,40 -tetrakis(o-chlorophenyl)-5,50 -bis(m,p-dimethoxyphenyl)-biimidazole, 2-mercaptobenzoxazole, and CH2Cl2, dried, a polypropylene cover sheet added, imagewise exposed, and the sheet charged to produce a latent image in the polymd. (image) areas. The latent image was then toned and the toned image transferred to a receptor paper to produce a quality image. This issued as U.S. 4,732,831 (3/22/1988).
Parallel to this effort, DuPont was in contact with Indigo, an Israeli company, which had pioneered the development of liquid electrophotographic toners. Indigo’s founder, Benny Landa, who was to be awarded the Land Award for His Contributions to Imaging Science in 2002, claimed that liquid toners offered some unique advantages over dry toners, such as improved density, better resolution, and ease in handling. Landa had licensed technology to Savin, who in turn sublicensed it to Xerox, DuPont, and Harris-Intertype. It was not long before a business relationship developed between DuPont and Xerox, who were asked to build a machine to demonstrate feasibility for using DuPont’s photopolymer masters. A collaboration was initiated in 1986 and resulted in a machine that was demonstrated on the floor of the Graph Expo exhibition in Chicago in November 1987. The relationship was sufficiently productive that a joint venture
242
THE INVENTION OF DYLUX
between DuPont and Xerox was organized as DXImaging, with its own management, at a site distinct from Wilmington, Delaware, and Webster, New York, in order to be able to assert its independence. DuPont’s mission was to develop liquid toners based on Landa’s work and to supply master materials. The partnership was to develop technology and devices for the graphic arts industry served by DuPont and for the office copy community, served by Xerox. We introduced a printout system, to demonstrate that the plates had been imaged, by incorporating a leuco dye in the formulation and discovered that there was improved back transfer and charge decay control. Photohardenable electrostatic master having improved backtransfer and charge decay. Blanchet-Fincher, Graciela B.; Fincher, Curtis R., Jr.; Cheung, Lawrence K. F; Dessauer, Rolf; Looney, Catharine E. (DuPont de Nemours, E. I. and Co.) U.S. US 4,818,660 (Cl. 430-281; G03G5/026), 04 Apr 1989, Appl. 117,189, 04 Nov 1987; 14 pp. Photohardenable electrostatic master comprising an electrically conductive substrate, e.g., aluminized polyethylene terephthalate, bearing a layer of a photopolymer comprising an organic polymeric binder, compound having at least one ethylenically unsaturated group, photoinitiator, an aromatic amino compound, as defined, and an acid, e.g., p-toluene sulfonic acid, or an oxidized substituted aromatic amino compound. The photohardenable electrostatic master having improved charge decay and backytransfer is used for electrostatic proofing, etc.
There were remarkable technical challenges. In the early phases of the program, the imaging materials were superior to the machinery that was used to form and transfer images. As the machines became better, the flaws in the materials needed to be overcome, and they were. There was constant improvement, and what was achieved two years after the Graph Expo demonstration demonstrated remarkable progress. The system as conceived involved a complex machine, ca. 15-ft: long, on which exposed masters were mounted on drums and toned, and the toner was transferred to a paper support. This system gave negative-type images, which were desired by the domestic market, but alternate masters were required to satisfy the positive-type proofs required in Europe and in other parts of the world. Two routes toward positive-type proofs were discovered. R. D. Kempf of Towanda R&D showed that the dual response system, in which a visible exposure gave rise to an inhibitor, followed by a flooding UV exposure yielded masters, which could be toned with our conventional liquid toning system, and gave positive-working proofs. This was an illustration of applying the photopolymer technology of HABI chemistry for yet another application. Method for preparing positive and negative images using photohardenable electrostatic master. Kempf, Richard J. (du Pont de Nemours, E. I., and Co.)
1984–2004
243
Eur. Pat. Appl. EP 315,121 (Cl. G03G5/026), 10 May 1989, US Appl. 116,656, 04 Nov 1987; 35 pp. The title method comprises an elec. conductive substrate bearing a photohardenable layer contg. a binder, a monomer, a photoinitiator combination, a photoinhibitor, and 1 visible light sensitizer, and the neg. or pos. image is produced by: (1) exposing the photohardenable layer to visible or UV light resp.; (2) electrostatically charging to form an image on the exposed area; (3) developing with a toner, and (4) transferring the toned image to a receptor surface. The method can be used in producing color proofs. Thus, a PET film was coated with a photo-imaging compn., and the compn. patternwise exposed with visible light, electrostatically charged, and then developed to obtain an excellent quality neg. image.
Another route using our standard HABI/leucodye system showed that the formation of cationic species on UV light exposure gave rise to sufficient conductivity to permit charging and toning to yield the desired positive images. It was just another form of Dylux—on a conductive support! Direct-positive photoimaging material for electrographic master preparation. Kempf, Richard J; Dessauer, Rolf; Freilich, Steven C. (DuPont de Nemours, E. I., and Co.) US 4,945,020 (Cl. 430 49; G03C13/32), 31 Jul 1990, Appl. 374,491 30 Jun 1989; 14 pp. A direct-pos. photoimaging material for theprepn. of electrog. masters for use in color proofing comprises an elec. conductive support bearing a layer of a photosensitive compn. consisting of >1 polymeric binder, a hexaaryl-biimidazole photooxidant a leuco dye that is oxidizable to an ionic species by the photo-oxidant, a nonionic halogenated hydrocarbon, and a plasticizer. The photosensitlve compn. may also contain a spectral sensitizer and a thermal stabilizer. The polymeric binder is preferably selected from the group consisting of poly(Me methacrylate), polystyrene, an cellulose acetate butyrate. The plasticizer is preferably selected from the group consisting of dioctyl phthalate, triacetin, (tert-butylphenyl)diphenyl phosphate, diethyleneglycol dibenzoate, and (2ethylhexyl)hexyl phthalate. By a single imagewise exposure of the photo-imaging material, an electrog. matter having a high-resoln. pos. image is obtained.
All systems offered us the ability to generate as many as a dozen proofs of good quality. A number of patents issued, and the technology was described in the following abstract: Photopolymers in electrostatic imaging applications. Blanchet, Graciela B.; Dessauer, R. (E. I. DuPont de Nemours, Cent. Res.. Dep., Wilmington, DE 19898 USA). J. Imaging Sci. Technol. 1993, 37(3), 238–45 (Eng). In this work the authors discuss the feasibility of using photopolymers as electrostatic masters in electrophotog. because they provide the advantage of permanent image formation necessary for multiple copy applications. For the compns. described here, multiple copies can be achieved without any measurable degrdn. in image quality or image resoln. In addn., the latent image of photopolymer masters can be
244
THE INVENTION OF DYLUX
precisely tailored by independently adjusting the elec. properties of the exposed and exposed regions, which, in turn, control the dot gain and dot range of the image on paper. A detailed description of the chem. and phys. characteristics of photopolymer electrostatic masters is presented here. The effects of temp. and humidity on the elec. properties and, consequently, on image quality are also described.
The development of the machines to achieve this level of automated proofing went along at a rapid pace, and the EMP proofing system was installed at a betatest site in New York, where we learned of the critical need to have the machines operate in a humidity controlled environment. One of the eight machines, which were built was demonstrated at DRUPA in 1990, and was favorably received by many of the visitors to the DuPont booth. Challenges toward commercialization lay ahead. The proofing machines, initially programmed to cost less than $200,000 each were found to cost more to manufacture, and the anticipated use of any paper supports for the proofs turned out to be difficult to meet. Nevertheless, the program moved along well until January 1991, when Xerox and DuPont decided to dissolve the partnership, and terminate further work. Those scientists who had been employed by Xerox and DuPont returned to their previous employers, and the remainder were laid off. I had retired from DuPont before joining DXImaging, and then began a new career as consultant.
B. Photoimaging, Ltd. During one of my trips to England to image our next-generation laser-exposable DXI masters, I met Gerry Pollack, an inventor of light sources and fixtures. We were looking for a lightpen with which we could correct our masters, i.e., achieve localized polymerization, where for some reason or other, the coating was not exposed. Pollock had some clever ideas and put a device together. Of course, we talked about Dylux proof paper, and he asked if we could use the technology to decorate wood. I told him that we had done that years earlier and had almost reached commercialization. Pollock suggested we get a sample of photodecorated wood, which I had stored in my garage, and take it to the Hanover Fair, to be held a few weeks later. Well, this seemed to go too fast, even for me. So I proposed that we consider this technology during my next visit to his lab, and he promised to line up some investors to back this operation. In turn, I contacted DuPont, which owned the technology, and got an agreement to see if a license could be arranged. Two Englishmen, who formed a new company, Photoimaging, Ltd. came to America in September 1990 and talked with the Manager of Intellectual Property and Licensing, and I was allowed to work on this program in my spare time. In March 1991, after the DXImaging operation was closed down, I had
1984–2004
245
lots of spare time. My new English associates were generous in time, money, trips to Europe and the like, and we made progress. I felt it was necessary to line up a manufacturer of lacquers; and in time, we identified one in York, England. They sent me to Italy to visit one of their subsidiaries that was in the furniture lacquer business; they were moderately enthusiastic, but had difficulty visualizing a way toward commercialization. Nevertheless, the concept continued to look attractive, and we speculated that it might be better to go after some sort of fashion opportunity, such as the photodecoration of leather, which would, if successful, generate some revenue rapidly. In the great style of the backer of this venture, we had a splendid luncheon at the Savoy Hotel in London, and outlined a business plan. This was in November 1992. All of a sudden, I heard no more from my colleagues. I could not figure out what had happened, and received no response to faxes, phone calls, and letters. It seems that the principal had been involved in a motoring accident on the way home from the Savoy, and struck a bicyclist. In time, he was to spend a year in jail. This ruined him financially, and regrettably, also killed photodecoration in Europe. Somehow, photodecoration seemed to be star-crossed.
C. Competition Dylux proof paper was fortunate that it never had any serious competition until long after its introduction. In 1969 we became aware of S. D. Warren’s Fotoproof, which involved clever chemistry but required a thermal aftertreatment, which necessitated installation of venting equipment to remove toxic off-gases. These proofs offered the user some real advantages; they were coated on a better-quality paper than Dylux, and gave higher image quality, as a result of a UV-absorber in the paper base, which reduced backscatter. In addition, the black-brown proof looked visually more attractive than the cyan color for Dylux 503. However, as environmental concerns began to become important in the 1970s, Warren ultimately withdrew its offering. During the period of the Dylux Venture, Eastman Kodak, Agfa-Gevaert, and Minnesota Mining and Manufacturing Companies were DuPont’s major competitors in the domestic imaging field. All the time, we were looking to find out if these companies had any developments based on HABI technology, which could have threatened our developments in the market. It was of interest to find out when there would be some patent activity by one of the major photographic companies. The first of these was issue to Agfa-Gevaert in 1970. It employed HABIs in combination with silver photographic materials. Photothermographic reproduction system and photoensitive materials. Poot, Albert L.; Van Besauw, Jan Fys; Theofiel H. (Agfa-Gevaert A.-G.) Ger. Offen.
246
THE INVENTION OF DYLUX
2,117,054 (Cl. G 03c), 21 Oct 1971, Brit. Appl. 09 Apr 1970; 47 pp. In a component thermog. copying system, 1 sheet is treated with a photosensitive biimidazolyl deriv., which may be spectrally sensitized with a cyanine dye, and a photooxidizable compd., such as a naphthol or a hydroquinone. The image-receiving sheet contains a compd. that will react with the photooxidizable cmpd. and a heat-sensitive reducing agent. Thus parchment is coated with 40 g/m2 of a mixt. contg. bis[2(o-chlorophenyl)-4,5-diphenylimidazole] 10, 4-methoxy-1-naphthol 1, and ethyl cellulose 10 g in 2-butanone 500 ml. After uv exposure through an original transparency the sheet was contacted for 5 sec at 125 with an image-receiving sheet of white paper coated with 0.2 g/m2 Ag as Ag behenate, 400 mg/ m2 phthalazinone, and 40 g/ m2 2,6-di-tert-butyl-4-methylphenol in poly(tert-butyl methylate) binder. A copy having a sharp pos. black image was obtained (U.S. Patent 3,734,733, May 22, 1973).
