J“«-. < --l. n-u‘ 11"_“.' 3'v59i13:<-;s_ ~ ,. , .- "‘- - ‘ H SPECTROSCOPIC ELUCIDATION OF THE NATURE OF THE TRIPLET STATES OF PHENYL ALKYL KET ONES Dissertation for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY MARY LOUISE MAY 1977 LIBRARY 1. -- Midway: Ste to L} ‘9', u I": H 2}, £51? I f ) .09.- K}. This is to certify that the thesis entitled SPECTROSCOPIC ELUCIDATION OF THE NATURE OF THE TRIPLET STATES OF PHENYL ALKYL KETONES presented by Mary Louise May has been accepted towards fulfillment of the requirements for Ph.D. Chemistry degree in @7‘” Major professor Date August 10, 1977 0-7 639 ABSTRACT SPECTROSCOPIC ELUCIDATION OF THE NATURE OF THE TRIPLET STATES OF PHENYL ALKYL KETONES BY Mary Louise May Phosphorescence and electron paramagnetic resonance spectroscopy were employed to study substituent, solvent, and temperature effects on the spectra and lifetimes of a wide variety of phenyl alkyl ketones. In order to evaluate these effects on the n,w* triplet without interference from a low-lying n,n* triplet, a study of substituent and solvent effects on the spectra of a series of similarly substituted benzonitriles was also undertaken. It is shown that the two component phosphorescence lifetime of many phenyl alkyl ketones should not be attri- buted to independent phosphorescence from the two close lying triplets since they will exist in thermal equili- brium and exhibit only a single intermediate lifetime. At 77°K ketones such as butyrophenone exhibit a higher energy spectrum in methylcyclohexane than in isopentane. It is suggested that the first formed excited state is Mary Louise May rigidly held by the viscous methylcyclohexane but is allowed to relax in the more fluid isopentane to a lower energy conformation. The phosphorescence spectra at 77°K and 300°K of aceto- phenone were obtained in polymethylmethacrylate. Some high energy emission which was observed only at 300°K is attri- buted to delayed fluorescence. The large splitting which was observed in the EPR AM = 2 signal of p-fluorobenzonitrile and the line broadenings which were found for the ortho and meta compounds are used to calculate the spin densities at the various carbons. SPECTROSCOPIC ELUCIDATION OF THE NATURE OF THE TRIPLET STATES OF PHENYL ALKYL KETONES BY Mary Louise May A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1977 To my parents for their inspiration and encouragement ii ‘~. 07057 ACKNOWLEDGEMENTS The author wishes to extend her sincere appreciation to Professor Peter J. Wagner for his patient and inspiring guidance throughout the course of this endeavor. The author also wishes to thank the Department of Chemistry and the National Science Foundation for providing the research assistantships administered by Professor Wagner. iii LIST OF TABLE OF CONTENTS TABLES O O O O O O O O O O C O O I 0 LIST OF FIGURES . . . . . . . . . . . . . . INTRODUCTION . . . . . . . . . . . . . . . . A. B. C. D. E. RESULTS A. The Triplet States of Phenyl Alkyl Ketones Phosphorescence Spectrosc0py . . . . Electron Paramagnetic Resonance . . . Research Objectives . . . . . . . . . Final Remarks . . . . . . . . . . . . Spectrosc0py of Phenyl Alkyl Ketones l. Solvent and Substituent Dependent Phosphorescence . . . . . . . . . 2. Temperature Dependent Phosphorescencw 3. Ketone EPR . . . . . . . . . . . Spectrosc0py of Benzonitriles . . . . 1. Nitrile Phosphorescence . . . . . 2. Nitrile EPR . O O O O O O O 0 O 0 DISCUSSION 0 O O O O O O O O O O O O O I O O A. B. C. Dual Phosphorescence of Phenyl Alkyl Ketones Temperature Dependent Phosphorescence . . . . Nitrile SpectrOSCOpy . . . . . . . . iv vi vii 01 re H 12 13 14 14 2.1- 21 26 26 26 41 41 51 52 EXPERIMENTAL O O O O O O O I I O O O O O O O O O A. Chemicals . . . . . . . . . . . . . . . . 1. 2. 3. 4. Ketones . . . . . . . . . . . . . . . Nitri les 0 O 0 0 O O O O O O O O O O SOlventS O O O O O O O O O O O O O O Methylmethacrylate . . . . . . . . . B. MethOds I O O C I O I O O O O O O O I O O 1. 2. 3. Phosphorescence Spectra and Lifetimes EPR Spectra . . . . . . . . . . . . . Preparation of Polymethylmethacrylate Samples . . . . . . . . . . . . . . . LIST OF REFERENCES 0 O O O O O O O O I O O O O C 61 61 61 63 64 64 65 65 66 69 71 Table LIST OF TABLES PHOSPHORESCENCE OF SOME PHENYL KETONES . . . TRIPLET STATE EPR OF SOME VALEROPHENONES TRIPLET STATE ENERGIES OF SOME BENZONITRILES TRIPLET STATE EPR OF SOME BENZONITRILES vi 15 26 27 31 Figure l. 2. 100 11. LIST OF FIGURES Modified Jablonski Diagram for phenyl ketones . An idealized EPR spectrum of a randomly oriented triplet . . . . . . . . . . . . . . . Phosphorescence of butyrophenone at 77°K in various mixtures of isopentane and methylcy- clohexane . . . . . . . . . . . . . . . . . . . Phosphorescence of valerophenone at 77°K in methylcyclohexane and isopentane . . . . . . . Phosphorescence of nonanophenone at 77°K in methylcyclohexane and isopentane . . . . . . . Phosphorescence of acetophenone at 77°K in methylcyclohexane and isopentane . . . . . . . Phosphorescence of butyrophenone at 77°K in methylcyclohexane at increasing concentrations Phosphorescence of acetophenone in poly- methylmethacrylate at 77°K and 900°K . . . . . Phosphorescence of p-methylacetophenone in polymethylmethacrylate at 77°K and 300°K . . . Phosphorescence of p-methoxyacetophenone in polymethylmethacrylate at 77°K and 300°K . . . Phosphorescence spectra of benzonitrile and the dicyanobenzenes in ethanol at 77°K . . . . . . vii 11 17 18 19 20 22 23 24 25 28 Figure 12. 13. 14. 15. 16. 17. 18. 19. Phosphorescence spectra of the fluorobenzo- viii nitriles in ethanol at 77°K . EPR spectra of benzonitrile in ethanol at 77°K EPR spectra of 1,4-dicyanobenzene in ethanol at 77°K . . . . . EPR spectra of p-fluorobenzonitrile in ethanol at 77°K . . . . . EPR spectra of o-fluorobenzonitrile in ethanol at 77°K . . . . . EPR spectra of m-fluorobenzonitrile in ethanol at 77°K . . . . . D vs. D vs. ET for benzonitriles o for benzonitriles O 29 32 35 38 39 40 54 INTRODUCTION The spectrosc0py of phenyl alkyl ketones is a field which has aroused much interest and controversy primarily because the two lowest triplets of these compounds, one mostly n,n* and the other mostly n,n* lie very close together. The relative energies of these states are quite solvent and substituent dependent and it is no easy matter to decide which state or "mixed" state is responsible for photoreactivity or phosphorescence. A. The Triplet States of Phenyl Alkyl Ketones An understanding of the origin of the triplet states of phenyl'ketones can best be achieved by referring to a modified Jablonski diagram.1 In this diagram (Figure 1) absorption and emission of light are shown by straight vertical lines, radiationless transitions are shown by wavy vertical lines, and intersystem crossing (ISC) and internal conversion (IC) are shown by wavy horizontal lines. Since phenyl ketones have a singlet ground state, the absorption of a photon results in an excited singlet. Any vibrational or rotational energy which the molecule might possess above the first excited singlet is rapidly lost (1012 sec-l) fl through vibrational and rotational relaxation.