Photochemical generation of stable fluorescent compounds (photofluorescence)

The mechanistic insights afforded by investigations of photochemicallygenerated stable fluorescent compounds are examined. Examples of the types of such photofluorescent processes found in the Iiterature are reviewed. The photochemical self-quenching rearrangement of furyl chromones, wavelengthdependent rearrangement of azirines, halogeo-dependent cycloelimination from halophenylnaphthalen~s and decarboxylation-decarbonylation of 1,2- dicarboxylic anhydrides are examined as photofluorescent systems. The efficient photoelimination of carbon monoxide and carbon dioxide from thermally stable 1 ,4-dicarboxylic anhydrides is reported and mechanistic information obtained from the formation of fluorescent products in this process is described. The photochemical generation of stable fluorescent products from nonfluorescent precursors, a process which we have termed photofluorescence, has been casually noted in the Iiterature for many years. There appears to have been no previous deliberate effort to create and examine this phenomenon, although there are very good reasons for doing so. I would first like to mention briefly some of the more cogent reasons and then proceed to examine the photochemistry that our efforts in this area have led to. The most fundamental reason for photogenerating fluorescent materials is to use them as a probe for examining the potential energy surface of the photochemical transformation. Professor Förster described his sturlies on diabatic and adiabatic processes in photochemistry 1 to the Third IUPAC Symposium on Photochemistry. He raised the question as to how molecule A is transformed photochemically to molecule B (Figure 1 ). Does deactivation ofthe electronic excitation take place in the reactant (as in 1) or in the product (as in 2) or in between (as in 3). Observation of characteristic emission from the fluorescent product excited state achieved through absorption by the precursor is good evidence for process 2 deactivation. We will explore the experimental requitements for such observations and report our results further on, but before that I'd like to go on to the applied areas and mention the photo-imaging possibilities with stable photofluorescent systems. We are weil aware that under most circumstances a fluorescent test can 389 A. ZWEIG A+hv----- A* A*-Avibr-------- B A*--·---B*- B A* B Excitation Deactivation (1) (2) (3) Figure 1. Alternative pathways in the photochemical conversion of A toB detect orders of magnitude less material than a colour test and this is why fluorescent tags and tracers are widely used in trace analytical applications2 . Indeed it is a common Iabaratory practice to examine a chromatographed reaction mixture with an ultra~violet lamp to detect separating fluorescent products. It should be evident that an efficient photor.hemical process that generates a stable and efficient fluorescent product from a non-fluorescent precursor will have an effective speed orders of magnitude greater than an equally efficient photochemical process that yields even a deeply-coloured product only detectable by differences in absorption. It should not be surprising, therefore, that as seen in Figure 2, the first photofluorescent imaging materiaP exhibited practical sensitivities comparable to that of photopolymer systems and much faster effective speed than that ofnon-polymerizing organic imaging materials, but, of course, still five orders of magnitudes lower than that of high speed silver photographic film. Practical exposure Relative System microjoulesjcm2 sensitivity Silver halide, Tri~X 0.001 3 X 107 Photofluorescer 100 300 Photoplastic, 216 film 150 200 Dry silver 800 43 UVIfilm 5000 6 Photachromie 30000 1 Kalvar 400000 0.075 Diazo 800000 0.033 Photofluorescer data from A. Zweig, Symposium 111 on Unconventional Photographie Systems (1971). Other data from J. E. Bigelow, 9th National Symposium on Information Display, May 1968. Figure 2. Relative sensitivity of various films Practical e~posure as used in Figure 2 refers to the exposure to radiation per unit area required to differentiate between exposed areas and background after suitable development. The high speed of silver halide film as compared to organic systems is due to the development step where the effect of only a few quanta makes a micron size grain develop. Professor Bird has pointed out that because several photons are needed to make a grain developable and because of other lasses, the detective quantum efficiency of even the most sensitive silver halide films is only of the order of one to two per cent4 . He has suggested how suitable organic photofluorescent materials could be utilized to capture and display an image with greater sensitivity than that available