3M was intensely involved in exploring Dry Silver1 technology, in which a silver-based coating was exposed to light and developed thermally. This gave high-contrast, high-quality prints, but the applications they pursued were not of any threat to us. One 3M HABI-patent appeared in the 1970s. Imaging composition. Streeper, Richard D. (Minnesota Mining and Mfg. Co.) Ger. Offen. 2,653,669 (Cl. G03C1/72), 26 May 1977, US Appl. 634,619, 24 Nov 1975; 40 pp. Imaging compns. capable of giving both pos. and neg. images when irradiated at 250–400 nm contain a dye in the leuco form, and a hexaaryl biimidazole. The addn. of an O2-sensitizing compd. to this compn. increases the sensitivity up to 700 nm. Thus a soln. contg. bis(4-dimethylamino-2-tolyl (4-dimethylamino2-chlorophenyl)methane 50, 2,20 ,4,40 ,5,50 -hexaphenyl-biimidazole 50, ethylated Vat Blue 18 (C.I. 59815) 2.5, p-toluenesulfonic acid 25, N-ethyl-4-toluenesulfonamide 750, Et cellulose 300, and Me2CO-BuOH (95:5) 14470 parts by wt. was coated on a photog. support under darkroom conditions at a wet thickness of 200m. A neg. image was obtained by imagewise exposure with radiation 400 nm (U.S. Patent 4,090,087, May 23, 1978)
Eastman Kodak discovered that HABI technology was of value in bleaching out of sensitizing dyes. Patents continued to issue for over 30 years in this area. The first were U.S. Patents 4,196,002 and 4,201,590; this interest has continued. As recently as 2003, there were bleach-out patents from Eastman Kodak. Heat-bleachable composition useful in photography. Levinson, Steven Roy; Adin, Anthony (Eastman Kodak Co.) Brit. UK Pat. Appl. 2,004,380 (Cl. G03C1/76), 28 Mar 1979, US Appl. 834,587, 19 Sep 1977; 15 pp. Heat-bleachable antihalation or filter layers for photog. films are manufd. from an oxidative arylimidazolyl dimer, e.g. I dimer (R and R0 ¼ H or C1-4 alkyl), and a formazan dye in a polymeric binder. Thus, 35 mg, 2,4,5-triphenylimidazole dimer was dissolved in 1 g
247
1984–2004
R′
R′
N
N
R
2
I
THF and 2 g 20% Me2CO soln. of polysulfonamide binder and 7 mg formazan dye were added. The soln. was coated 0.15-mm-thick onto poly(ethylene terephthalate) film and dried to give a coating which was completely bleached when contacted a few s. with a metal block at 150 . Heat-sensitive reactive products of hexaarylbiimidazole and antihalation dyes. Adin, Anthony; Levinson, Steven R. (Eastman Kodak Co.) U.S. 4,201,590 (Cl. 430-617; G03C1/02), 06 May 1980, Appl. 834,587, 19 Sep 1977; 18 pp. cont.-in-pat. of U.S. Ser. No. 83. A heat bleachable compn. is described which consists of 1 hexaarylbiimidazole with 1 dye, esp. an antihalation or filter dye, that is reactive with the product of the hexaarylbiimidazole formed upon heating the material to 90 . This provides a material that becomes 40% colorless within 20 min, typically within 30 s, upon heating to the above-mentioned temp. The dye-contg. material is esp. useful for antihalation and filter purposes in a photog. element. The material can also be a heat-bleachable dye layer of a photothermog. or thermog. element. Thus, a Ag behenate-based photothermog. material was coated on the backside with an antihalation layer contg. poly-(Me methacrylate) 300.0, 1,2-bis(2,4,5-triphenylimidazole) 75.0, and 1,5-diphenyl-3-(p-nitrophenyl)formazan 7.5 mg/929 m2 of support. The resulting photothermog. material was then exposed through a line image (opaque 100 m lines 5 mm apart on a clear background) and then developed by contacting the side contg. the antihalation layer with a heated metal block for 5 s. at 130 . A developed image was produced in the photothermog. layer and the antihalation layer changed from colored to colorless. A control without the antihalation layer gave a less sharp image.
In the early 1990s, Fuji Photofilm Ltd. Introduced Copiart13 proofing materials, which also depended on HABI chemistry. The chemistry used the lower cost o-Cl-HABI and Leuco Crystal Violet along with tribromomethyl phenyl sulfone, which were microencapsulated in selected vehicles. The image stabilization was achieved by a thermal aftertreatment, which ruptured the microcapsules, which then interacted with phenidone, or other reducing agents to prevent subsequent color formation.
248
THE INVENTION OF DYLUX
I had always felt that microencapsulation offered imaging scientists an excellent route toward improved and innovative product. The work on Cycolor1 by Mead Corporation confirmed that. Although some Dylux formulations were successfully microencapsulated during the Venture period, our management was not too excited about this approach, concluding that an additional step would add cost to our products. Now, Fuji was taking it further. Some of the advantages of the Fuji offering were apparent. The intense color, a deep purple, of the oxidized LCV gave a near black image in the yellow-lighted environment, common to printing operations. As the entire HABI absorption spectrum could be utilized to initiate the imaging process, no restriction on the imaging source was required. Copiart was thus considered a major threat to a business, which had expanded rapidly, and successfully in the absence of any serious competition. The Fuji technology is described in a series of U.S. patents, which issued between 1990 and 1991. The patent abstracts describe the extent of the technology. U.S. Patent 4,929,530 An optical image forming material comprising a support (A) at least one member selected from the group consisting of oxidative-developable leuco dyes, (B) at least one member selected from the group consisting of photooxidizing agents, (C) at least one reducing agent and (D) at least one organic sulfonamide compound and/or at least one member selected from the group consisting of hydroxy compounds represented by the following formula (I) or (II) wherein R1 to R11 are as described herein, the leuco dye (A) and the photo-oxidizing agent (B) being present in microcapsules, and the reducing agent (C) and the organic sulfonamide compound and/or the hydroxy compound (D) being present outside of the microcapsules. OH R1
R5 R4
R2 R3 I R9
R10
C OH R11 II
U.S. Patent 4,942,107 An image-forming material and an image recording method for using the same are disclosed. The image-forming material comprises microcapsule and a reducing agent which is present outside of the microcapsules, wherein the microcapsules contain (a) a leuco dye capable of oxidatively
1984–2004
249
developing a color, (b) a photo-oxidizing agent, and (c) a phenolic compound. The image-forming material exhibits excellent image reproducibility, preservability before use, and image preservability. U.S. Patent 4,962,009 An image-forming material comprising: (a) microcapsules, in which at least one leuco dye capable of developing a color by oxidation and at least one photooxidizing agent (preferably a combination of a lophine dimer and an organic halogeno-compound) are enclosed together, and (b) at least one reducing agent not enclosed in the microcapsules (preferably present in the form of an emulsified dispersion), thereby achieving image formation in a completely dried condition, and ensuring excellent freshness keeping property, image reproducibility and fixability. U.S. Patent 4,985,331 Multi-color recording materials where the color forming layers are established on the same side of the support are disclosed. The multi-color recording system may be either of: (1) a two-color recording material comprising a support having provided thereon: (a) a light-sensitive heat-sensitive color forming layer comprising (i) microcapsules containing a leuco dye which is capable of forming a color by oxidation and a photo-oxidizing agent and (ii) a reducing agent; and (b) a heat-sensitive color forming layer having a color forming temperature differing from the glass transition temperature of the microcapsule walls, and which comprises an electron donating leuco dye capable of forming a color of a hue differing from that of the light-sensitive heat-sensitive layer and an electron accepting compound capable of reacting with the electron donating lueuco dye to form a color; (2) a multi-color recording material comprising a support having provided thereon: (a) a light-sensitive heat-sensitive color forming layer comprising (i) at least two types of microcapsules each containing a leuco dye capable of forming a color by oxidization and a photo-oxidizing agent and (ii) a reducing agent, wherein the polymer walls of the at least two microcapsules have different glass transition temperatures from each other, and the leuco dyes contained in the microcapsules have different hues from each other; and (b) a heat-sensitive color forming layer as described above; or (3) a multi-color recording material comprising a support having provided thereon: (a) a light-sensitive heat-sensitive color forming layer comprising (i) at least two types of microcapsules each containing a leuco dye capable of forming colors by oxidation and a photo-oxidizing agent and (ii) a reducing agent, wherein the leuco dyes contained in the microcapsules have different hues from each other, and the photo-oxidizing agents contained in the microcapsules are activated by actinic light of different wavelengths; and (b) a heat-sensitive color forming layer as described above U.S. Patent 5,035,974 A light-image forming material comprising a support having provided thereon a light-image forming layer comprising microcapsules containing an oxidative-developable leuco dye and a photo-oxidizing agent, and a reducing agent as essential ingredients, wherein the material also includes a covering layer or intermediate layer containing a film-forming high-molecular binder and/or an inorganic or organic pigment. U.S. Patent 5,0535,375 A multicolor recording material is disclosed, which comprises a support having provided thereon a coating layer comprising two or
250
THE INVENTION OF DYLUX
more kinds of microcapsules each containing at least a leuco dye capable of being colored by oxidation and a photo oxidizing agent outside said microcapsules, and a reducing agent as necessary components, with two or more leuco dyes capable of forming different colors from each other being retained in different kinds of said microcapusles respectively composed of walls of high polymers having different glass transition points.
A research effort by Drs. William Mooney and Jeffrey Patricia at the Towanda R&D Laboratory resulted in work that gave rise to experimental proofing product that matched the appearance of the Copiart product but still incorporated the advantages of conventional Dylux chemistry, for example, photodeactivation. U.S. Patent 5,744,280, in which storage-stable photoimageable leuco dye/photooxidant compositions and imaging processes utilizing these compositions are disclosed, was issued. U.S. Patent 5,744,280 Storage-stable photoimageable deutero leuco dye/photooxidation compositions with improved leuco dye Mooney, III; William Frank; Patricia, Jeffrey J. E. I. du Pont de Nemours and Company Appl. No.: 711764 April 28, 1998 / September 5, 1996 Current U.S. Class:430/270.1; 430/320; 430/325; 430/ 915 Intern’l Class: G03C 001/492; G03F 005/00 Storage-stable photoimageable leuco dye/photooxidant compositions and imaging processes utilizing these compositions are disclosed. These compositions comprise deuterated leuco compounds, e.g., deuterated aminotriarylmethanes, wherein the extent of deuteration is preferably at least 60%. These new compositions and the processes utilizing these new compositions afford improved monochrome images having enhanced film stabilities and image contrast characteristics, wherein desirable higher optical densities in imaged areas (Dimaged) and desirable lower optical densities in unimaged areas (Dunimaged, Dfixed, and D) are simultaneously both achieved in accordance with the compositions and processes of the instant invention.
It was found that Dylux 503B, the commercial product withstood the assault of Copiart and the research effort was not commercialized. There were some concerns on the part of DuPont that the Fuji offering was dominated by the Dessauer and Firmani patent U.S. Patent 4,247,618 (Photoimaging Systems with Cyclic Hydrazides), describing the advantages of phenidone as a fixing agent, which was in force until 1998, but these issues were resolved. Inevitably, during the manufacturing processes problems arise due to variation in one or more components of coating compositions. Research at the Towanda R&D Laboratory responded to such. An improvement in imaging performance was reported in a recent patent application. Elements for forming print-out images. Mooney, William Frank; Logrando, David Raymond (USA). U.S. Pat. Appl. Publ. (2002), 164,634 (Cl. 430-334,
1984–2004
251
G03F7/021) 10 Oct 2002, Appl. 521,536 8 Mar 2000; 12 pp., Cont.-in-part of U.S. Ser. No. 521,536, abandoned. (Eng.) [100486-97-3] The present invention relates to element for forming a print-out image contg. (1) a substrate, which may be cellulose, having a first surface and a second surface; (2) a dye forming compn. on the first surface of the substrate; and (3) a non-dye forming compn. on the second surface of the substrate having at least one hydrogen donor compd.
Another competitive threat was the emerging digital printing revolution. During the 1990s, CTP (computer to plate) became a reality, as a number of manufacturers produced imaging engines that utilized near-infrared radiation, directed by electronic means, to expose printing plates and bypass the need for lithographic negatives. Could we adapt Dylux technology to meet this challenge? Would paper proofs be a continued necessary in an era where colored ink-jet could deliver high-quality image on nonphotosensitive substrates? Or would soft proofs, as depicted on a CRT-monitor eliminate the need for paper proofs altogether?
D. Commercial Products During most of the 1980s and beyond, there were several commercial products: the cyan-forming single- and double-side coated Dylux 503B; the single and the neutral-shade single and double-side coated Dylux 535; and a film product, Dylux 608, that was found useful as a registration master in connection with prepress proofing applications. It was coated on a 8-mil polyester base. The double-side coated cyan-forming product was the most successful and satisfied an ever-increasing market. The paper products were supplied in a variety of sizes as well as rolls. No special effort was made by DuPont to manufacture exposure equipment, but there were many companies that offered such. These products were still very successful some 35 years after their introduction into the commercial market place. But there appeared new opportunities.
E. Watermarks Sometimes a discovery in one field can have significant impact on a totally different activity. In 1970, I met a remarkable stamp collector, Thomas M. Gravell, who wondered if Dylux proof paper could be employed to record watermarks of stamps. Watermarks are caused by differential transmissivity of paper and can frequently be observed by holding the paper up to light, or as was the practice until the 1970s, wetting the paper with benzene or carbon tetrachloride before viewing it by transmitted light. However, no facile way of recording this information existed. Gravell asked me for some Dylux 503 proof paper, borrowed
252
THE INVENTION OF DYLUX
one of the contact printers built by Harold Wilbur, and soon established himself as an authority in the field of watermark recording (T. L. Gravell, Stamps 156(4), July 24, 1971, p. 646). After he sold his stamp collection, he turned his attention to the recording of watermarks of historic documents as well as literary documents and published a series of books on the subject. His technique was to expose the watermarked paper onto Dylux 503 proof paper with visible light and then develop the color with UV light. To achieve a permanent record, he had the watermarks then photographed with conventional silver film. His work was highly regarded by a number of research libraries, including the Library of Congress and the Folger Shakespeare Library. His publications include the following: Reproducing Watermarks for Study, Restaurateur 1978, 2, 95 (with G. Miller). A Safe, Inexpensive Way to Make Prints from Glass Negatives, Restaurateur 1978, 2, 185. A Catalogue of American Watermarks 1690–1835, Garland Reference Library of the Humanities, Vol. 15, Garland Publishing, New York and London, 1979. Catalogue of Foreign Watermarks Found on Paper Used in America 1700– 1835, Garland Reference Library of the Humanities, Garland Publishing, New York and London, 1979. In the 1990s, he donated his collection of several thousand watermarks to the Virginia Technical University, where the Gravell Archives are named in his honor. He asked me to present a paper describing his work at the Jacques Cartier Conference on Digitalization of Documents in Lyon, France, in 1999 (Actes du colloque Vers une nouvelle e´rudition: nume´risation et recherche en histoire du livre, Rencontres Jacques Cartier, Lyon, De´ c. 1999).
F.