‘ Fluorescence Figure l. O u C O U a O I. 0 Ground State Modified Jablonski Diagram for phenyl ketones. Spin orbit coupling permits this lowest singlet, which is an n,n* singlet, to cross over to the triplet manifold. The extent of spin orbit coupling, and hence the ease with which intersystem crossing is accomplished, is inversely related to the energetic separation of the singlet and the triplet.3 It is also enhanced if one of the electrons involved in the transition is close to a heavy atom. Perhaps the most important consideration, how- ever, is the type of triplet state to which the crossing is occurring. Crossing from a n,n* singlet is estimated to be about 103 times faster to a n,n* triplet than to a n,nf triplet.4'5 However, if the two triplets are close together, then they are probably vibronically mixed6 and the situation becomes more complicated. It might be well, at this point, to look a little more carefully at what is known about these two close-lying triplet states. The 3n,n* state is formally regarded as arising from the promotion of an oxygen non-binding, n, electron into a n antibonding, n*, orbital. 0.0 .0 II II* C-——-R -————9 C -— R Ground State 3n,w* The net shift in electron density away from the oxygen leaves an electron deficient oxygen which has frequently been compared to an oxy radical. For example, benzophenone, which definitely has a lowest n,u* triplet, has been found to behave in a qualitatively similar fashion to the t—butoxy radica1.7'8'9 The n,n* triplet, on the other hand, is formally regarded as arising from the promotion of a n electron into a n antibonding, w*, orbital. Ground State ?' 9 g—R Lamolalo has shown that the n,n* triplet of phenyl alkyl ketones corresponds to the lowest triplet state of benzene (BLa)‘ Murrell11 views it as being basically an excited benzene ring with the carbonyl providing a strong perturba- tion. Nagakura and Tonaka12 view it as a charge transfer state. In any case, the n,w* triplet is very different from the n,w* triplet in that unlike the n,n* triplet, it has most of its excitation localized in the aromatic ring and it also has an electron rich carbonyl oxygen.13 In spite of these differences, it is still frequently difficult to distinguish one from another. While it can be said that an unsubstituted phenyl alkyl ketone in a non- polar solvent possesses a lowest 3n,w* state and that such things as electron donating ring substituents and polar 3 solvents will lower the n,n* state relative to the 3n,n* 10,14,15 state, it is frequently not at all obvious which state is actually lowest. The close proximity in which they lie leads to the possibilities of vibronic coupling16 and also thermal equilibrium. B. Phosphorescence Spectroscopy Phosphorescence emission spectroscopy can provide information concerning the lifetime, energy, and electronic configuration of the triplet state. It is usually presented as a plot‘of relative emission intensity versus frequency or' wavelength. In lifetime determinations, the emission intens- ity at a particular frequency (usually the O - 0 band) is plotted against time. The Franck-Condon principle tells us that electronic transitions are so fast (10".15 12 sec) in comparison to molecu- lar motion (10- sec) that the position of the nuclei are virtually unchanged during the transition.3 For this reason, a phosphorescence spectrum frequently displays a vibrational frequency progression which is a reflection of the various vibrational levels of the ground state to which phosphorescence occurs. The presence of the carbonyl vibrational progression reflected in a series of peaks with 1650 cm-1 spacing is a strong indication that the emitting species is a n,n* triplet. The vibrational band structure, which is sharp in a non-polar solvent, becomes quite broad in a polar solvent. It has long been accepted that a polar solvent or an electron donating substituent will shift a 3n,n* state to shorter wavelengths and a 3n,n* state to longer wave- lengths (and to a greater extent).17'18'44 19 _Arnold has demonstrated that for benzophenone, which has a lowest 3n,n* quite far removed from the 3n,n* level, there exists a good correlation between the triplet energy and the common Hammett 0 values of substituents. Presumably the n,n* triplet levels of phenyl alkyl ketones are similarly affected, but this is difficult to ascertain directly since most electron donating groups result in the 3n,n* state becoming nearly isoenergetic with or lower in energy than, the 3n,n*. Hochstrasser and Marzzacco16 believe that whenever the 3n,n* and %nm* states are in close proximity, vibronic coupling will result in a "mixed" state. They represent this "mixed" state as being a linear combination of the two unperturbed states. ‘P = alb<fl.1r*)+b\b(n.n*) T1 tw mat par ing fir the] The mixing coefficients a and b are dependent on the energy gap between the two "pure" states and also upon the ability of vibrations of the proper symmetry to mix. No qdantitative evaluation of these coefficients has been tried, but several workers have invoked the theory to explain the emission spectra which they have observed.10'15’20'21'22’23'24 The n,n* triplet also derives n,n* character from spin orbit coupling; however, the amount would be very small. The measurement of phosphrescence lifetimes is another standard procedure which is used to distinguish the 3n,n* from the 3n,n* state in phenyl ketones. The n,n* triplet is considered to be short lived (T 0.1 sec).25’26 If the triplets are fairly well separated, this distinction will hold. However, when the states are close lying, intermediate and two component lifetimes are frequently observed.10’l4'21'25'27 C. Electron Paramagnetic Resonance Since the triplet state, by definition, possesses two unpaired electrons and is therefore paramagnetic, its magnetic properties can be profitably studied by electron, paramagnetic resonance spectroscopy (EPR). Briefly speak- ing, the EPR experiment for triplet states consists of first obtaining the triplets by excitation with light and then subjecting them to an increasing magnetic field and and observing the absorption of electromagnetic energy when the resonance condition is met. To a first approximation, the change in energy levels is given by: AE(cm-1) = hv - gBH where g is the Lande g-factor, B is the Bohr magneton, and H is the magnitude of the magnetic field. Forbidden transiti Magnitude of applied radiofrequency Only one absorption would be observed in this simplified scheme since the two allowed transitions for which AMS = 11 are of equal energy and the transitions for which AMs = :2 are forbidden. The introduction of the dipole-dipole interaction of the spin magnetic moments removes this degeneracy. The resonance conditions are now depicted by: Hr" In addition to the two nondegenerate Am = 1 transitions, the Am = 2 transition, which had been previously magnetic dipole forbidden, is now observed. The constants D and E are referred to as the Zero Field Splitting (ZFS) parameters. The name arises because even at