from silver halide film4 . In his concept, as illustrated in Fiqure 3, the photofluorescent material captures the image in a molecular transformation in the 390 PHOTOCHEMICAL GENERATION OF STABLE FLUORESCENT COMPOUNDS Imaging light ~nurce. h1• I -G Molecular fluorescence precursor Molecular fluorescer (latent image) .Qeveloping light source hv' aB h•·~ Signal averaging fluorescence emission collection Figure 3. Highspeed photography with photofluorescent materials first step and then in a developing step the presence of fluorescing molecules is detected and amplified by a developing beam of light. The development is effected by employing the initially produced stable fluorescer as a sort of catalyst which reveals its presence through the performance of its function. Thus conceptually, if not yet in practice, photofluorescent systems offer the opportunity of even greater sensitivity than silver halide-based materials. One of the first descriptions of the photolytic generation of an identified fluorescent product was made by Lewis, Magel and Lipkin5 , who observed development of a blue fluorescent product when non-fluorescent cis-stilbene (1) was photolysed. The photolytic change induced was shown to be the isomerization of 1 to trans-stilbene (2). The authors utilized this photolytic 1 cis-stilbene non-fluorescent hv ~ ~ O H'c=cJVJ 'H I 2 trans-stilbene hlue fluorescence generation of a fluorescent product to help clear up the existing confusion regarding this cis-trans isomerism. Cis to trans photoisomerisms are an important dass of photofluorescers. In Figure 4 we have listed some familiar examples such as substituted stilbenes6, indigo dyes 7 , and ß-styryl compounds8, as weil as the remarkable photoconversion of a non-fluorescent all cis 1,2,4,7-tetraphenylcyclooctatetraene to a fluorescent trans containing isomer9 , illustrating the apparent thoroughness of the relationship of cis-trans photoisomerization to photofluorescence. Despite this ubiquity, these are not the best reactions to choose for an initial detailed investigation ofthe photofluorescence process because: (1) there are relatively small differences in absorption spectra between product and precursor, (2) under some conditions the cis isomer becomes fluorescent 90, and, (3) photoreversibility, modest photoefficiency and occurrence of side reactions cloud attempts at quantitative evaluation. There are certainly many other types ofreactions described in the Iiterature 391 A. ZWEIG ~R' c-c oc=c~ -o~ hv ~~ ~ hv ~ R ~ R X= S, NHorNR ~co; VoH hv -hvhv _h_v_ Figure 4. Photofluorescent cis-trans photoisomerism \ where stable fluorescent products are photogenerated. Examples can be found in most of the broad classes of transformations that have been established. Foraframe of reference, I've listed in Figure 5 some of these classes and have selected in each an example to illustrate the point. As in the case of cis-trans isomerisms, the fact that the product may be fluorescent while the precursor is not, has, except in some very specialized situations, not received more than perfunctory notice. One exception to this is the photodedimerization process to yield stable fluorescent monomers. Here Förster1 reported that anthracene is not photoproduced in the fluorescent state in greater than 392 PHOTOCHEMICAL GENERATION OF STABLE FLUGRESCENT COMPOUNDS Oxidative cyclization" au Dedimerization 11 Tautomerization 12 Rearrangement 13 Elimination 14 Oxidation 1 s Addition 16 Substitution 17 Cycloelimination 18 ~ 0 ~ H H ~CH, ~ H~ I N~ o- ~ c6Hs .yl ~ 0 0 ~' ~ NO:z + OH AcO H ~C6Hs V ö hv -o;- ~2oco OH ., cCo OH hv o6COCH3 co; hvjdye, 0 2 0 hv hv, CNFigure 5. Photochemical reactions generating stable fluorescers 393 A. ZWEIG one-hundredth of one per cent. More recently10, he has reported that di-9-methylanthracene directly photoproduces 9-methylanthracene emission with a fluorescence quantum yield of 0.04 ± 0.005 per cent. Our own awareness of photochemically generated stable fluorescers resulted from studies of photochemical rearrangements of 3-aroyl-2-(2-furyl) chromones such as the 3-p-anisoyl compound shown in Figure 619. We 0 0 0 295nm 1.5 ~ 0.20 + coz + co 254nm C6H 5CH= CHC6H5 cp ~ 0.20 + C02 +CO 254nm C6H 5N = NC6H5 4> -0.1." + C02 +CO A~?: 234nm (e = 15900): 266s (1190); 271s {670) C6Hso H,cft /N-i HsC6 0 254 nm (C6H 5 hC =NC6H 5 4> ~ 0.41 + coz + co ).~~ic12 258s ( 1280): 263 (1 040); 270 (636) Figure 12. Photoelimination from aryl st11.:dnic anhydrides and aza derivatives 399 A. ZWEIG sensitized with benzene but not with acetone and could not be quenched by compounds with low triplet states, such as perylene (ET = 36 kcal/mole)30 and thus the evidence indicates that they proceed from an excited singlet state. Insight into the stereochemistry of the photoelimination of diphenyldi