Thermal Dylux1 Proofpaper
Dr. Harry Zwicker, an electrical engineer who was involved in evaluating plates made by DuPont in England for CTP applications asked me one day in 1993 whether it would be possible to make proofs using near-infrared radiation. Zwicker had access to one of CREO’s thermal-imaging engines, and we prepared coatings using near-infrared absorbing dyes to transfer energy into a HABI/leuco dye system successfully to form color. For some time I had begun to interact with Dr. Jon V. Caspar, a very talented physical chemist, who was at Central Research Department doing exploratory organic photochemistry. Believe it or not—someone was actually trying to understand HABI chemistry with instrumentation that was not available during the earlier venture phases or before. I asked Jon to help me in the task of identifying chemistry that would permit the formation of an infrared-sensitive imaging system. Caspar and I
1984–2004
253
concluded that we needed maximum sensitivity, in terms of identifying a HABI that would cleave thermally while possessing adequate stability under ambient conditions. I had already concluded that we should start designing the system around the most sensitive triarylmethane dye in our repertory, Leuco Crystal Violet (LCV). The Fuji workers had already shown that adequately stable imaging systems could be built around o-Cl-HABI/LCV/tribromomethyl phenyl sulfone; and thus we had the skeleton of an imaging system. We suspected and Caspar showed that o-Cl-HABI was thermally too stable, and of the available HABIs, o-EtO-HABI, which had been developed at Neu Isenburg more than a decade ago, would be a good compromise between stability and ease ofthermolysis. Caspar reasoned that the inclusion of UV-absorbers would not significantly interfere with the system, and I had retained a collection of these materials from my earlier work. I favored the use of 5,50 -di-t-butyl-2,20 -dihydroxy-4,40 dimethoxybenzophenone, invented by Gunther Hoeschele of DuPont in the 1960s (G. Hoeschele, U.S. Pat. 3,105,094, Sept. 24, 1963). As this was not commercially available, we did our development work around the commercial 2,20 ,4,40 -tetrahydroxybenzophenone. We investigated alternate binders, in the hope of identifying a thermoplastic material that would give us some of the advantages of the technology proposed by Looney three decades earlier but were not especially successful. Critical in all that was the identification of the right NIR-sensitizing dye. A squarilium dye, available from earlier DuPont work gave an excellent imaging system in which there was relatively little color in the unimaged coating; but this dye, which contained sulfur, unfortunately produced an unattractive odor, which could not be removed. I suggested adding an odor-masking additive, but that concept was rejected. Research on antihalation dyes by Dr. Dietrich Fabricius, an extremely productive chemist located at DuPont’s Brevard plant gave us materials that also possessed near-infrared absorbing properties. Regrettably, midway during this program, DuPont sold the Brevard facility to Agfa-Gevaert, and we could no longer cooperate with Fabricius. Our program made good progress, interacting with a group at Towanda, which included Rick Covaleskie, Bill Mooney, and Jeffrey Patricia. We were all tuned to a common goal, and it was a pleasure to be part of this team. Thermally imageable monochrome proofing product with high contrast and fast photospeed. Dessauer, Rolf; Caspar, Jonathan V. (E. I. Du Pont De Nemours and Company, USA) U.S. US 5,858,583 (Cl. 430-17; G03CI/675),12 Jan 1999, Appl. 888,266, 3 Jul 1997; 19 pp. (Eng). Novel thermally imageable monochrome product compns., elements, and processes are disclosed. These comprns. and elements characteristically have high contrast and fast imaging speeds. The thermally imageable compns. of this invention comprise (a) at least one hexaarylbiimidazole compd., (b) at least one leuco dye, (c) at least one acid-generating compd., (d) a polymeric binder, (e) optionally at least one UV stabilizer and/or at least one
254
THE INVENTION OF DYLUX
inhibitor of color formation, and, in certain embodiments, (f) at least one near IR-absorbing dye. These compns. have the propensity for affording, upon thermal imaging, highly colored images having high optical d. values. At the same time, background color is low in preferred compns. even after extensive exposure to ambient light. These compns. can be imagewise heated to effect color formation (i.e., generation of an image) or, in case of compns. contg. at least one near IRabsorbing dye, can be imagewise exposed to near IR radiation from a laser or other device to effect color formation (i.e., generation of an image).
A DuPont patent (M. R. McKeever, U.S. Pat. 4,298,678, Nov. 3, 1981) taught that the addition of diethylhydroxyamine stabilized photoresist compositions in which the printout system o-Cl-HABI/LCV was employed. We found that the thermal Dylux system that we developed was similarly stabilized, so that little premature violet color was formed. However, the coatings underwent slow color changes, which we attributed to reaction of the halogen of the NIR dye with the amine. Several consultations with John Neumer and others resulted in our selecting dibenzylhydroxylamine vs. the aliphatic hydroxylamine, and the problem appeared solved. We identified some exudation problems with this system, which contained no plasticizer so as to maximize the amount of energy to be absorbed by the imaging components. By careful control of the stoichiometry, we were able to develop attractive greenish coatings that had adequate imaging speed and adequate ambient light stability in indoor light. The coating had to work in the Creo Trendsetter1, the machine of choice; and by selection of a stiff base sheet, we were able to produce beautiful deep purple/light green proofs, which had excellent resolution and attractive visual appearance. Still, we recognized that the system had limitations, such as speed, ambient light stability, and the lack of a white background.
Nonphotosensitive, thermally imageable element having improved room light stability. Dessauer, Rolf, Caspar, Jonathan V. (E. I. du Pont de Nenmours and Company, USA) Eur. Pat. Appl. EP 941,866 (Cl. B41M5/30), 15 Sep 1999, US Appl. 37,403, 10 Mar 1998; 28 pp. (Eng). Novel thermally imageable monochrome compns., elements, and processes are disclosed. These compns. and elements characteristically have high contrast and fast imaging speeds. The thermally imageable compns. of this invention contain at least one polymeric binder, a specified leuco dye, and a specified hydroxylamine compd. These compns. have the propensity for affording, upon thermal imaging, highly colored images having high optical d. values. At the same time, background color is low in preferred compns. even after extensive exposure to ambient light. These compns. can be imagewise heated to effect color formation or, in case of compns. contg. at least one near IR-absorbing dye, can be imagewise exposed to near IR radiation from a laser or other device to effect color formation. (U.S. 6,251,571, June 26, 2001).
1984–2004
255
Our marketing group agreed that the system, as conceived, was ready for market testing, and sufficient amounts were manufactured for this purpose. Initital response was favorable. Thermal Dylux 800 was introduced to the market at Graph Expo in 1996, with a demonstration on the floor of McCormack Hall in Chicago. We had gained valuable experience in developing thermal imaging systems and interacting with Creo, the predominant manufacturer of NIR-exposure devices for the graphic arts industry. By 1998, it became evident that a thermal four-color proofing system was required, i.e., a system in which a four-color proof could be formed using thermal imaging engines, as for example the Creo Trendsetter. G. B. Blanchet proposed a thermal-imaging system, which required transfer of colorants from four donor sheets to a single receptor sheet, with subsequent transfer to a paper support. This development consumed large technical and marketing efforts, and significantly reduced manpower for work on improving the performance and sales of Thermal Dylux proofpaper.
G. HABIs in Photopolymer Products A major effort was undertaken in the late 1960s and early 1970s at Parlin under the leadership of Dr. A. B. Cohen to produce a series of photopolymer products. The office copy system, described earlier, failed to reach commercialization because electrophotography offered a better approach to this market. But there were three major areas that could be and were successfully entered: color proofing, photoresists, and lithographic plates. The photopolymer office copy system, however, had shown that HABIs could be effective photoinitiators in polymeric systems and provided stable, almost colorless and readily modifyable components. Exploratory work by G. R. Nacci, J. E. Pazos, and others showed just what possible ways there were to utilize effectively these then novel photoinitiators. This work inevitably focused on the use of o-Cl-HABI, which was the least expensive hexaarylbiimidazole that possessed the requisite stability and absorption spectrum. By the time that the photopolymer applications became commercially important, the Dylux Venture had been terminated, and there was essentially little interest in synthetic work to produce still more effective photoinitiators of this family. Thus, in some respects, the use of the research on photopolymer products became essentially an investigation of changes in properties resulting from exposure, and this could be categorized as follows: Exposed areas were less tacky than unexposed areas, which would give rise to areas that could be selectively toned.
256
THE INVENTION OF DYLUX
Exposed areas were less conductive than unexposed areas, which would give rise to electrophotographic masters in a xeroprinting process. Exposed areas had different adhesion relative to unexposed areas, which would give rise to peel-apart imaging systems, in which a colored sandwich structure would be separated to give a positive and a negative image. Exposed areas were less soluble than unexposed areas, which would give rise to photoresists and lithographic plates, where the unexposed areas could be washed off. In these areas, addition of leucodyes allowed the user to see a printout image, which established the level of exposure, and was helpful when multiple exposures on the photosensitive surface were made. Common to these photopolymer products was the use of coinitiators, such as 2-MBO, diacrylate and triacrylate monomers, binders, and other additives as required. In general, the film products were coated on a polyester support and a polyolefin sheet laminated to the coated material, to permit exposure of the films in an oxygen-free environment, as was required to obviate the chain-stopping effect of peroxide formation. Almost all of these systems employed o-Cl-HABI as photoinitiators. No serious effort was ever made to address the issue of what could be done to make HABIs even better. Could it be that the o-Cl-HABI was the perfect photoinitiator system? I doubt it. It was simply a matter of expediency—this photoinitator was available and finding a better one would be difficult. I found that the philosophy behind making the TCM-HABI—Could we use less costly ingredients, shift spectra, increase stability, etc.—required an approach toward chemistry that depended on synthesis to solve problems. Most industrial photochemists are better at formulating around deficiencies, and so it was at DuPont. Once o-Cl-HABI was available, the formulators worked around it. Only when a special problem arose, such as the need for a less colored photoinitiator, was an effort made to develop new compounds; After all, one cannot readily add a compound that would make a yellowish compound white! Hence, the work that resulted in the 2-(alkyloxy)-4,5-diphenylimidazole dimers. DuPont developed a series of very successful photopolymer products in which HABI played an important role. HABI-based photopolymer products, as well as other HABI-chemistry, not strictly related to color-printout formation will be treated in a subsequent chapter. For the record, the applications in which HABI-chemistry played a dominant role include proofing, in which surprints were generated with Cromalin1; peel-apart colored sheets with Cromacheck1; printing plates with Lydel1; photoresists with Riston, and holography with Omnidex1 films. A number of applications that were not commercialized included Xeroprinting with EMP Masters and nonsilver lithographic films.
1984–2004
H.
257
The HABI Literature
It has been an interesting experience to keep track of the HABI literature. Initially this involved collecting the Chemical Abstract references as these appeared; later search engines such as STN have been helpful, though these unfortunately sometimes make identification of the chemical structures difficult. The U.S. Patent Office provides the user with a relatively easy way of locating HABI references, but what is easy to do for patents issued after 1976 is very difficult for patents before that year. I have tracked all this from the beginning; but, of course, depend on the abstractors as well as the patent office for accuracy and reliability. Still, it has been an interesting experience to determine the extent to which HABIs have become widely accepted as photoinitiators: as of Dec. 28, 2004, there were over 950 Chemical Abstracts references and more than 1050 U.S. patents that include the terms HABI, hexaarylbiimidazoles, and other identifiers. Curiously, Chemical Abstracts in earlier years did not include all issued U.S. patents; thus I have never been able to locate the important MacLachlan patent U.S. Patent 3,390,996, dealing with photodeactivation, in Chemical Abstracts (A. MacLachlan, U.S. Patent 3,390,996, July 2, 1968). Dr. Vadim Krongauz, an extremely talented chemist, who worked at DuPont from 1987 to 1992 undertook many studies in which he investigated the role of HABIs in various photopolymer systems. In 1995, his book Processes in Photoreactive Polymers (Processes in Photoreactive Polymers, V. V. Krongauz and A. D. Trifunac, Eds., Chapman & Hall, New York) was published. Regrettably, some of the information relating to Dylux was not entirely correct, but this book includes an informative article by Koko Maeda, in which he described HABIs as ‘‘the choice initiators for photopolymers.’’
I. Could We Have Done Better? Having the benefits of considerable hindsight, one can now predict how we could have done better! As scientists, we had a fine group of dedicated individuals. At the management level, we had hardworking and intelligent people. However, we had almost no one at a management level who had a grounding in the technology that we wanted to commercialize. The company’s structure was more designed to encourage people to think of their careers as remote from what they were working on. Had their future really and directly been connected with the research and marketing efforts, our effort might have been far more successful. The venture manager once pointed out to us that our task was to determine if there is a business for this technology and not necessarily to make a business out of it. Certainly the compensation for almost all my colleagues
258
THE INVENTION OF DYLUX
was unrelated to the success of what we were doing, and those who were to advance in the company inevitably did it by being promoted to other business areas. Had George Coraor remained as research manager, we could well have invented other systems for other applications. Had the work at Radiation Physics Lab continued to elucidate the photopolymerization mechanism in such a way that we would have developed a stronger patent position and dominated the field more effectively. We had technology that fit into many markets. The success in the proofing and photoresist field was significant, but utilization of our technology to achieve a similarly dominant position in other areas never really happened. One of the problems of success is that managements, which are driven by earning, will want to commercialize a product and then maximize the marketing effort in the first direction that looks good. In the case of Dylux proof paper, we had a relatively hard time finding someone to recognize the prepress proofing opportunity. Once this was shown to be a viable business, it was difficult to find anyone to expand into other areas, because inevitably, it may be easier to achieve better market penetration than to pioneer yet another field. Those who really understood the technology knew that making products, which were required for other end-uses, required but modest changes in the stoichiometry, substrate, or additives. However, to management, a new product area brings about a new set of head-aches, and worse, reduces earnings. Structurally, when the chemistry fell under the mantle of the proofing group, they had little incentive to use their resources to expand into another, unrelated business area. Our abdication of the bar-code label business, for example, was a failure to see an emerging opportunity. Going beyond that, one might have considered UV-curable coatings, where the unique properties of HABIs would have been important. We never considered applying some of this chemistry to agricultural products, pharmaceuticals, etc. Scientists must be optimists! After some 40 years of thinking about HABIs, it is hard not to imagine that there must be better HABIs than the almost universally used o-Cl-HABI. Instrumentation has advanced immeasurably since we first oxidized leucodyes in 1960. It was disappointing to me personally that after the initial successes with HABIs in the 1960s, no synthetic effort to develop better materials was made for nearly 40 years; it was never an ongoing effort, and the invention of TCTM-HABI, was almost an accident. Except for an all too brief interlude in the late 1990s, no effort to develop new HABIs was undertaken. Other than for my personal interest, no one ever really followed the HABI patent estate, defended it against competitors who encroached on it, and—at the management level—really understood it. Yet, there are publications from various sources, as for example those of J. V. Caspar et al., that describe investigations into the chemistry of these interesting materials.