zero field, there still ex- ist three different magnetic energy levels. The difference between these three levels depends only upon D and E. The parameter D is a measure of the average dipolar interaction between the spins when the magnetic field is perpendicular to the plane of the molecule. D can be either a positive or negative quantity but its magnitude is prOpor- tional to the inverse third power of the average distance 10 It is easy to under- between the two unpaired electrons. stand why this is such a valuable parameter to be able to calculate. Since the unpaired electrons in a 3n,n* are I O 3 I I much more delocalized than in a n,n*, it is to be expected 10 that molecules whose phosphorescent state is 3n,n* will have much smaller D values than ones whose phosphorescent state is 3n,n*. The parameter E gives a measure of the distinction between the X and Y axes of the molecule. If these two directions are equivalent, then B = 0. E measures the devi- ation of the molecule from threefold or higher axial sym- metry. The preceding discussion is based on the assumption that all the molecules are oriented in the same direction with regard to the static magnetic field. It is possible, however, to observe the EPR transitions even when the molecules are randomly oriented in a frozen solution. The, transition between the nonadjacent energy levels (Am = 2) is remarkably less anisotropic than many people expected and gives rise to a strong signal. The transitions between adjacent energy levels (Am = 1) are more anisotropic and give weaker signals. Usually these signals are observable for a n,w* triplet, but they are frequently not for a n,n* triplet. There are a total of six Am = l transitions--a pair for each of the three axes, x, y, and z, with which the field can be parallel. Of course, if E = 0, then two sets of these lines will coincide and only four Am = 1 lines will be observed. An idealized spectrum for a typical randomly-oriented triplet is shown in Figure 2. 11 .uwamwuu vousofiuo saaovsmu m we samuumnm mmm monwflwmvw d< .N muawwm l1 radical regio Tc en. ef ke 12 Proton hyperfine splitting is generally not observed in triplet EPR as the coupling constants are considerably smaller than the line-widths. .However, from 28 it is known work on free radicals containing fluorine, that the fluorine coupling constants are 2 to 3 times larger than the corresponding proton couplings. Since the fluorine has little effect on the overall spin density distribution of the molecule, the study of fluorinated molecules can lead to an estimate of spin distribution in 30 and Fisher58 have the unsubstituted molecule. Mispelter estimated the spin density of triplet biphenyl by studying the splitting of the Am = 2 and Am = 1 Z peaks of fluorine substituted biphenyl. D. Research Objectives Better understanding of the photochemical and photophysical processes of phenyl ketones can be achieved if we have a clearer understanding of their triplet states. To this end, phosphorescence and EPR spectroscopy were employed to study substituent, solvent, and temperature effects on the spectra and lifetimes of a wide variety of ketones. Additionally, in order to evaluate these effects on the n,n* triplet without interference from a low-lying n,n* triplet, a study of substituent and solvent effects on the phosphorescence and EPR spectra of a series of substituted benzonitriles was also under- taken. 13 E. Final Remarks This thesis has been a long time in preparation. The majority of the studies were conducted during 1969-1972 and the summers of 1973 and 1974. Preliminary results were published in 1970,49 1972,50 and 1976.85 The ideas, which we first expressed in 1970 and l972, of thermal equilibrium between triplet states and conformationally- different triplets are now fairly common and well accepted.78.79,81,82,33 RESULTS A. Spectrosc0py of Phenyl Alkyl Ketones 1. Solvent and Substituent Dependent Phosphorescence Phosphorescence spectra and lifetimes at 77°K in various solvents were obtained for many phenyl alkyl ketones of the general formula PhCOR as well as for several ring substituted ones. These results are summarized in Table 1. In many cases, the overall phosphrescence decay was non- exponential consisting of both a short- and a long-lived component. For the a,a - dimethyl substituted ketones, however, only a short-lived emission was observed. Their spectra resembled that of benZOphenone, having classic n,n* structure. Several of the straight chain ketones were observed to have a spectrum which depended strongly on the viscosity of the solvent. Spectra were obtained at 77°K in both isopentane (IP) and methylcyclohexane (MCH). These solvents vary over some six orders of magnitude in viscosity. Figure 3 shows a series of buterphenone spectra which were taken in various mixtures of isopentane and methylcyclohexane. The spectra are very ketone-like, consisting of three prominent bands with the 1650 cm"1 spacing characteristic of the carbonyl stretch. However, in IF, the spectra are at lower energy than in NC”. Three bands appearing at 14 15 Table 1 PHOSPHORESCENCE OF SOME PHENYL KETONES a . l Ketone Solvent ET (kcal) Tshort(msec) Tlong(mseq %Tlong PhCO(CHz)2CH3 1pc 71.8 9.6 270 3 PhCO(CH2)3CH3 Z-MPd 74.5 6.4 27 2 EtOH 74.3 34 260 6 EPTe 74.4 27 74 17 o-MeO-PhCO(CH2)3CH3 EtOH 72.2 122 400 42 m-MeO-PhCO(CH2)3CH3 EtOH ----g 230 700 50 p-MeO~-PhCO(CH2)3CH3 EtOH 70.1 208 520 68 m-Me-PhCO(CH2)3CH3 Eton 72.4 90 290 56 p-Me—PhCO(CH2)3CH3 EtOH 73.0 175 940 6 IP ---—g 2 11 l o-Cl-PhCO(CH2)3CH3 EtOH 72.8 32 4 IP ----g 6 28 37 m-Cl-PhCO(CHZ)3CH3 EtOH 73.3 41 175 8 p-Cl-PhCO(CH2)3CH3 BtOH 72.4 40 99 50 IP 70.6 5 75 7 o-CF3-PhCO(CH2)3CH3 LtOH 71.8 IP 71.3 9 107 4 m-CF3-PhCO(CH2)3CH3 EtOH 74.1 IP 71.7 3 37 2 p-CF3-PhCO(CH2)3CH3 EtOH 73.0 IP 71.0 S 165 14 o-F-PhCO(Cll2)3CH3 EtOH 74.5 13 ,200 2 IP 72.0 m-F-PhCO(CH2)3CU3 LtOH 74.1 1P 70.6 8 132 12 Table 1--continued 16 a . b t T T T Ketone Solven T (kcal) short(msec) long(msec) % long p-F-PhCO(CH2)3CH3 EtOH 75.3 I? 72.4 ..39 127 19 PhCOCMe2(CH2)2CH3 EPT '72.5 5.5 None PhCOCMe3 MCHf 72.0 . None EtOH 72.6 . None IP 71.9 . None PhCOPh MCH 68.6 . None EtOH 69.5 . None a - 0-0 band, t.2 kcal b - Percentage of total emission (at 0-0 band) c - Isopentane d - 2 - Methylpentane e - 5:5:2 ether:isopentane:triethylamine f - Methylcyclohexane g - Spectra too broad for accurate triplet determination 397, 425, and 457 nm can be seen to replace the bands which appeared at 384, 411, and 440 nm. Valerophenone and nonano- phenone were found to display a similar type of behavior (Figures 4 and 5) except that they have a stronger tendency to retain the high-energy profile even in neat isopentane. The spectrum of acetophenone (Figure 6) does not show any change of appearance when the solvent is changed from IP to MC” . 17 50:50 IPzMCH IP: MCH 425 ,l l 33'! 411 457 440 384 8020 90: 10 IP:MCH leMCH Figure 3. Phosphorescence of butyrophenone at 77°K in various mixtures of isopentane and methylcyclohexane. 18 .mawusmmomH vsm.oamxo£oaomoahnuma ea Mons um maosemmonmaw> mo museummuonemonm .c ouswam. _/ m_ _&U<< .E bwrxaxew M. e...