1984–2004
259
ESR Study Of Lophyl Free Radicals In Dry Films. Caspar, Jonathan V.; Khudyakov, Igor V.; Turro, Nicholas J.; Weed, Gregory C. Central Research and Development Department, E.I. DuPont de Nemours and Company Macromolecules (1995), 28(2), 636–41 Photolysis of 2-(20 -chlorophenyl)-20 -(20 -chlorophenyl)-4,40 5,50 - tetraphenylbis-imidazole (D) into corresponding lophyl free radicals (R ) was studied by steady-state and time-resolved (TR) CW ESR in dry films of PMMA at 6-400 K and in a frozen liq. soln. at 77 K. Steady-state ESR expts. demonstrate the formation of spatially proximate electronically interacting radical pairs (RPs) R . . .R as well as the formation of spatially isolated free radicals R . The ratio of amts. of photogenerated R to R . . .R depends upon the temp. and the pro-perties of the film. This ratio increases with increasing temp., and at the same temp. the ratio is larger in films contg. a plasticizer (triacetin). At low temps. (6200 K) RPs are obsd. and are characterized by zero-field splitting (ZFS) values of 2D 250 G (corresponding to intraradical distance of r 6. ANG.) and E 10 G. Films which do not contain a plasticizer demonstrate ESR spectra of RPs with ˚ ) at room, and even at elevated, temps. (up to 400 K) as well 2D 100 G (r 8 A as ESR spectra of RPs of two types (r 6 and 8. ANG.) at re-duced temps. Thus, the measurement of ZFS parameters of RPs produced by the photolysis of D in films can be used to characterize photopolymer film. TR ESR spectra of the triplet state of D were obsd. in films at 53 K. Simulation of the spectra leads to ests. of 2D 2200 G and E 165 G. Decay kinetics of the lophyl radicals in films was satisfactorily described by polychromatic second-order kinetics with a rectangular distribution of reactivity. The estd. max. difference in activation energies for a film without a plasticizer is ca. 4 kcal/mol, suggesting a range of recombination rate consts. of ca. 103. ESR study of lophyl radicals in dry films. Khudyakov, Igor V.; Turro, Nicholas J.; Weed, Gregory C.; Caspar, Jonathan V. (Department Chemistry, Columbia University, New York, NY USA). ICPS 94 Phys. Chem. Imaging Syst., IS&T’s 7th Annu. Conf. 1994, 2, 682–6 (Eng). IS&T - The Society for Imaging Science and Technology: Springfield, Va. 2-(20 -Chlorophenyl-20 -(20 -chlorophenyl)-4,40 ,5,50 -tetraphenylbisimidazoles (o-Cl-HABI and other hexaarylbisimidazoles are commonly termed HABls) have been used as free radical initiators in photopolymer imaging for more than 2 decades. In the present work we investigated the photochem. of a representative HABI, o-Cl-HABI. HABI undergo thermolysis, direct and sensitized photolysis in both the liq. and the solid phase with the formation of corresponding lophyl radicals. ESR was used to study formation and decay kinetics of these radicals in photoexposed E’MMA films contg. triacetin, TMPTA, N-phenylglycine, 2-mercaptobenzoxazole Leuco Crystal Violet and o-Cl-HABI.
Will a synthetic effort be made to take advantage of these findings?
J.
HABI Statistics
One of the reasons for the commercial success of this technology was the patent coverage that we enjoyed. Even though, ultimately, the greater reward came
260
THE INVENTION OF DYLUX
from the utilization of HABIs as photoinitiators, the patent estate was primarily based on the early Dylux work. It is of interest to examine this area. Forty-three U.S. patents were issued to DuPont covering HABIs and leucodye oxidation systems, of which 38 issued between 1968 and 1986. Overall, DuPont obtained over 400 U.S. patents for HABI chemistry; the next most active company was Fuji Photo Film Ltd., which received 194 patents; and Eastman Kodak received 50 patents. As of December 31, 2004, there were 1070 U.S. patents that referred to HABI chemistry. The field is still very active, with 67 issues in 2003 and 89 in 2004.
ACKNOWLEDGMENTS Over 200 DuPont scientists, managers, marketers, technicians contributed to the success of the Dylux program. I have tried to name as many as appeared appropriate in the framework of this chapter. It should be pointed out that invention sometimes is the easy process, and the efforts behind commercialization and manufacturing are difficult and overlooked. With that in mind, I would especially like to acknowledge my colleagues at the Towanda Plant who contributed so much to keep the coaters rolling successfully: they were the late Thomas D. Sheets as well as David R. Logrando, Dr. William F. Mooney II, and Dr. Jeffrey J. Patricia—and their associates.
REWARDS AND AWARDS DuPont had a bonus system that rewarded scientists who had made significant contributions, and it was based on a formula that related the earnings from the invention over a period of time. Cescon, MacLachlan and I each received a modest bonus award in 1974, some 5 years after Dylux was commercialized. It was not particularly large, and considering the amount of dedication and effort that I had given to the program, I considered it quite small. Part of the problem of course was that Dylux proof paper growth depended on the availability of exposure equipment, and as time went on, more equipment was in the field, and for many years, proof paper sales increased every year, to the point where it reached a substantial level sometime in the early 1980s. It was an excellent money earner, or cash cow, as these are now called—no major investment, no significant competition, the ingredients were manufactured and coated elsewhere. For many years it was the most profitable product in the Photo Products line! In 1986 I raised the awards issue with my director, T. D. Smith, who allowed that I should have gotten a significant award, for after all, not only Dylux but all
REWARDS AND AWARDS
261
the HABI-based photopolymer products came out of this work. He and V. C. Chambers presented my case to the upper management, and in time I was given a rather substantial award. Smith had asked me how much I should get and I suggested that 0.1% of the cumulative net earnings of the HABI products would seem a nice formula, but he thought that no one had ever gotten bonuses that reached seven figures, and I did not, either. However, this time, the bonus involved a very respectable sum. In 2000, I-Technologies, which was the new name for Photo Products Department, celebrated its 50th anniversary, and I received a plaque ‘‘In Recognition of Discovery Efforts in HABI Chemistry’’ and in 2001 I was the second DuPont scientist to receive the Plambeck Award, ‘‘In Recognition of Contributions to Photopolymer Technology.’’ Also in 2001, I was the 28th DuPont scientist to receive the Pedersen Award, ‘‘For Contributions to Science and Efforts to Establish the Commercial Viability of Hexaarylbiimidazole Photochemistry, Which Led to the Creation of Several Profitable Du Pont Products.’’ I was asked to give a lecture reviewing the work, which lead to the Pedersen Award, and did so to several groups. I had always hoped that some time there would be a study of what we did right, what we might have done better, and where we went wrong, and that one could learn more about this program than its successful chemistry. This chapter is an attempt to leave a permanent record behind.
INDEX
Ab initio calculations: hydrogen-atom translocation, 19–22 hypericin ground state heterogeneity, 9–10 thiophosgene electronic states: carbon-sulfur double bonds, 32–34 first triplet electronic state (T1), 52–56 second singlet excited state (S2), 57–64 Add-on toning, Dylux technology and, 237–238 Adduct formation, Paterno`-Bu¨chi reaction, furan derivatives, 88–92, 98–103 Agfa-Gevaert, competition to Dylux from, 245–246 Aldehydes, Paterno`-Bu¨chi reaction: asymmetric furan reactions, 111–116 2,3-dihydrofuran, 93–95 furan compounds, 84–92 Allylic alcohols, Paterno`-Bu¨chi reaction, asymmetric furan reactions, 114–116 Analytical Hessian calculations, hypericin ground state heterogeneity, 9–10 Antibonding electron density, thiophosgene molecules, first excited singlet state (S1 (n,p*)), 45–47 Aromaticity, Paterno`-Bu¨chi reaction, non-furan compounds, 122–123
Asteltoxin formation, Paterno`-Bu¨chi reaction, furan derivatives, 98–103 Asymmetric reactions, Paterno`-Bu¨chi reaction, furan derivatives, 103–116 Atomic coefficients, Paterno`-Bu¨chi reaction, furan compounds, 89–92 Automatic clinical analyzer, Dylux technology and, 215–217 Avenaciolide, Paterno`-Bu¨chi reaction, furan derivatives, 100–103 Azaindole, Paterno`-Bu¨chi reaction, 122 Balon, Walter, 149–150 Band contour analysis, thiophosgene molecules: first triplet electronic state (T1), 53–56 second singlet excited state (S2), 59–64 BBO crystals, hypericin ground state heterogeneity analysis, 12–15 Benzaldehyde, Paterno`-Bu¨chi reaction: asymmetric furan reactions, 112–116 2,3-dihydrofuran, 93–95 furan compounds, 88–92 Benzene, Paterno`-Bu¨chi reaction: 2,3-dihydrofuran, 93–95 furan derivatives, 87–92 Benzofuran, Paterno`-Bu¨chi reaction, 92–93 Benzophenone, thiophene, Paterno`-Bu¨chi reaction, 117
Advances in Photochemistry, Volume 28 Edited by Douglas C. Neckers, William S. Jenks, and Thomas Wolff # 2005 John Wiley & Sons, Inc.
263
264
INDEX
Biimidazoles: HABI chemistry and synthesis of, 139–141 permanent color and, 145–147 TCTM-HABI chemistry and, 223–224 ultraviolet light and, 160–161 Biradical intermediates, Paterno`-Bu¨ chi reaction: asymmetric reactions, 107–116 2,3-dihydrofuran, 93–95 furan compounds, 85–92 Black Dylux 535 product, development and patenting of, 234–237 Bleistine, Cy (Mrs.), 139 Blue-line technology, Dylux papers and, 196–197 Boeing project, Dylux technology and, 183–184 Booth, Bruce (Dr.), 216–217 Born-Oppenheimer states, thiophosgene molecules, first excited singlet state (S1 (n,p*)), 64–72 Botsolas, Philip, 212–213 Brannock gauge, development of, 166–167 Breslow, Ron, 144 Business planning, Dylux development and, 184–185 Butacite1, development of, 135, 143 Carbon-sulfur double bonds, thiophosgene molecules, 28–29 electronic states, 32–34 first triplet electronic state (T1), 50–56 second singlet excited state (S2), 56–64 Carbonyl compounds: 1,2-cycloaddition reaction: Paterno`-Bu¨ chi reaction: furan derivatives, 83–116 asymmetric reactions, 103–116 benzofuran, 92–93 2,3-dihydrofuran, 93–95 furan, 83–92 synthetic applications, 96–103 non-furan pentaatomic heterocycles, 116–123 azaindole, 122 imidazole, 119–120 indole, 120–121 pyrrole, 117–118 reactivity observations, 122–123 selenophene, 118–119 thiazole, isoxazole, and isothiazole, 120
thiophene, 116–117 research background, 82–83 hypericin interaction with, 17–19 thiophosgene electronic states, second singlet excited state, 75–76 Cathode ray tube (CRT) imaging: Boeing project involving, 183–184 HABI chemistry and, 176–177 Cavity ring-down (CRD) experiments, thiophosgene molecules, first triplet electronic state (T1), 49–56 CDM-HABI compound: development of, 165–166 TCTM-HABI development and, 223–227 Cescon, L. A., 137–139, 145–146, 150, 155, 168–169, 173, 260 Chilton, Tim, 151–152 China, Dylux technology marketing in, 239–240 Chiral phenylglyoxylates, Paterno`-Bu¨ chi reaction, furan derivatives, 104–116 Chloranil (tetrachlorobenzophenone), spirit duplication systems and, 147–148 o-Cl-HABI: chemistry of, 223–224 Photomarker1 Corp. technology and, 228 photopolymerization and, 255–256 Thermal Dylux1 proofpaper and, 253–255 Chlorine isotope effects, thiophosgene molecules, 29 Cis isomers, Paterno`-Bu¨ chi reaction, asymmetric furan reactions, 113–116 Cohen, Robert (Dr.), 160, 165, 168–169, 173 Color-overlay film, Dylux technology for, 218–221 Commercial product development, Dylux technology and, 251 Competition, for Dylux technology, 245–251 Computer simulation, Dylux technology and, 207–209 Computer to plate (CTP) technology: development of, 251 Thermal Dylux1 proofpaper, 252–255 Condon approximation, thiophosgene molecules, first excited singlet state (S1 (n,p*)), 64–72 Continuum generation, hypericin ground state heterogeneity analysis, 11–15 Copiart product, patents for, 248–250 Coraor, George R., 138–145, 150, 155, 157–160, 172, 175, 202–203
INDEX
Corfam1, 166–167, 203 Cronapaque1, photo-oxidation and, 153 1,2-Cycloaddition reaction, carbonyl and pentaatomic heterocyclic compounds: Paterno`-Bu¨ chi reaction: furan derivatives, 83–116 asymmetric reactions, 103–116 benzofuran, 92–93 2,3-dihydrofuran, 93–95 furan, 83–92 synthetic applications, 96–103 non-furan pentaatomic heterocycles, 116–123 azaindole, 122 imidazole, 119–120 indole, 120–121 pyrrole, 117–118 reactivity observations, 122–123 selenophene, 118–119 thiazole, isoxazole, and isothiazole, 120 thiophene, 116–117 research background, 82–83 Dark state analysis, thiophosgene molecules, second singlet excited state (S2), 73–76 Dayton, Herb, 139 Deactivation speed, imaging technology and, 168–173 de Campi, John, 187–188 Decay curve lifetime measurements, thiophosgene molecules: first excited singlet state (S1 (n,p*)), 68–72 second singlet excited state (S2), 73–76 Deuterium isotope effect: absence of, in excited-state intramolecular H-atom transfer, hypericin and hypocrellin, 6–7, 17–19 hydrogen-atom translocation and, 19–22 Deutsch, Albert, 155 DFF filter technology: Dylux applications for, 213–214 portable copiers and, 217–218 Diastereoisomeric excess, Paterno`-Bu¨ chi reaction, asymmetric furan reactions, 104–116 Dichroic beam splitter techniques, hypericin ground state heterogeneity analysis, 12–15 2,3-Dihydrofuran, Paterno`-Bu¨ chi reaction, 93–95
265
Dihydropyridines (DHP), add-on toning technology and, 237–238 Dimethylsulfoxide (DMSO), hypericin ground state heterogeneity analysis, 13–15 Dipole moment operator, thiophosgene molecules, first excited singlet state (S1 (n,p*)), 36–47 Double minimum potential, thiophosgene molecules, first excited singlet state (S1 (n,p*)), 42–47 Dual response imaging systems: patents for, 192 updating of, 233 DuPont Company: Apex programs at, 134 development of Dylux1 system at, 133–135 Orchem division at, 134–135 Photo Products Department at, 134–135, 202–204, 233–234 Radiation Physics Laboratory, 157–163 venture concept at, 134, 177–178 Dylux 503: limitations of, 200–201 watermark applications, 251–252 Dylux 503B, Copiart technology and, 250 Dylux1-4C overlay films, development of, 218–221 Dylux1 instant-access imaging materials: add-on toning system, 237–238 Black Dylux1 535 proofpaper, 234–237 Botsolas’ contributions to, 212–213 cathode ray tube imaging, 176–177 Cescon’s work on, 137 Chinese operations, 240 commercial product development, 251 competition against, 245–251 computer simulation techniques, 207–209 DFF filter development, 213–214 dual response systems, 233 DuPont research on, 133–135 European operations, 197–198, 215 foot imaging technology, 166–167 GTE-Sylvania home office copier, 217–218 higher speed systems, 168–173 polymerization initiation, 171–172 historical assessment of, 257–259 historical evolution of, 135–137 J.C. Penney inventory control project, 229–230 Kalography development, 186–187
266
INDEX
Dylux1 instant-access imaging materials (Continued) large wall-screen displays, 183–184 limitations of, 200–201 literature sources on, 141–143, 257 management changes and, 172–173 marketing efforts, 178–189 business planning, 184–185 de Campi’s contributions, 187–188 Mattell’s involvement with, 188–189 point-of-sales opportunities, 179 Wilbur’s contributions, 178–179 medical imaging systems, 215–217 microfilm market for, 173–176 military applications, 199–200 operations restructuring, 202–204 optical printing, 167–168 opto-magnetic printing, 185–186 origin of name, 181–182 overlay film development, 218–221 patent history and documentation, 138–145 patent statistics, 259–260 permanent color development, 145–165 applications, 151–152 Looney’s research on, 155–157 optimization experiments, 164 patents for, 164–165 radiation physics laboratory research, 157–163 stabilization problems, 152–155 phenidone technology, 222–223 photodecoration applications, 204–207, 218 Photoimaging, Ltd. formed, 244–245 Photomarker1 Corp., 227–228 photopolymerization patent, 238 photopolymer products, 255–256 Photo Products difficulties, 228–229 phototropic materials, 137–138 proofing technology, 168, 190–191 proofpaper properties, 193–197 research background, 132–133 rewards and awards for, 260–261 SPSE meeting introduction of, 191–192 SPSE Tokyo meeting, 230–232 TCTM-HABE process, 223–227 technological improvements in, 165–168 thermal Dylux1 proofpaper, 252–255 transfer to Photo Products, 233–234 Universal Product Code operations, 209–212 UVI movie production, 179–181
venture capital for, 177–178 W. H. Brady Co. and, 186–187 Wartell’s contributions to, 189–198 watermarks, 251–252 xeroprinting technology, 241–244 X-Raylux invention, 232–233 Eastman Kodak, competition to Dylux from, 246–247 Electron configurations, thiophosgene electronic states, 29–34 Electronic integral, thiophosgene molecules, first excited singlet state (S1 (n,p*)), 65–72 Electronic matrix elements, thiophosgene molecules: first excited singlet state (S1 (n,p*)), 65–72 second singlet excited state (S2), 72–76 Electronic states, thiophosgene molecules, 29–64 basic properties, 28–29 first excited singlet state (S1 (n,p*), 36–47 first triplet electronic state (T1), 47–56 ground electronic state (S0), 34–36 molecular orbitals and electron configurations, 29–34 second singlet excited state (S2), 56–64 Electro-optical printing, UPC-Dylux technology for, 211–212 Electrophotography, xeroprinting and, 241–244 Ellefson, Bob, 203–204 Endo selectivity, Paterno`-Bu¨ chi reaction, 2,3dihydrofuran, 94–95 Energy level diagram: thiophosgene electronic states: first excited singlet state (S1 (n,p*)), 40–47 molecular structure, 33–34 second singlet excited state (S2), 73–76 Equilibrium geometry, thiophosgene molecules, variation in, 77–78 Esters, Paterno`-Bu¨ chi reaction: asymmetric reactions, 107–116 furan compounds, 88–92 o-EtO-HABI, add-on toning technology and, 237–238 European expansion, of Dylux technology, 197–198, 215 Excited-state intramolecular H-atom transfer: hydrogen-atom translocation, 19–22 hypericin-like perylene quinones, 2–7
INDEX
Excited-state kinetics, multiple hydrogen-atom transfers in perylene quinones, 15–20 Exo stereoselectivity, Paterno`-Bu¨ chi reaction: 2,3-dihydrofuran, 95 furan derivatives, 86–92 Exposure equipment, for Dylux development, 178–179 Fabricius, Dietrich, 253 Fermi correlation, thiophosgene molecules, first triplet electronic state (T1), 50–56 Firmani, Ray, 173 First excited singlet state (S1 (n,p*)), thiophosgene molecules: basic properties, 36–47 Herzberg notation, 31–34 photophysical properties, 64–72 First triplet electronic state (T1), thiophosgene molecules, 47–56 Flexible filter, Dylux technology and, 213–214 Fluorescence breakoff, thiophosgene electronic states, first excited singlet state (S1 (n,p*)), 70–72 Fluorescence depletion (FD), thiophosgene molecules, second singlet excited state (S2), 74–76 Fluorescence spectra: excited-state intramolecular H-atom transfer, hypericin and hypocrellin, 2–7 multiple hydrogen-atom transfers in perylene quinones, 17–19 thiophosgene electronic states: first excited singlet state, 66–72 first triplet electronic state (T1), 54–56 ground electronic state, 36–37 thiophosgene molecules: first excited singlet state (S1 (n,p*)), 45–47 second singlet excited state (S2), 57–64 Foot imaging, HABI chemistry and, 166–167 Fourier transform analysis: thiophosgene electronic states, first triplet electronic state (T1) lifetimes, 55–56 thiophosgene molecules, first excited singlet state (S1 (n,p*)), 67–72 Franck-Condon factors, thiophosgene molecules: basic properties, 29 first excited singlet state (S1 (n,p*)), 38–47, 65–72 first triplet electronic state (T1), 52–56
267
Frequency calculations, hydrogen-atom translocation, 19–22 Friar, Larry, 168 Fuji Photofilm, competition to Dylux from, 247–250 Furan derivatives, Paterno`-Bu¨ chi reaction, 83–116 asymmetric reactions, 103–116 benzofuran, 92–93 2,3-dihydrofuran, 93–95 furan, 83–92 synthetic applications, 96–103 Furfural derivatives, Paterno`-Bu¨ chi reaction, furan compounds, 88–92 2-Furylmethanol, Paterno`-Bu¨ chi reaction, furan compounds, 91–92 Gardner-Kasha supposition, thiophosgene electronic states, first excited singlet state, 66–72 Gauche interactions, Paterno`-Bu¨ chi reaction, 2,3-dihydrofuran, 94–95 Gaussian terms, thiophosgene molecules, first excited singlet state (S1 (n,p*)), 44–47 Glyoxylates, Paterno`-Bu¨ chi reaction, furan derivatives, 103–116 Greenewalt, Crawford, 167 Griesbeck rule, Paterno`-Bu¨ chi reaction, asymmetric furan reactions, 114–116 Ground electronic state (S0): thiophosgene electronic states, first triplet electronic state (T1) lifetimes, 55–56 thiophosgene molecules, 34–36 Ground-state heterogeneity, hypericin-like perylene quinones, 7–15 Gruebele Hamiltonian, thiophosgene electronic states, second singlet excited state, 75–76 GTE-Sylvania home office copier, Dylux technology and, 217–218 Half-wave plate techniques, hypericin ground state heterogeneity analysis, 11–15 Hammett free energy correlation, Paterno`-Bu¨ chi reaction, asymmetric furan reactions, 115–116 Hammond, George (Dr.), 144 Hanson, Victor F., 157–158 Haven, Alfred C., 150–151 Herzberg notation, thiophosgene electronic states, 30–34
268
INDEX
Herzberg-Teller coupling, thiophosgene molecules, first excited singlet state (S1 (n,p*)), 38–47 Hessian matrices: hydrogen-atom translocation, 19–22 hypericin ground state heterogeneity, 9–10 Hexaarylbimidazole (HABI) chemistry: Dylux instant-access imaging using: add-on toning system, 237–238 Black Dylux1 535 proofpaper, 234–237 Botsolas’ contributions to, 212–213 cathode ray tube imaging, 176–177 Cescon’s work on, 137 Chinese operations, 240 commercial product development, 251 competition against, 245–251 computer simulation techniques, 207–209 DFF filter development, 213–214 dual response systems, 233 DuPont research on, 133–135 European operations, 197–198, 215 foot imaging technology, 166–167 GTE-Sylvania home office copier, 217–218 historical assessment of, 257–259 historical evolution of, 135–137 imaging speed technology, 168–173 polymerization initiation, 171–172 J.C. Penney inventory control project, 229–230 Kalography development, 186–187 large wall-screen displays, 183–184 limitations of, 200–201 literature sources on, 141–143, 257 management changes and, 172–173 marketing efforts, 178–189 business planning, 184–185 de Campi’s contributions, 187–188 Mattell’s involvement with, 188–189 point-of-sales opportunities, 179 Wilbur’s contributions, 178–179 medical imaging systems, 215–217 microfilm market for, 173–176 military applications, 199–200 operations restructuring, 202–204 optical printing, 167–168 opto-magnetic printing, 185–186 origin of name, 181–182 overlay film development, 218–221 patent history and documentation, 138–145 patent statistics, 259–260
permanent color development, 145–165 applications, 151–152 Looney’s research on, 155–157 optimization experiments, 164 patents for, 164–165 radiation physics laboratory research, 157–163 stabilization problems, 152–155 phenidone technology, 222–223 photodecoration applications, 204–207, 218 Photoimaging, Ltd. formed, 244–245 Photomarker1 Corp., 227–228 photopolymerization patent, 238 photopolymer products, 255–256 Photo Products difficulties, 228–229 phototropic materials, 137–138 proofing technology, 168, 190–191 proofpaper properties, 193–197 research background, 132–133 rewards and awards for, 260–261 SPSE meeting introduction of, 191–192 SPSE Tokyo meeting, 230–232 TCTM-HABE process, 223–227 technological improvements in, 165–168 thermal Dylux1 proofpaper, 252–255 transfer to Photo Products, 233–234 Universal Product Code operations, 209–212 UVI movie production, 179–181 venture capital for, 177–178 W. H. Brady Co. and, 186–187 Wartell’s contributions to, 189–198 watermarks, 251–252 xeroprinting technology, 241–244 X-Raylux invention, 232–233 fatigue properties of, 144 oxidation systems patent, 148–149 spectral properties, 153–154 Highest occupied molecular orbital (HOMO), Paterno`-Bu¨ chi reaction: 2,3-dihydrofuran, 95 furan compounds, 89–92 Human serum albumin, hypericin interaction with, 17–19 Hydrogen atom transfer rate: excited-state intramolecular H-atom transfer, hypericin and hypocrellin, 5–7 excited-state intramolecular transfer in hypericin, 2–7
INDEX
hypericin ground state heterogeneity, 7–15 multiple transfers, perylene quinones, 15–19 Hydrogen atom translocation, vibrational calculations, 19–22 Hydrogen bonding, Paterno`-Bu¨ chi reaction, asymmetric furan reactions, 113–116 Hydroquinone derivatives, HABI chemistry and, 153–154 5-Hydroxytropolone, hydrogen-atom transfer in, 7–8 Hyperfine coupling (HFC), Paterno`-Bu¨ chi reaction, furan derivatives, 86–92 Hypericin-like perylene quinones: excited-state intramolecular H-atom transfer, 2–7 ground-state heterogeneity, 7–15 hydrogen atom translocation, 19–22 multiple hydrogen-atom transfers, 15–19 Hypocrellins: excited-state intramolecular H-atom transfer in, 2–7 ground-state heterogeneity in, 7, 9 hydrogen-atom translocation, 19–22 multiple hydrogen-atom transfers in, 15–19 Hypomycin B, multiple hydrogen-atom transfers in, 15–19 Image-forming materials, patents for, 248–250 Imaging speed technology, HABI chemistry and, 168–173 polymerization initiation, 171–172 Imidazole, Paterno`-Bu¨ chi reaction, 119–120 Imidazolyl radical: HABI chemistry, 144–145 thermally activated imaging, 156 Indole, Paterno`-Bu¨ chi reaction, 120–121 Infrared (IR) spectroscopy: hypericin-like quinones, hydrogen atom translocation, 20–22 thiophosgene electronic states, ground electronic states, 36 In-plane totally asymmetric mode, thiophosgene molecules, first excited singlet state, 65– 72 In-plane totally symmetric modes, thiophosgene molecules, first excited singlet state, 65– 72 Internal conversion (IC) process, thiophosgene molecules: first excited singlet state (S1 (n,p*)), 65–72
269