“ , VJ h 21““ .mww We w M___ _~‘. A. r» 5.5. .. s... v _ _ 1 nm¢ ~ee uee ”an N—c “a”... - «we mm 1. «me 19 .osmudmaomH was osmxosoauhuamnuoa aw.Monn um osonmsmoqmsoa we museumouosnmonm .m muswwm m. can «an «we ~ee -e awe omm at .IU¢< coo 20 .wsmuameomfi was mamxmnoau%uamnums aw Mons um macadamoumum mo museuwuuonnmoam .o musmwm J m. ..IU¢< mwm a: man :0 a: a; 21 In another series of experiments, it was observed that increasing the ketone concentration in MCH has the same effect on the spectra as adding IP to the solvent system. This is exhibited in Figure 7 for butyrophenone. 2. Temperature Dependent Phosphorescence' The phosphorescence spectra of several ketones in polymethylmethacrylate were obtained at 77°K and at 300°K. Methylmethacrylate was first purified to remove inhibitors and then the chosen ketones-~acetophenone, p-methylaceto- phenone, and p-methoxyacetophenone--were introduced. The samples were degassed, sealed, and allowed to polymerize for three days at 40°C and one day at 70°C. The same samples were used to obtain spectra at both 77°K and 300°K. These are exhibited in Figures 8, 9, and 10. At 77°K, the spectra are generally sharp and strong. Acetophenone shows a triplet energy of 73.9 kcal; p-methyl and p-methoxyacetOphenone show energies of 72.8 and 71.9 kcal, respectively. At300°K, the spectra are less intense by an order of magnitude, but do show, in each case, some higher energy emission (~2 to 3 kcal higher) than was observed at 77°K. 3. Ketone EPR Valerophenone, o-, m-, and p-methoxyvalerophenone, and m- and p-methylvalerophenone EPR spectra were obtained at 77°K randomly oriented in ethanol. Only the m = 2 transition was observed and hence only D* could be measured. These values are given in Table 2. 22 .msowumnusoomou wswmmmuosw um mswxmnoaumu Hugues a“ Menu um ososose0HMusn mo muswummuonnmonm .n ouswwm :- mug—I. l ENL= .( New sen _ sm 5 saw A ace mNe mwe nun .Mooom use Mann um ouwahuomeumEthumsaaoe aw wsosonmoum-w mo muamomouonamonm .m muswwm v. 023 23 3e 8..” so: as s can 5m 3m e3 . 3e 3v 24 .xooom was Mosh um ouwamuomnumaameumsmaoe cw encamnmoumumm xoeem can can man So NJ» — ODI‘. 9‘ were we museummuonmmonm ¥o- a: So can man .a munwwm 25 .Meoom.osm.uonn um mumamuumnumaahsuoshaoa aw ososmnnoumomexonumeue mo ousmommuonmmoem goeom 8m v.0: «an mwe m3 .oa muswwm can 26 Table 2 TRIPLET STATE EPR OF SOME VALEROPHENONESa Ketone Hminlg) D* PhCO(CH2)3CH3 1423 .1347 o-MeO-PhCO(CH2)3CH3 1429 .1331 m-MeO-PhCO(CH2)3CH3 1392 .1429 p-MeO-PhCO(CH2)3CH3 1479 .1178 m-Me-PhCO(CI-12)3CH3 1420 .1365 p-Me-PhCO(CH2)3CH3 1450 .1269 8All spectra were obtained using ethanol as a solvent and a ketone concentration of 10'3 M. The microwave frequency was 9.239 GHz. B. Spectroscopy of Benzonitriles l. Nitrile Phosphorescence Phosphorescence spectra of benzonitrile and methyl, methoxy, fluoro, chloro, trifluoromethyl, and cyano substi- tuted benzonitriles were obtained at 77°K. Most spectra . were observed in ethanol, isopentane, methylcyclohexane, and heptane. The triplet energies are shown in Table 3. Some examples of typical spectra are shown in Figures 11 and 12. 2; Nitrile EPR Due to the longer lifetimes and stronger emissions of the nitriles, it was possible to observe both Am = 2 and" Am = 1 transitions for them. The values of X, Y, and z and 27 Table 3 TRIPLET STATE ENERGIESa of SOME BENZONITRILES Solvent Nitrile EtOH (Lit. Value) MCH IP Hept. o-Cl-PhCN 74;3 kcal 74.3 kcal 73.7 kcal 74.1 kcal erl-PhCN 74.9 74.9 74.5 74.5 p-Cl-PhCN 74.1 73.7 74.1 74.5 o-CF3-phCN 75.1 74.7 74.9 74.7 m-CF3-PhCN 76.5 76.3 76.1 i 76.3 p-CF3-PhCN 76.1 75.7 75.7 75.9 o-F—PhCN 77.1 (76.8c) 76.9 -—--b m-F-PhCN 76.3 (76.1°) 75.6 ----b p-F-PhCN 77.7 (77.86) 77.7 ----b d o-MeO—PhCN 74.9 (74'3 ) 73.9 73.9 74.3 (74.56) d (72.0 ) b b b m-MeO-PhCN 72.1 (75 0e) ---- ---- ---- (75.26) p-MeO-PhCN 75.5 (75.3e) 75.1 75.3 75.9 m—Me-PhCN 75.5 (75.38) 75.1 75.1 74.9 p—Me-PhCN 75.9 (75.8e) 75.7 75.5 76.3 o-CN-PhCN 72.9 (72.89) m-CN-PhCN 75.5 (75.49) p-CN-PhCN 70.9 (70.1e) PhCN 77.3 (77.09) I All energies determined at the 0-0 band and are accurate to 0.2 kcal. Spectra too broad for accurate triplet energy determination. - Lui and McGlynn, J. of Lumimasovnon, 2, 44‘) (1975). - Lui and McClynn, déupf Mnlvc. SQ££;J 55, 163 (1975). ° - w--m~ . o--.- o - Unpublished results communicated by Prof. D. Arnold. QCC’Q’ I 28 416 392 410 , 431 370 392 l 480 380 369} Benzonitrile - 1,2-Dicyanobenzene 404 I 434 ;475r 447 3R9 500 ‘ .212} 1,3-Dicyanobenzene .—___§29} 1,4-Dicyanobenzene Figure 11. PhoSphorescence spectra of benzonitrile and the dicyanobenzenes in ethanol at 77°K. 29 .Mo: um Hocmfim CH woafiuufisoncmpouosa 93 mo muuomem museumeuocemonm . NH enema m3 nuasousm non—03min 6.3.3 accuse pouosamia maauuaaonsoaouosamuo on am can Sm . e _ 8m mun can Z.” 23 .i. 63 can mac :3 u :0 as can 30 the zero field splitting parameters D and B were calculated for all the nitriles and are given in Table 4. Some repre- sentative spectra, those of benzonitrile and 1,4 dicyano- benzene, are shown in Figures 13 and 14. The peak locations are those obtained after all necessary corrections have been applied. p-Fluorobenzonitrile was found to display a split- ting of the Am = 2 transition while o- and m- fluorobenzo- nitrile showed a significant line broadening. These Am = 2 transitions are exhibited in Figures 15, 16, and 17. 31 Table 4 TRIPLET STATE EPR OF SOME BENZONITRILES Am = Am 1 Hmin D* X Y Z D E D* PhCN 1399 .1389 .0392 .0517 -.0908 .1363 .0062 .1367 o-CN-PhCN 1368 .1467 .0172 .0730 -.0901 .1351 .0279 .1435 erN-PhCN 1353 .1503 .0290 .0687 -.0959 .1438 .0204 .1481 p~CN-PhCN 1424 .1320 .0248 .0603 -.0846 .1269 .0177 .1306 o-CFBPhCN 1405 .1373 .0305 .0576 -.0902 .1353 .0135 .1373 erF3PhCN 1393 .1404 .0410 .0513 —.0925 .1387 .0056 .1390 p-CF3PhCN 1393 .1404 .0401 .0530 -.0927 .1390 .0065 .1395 o-MePhCN 1405 .1373 .0308 .0580 -.0887 .1331 .0136 .1352 mPMePhCN 1400 .1383 .0210 .0669 -.0870 .1350 .0230 .1364 p-MePhCN 1420 .1332 .0377 .0504 -.0872 .1308 .0063 .1313 o-MeOPhCN 1412 .1357 .0112 .0714 -.0826 .1239 .0301 .1344 m-MeOPhCN 1410 .1363 .0057 .0748 -.0805 .1207 .0346 .1348 p-MeOPhCN 1422 .1329 .0206 .0634 -.0840 .1260 .0214 .1313 o-CIPhCN 1362 .1482 .0215 .0715 ".0916 .1374 .0256 .1444 m-ClPhCN 1394 .1402 .0262 .0651 -.0913 .1370 .0195 .1411 p-ClPhCN 1410 .1363 .0400 .0483 -.0883 .1324 .0041 .1325 o-FPhCN 1394 .1402 .0354 .0561. -.0915 .1372 .0103 .1384 m-FPhCN 1397 .1394 .0298 .0600 -.0898 .1347 .0151 .1372 p-FPhCN 12:: :135: .1423 .0439 .0503 -.0941 .1412 .0032 .1413 32 .Monn um fiosm:um :4 oflfiuuwscnson we :uuocem up“ .mH onswwm. 1441—4144—4144ddqnd—1411—Jqqanfiaddlfidqiiinq.«fiiwddad—lqa‘add-«4H11ddfimzouuwz .oooc mwsmm snow comes washeem eases DFVPrFPFFPrFVPLPAbpbPl-h—DibbP-D~>b$hbb—bbrlblbbbhbb.—L[FpPFPPhppirhr‘bbrbILhhrbhbbrbrhhbbbhtrrbphbrbbrb 33 I I r' it *T"'*_Ij "Trf'*l""1""_r" I rF‘I‘ W '_l'_' 1* 1 Field Setting 2100C Scan Range 10000 Microwave Freq. 9.197GH2 Hz=l783G "dark" run 75..11.--.111111....a-..111..-L.111L1.11141411.11.11.11111.111...1.