second singlet excited state (S2), 72–76 symmetry absence, 77 Intersystem crossing (ISC), thiophosgene molecules: first excited singlet state (S1 (n,p*)), 71–72 first triplet electronic state (T1), 51–56 Intramolecular excited-state intramolecular Hatom transfer, in hypericin, 2–7 Intramolecular vibrational energy redistribution (IVR), thiophosgene electronic states: first excited singlet state (S1 (n,p*)), 70–72 first triplet electronic state (T1) lifetimes, 55–56 ground electronic state, 36 Inventory control, Dylux technology and, 229–230 Inversion doubling splittings, thiophosgene molecules, 29 first excited singlet state (S1 (n,p*)), 44–47 first triplet electronic state (T1), 50–56 second singlet excited state (S2), 59–64 Isatine derivatives, Paterno`-Bu¨ chi reaction, asymmetric furan reactions, 110–116 Isothiazole, Paterno`-Bu¨ chi reaction, 120 Isoxazole, Paterno`-Bu¨ chi reaction, 120 James, Dan (Dr.), 163 Japan, Dylux research presented in, 230–232 J.C. Penney, Dylux technology applications at, 229–230 Johnson, Donald R., 215–217 Kalography, Dylux solutions and, 187 Kellogg, Reid (Dr.), 158 Ketones, Paterno`-Bu¨ chi reaction: 2,3-dihydrofuran, 93–95 furan compounds, 84–92 Labeling industry, UPC-Dylux technology for, 209–212 Large wall-screen displays, Dylux technology and, 183–184 Laser-induced fluoroscence (LIF), thiophosgene molecules, 28–29 first excited singlet state (S1 (n,p*)), 46–47 lifetime measurements, 68–72 second singlet excited state (S2), 56–64 photophysical properties, 72–76 Leuco Crystal Violet (LCV), Thermal Dylux1 proofpaper, 253–255
270
INDEX
Leuco dye photocopy system. See also Dylux1 instant-access imaging materials color-overlay film development and, 221 development and patenting of, 149–150 proofpaper containing, 200–201 ultraviolet and visible light in, 160–162 Lewis acid catalysis, Paterno`-Bu¨ chi reaction, furan derivatives, 96–103 Lifetime results, thiophosgene electronic states, first excited singlet state, 66–72 Literature sources. See also Publications on Dylux technology, 202–203 HABI chemistry, 141–143, 257 Long path length absorption, thiophosgene molecules, first triplet electronic state (T1), 49–56 Looney, Catherine, 155–157, 170, 175, 207, 214 Lowest unoccupied molecular orbital (LUMO), Paterno`-Bu¨ chi reaction, 2,3dihydrofuran, 95 MacLachlan, Al, 157–163, 260 Magnetic rotation spectra (MRS), thiophosgene molecules, first triplet electronic state (T1), 49–56 Mahler, Walter, 202–204 Manos, Phil, 153–155 Marketing efforts, Dylux1 instant-access imaging materials, 178–189 business planning, 184–185 de Campi’s contributions, 187–188 Mattell’s involvement with, 188–189 point-of-sales opportunities, 179 Wilbur’s contributions, 178–179 Marking transfer sheets, patents for, 212 Mark Systems, Dylux technology and, 186 Markus, Bernard, 186 Mattell Corporation, Dylux technology and, 188–189 Medical imaging systems, Dylux technology and, 215–217 Mesonaphthobianthrone, hypericin-like fluorescence in, 2–7 2-Methylfuran, Paterno`-Bu¨ chi reaction: asymmetric reactions, 105–116 furan compounds, 84–92 5-Methyl-2-furyl derivatives, Paterno`-Bu¨ chi reaction, asymmetric furan reactions, 115–116
N-Methylimidazole, Paterno`-Bu¨ chi reaction, 119–120 Microencapsulation technology, Dylux applications in, 248–250 Microfilm: blowback, Dylux technology and, 192 HABI chemistry applications in, 173–176 Military applications, of Dylux technology, 199–200 Mirror image symmetry, hypericin and hypocrellin, 7 ground-state heterogeneity, 7–15 Molecular orbitals. See also Highest occupied molecular orbital; Lowest unoccupied molecular orbital thiophosgene electronic states, 29–34 Moyer, Robert, 206 Naphthaldehydes, Paterno`-Bu¨ chi reaction: 2,3-dihydrofuran, 95 thiophene, 117 Naphthazarin, hydrogen-atom transfer in, 7–8 o-Nitroaromatic photoinhibitor, patent for, 233 Nonradiative lifetime, thiophosgene molecules, first triplet electronic state (T1), 52–56 ‘‘Normal’’ form hypericin, ground state heterogeneity in, 10–11 N-pyridylsydnone (N-PS), photochromic properties of, 136–137 Nuclear kinetic energy operator, thiophosgene molecules, first excited singlet state (S1 (n,p*)), 64–72 Nuclear magnetic resonance spectroscopy: hypericin-like quinones, ground-state heterogeneity, 9 Paterno`-Bu¨ chi reaction, furan compounds, 83–92 Nuclear Overhauser effect (NOE) analysis, Paterno`-Bu¨ chi reaction: asymmetric furan reactions, 115–116 furan compounds, 84–92 One-photon excitation, thiophosgene molecules, 77 Optical-optical double rersonance (OODR) spectroscopy, thiophosgene molecules: basic properties, 28–29 first excited singlet state (S1 (n,p*)), 46–47 lifetime results, 66–72 first triplet electronic state (T1), 49–56
INDEX
second singlet excited state (S2), 56–64 fluorescence depletion, 74–76 Optical parametric amplifier pump, hypericin ground state heterogeneity analysis, 10–15 Optical printing, HABI chemistry and, 167–168 Optimization, HABI chemistry and, 160 Opto-magnetic printing, Dylux technology and, 185–186 Out-of-plane motion, thiophosgene molecules: first excited singlet state (S1 (n,p*)), 44–47, 65–72 second singlet excited state (S2), 56–64 Overlaid film, Dylux technology for, 218–221 Overlap integral, thiophosgene molecules, first excited singlet state (S1 (n,p*)), 65–72 Oxetane formation, Paterno`-Bu¨ chi reaction: furan derivatives, 96–103 pyrrole, 117–118 Oxetanocin formation, Paterno`-Bu¨ chi reaction, furan derivatives, 102–103 Panar, M., 202–204 Paris, Jean, 143 Patents: automatic inventory control technology, 230 biimidazoles, 140–141 co-irradiation method for producing positive images utilizing phototropic spiropyrans or indenone oxide or dual response photosensitive composition, 192 co-irradiation system for producing positive images, 192 dimers derived from unsymmetrical 2,4,5triphenylimidazole compounds as photoinitiators, 226–227 direct-positive photoimaging material for electrographic master preparation, 243 dry silver-free photographic process, 158 dual-response photosensitive compositions containing acyl ester of triethanolamine, 195 dual-response photosensitive compositions containing alkylbenzenesulfonic acid and arene sulfonamide, 196 elements for forming print-out images, 250–251
271 flexible ultraviolet radiation transmitting filters, 214 heat-bleachable composition useful in photography, 246–247 heat-sensitive reactive products of hexaarylbiimidazole and antihalation dyes, 247 hexaarylbiimidazole oxidation systems, 148–150 hexaarylbisimidazole oxidation systems, 149–150 image-producing layer of nearly uniform thickness for holography, 175–176 imaging and fixing radiation-sensitive compositions by sequential irradiation, 169–170 imaging composition, 246 leuco compounds oxidizable to dyes, 235 leuco dye/hexaarylbiimidazole thermally activated imaging process, 156 leuco dye photocopy system, 149 leuco triarylmethane/hexaarylbiimidazole color forming system with deactivator, 159–160 light-sensitive composition of organic colorgenerator, photooxidant and organic thermally activable reducing agent progenitor, 154–155 marking transfer sheets, 212 method for preparing positive and negative images using photohardenable electrostatic master, 242–243 microencapsulation technologies, 248–250 nonphotosensitive, thermally imageable element having improved room light stability, 254 photoactivatable hexaarylbiimidazolecoumarin compositions, 160 photodecorating sheet material with matched colored designs, 206–207 photohardenable electrostatic master having improved backtransfer and charge decay, 242 photoimaging systems with cyclic hydrazides, 222 photopolymerizable composition containing o-nitroaromatic photoinhibitor, 233 photopolymerizable compositions and elements, 171–172 photosensitive composition, 148
272
INDEX
Patents (Continued) photosensitive composition comprising organic nitrogen-containing colorgenerator, photo-oxidant, and redox couple, 159 photothermographic reproduction system and photosensitive materials, 245–246 portable copier using flash lamp article, 217–218 preparation of papers coated on both sides with photosensitive composition, 194 research on HABI and searching of, 164–165 storage-stable, light-sensitive composition, aminotriaryl-methane and organic photooxidant, 149 storage-stable photoimageable deuteroleuco dye/photooxidation compositions with improved leuco dye, 250 thermally imageable monochrome proofing product with high contrast and fast photospeed, 253–254 triphenylmethane dyes, 156 2,4,5-triphenylimidazole dimers, 140 universal product code marking composition containing photosensitive dye former, pigment, and binder, 210 visible light-activated phototropic compositions comprising a hexaarylbiimidazole and bis(alkylamino)acridine dye, 163 visible light-activated phototropic compositions comprising a hexaarylbiimidazole and hydroxyphthalein dye, 163 visible light-sensitive phototropic compositions comprising hexaarylbiimidazole and carbocyanine dye, 160, 163 xeroprinting with photopolymer master, 241–244 Paterno`-Bu¨ chi reaction: furan derivatives, 83–116 asymmetric reactions, 103–116 benzofuran, 92–93 2,3-dihydrofuran, 93–95 furan, 83–92 synthetic applications, 96–103 non-furan pentaatomic heterocycles, 116–123 azaindole, 122 imidazole, 119–120
indole, 120–121 pyrrole, 117–118 reactivity observations, 122–123 selenophene, 118–119 thiazole, isoxazole, and isothiazole, 120 thiophene, 116–117 Pentaatomic heterocyclic compounds, 1,2cycloaddition reaction, Paterno`-Bu¨ chi reaction: furan derivatives, 83–116 asymmetric reactions, 103–116 benzofuran, 92–93 2,3-dihydrofuran, 93–95 furan, 83–92 synthetic applications, 96–103 non-furan pentaatomic heterocycles, 116–123 azaindole, 122 imidazole, 119–120 indole, 120–121 pyrrole, 117–118 reactivity observations, 122–123 selenophene, 118–119 thiazole, isoxazole, and isothiazole, 120 thiophene, 116–117 Permanent color, HABI chemistry and, 145–151 Perturbation gaps, thiophosgene molecules: first triplet electronic state (T1), 47–56 second singlet excited state (S2), 62–64 Perylene quinones. See also Hypericin-like perylene quinones multiple hydrogen atom transfers in, 15–19 Phantom states, thiophosgene molecules, second singlet excited state (S2), 73–76 Phenanthrenequinone: DYLUX 503 development and, 163 visible light and, 160, 162 Phenidone, development and patents for, 221–222 Phosphorescence emission, thiophosgene molecules: absence of, 77 first triplet electronic state (T1), 53–56 Photochemical research, color-overlay film development and, 218–221 Photochromic sunglasses, HABI chemistry and, 144–145 Photodecoration: Dylux tetchnology and, 204–209 HABI chemistry and, 207–209 termination of, 218
INDEX
Photoelastic modulator (PEM), thiophosgene molecules, second singlet excited state (S2), 63–64 Photofix system, patents for, 195–196 Photogenerated protons, excited-state intramolecular H-atom transfer in hypericin-like perylene quinones, 2–7 Photohorizons technology, military application of, 199–200 Photoimaging, Ltd., Dylux technology and, 244–245 Photomarker1 Corp., Dylux technology and, 227–228 Photo-oxidation, research and patenting for, 148–151, 223–227 Photophysical properties, thiophosgene molecules, 64–76 first excited singlet state (S1), 64–72 first triplet electronic state (T1), 51–56 second excited singlet state (S2), 72–76 Photopolymerization process: HABI chemistry and, 170–172, 255–256 o-nitroaromatic photoinhibitor, 233 xeroprinting and, 241–244 Photosensitive composition: coatings technology for, 193–197 patents, 148, 194–196 UPC development and, 211–212 Phototropic technology, HABI chemistry and, 137–138 Picosecond temporal resolution, thiophosgene molecules, second singlet excited state (S2), 73–76 Planar conformation: thiophosgene molecule electronic states: carbon-sulfur double bonds, 32–34 ground electronic state, 34–36 thiophosgene molecules: first excited singlet state (S1 (n,p*)), 40–47 second singlet excited state (S2), 57–64 Platelet-activating factor formation, Paterno`Bu¨ chi reaction, furan derivatives, 100–103 Point-of-sales opportunities, for Dylux, 179 Polarization-dependent OODR spectra, thiophosgene molecules, second singlet excited state (S2), 63–64 Polymeric HABIs, permanent color production, 145–151
273
Polymerization, HABI chemistry and, imaging speed and initiation of, 171–172 Portable copiers, Dylux technology and, 217–218 Porter, George, 158 Potential energy functions: hypericin ground state heterogeneity, 9–10 thiophosgene molecules: first excited singlet state (S1 (n,p*)), 41–47 first triplet electronic state (T1), 52–56 second singlet excited state (S2), 56–64 Predissociation, thiophosgene electronic states, first excited singlet state (S1 (n,p*)), 70–72 Printing industry, HABI chemistry applications in, 151 Project Apollo, Dylux technology and, 183 Proofing technologies: Dylux development and, 190–191 HABI chemistry and, 168 Proofpaper: black Dylux 535 product, 234–237 limits of Dylux technology for, 200–201 properties of, 193–197 TCTM-HABI development and, 223–227 Thermal Dylux1 proofpaper, 252–255 Proton-transfer species, hypericin and hypocrellin, 7–8 10-ps component, excited-state intramolecular H-atom transfer, hypericin and hypocrellin, 4–7 Publications. See also Literature sources computer simulation with Dylux technology, 207 on Dylux technology, 202–203 HABI chemistry, 141–143, 257–259 miscellaneous, 199 photopolymerization systems, 239 photopolymers in electrostatic imaging applications, 243–244 on photosensitive and heat sensitive recording materials, 231–232 watermarks technology, 252 Pump wavelength, hypericin transient absorption kinetics and, 13–15 Pyramidal conformation, thiophosgene molecules: first excited singlet state (S1 (n,p*)), 41–47 second singlet excited state (S2), 57–64 Pyrrole, Paterno`-Bu¨ chi reaction, 117–118
274
INDEX
Quadratic terms, thiophosgene molecules, first excited singlet state (S1 (n,p*)), 44–47 Quantum beat-modulated fluorescence decay, thiophosgene molecules: electronic states, 77 first excited singlet state (S1 (n,p*)), 67–72 Quantum mechanical calculations: hypericin ground state heterogeneity, 9–10 multiple hydrogen-atom transfers in perylene quinones, 17–19 Quenching properties, Paterno`-Bu¨ chi reaction, non-furan compounds, 122–123 Radiationless transitionis, thiophosgene molecules, first excited singlet state (S1 (n,p*)), 64–72 Radiation physics, HABI chemistry and, 157–163 Raman spectroscopy, thiophosgene electronic states, ground electronic states, 36 Reactivity studies, Paterno`-Bu¨ chi reaction, non-furan compounds, 122–123 Recording materials, Dylux technology for, 231–232 Regioiselectivity, Paterno`-Bu¨ chi reaction: asymmetric furan reactions, 112–116 benzofuran, 92–93 furan compounds, 84–92 Restricted Harterr-Fock (RHF) theory, hypericin ground state heterogeneity, 9–10 Riehm, Roland (Dr.), 158 Ring-opening reactions, Paterno`-Bu¨ chi reaction, furan derivatives, 97–103 Rotational analysis, thiophosgene molecules, first excited singlet state (S1 (n,p*)), 45–47 Rydberg states, thiophosgene molecules, first excited singlet state (S1 (n,p*)), 39–47 Scharf’s hypothesis, Paterno`-Bu¨ chi reaction, furan compounds, 92 Schultz, D. (Dr.), 196–197 Second singlet excited state (S2), thiophosgene molecules, 56–64 Herzberg notation, 31–34 photophysical properties, 72–76 spin sublevel transitions, 78 Selection rules, thiophosgene molecules, 29 first excited singlet state, electronic matrix elements, 65–72
first triplet electronic state (T1), 47–56 second singlet excited state (S2), 60–64 Selenophene, Paterno`-Bu¨ chi reaction, 118–119 Seminumerical Hessian calculations, hypericin ground state heterogeneity, 9–10 Sern-Volmer plot, Paterno`-Bu¨ chi reaction, nonfuran compounds, 122–123 Signaigo, Frank (Dr.), 151 Silversmith, Ernest (Dr.), 139, 160 Silyl furan derivatives, Paterno`-Bu¨ chi reaction, 86–92 Singlet-triplet transitions, thiophosgene molecules, first triplet electronic state (T1), 47–56 Sinnott, Richard, 185–186 Skeleton coordinates, hydrogen atom translocation, 21–22 Society of Photographic Engineers and Scientists (SPSE): Dylux presentations to, 191–192 Tokyo meeting of (1977), 230–232 Spectrographic analysis, thiophosgene molecules, first excited singlet state (S1 (n,p*)), 43–47 Spectroscopy, HABI chemistry and, 155–157 Spin-orbit coupling (SOC), Paterno`-Bu¨ chi reaction, furan derivatives, 86–92 Spin-orbit operator, thiophosgene molecules: first triplet electronic state (T1), 47–56 second singlet excited state (S2), 61–64 Spirit duplication, history of, 147–148 Stannyl furan derivatives, Paterno`-Bu¨ chi reaction, 86–92 synthetic applications, 101–103 Statistical analysis, in HABI chemistry, 259–260 Statistical limit of radiationless transitions, thiophosgene molecules, first excited singlet state (S1 (n,p*)), 64–72 Stereochemical behavior, Paterno`-Bu¨ chi reaction: asymmetric reactions, 107–116 furan compounds, 91–92 Stereoselectivity, Paterno`-Bu¨ chi reaction, asymmetric furan reactions, 111–116 Stick spectra, thiophosgene molecules, first excited singlet state (S1 (n,p*)), 43–47 Stimulated emission pumping (SEP), thiophosgene electronic states: ground electronic state, 36 second singlet excited state (S2), 75–76
INDEX
Strilko, Peter (Dr.), 159, 164 Symmetry axes: thiophosgene electronic states, 29–34 first excited singlet state (S1 (n,p*)), 38–47 Synthetic applications, Paterno`-Bu¨ chi reaction, furan derivatives, 96–103 Tautomeric forms, of hypericin and hypocrellin, 7–8 TCM-HABI, photopolymerization and, 256 TCTM-HABI: Black Dylux 535 product and, 235–237 development and patenting of, 223–227 Terss, Robert, 172–173 Tertiary amines, photopolymerizable compositions and elements, 171–172 Tetracyanoethylene (TCNE), HABI chemistry and, 208–209 Thermal Dylux1 proofpaper, 252–255 Thiazole, Paterno`-Bu¨ chi reaction, 120 Thin-film polarizer (TFP), hypericin ground state heterogeneity analysis, 11–15 Thiophene, Paterno`-Bu¨ chi reaction, 116–117 Thiophosgene: basic properties, 28–29 electronic states, 29–64 first excited singlet state (S1 (n,p*)), 36–47 first triplet electronic state (T1), 47–56 ground electronic state (S0), 34–36 molecular orbitals and electron configurations, 29–34 second singlet excited state (S2), 56–64 future research issues, 77–78 photophysical properties, 64–76 first excited singlet state (S1), 64–72 first triplet electronic state (T1), 51–56 second excited singlet state (S2), 72–76 3M, competition to Dylux from, 246 Time constants, multiple hydrogen-atom transfers in perylene quinones, 17–19 Time delay measurements, thiophosgene molecules, first triplet electronic state (T1), 53–56 TLA-454, 149–150 DYLUX 503 development and, 163 proofpaper containing, 200–201 Trademark registration, for Dylux, 181–182
275
Transient absorption kinetics, hypericin ground state heterogeneity analysis, 13–15 Triarylimidazoles, HABI chemistry, 139–141 2,4,5-Triphenylimidazole dimers, patent, 140 Triplet biradical intermediates, Paterno`-Bu¨ chi reaction, asymmetric furan reactions, 110–116 Tunable pump laser analysis: hypericin ground state heterogeneity, 10–15 thiophosgene electronic states, first triplet electronic state (T1), 54–56 Turro, Nick, 144 Universal Product Code (UPC), Dylux technology and, 209–212 Urey-Bradley molecules, thiophosgene electronic states, 36 UVI movie, for Dylux promotion, 179–181 UVI-Ultraviolet imaging, patenting for, 148–150 Vibrational deficiency, thiophosgene molecules, first excited singlet state, 65–72 Vibrational modes: hydrogen-atom translocation, 19–22 multiple hydrogen-atom transfer in perylene quinones and, 18–19 thiophosgene molecules, 28–29 electronic states, ground electronic state, 35–36 first excited singlet state (S1 (n,p*)), 38–47 first triplet electronic state (T1), 49–56 second singlet excited state (S2), 58–64 Vibronic coupling integral, thiophosgene molecules, first excited singlet state (S1 (n,p*)), 65–72 Vibronic interactions, thiophosgene molecules: first excited singlet state (S1 (n,p*)), 38–47 second singlet excited state (S2), 60–64 Viscosity dependence, excited-state intramolecular H-atom transfer, hypericin and hypocrellin, 5–7 W. H. Brady Co., Kalograph technology and, 186–187 Walsh correlation diagram: thiophosgene electronic states, 32–34 thiophosgene molecules: first excited singlet state (S1 (n,p*)), 40–47 second singlet excited state (S2), 56–64
276 Wartell, William S., 189–198 Watermarks, Dylux technology and, 251–252 Wheeler, Phil, 144 Whitman, G. M., 151 Wilbur, Harold, 178–179, 203–204
INDEX
Xeroprinting, Dylux technology and, 241–244 X-Raylux, development of, 232–233 X-ray prints, Dylux technology for, 232–233 Yembrick, Charles (Dr.), 159, 164–165, 168 Zwicker, Harry, 252–255
CUMULATIVE INDEX VOLUMES 1-28
Addition of Atoms to Olefins, in Gas Phase (Cvetanovic) . . . . . . . . . . . Advances in the Measurement of Correlation in Photoproduct Motion (Morgan, Drabbels, and Wodtke) . . . . . . . . . . . . . . . . . . . . . . . . . . AFM and STM in Photochemistry Including Photon Tunneling (Kaupp) Alcohols, Ethers, and Amines, Photolysis of Saturated (von Sonntag and Schuchmann) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alkanes and Alkyl Radicals, Unimolecular Decomposition and Isotope Effects of (Rabinovitch and Setser) . . . . . . . . . . . . . . . . . . . . . . . . Alkyl Nitrites, Decomposition of and the Reactions of Alkoxyl Radicals (Heicklen) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alternative Halocarbons, Atmospheric Photochemistry of (Francisco and Maricq) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Anthracenes, Excited State Reactivity and Molecular Topology Relationships in Chromophorically Substituted (Becker) . . . . . . . . . . . . . . . Anti-Stokes Fluorescence, Cooling of a Dye Solution by (Zander and Drexhage) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Aromatic Hydrocarbon Solutions, Photochemistry of (Bower) . . . . . . . . Asymmetric Photoreactions of Conjugated Enones and Esters (Pete) . . . Atmospheric Reactions Involving Hydrocarbons, FTIR Studies of (Niki and Maker). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
VOL. PAGE 1 115 23 19
279 119
10
59
3
1
14
177
20
79
15
139
20 1 21
59 23 135
15
69
Benzene, Excitation and Deexcitation of (Cundall, Robinson, and Pereira) Biocatalysis and Biomimetic Systems, Artificial Photosynthetic Transformations Through (Willner and Willner) . . . . . . . . . . . . . . . . . . . Biochromophoric Systems, Excited State Behavior of Some (De Schryver, Boens, and Put) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
10
147
20
217
10
359
Cancer Treatment, Photochemistry in (Dougherty) . . . . . . . . . . . . . . . . Carbonyl Compounds, The Photocycloaddition of, to Unsaturated Systems: The Syntheses of Oxetanes (Arnold). . . . . . . . . . . . . . . . . . .
17
275
6
301
Advances in Photochemistry, Volume 28 Edited by Douglas C. Neckers, William S. Jenks, and Thomas Wolff # 2005 John Wiley & Sons, Inc.
277
278
CUMULATIVE INDEX VOLUMES 1-28
VOL. PAGE Catalysis of Photoinduced Electron Transfer Reactions (Fukuzumi and Itoh). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cobalt (III) and Chromium (III) Complexes, the Photochemistry of, in Solution (Valentine, Jr.). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Complexes, Photoinitiated Reactions in Weakly Bonded (Shin, Chen, Nickolaisen, Sharpe, Beaudet, and Wittig) . . . . . . . . . . . . . . . . . . . Cyclic Ketones, Photochemistry of (Srinivasan) . . . . . . . . . . . . . . . . . . 1,2-Cycloaddition Reaction of Carbonyl Compounds and Pentaatomic Heterocyclic Compounds (D’Auria, Emanuele, and Racioppi). . . . . Cyclobutanones, Solution Phase Photochemistry of (Morton and Turro) . Cyclometallated Complexes, Photochemistry and Luminescence of (Maestri, Balzani, Deuschel-Cornioley, and von Zelewsky) . . . . . . .
25
107
6
123
16 1
249 83
28 9
81 197
17
1
8
77
14
1
16
1
3 8
241 1
9
1
13 6
237 425
22
197
20 11
1 489
16
119
Flash Photolysis with Time-Resolved Mass Spectrometry (Carr) . . . . . . Free Radical and Molecule Reaction in Gas Phase, Problems of Structure and Reactivity (Benson) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . FTIR Emission Studies, Time Resolved, of Photochemical Reactions (Hancock and Heard) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Molecular Glasses: Building Blocks for Future Optoelectronics (Fuhrmann and Salbeck) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
25
1
2
1
18
1
27
83
Gas Phase, Addition of Atoms of Olefins in (Cvetanovic) . . . . . . . . . . . Gas Phase Reaction, Photochemical, in Hydrogen-Oxygen System (Volman) Gas Phase Reactions, Involving Hydroxyl and Oxygen Atoms, Mechanisms and Rate Constants of (Avramenko and Kolesnika) . . . . . . .
1 1
115 43
2
25
2
137
a-Dicarbonyl Compounds, The Photochemistry of (Monroe) . . . . . . . . . Diffusion-Controlled Reactions, Spin-Satistical Factors in (Saltiel and Atwater). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Electron and Energy Transfer, Mimicking of Photosynthetic (Gust and Moore). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Electron Energy Transfer between Organic Molecules in Solution (Wilkinson) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Electronically Excited Halogen Atoms (Hussain and Donovan) . . . . . . . Electron Spin Resonance Spectroscopy. Appliction of to Photochemistry (Wan). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Electron Transfer, Photoinduced in Organic Systems, Control of Back Electron Transfer of (Fox). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Electron Transfer Luminescence in Solution (Zweig) . . . . . . . . . . . . . . Elementary Photoprocesses in Designed Chromophore Sequence on aHelical Polypeptides (Sisido). . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ethylenic Bonds, Present Status of the Photoisomerization Abut (Arai and Tokumaru) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Excimers, What’s New in (Yakhot, Cohen, and Ludmer) . . . . . . . . . . . . Excited Electronic States, Electronic and Resonance Raman Spectroscopy Determination of Molecular Distortions in (Zink and Shin). . . . . . .