-..1...11. “Lulliilainulairfj vv'rvvr'v;rvrvvirv'rrvvrva—‘V—rvvvafwvrjlvv vtvvvv'vwvvtrvvvv'vvv Field Setting 2200C Scan Range 10000 Hy = 23166 Microwave Freq. 9.197GHz ~~§¢~¢ . vl'rfivrjVVVIYYVTY‘f‘VvaTrVV'eri‘v "Figure 13. (cont'd.) vv 34 Irvvv—rrvvv'yvvv'v—v—rvyvvvv‘IvawthVVIVVfi‘vtuifi‘vvvrvr‘rvvvhvvfwt vvvivvvvrvvwv'v VVIYfi‘VV'VV VI, Field Setting 40100 Scan Range lOOOG Microwave Freq. 9.199GHz H" = 38126 44y==403143 -.§ 0.! O 11--1-1L.1.LAL1.Alxl.--.L441-1_11JALAA_1.14_LAJLL---'..ALIAALAJ_LAAAILA1.1...-l.x..14_i4-la.a.LlithLhe. Figure 13. (cont'd.) 35 .wonm um Hosmsum ca ocmucmnoswkoflolq .H mo mnuumam mum («a musuwm adddldlfiqidddldq.«a:-<1414d<44q4«ji—i1Jdn—dq14l—1411—rF~PLrA~PlPPrbyb Fr+rerFlPhlP> 632 u 58: o uP-l-hh.bhbbrbhhnhr—lr*. P p P h b .J11dl—dl11ddilfidq—1-T1-u114411-4d14l1l41ql1414‘ b . Nmuoma.m .vmum o>maouuaz oooe mwsmm scum comes mewuuem ashes I F V D _L7L+PPL¥FL P PLlPhbhybbt-bhhlthLrPt-ih m..- 36 ‘Tfi ‘V ,1......,.'.'~...,.'.,.._,.,.' .w-v—I'w-v-v'r,rvVVIVVVY'Tfi-1+CWV-- I--- Y' I"' I -1 Field Setting 20100 Microwave Freq. 9.1970112 H =169BG -y / 2 0 314411111111111--- n “11.- -iia-11.11.1..HL---AI---.1-..11.111111-11.--.1.-..1... 1111111144111): 'Tfir'T""I""I—'_'fivr"'fi"""r'IUV'rfi'fVYV' 1I *Tififil'fififr'fi'Ififi'1"fi"j"U_Tr .Field Setting 26000 Scan Range 10000 IMicrowave Freq. 9.1960Hz W3 ‘1 . 'M l I I l I -1‘4LLJ - -ll-l-l_‘.--l--411---ALA--lAquA---L.AA_11 A; l l l J A .. Figure 14. (cont'd.) 37 Field Setting 36000 Scan Range 10000 Microwave Freq. 9.197GHz' 41-x].-..1-..-l...1l...11...11...-l.;-.l...111114Y-..111.111.-..1.1.-111.L1...11....11..114..-l....! . ”2:48046 Field Setting 44000 . Scan Range 10000 Microwave Freq. 9.1970Hz .’ 1 J. 1.- l . -1--.-J.- . .._’-.--l -.l l .9 Figure 14. (cont'd.) .Mofim um Hosmnum a“ mHHuuHsoncopouosamim mc.muuumem mam .mH ouswwm 38 Al—J41A1 ml. 414dd—7FbFrLbhbh-.brlPPrFFrprDFFWL D_ inrkoir 39 .Monn um Hocmnum cw mflwuuwscncmpouosHulo mo muuummw Mmm .mH muswwm 0.... JJ‘ddi—dd‘41-.Idddqdiqd‘l-ll—ld‘441‘4duddlld14191—q- PFDFFPFPP-PLrb—DFDPFPPPbbthbbPFPP—th iddifldid1ddqdfii—didddd4dl. o o 32 ..... es... IPFDPV—rbbthPP’Fi—D>>b udddqdddidqddjid m... diddduudl-‘411iqddd4-4I Nmuoma.m .veum o>waouuez oooq manna scum some” waeuuem weeds -.rhn—nbthpPPb—thbPthb—bhb>->l—rbbipbt. 40 .eoee he seaweee as eHHweHeoueehowoaaeis mo wueoeee mam .AH unease 4‘71‘1‘4d‘41-dfi4ddddnddd—dd-Jid.d.d1444:41-1d‘1‘4dl+~1didiflq1ddtd“dNId—‘flId-‘d‘dqd‘4d.dd1dd.1‘4‘1‘11q1 at Q... . o can u 55: emcee." . m . dosh 96333: oooe omens atom wowed weauuum usage -P>PrprbhbhbDrbt—erFP-DFFD—y Fb-FPDb-VbPFPPLFP-PhbPFFDVbhrtPDF'bFFPP—FPFF-PPhD-PbPP—FFFLpr-FFPPrh DISCUSSION A. Dual Phosphorescence of Phenyl Alkyl Ketones The observation of dual phosphorescence in aromatic carbonyl compounds has generated considerable discussion 10,14,21,27,33 Yang and Murov27 in the recent literature. were among the first to attempt an explanation. They observed that l—indanone exhibited two phosphorescences of different lifetimes and attributed the shorter-lived phosphorescence to the n,n* triplet state and the longer- 1ived emission to a mixed n,n* and n.n* state. Several other authorslo'm'27'33 have since reported that in moder— ately polar solvents, the overall phosphorescence decay of many pheny alkyl ketones is non~exponentia1, consisting of both a long—lived and a shortnlived component. These reports similarly suggest that the short-lived component is the 3n,n* state, while the long-lived component is a state of mostly 3n.n* character. It is generally accepted that the Lawn,n*, triplet of these ketones normally lies only a few hundred recipro- cal centimeters above the lowest, mmi, triplet in nonpolar 10 10,14,21 and that a polar medium or the addition of 20,32 media electron-donating substituents can invert the order- ing. We would like to point out that the close proximity of these two states makes thermal equilibrium between them 41 42 an important consideration. A non-equilibrium distribution would not be expected to occur unless internal conversion between the states was slower than phosphorescence which seems highly improbable. As an example, even if the energy difference between the two lowest triplets was 250 cm-1, 1 percent of the triplets would still exist in the upper state at 77°K. This can be illustrated as follows TZ/Tl = e-AE/RT = e-700/1.99-77 .0092 where'T1 is the fraction of triplets in the lowest state, T2 is the fraction of triplets in the next state, E is the energy gap between the two triplets (i.e., 250 cm"1 or 700 cal), R is 1.99 cal/mole °C, and T is measured in degrees Kelvin. If the energy gap is only 350 cal instead of 700 cal, then ~10 percent of the triplets would exist in the upper state. Such an equilibrium situation would not affect phosphorescence when the 3n,0* state is lower but it would when the 3n,n* state is lower because of the much faster 3 emission rate of the n,n* state. Consider the following scheme: 43 -1 10% 3n,n* =322=2§g===0 m80% n,n* emission 1L 3 5 sec”1 90% n,n* > N20% n,n* emission In this instance, about 80 percent of the phosphorescence would come from the upper 3n,n*. The observed lifetime in an equilibrium situation such as this would be intermediate between those of the two triplet states. 1/1 = T k + T k We believe that the spectra which we observed for straight chain phenyl alkyl ketones as well as those observed by others can best be interpreted as phosphorescence from an equilibrium mixture of both triplets. Several obser- vations seem to support this. There is no significant polar solvent effect on the emission lifetime of benzophenone, whose lowest triplet is clearly n,n*.35 For valerophenone, and many of the substituted ketones which we studied, there is a substantial effect, such that the faster-emitting component in ethanol is too slow to originate solely from an n,n* state. Such is also the case for some of the ketones in isopentane. Intermediate phosphorescence life— times for p-chlorophenyl ketones have been reported by other workers also.36’37 Additional evidence for an equilibrium assignment is provided by the heavy atom effect which Kearns20 observed 44 on the lowest 8 - T* absorption of several aromatic ketones. No such effect would be expected if the emission came solely from the 3n,n*. The n,n* polarization of the highest 36 and the overall 20,36 energy (supposed 0-0) phosphorescence band n,n*-like appearance of the phosphorescence spectrum further substantiate our beliefs. Equilibrium mixtures of n,n* and n,n* triplets in the phosphorescence of a compound containing two separate, nonconjugated chromophores with similar triplet excitation energies have been observed by Lamola.38 If the shorter-lived component from the halo ketones and from valerophenone represents an equilibrium mixture of n,n* and n,n* triplets, it is still necessary to address the question as to the source of the longer-lived component, not only in these ketones but in all cases. It seems cer- tain that the long-lived emission does not come directly from the n,n* triplet.34 Not only is it awkward to postulate that some of the n,n* triplets equilibrate with the n,n* triplets while others do not, but it is very difficult