Halogenated Compounds, Photochemical Processes in (Major and Simons) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
279
CUMULATIVE INDEX VOLUMES 1-28
VOL. PAGE Heterogeneous Catalysts, the Question of Artificial Photosynthesis of Ammonia on (Davies, Boucher, and Edwards) . . . . . . . . . . . . . . . . Hydrogen-Oxygen Systems, Photochemical Gas Phase Reactions in (Volman) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hydroxyl and Oxygen Atoms, Mechanisms and Rate Constants of Elemen tary Gas Phase Reactions Involving (Avramenko and Kolesnikova) Hydroxyl Radical with Organic Compounds in the Gas Phase, Kinetics and Mechanisms of the Reactions of (Atkinson, Darnall, Winter, Lloyd, and Pitts) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hypericin and Its Perylene Quinone Analogs: Probing Structure, Dynamics, and Interactions with the Environment (Das, Halder, Chowdhury, Park, Alexeev, Gordon, and Petrich) . . . . . . . . . . . . . . Hypophalites, Developments in Photochemistry of (Akhtar) . . . . . . . . .
19
235
1
43
2
25
11
375
28 2
1 263
Imaging Systems, Organic Photochemical (Delzenne) . . . . . . . . . . . . . . Intramolecular Proton Transfer in Electronically Excited Molecules (Klo¨ pffer). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Invention of Dylux1 Instant-Access Imaging Materials and the Development of Habi Chemistry—A Personal History (Dessauer). . Ionic States, in Solid Saturated Hydrocarbons, Chemistry of (Kevan and Libby) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Isotopic Effects in Mercury Photosensitization (Gunning and Strausz) . .
11
1
10
311
28
129
2 1
183 209
Ketone Photochemistry, a Unified View of (Formosinho and Arnaut) . . .
16
67
23
213
25
173
1
209
1
183
1 8
209 227
Lanthanide Complexes of Encapsulating Ligands at Luminescent Devices (Sabbatini, Guardigli, and Manet) . . . . . . . . . . . . . . . . . . . . . . . . . Laser Trapping-Spectroscopy-Electrochemistry of Individual Microdroplets in Solution (Nakatoni, Chikami, and Kitamura) . . . . . . . . . . . . Mechanism of Energy Transfer, in Mercury Photosensitization (Gunning and Strausz) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mechanistic Organic Photochemistry, A New Approach to (Zimmerman) Mercury Photosensitization, Isotopic Effects and the Mechanism of Energy Transfer in (Gunning and Strausz) . . . . . . . . . . . . . . . . . . . Metallocenes, Photochemistry in the (Bozak) . . . . . . . . . . . . . . . . . . . . Methylene, Preparation, Properties, and Reactivities of (De More and Benson) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Molecular Distrortions in Excited Electronic States, Electronic and Resonance Raman Spectroscopy Determination of (Zink and Shin)
2
219
16
119
Neutral Oxides and Sulfides of Carbon, Vapor Phase Photochemistry of the (Fileeth) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nitric Oxide, Role in Photochemistry (Heicklen and Cohen) . . . . . . . . . Noyes, W., A., Jr., a Tribute (Heicklen) . . . . . . . . . . . . . . . . . . . . . . . . Nucleic Acid Derivatives, Advances in the Photochemistry of (Burr) . . .
10 5 13 6
1 157 vii 193
14
135
17 24
313 1
Olefins, Photolysis of Simple, Chemistry of Electronic Excited States or Hot Ground States? (Colling) . . . . . . . . . . . . . . . . . . . . . . . . . . . . Onium Salts, Photochemistry and Photophysics of (DeVoe, Olofson, and Sahyun) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Optical Single-Molecule Detection at Room Temperature (Meixner) . . .
280
CUMULATIVE INDEX VOLUMES 1-28
Organic Molecules, Photochemical Rearrangements of (Chapman) . . . . Organic Molecules in Adsorbed or Other Perturbing Polar Environments, Photochemical and Spectroscopic Properties of (Nichollas and Leermakers) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Organic Molecules in their Triplet States, Properties and Reactions of (Wagner and Hammond) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Organic Nitrites, Developments in Photochemistry of (Akhtar) . . . . . . . Organic Photochemical Refractive-Index Image Recording Systems (Tomlinson and Chandross) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Organized Media on Photochemical Reactions, A Model for the Influence of (Ramamurthy, Weiss, and Hammond) . . . . . . . . . . . . . . . . . . . . Organo-Transition Metal Compounds, Primary Photoprocesses of (Bock and von Gustorf) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Perhalocarbons, Gas Phase Oxidation of (Heicklen) . . . . . . . . . . . . . . . Phenyl Azide, Photochemistry of (Schuster and Platz). . . . . . . . . . . . . . Phosphorescence and Delayed Fluorescence from Solutions (Parker) . . . Phosphorescence-Microwave Multiple Resonance Spectroscopy (El-Sayed) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Photoassociation in Aromatic Systems (Stevens) . . . . . . . . . . . . . . . . . . Photochemical Mechanisms, Highly Complex (Johnston and Cramarossa) Photochemical Oxidation of Aldehydes by Molecular Oxygen, Kinetics and Mechanism of (Niclause, Lemaire, and Letort) . . . . . . . . . . . . Photochemical Reactivity, Reflections on (Hammond) . . . . . . . . . . . . . . Photochemical Rearrangements of Conjugated Cyclic Ketones: The Present State of Investigations (Schaffner) . . . . . . . . . . . . . . . . . . . Photochemical Transformations of Polyenic Compounds (Mousseron) . . Photochemically Induced Dynamic Nuclear Polarization (Goez) . . . . . . Photochemistry in Cyclodextrin Cavities (Bortolus and Monti) . . . . . . . Photochemistry of Triarylmethane Dye Leuconitriles (Jarikov and Neckers) Photochemistry of Conjugated Dienes and Trienes (Srinivasan) . . . . . . . Photochemistry of Rhodopsins, The (Ottolenghi) . . . . . . . . . . . . . . . . . Photochemistry of Simple Aldehydes and Ketones in the Gas Phase (Lee and Lewis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Photochemistry of the Troposphere (Levy) . . . . . . . . . . . . . . . . . . . . . . Photochemistry of Vitamin D and Its Isomers and of Simple Trienes (Jacobs and Havinga) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Photochemistry, Vocabulary of (Pitts, Wilkinson, Hammond). . . . . . . . . Photochromism (Dessauer and Paris) . . . . . . . . . . . . . . . . . . . . . . . . . . Photodissociation Dynamics of Hydride Molecules: H Atom Photofragment Translational Spectroscopy (Ashfold, Mordaunt, and Wilson) Photo-Fries Rearrangement and Related Photochemical (l.j) Shifts of (j ¼ 3,5,7) of Carbonyl and Sulfonyl Groups (Bellus) . . . . . . . . . . . Photography, Silver Halide, Chemical Sensitization, Spectral Sensitization, Latent Image Formation (James) . . . . . . . . . . . . . . . . . . . . . . Photo-Induced and Spontaneous Proton Tunneling in Molecular Solids (Trommsdorff) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Photoionization and Photodissociation of Aromatic Molecules, by Ultra violet Radiation (Terenin and Vilessov) . . . . . . . . . . . . . . . . . . . . .
VOL. PAGE 1 323
8
315
5 2
21 263
12
201
18
67
10
221
7 17 2
57 69 305
9 8 4
311 161 1
4 7
25 373
4 4 23 21 26 4 12
81 195 63 1 1 113 97
12 9
1 369
11 1 1
305 1 275
21
217
8
109
13
329
24
147
2
385
281
CUMULATIVE INDEX VOLUMES 1-28
VOL. PAGE Photoluminescence Methods in Polymer Science (Beavan, Hargreaves, and Phillips) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Photolysis of the Diazirines (Frey) . . . . . . . . . . . . . . . . . . . . . . . . . . . . Photooxidation Reactions, Gaseous (Hoare and Pearson) . . . . . . . . . . . . Photooxygenation Reactions, Type II, in Solution (Gollnick) . . . . . . . . . Photophysical Probes of DNA Sequence-Directed Structure and Dynamics (Murphy) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Photophysics of Gaseous Aromatic Molecules: Excess Vibrational Energy Dependence of Radiationless Processes (Lim) . . . . . . . . . . . . . . . . Photopolymerization, Dye Sensitized (Eaton) . . . . . . . . . . . . . . . . . . . . Photoreactive Organic Thin Films in the Light of Bound Electromagnetic Waves (Sekkat and Knoll) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Photosensitized Reactions, Complications in (Engel and Monroe) . . . . . Photosynthetic Electron and Energy Transfer, Mimicking of (Gust and Moore). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Photothermal Studies of Photophysical and Photochemical Processes by the Transient Grating Method (Terazima). . . . . . . . . . . . . . . . . . . . Phytochrome, Photophysics and Photochemistry of (Schaffner, Branslavsky, and Holzwarth) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Polymers, Photochemistry and Molecular Motion in Solid Amorphous (Guillet) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pressure-Tuning Photochemistry of Metal Complexes in Solution (Eldik and Ford) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Proton-Transfer Reactions in Benzophenone/N,N-Dimethylaniline Photochemistry (Peters) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Primary Processes and Energy Transfer: Consistent Terms and Definitions (Porter, Balzani and Moggi) . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11 4 3 6
207 225 83 1
26
145
23 13
165 427
22 8
117 245
16
1
24
255
15
229
14
91
24
61
27
51
9
147
Quantized Matter, Photochemistry and Photoelectrochemistry of: Properties of Semiconductor Nanoparticles in Solution and ThinFilm Electrodes (Weller and Eychmu¨ ller). . . . . . . . . . . . . . . . . . . . Quantum Theory of Polyatomic Photodissociation (Kreslin and Lester)
20 13
165 95
Radiationless Transitions, Isomerization as a Route for (Phillips, Lamaire, Burton, and Noyes, Jr.) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Radiationless Transitions in Photochemistry (Jortner and Rice) . . . . . . .
5 7
329 149
19 26
179 93
17
145
15 7 17 11
279 311 217 105
4
49
Semiconductor Nanoclusters, Photophysical and Photochemical Processes of (Wang). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Semiconductor Photocatalysis for Organic Synthesis (Kisch) . . . . . . . . . Silver Halides, Photochemistry and Photophysics of (Marchetti and Eachus) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Single Crystals, Photochemical Mechanism in: FTIR Studies of Diacyl Peroxides (Hollingsworth and McBride) . . . . . . . . . . . . . . . . . . . . Singlet Molecular Oxygen (Wayne) . . . . . . . . . . . . . . . . . . . . . . . . . . . Singlet Molecular Oxygen, Bimolecular Reactivity of (Gorman) . . . . . . Singlet Molecular Oxygen, Physical Quenchers of (Bellus) . . . . . . . . . . Singlet and Triple States: Benzene and Simple Aromatic Compounds (Noyes and Unger) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
282
CUMULATIVE INDEX VOLUMES 1-28
Small Molecules, Photodissociation of (Jackson and Okabe) . . . . . . . . . Solid Saturated Hydrocarbons, Chemistry of Ionic States in (Kevan and Libby) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Solvation, Ultrafast Photochemical Intramolecular Charge Transfer and Excited State (Barbara and Jarzeba). . . . . . . . . . . . . . . . . . . . . . . . Spectroscopy and (Photochemistry of Polyatomic Alkaline Earth Containing Molecules (Bernath) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Spin Conservation (Matsen and Klein) . . . . . . . . . . . . . . . . . . . . . . . . . Stilbenes, Bimolecular Photochemical Reactions of (Lewis) . . . . . . . . . Stilbenes and Stilbene-Like Molecules, Cis-Trans Photoisomerization of (Go¨ rner and Kuhn) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Structure and Reactivity of Organic Intermediates as Revealed by TimeResolved Infrared Spectroscopy (Toscano) . . . . . . . . . . . . . . . . . . . Sulfur Atoms, Reactions of (Gunning and Strausz) . . . . . . . . . . . . . . . . Sulfur and Nitrogen Heteroatomic Organic Compounds, Photochemical Reaction of (Mustafa) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Supramolecularly Organized Luminescent Dye Molecules in the Channels of Zeolite L (Calzaferri, Maas, Pauchard, Pfenniger, Megelski, Devaux) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Surfactant Solutions, Photochemistry in (von Bu¨ nau and Wolff) . . . . . . The EPR Spectroscopic D Parameter of Localized Triplet Diradicals as Probe for Electronic Effects in Benzyl-type Monoradicals (Adam, Harrer, Kita, and Nau) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Photochemistry of Indoles (Weedon) . . . . . . . . . . . . . . . . . . . . . . . Theory and Applications of Chemically Induced Magnetic Polarization in Photochemistry (Wan). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Thiophosgene: A Tailor-Made Molecule for Photochemical and Photophysical Studies (Moule, Fujiwara, and Lim). . . . . . . . . . . . . Transition Metal Complexes, Primary Processes in (Forster) . . . . . . . . . Triatomic Free Radicals, Spectra and Structures of (Herzberg). . . . . . . .
VOL. PAGE 13 1 2
183
15
1
23 7 13
1 1 165
19
1
26 4
41 143
2
63
27 14
1 273
24 22
205 229
12
283
28 16 5
27 215 1
3
157
22
1
Ultraviolet Photochemistry, Vacuum (McNesby and Okabe) . . . . . . . . . Ultraviolet Photodissociation Studies of Organosulfur Molecules and Radicals: Energetics, Structure Identification, and Internal State Distribution (Cheuk-Yiu Ng) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ultraviolet Radiation, Photoionizational and Photodissociation of Aromatic Molecules by (Terenin and Vilessov) . . . . . . . . . . . . . . . Up-Scaling Photochemical Reactions (Braun, Jakob, Oliveros, Oller do Nascimento) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2
385
18
235
Velocity Mapping of UV Multiphoton Excited Molecules (Chandler and Parker) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
25
59
Weakly Bonded Complexes, Photoinitiated Reactions in (Shin, Chen, Nickolaisen, Sharpe, Beaudet, and Wittig) . . . . . . . . . . . . . . . . . . .
16
249
Xanthine Dyes, Photochemistry of the (Neckers and Valdes-Aguilera) . .
18
315