to rationalize the lower energy state affording the minor emission component. Lim and coworkers34 offered one possible explanation. They observed the same two component phosphorescence from l-indanone as originally reported by. Yang and Murov.27 However, they found that 2,2-dimethyl-l- indanone, which posseses no protons a to the carbonyl, displays only short-lived emission and suggested that the long-lived emission from indanone arose from an excited 45 enolate ion. In agreement with this, all of the various straight-chain phenyl alkyl ketones which we examined dis- play both short- and long-lived phosphorescence. The a,o-dimethy1 substituted ones display only short—lived emission. However, Chu and Kearns80 have now shown that the long-lived emission observed from l-indanone is actually due to photoproducts which are formed at room temperature and then observed in phosphorescence at 77°K. So it would seem that the enolate ion is not necessarily responsible for the long-lived emission. This is not totally inconsis— tent with the absence of long-lived emission in the a,a— dimethyl substituted compounds. The dimethyl substitution lowers the triplet n+n* transition energy by 1-2 kcal but does not affect Amax for the 1A+1La transition, so that the triplet n+n* transition energy presumably remains constant at 75-76 kcal.10 Moreover, results with pyridyl ketones39 agree with those of Yang with trifluoromethylphenyl ketones;4o neither produces any long-lived emission in hydrocarbon glasses. In both kinds of ketones, the 3n,n* state is stabilized relative to the 3“,N* state. Conse- quently, the amount of long-lived emission correlates better 3 with the n,n* — 3n,n* energy separation than it does with the enolizability of the a protons. It would seem, then, that the observation of a long-lived component, even when the 3n,n* state is lowest, is diagnostic of the proximity of a 3n,n* state. We are still unable to say, however, just exactly what is responsible for this emission. 46 We do feel, however, that we can explain the viscosity-dependent dual phosphorescence of certain ketones. Phosphorescence at 77°K of ketones of the structure CGHSCOCHZR where R = ethyl, propyl, or n-heptyl was found to consist of two components in either methylcyclohexane (MCH) or isopentane (IP), with the minor component (T = 20-100 msec) amounting to only NS percent of the total emission (measured at the 0-0 band). This minor component is the same long-lived component which was discussed in the previous section. The lifetime of the major component varies from 4-5 msec-in IP, and from 5-10 msec in MCH. In both solvent systems, quite apart from the common problem of minor, long-lived emission, the major emitter is clearly a state of predominately n,n* character. It would appear that there are two possible conformations of the n,n* triplet, the lower-energy one (i.e., that in IP) being slightly shorter lived. In degassed benzene solutions at 25°C, both propio— phenone (R = methyl), and y,Y-dimethylvalerophenone (R = neopentyl) display phosphorescence of comparable intensity (¢ % 3 x 10-4)'and form to that already reported for benzo- 41'42 The 0-0 bands occur at 397 phenone and acetophenone. nm, the same as observed in isopentane at 77°K. These results suggest that solvent viscosity deter- mines the conformation of the emitting n,n* triplet. The butyrophenone spectra which were measured in several IP-MCH mixtures varying over some six orders of magnitude in 47 31 showed, in a 50:50 mixture, emission occurring viscosity primarily from a higher energy conformation with just a trace of lower energy spectrum. In a 70:30 IP:MCH mixture, emission came primarily from the lower energy conformation; and in 90 percent IP only a trace of the higher energy 0—0 band was detectable. It was also found that valerophenone and nonano- phenone display a similar type of behavior, the only difference being an increased tendency to remain in the high-energy conformation, i.e., the spectra in neat IP resemble butyrophenone in 80:20 IP:MCH. Lim and coworkers43 have found a similar mixed emission for propiophenone and butyrophenone in 3-methylpentane (3MP) glass. However, neither acetOphenone nor benzophenone display any differ- ‘ence in the energies of their n,n* phosphorescence in MCH compared to 1?. Consequently, the two-component n,n* emission observed when R is larger than hydrogen most likely does not involve salvation differences intrinsic to the carbonyl system. These results are very similar to the solvent~ viscosity-dependent dual phosphorescence reported for p-dimethoxybenzene44 and a similar interpretation seems appropriate.' These solvent effects probably arise from a slight difference between rotationally relaxed levels of ground and n,n* state. The lower-energy spectrum, which occurs both in solution and in the relatively-fluid isopentane at 77°K, presumably arises from the 48 conformationally-relaxed n,n* state. The higher-energy emission probably arises from an n,n* state held rigidly (by solvent) in the most favorable ground state conformation. Since the difference between 0-0 bands of the two spectra is 870 cm-1, part of which must reflect the energy difference between the two ground state conformations, the energy difference between the two excited conformations can only be estimated and is probably about half (1.3 kcal/mole). An appealing stereochemical interpretation of these results, based on spectrosc0pica11y determined conforma- tional preferences of ground state ketones, is as follows: I II II* III* In aldehydes and methyl ketones, the preferred ground state conformation is I (looking down the axis of the C-C bond between carbonyl and a carbon), with R eclipsing the 45 carbonyl group. In phenyl ketones, however, II is more. likely preferred. The carbonyl is not exactly coplanar 46 such that there would be with the benzene ring, serious nonbounded interactions between an ortho-hydrogen and an a-hydrogen in conformation I. In the excited state, 49 the benzoyl chromophore probably attains internal coplanar- ity47 (thus accounting for the N700 cm-1 Stokes shift observed for S ~+T transitions of phenyl ketones)20 so that 111* (equivalent to I) would be the favored conformation. That 11* and 111* should differ by approximately 1.3 kcal in the extent of nonbonded interactions between ortho- and o-hydrogens is a quite reasonable conclusion. These results suggest that, whenever R is larger than hydroqen, the rota- tion from first-formed II* to 111* is impeded if the solvent is rigid enough. It is not possible, of course, to distinguish between a pure viscosity effect and an effect of the solvent size. That the latter may be involved is suggested by our observation of only higher-energy spectrum in MCH glass and Lim's observation of an equal mixture of both spectra in the equally rigid 3MP glass.43 The observation that an increase in butyrophenone 2 to 10-1) is found to have the same concentration (from 10- effect on the spectra as an increase in the amount of IP is consistent with the preceding discussion. Increasing the ketone concentration will decrease the rigidity of the MCH glass and thereby favor the more relaxed conformation. It is also possible that at high concentration the ketone crystallizes out at 77°K, and that it is the spectrum of the unsolvated ketone which is being observed. It must be emphasized that all of these results are not related to the first mentioned problem of determining 50 the source of the small amount of long-lived (> 20 msec) phosphorescence which occurs in most phenyl alkyl ketones regardless of the conformation or energy of the emitting n,n* triplet. However, these results do bear on another problem. There is much interest in correlating the room temperature, solution-phase chemical behavior of excited ketones with their 77°K rigid-matrix phosphorescence. These results exemplify one real danger in such attempts, especially when small energy differences between excited states are important.40 On the plus side, the finding of identical energy spectra in IP at 77°K and in benzene at 300°K offers the hope that wider use of isopentane as a solvent for liquid nitrogen spectroscopy might allow more exact correlations between 77°K and 300°K spectroscopy. 83 have studied xanthone Pownall, Connors and Huber in 3-methylpentane and reached a conclusion similar to ours. They find that the dual phosphorescence of xanthone is due to emission from two n,n* states and believe that the two triplet states, as well as their respective ground states, are distinguished by their different molecular geometries. 81 have also reached similar con- Long, Li, and Lim clusions to ours in their studies of propiOphenone and butyrophenone. Additionally, in the case of benzaldehyde, they find that dual emission can result from interactions of the ground state with the solvent matrices leading to the formation of two different solute-solvent cage configu- rations. In a later paper, Kanamaru, Long, and Lim82 51 further expand their ideas. They theorize that a conforma- tional change may be due to pseudo Jahn-Teller distortion of the lowest triplet state resulting from vibronic inter- actions between closely spaced n,w* and n,n* triplet states. A.matrix potential which hinders this conformational change can give rise to dual phosphorescence. B. Temperature Dependent Phosphorescence The concept of studying phosphorescence in lucite is an appealing one as the temperature can be varied over a wide range while the ketone remains in a rigid-solvent matrix uninfluenced by quenchers. In the studies which we per- formed, acetophenone, p-methyl-, and p-methoxyacetophenone all showed the expected68 decrease in phosphorescence intensity and spectral sharpness as the temperature was raised from 77°K to 300°K. What was most interesting was the appearance of some higher-energy emission at 300°K. The source of this emission was not ascertained but it probably is the result of delayed fluorescence or emission from an upper triplet. Saltiel, et a1.42 studied substituted benZOphenones in carbon tetrachloride and also observed a high energy band which increased in intensity with temperature. They attributed this band to delayed fluorescence and pointed out the great photochemical importance of having a significant number of 81 molecules in equilibrium with T1 molecules. 1 . . The T state may serve as a reserVOir of exc1ted molecules 52 whose photoreactions proceed from 81 states. Effective reaction rates then depend on the Boltzmann population of 81 and should be strongly temperature dependent. For molecules with small 81 - T1 energy gaps, quenching of a photoreaction by triplet excitation acceptors need not signify that the reactive state is the triplet state of the donor. C. Nitrile Spectroscopy The benzonitriles were studied in an effort to better understand the behavior of the w,n* triplet independent of any interaction with a low-lying n,n* triplet. It would appear, from an examination of Table 3, that the conjuga- tive effect of a substituent is more important than its inductive effect in influencing the triplet energy of the n,n* state. It is observed that both the electron-donors and elecaron—acceptors lower the energy of the n,u* triplet relative to the unsubstituted benzonitriles. Neither change from a polar (EtOH) to a non-polar (MCH) solvent nor change from a rigid (MCH) to a fluid (IP) solvent appears to have any significant effect. It is interesting to compare the EPR data obtained on Me and MeO substituted valerophenones with the corres- ponding substituted benzonitriles. Since the M = 2 transi- tion was observed for the ketones, it is only possible to calculate D* for them. If these values are compared with the corresponding D* values for the benzonitriles, a correlation is observed. 53 Nitrile Ketone unsub .1367 .1347 m-CH3 .1364 .1365 p-CHS .1313 .1269 o-OCH3 .1344 .1331 m-OCH3 .1348 .1429 p-OCH3 .1313 .1178 This is not surprising since the cyano and carboxy groups have a similar effect on the n,n* transition of benzene.51 It has been suggested by several authorsE-M'SS'SG'86 that for triplets there should be a correlation between the ZFS parameter D and the triplet energy E They reason Ti that the smaller the D value, i.e., the further the electrons can get from one another, the smaller the triplet energy. This assumes a greater influence due to electron delocaliza- tion on the excited state than on the ground state. The applicability of this theory to our work was tested by plotting D vs. ET for all of the benzonitriles which were studied (see Figure 18). There does appear to be a crude correlation which is better for some substituents-- -CN, -0Me, and -F—-than it is for others. It is difficult to interpret this plot any further. Smaller86 has suggested that there might also be a correlation between the Hammett a factor and D. Such a plot was tried and is displayed in Figure 19. Again, the correlation is minimal at best. The lack of a simple interpretation of these data probably attests to the importance of spin—orbit interaction in zero field splittings. ET (kcal) S4 “I? /Op-F / I OH / 77 __ O’O-F / II / CHI-CF; om-F / 76 _ /O p-CF, e I 00-Me PlMo / rp-MeO O‘Fn—Me // O m-CN / / I 75_ I eo-CF, O o-MeO om-Cl / I l, I l I / / I / .O-CI / 74 )- l . p'CI / / / / / / 73 l/ / e 0 -CN I // I I O m-MeO / 72 "' / / -/ l / / 71' 06 -CN .120 .125 .130 .135 .150 .155 1:50 D Figure 18. D vs. E for Benzonitriles. T 55 .150 P I, em'CN I F I, .Ip I .140 ' 1 ’ I M'Cfae I ’P'CFs I // I m-Cle I I/' H [I I/ m-F ’ I l / I op-CI ” ’ / P'CH37’ .m-CH / .. 3 / .130 I I I II I, I, cp-CN p£Me0 // I I I .m-MeO .120 ~ 1 1 1 1 1 - 4 ' 2 0 .2 4 6 a- Figure 19. D vs. 0' for Benzonitriles. S6 Probably the most interesting piece of information to come out of the work on benzonitriles is the observation of fluorine hyperfine splitting. As can be seen by refer— ring to Figures 13, 15, 16, and 17, the AM = 2 signals for the fluorine-containing compounds are broader than that for benzonitrile itself (and those of many other substituted benzonitriles which were studied). With p—fluorobenzonitrile, the AM = 2 signal is a doublet with 54G splitting. Its Hx and By AM = l signals were observable, but the H2 signals were too weak to observe. Fluorine hyperfine splitting has previously been observed in the EPR spectra of triplet aromatic330’57'58 69 Fluorine and in PMDR spectra of phosphorescent ketones. substitution apparently does not significantly alter the electronic distribution of excited aromatics30 and there- fore the hyperfine splitting provides a simple monitor of electron density in triplets. There are two aspects of our results which seem important: the small apparent spin densities at the ortho and meta positions, and the actual size of the para.split- ting. In triplet naphthalene, the free spin density at. 70 each a carbon is 0.44 (total 0 = 2). The hyperfine splitting of the AM = 2 line for l-fluoronaphthalene has 57 30 been reported as 21G and 256. There is some disagree- ment over the quantitative dissection of the hyperfine 30,57 splitting of AM = 2 lines, so we shall simply apply 57 the empirical relationship3O aF = szfp: . Application of this approximate McConnell relationship yields a ngf value of W 57G, much the same as the value for the isotropic contribution to fluorine hyperfine splitting in a planar (spz) FCR2 radical.30'71 Furthermore, analysis of the SOG D splitting of the H2 AM = 1 transition in 4-fluoro- and 4,4'- difluorobiphenyl indicates a spin density of 0.52 at each para-carbon in triplet biphenyl (again on the basis of a 0 total spin density of 2).3 The corresponding splitting of the AM = 2 transition was 31G3O . F Wthh leads to a Qeff of 59G. If we take ngf = 58G, we deduce the following absolute values for the spin distribution in triplet benzo- nitrile. N N m 0.90 I" o. 37 ‘ C C 0.12 (L15 0.034- 0.013 0.93 0.36 Triplet Radical anion The ortho and meta values are maximum values calcu- lated by assuming that a equals the line broadening (rela- F 'tive to the width of Hm. for benzonitrile which is presumably 1n Eilready slightly broadened by its para-hydrogen). The amount 58 of spin on the C-CN unit was calculated on the assumption of negative spin density in the meta positions.72 The actual spin densities at the ortho and meta carbon in the triplet are similar to those reported for the radical anion73 but the para value is 2.5 times higher in the triplet. The behavior of radical anions has been ex- plained at least qualitatively by a simple MO picture whereby the electron-attracting cyano group destroys the degeneracy of the two lowest n* orbitals of benzene by preferentially stabilizing the symmetric n* orbital, which has most of its electron density at the l, 4 positions.74 Although a n,n* excitation can place an electron in the same orbital, it would be fortuitous for spin densities to, be the same in the triplet and in the radical anion, since the triplet has an extra unpaired electron in a n orbital. However, the fact that the para/«ortho + meta) spin density ratio in the triplet is almost three times what it is in the radical anion indicates that both electrons are in symmetric orbitals in the triplet. It is generally agreed that the lowest w,n* triplets of simple substituted benzenes have Platt's La configuration. It is well known that para substituents shift the La UV band of substituted benzenes to much lower energy than do ortho and meta substituents7S so our results fit in very nicely with the La assignment for the lowest triplet. The La n+n* transition of benzene is represented in MO terms by a combination of ns+ns* and 54,76 * C O n +n tran31tions. A A .1 3 1/1 1/12 + “8* 1/1 , ‘ 1/12 1/3 1/3 1/12 /12 . e +1. /12 /12 l l "1 1/3 1/4 1/4 1/4 If an electron-withdrawing substituent such as cyano stabilizes "5* and destabilizes "s' then the La state would be predominantly ws+ns* in character. That is, both the unpaired electrons are in orbitals with high electron density at the l and 4 carbons, as shown in the MO diagram. The most important valence bond structure for the 3La state is a l, 4 diradical, often referred to as form of the triplet. It is interesting to compare our results and conclu- sions with those recently published by Hirota and Nishimoto. 77 the "quinoidal" C:==N0 These authors studied the EPR spectrum of benzonitrile as guest in p-dibromobenzene single crystals. para—proton hyperfine splitting indicated a para-spin density The observed 77 60 of 0.68. The authors also concluded that the La triplet must be predominantly fls+fls*. Although no splittings due to ortho—protons were observable, slight line broadenings were interpreted to indicate an ortho-spin density as high as 0.30. This high value was used.tb support the sugges— tion72 that there should be similar spin distributions in the triplet and in the radical anion. The much higher sensitivity of fluorine as a probe indicates that the triplet spin distribution is significantly different from that of the radical anion. Although similar spin distri- butions in both radical cations and radical anions of 72 simple aromatic hydrocarbons are expected and commonly observed, such often is not the case in aromatics substi- tuted with strongly electron-donating or -accepting groups.7l Perhaps the following VB forms best describe the half- occupied n and m*orbitals in triplet benzonitrile. N N III III III C C (-----3 4- e 4. o n* n EXPERIMENTAL A. Chemicals 1. Ketones a. Acetophenone (Matheson Coleman-Bell) was distilled at reduced pressure by Dr. D. Ersfeld and a middle fraction taken for use. b. p-Methylacetophenone was purified by E. Harris.60 c. p-Methoxyacetophenone (Aldrich) was twice recrystallized from ligroin by Dr. H. Schott.61 d. Piva10phenone was prepared by adding 0.1 mole benzonitrile to 0.12 mole of t-butyl magnesium bromide in 200 ml of ether. The resulting solutions were refluxed one hour, cooled, and poured onto 3009 of ice to which 30 ml of concentrated HCl had been added. After all solids were dissolved, the two phases were separated and the cold aqueous phase was extracted with ether to remove any organic residues. The reaction mixture containing the imine hydrochloride was then warmed on'a steam bath for one hour, the crude product removed, and ether extractions added to it. After drying and removal of ether, the product was distilled at 45 mm and a middle fraction boil- ing at 132-1350C collected. e. Huterphonone (Aldrich) was purified by Dr. I. Kochevar.62 61 62 f. Valerophenone (Eastman) was purified by Dr. A. 63 Kemppainen. g. a,a-Dimethylvalerophenone was prepared by Dr. 64 'J. McGrath. h. o-Methoxyvalerophenone was prepared by Dr. A. 63 Kemppainen. i. m-Methoxyvalerophenone was prepared by Dr. A. 63 Kemppainen. j. p-Methoxyvalerthenone was prepared by Dr. A. Kemppainen.63 k. m-Methylvalergphenone was prepared by Dr. A. 63 Kemppainen. l. p-Methylvalerophenone (Pfaltz & Bauer) was puri- 63 fied by Dr. A. Kemppainen. m. o-Chlorovalerophenone was prepared by Dr. A. Kemppainen.63 n. m-Chlorovalerophenone was prepared by E. Harris.60 0. p-ChlorovalerOphenone (Columbia Organic Chemicals) 63 was purified by Dr. A. Kemppainen. p. o-Trifluoromethylvalerophenone was prepared by Dr. P. J. Wagner.63 q. m-Trifluoromethylvalerophenone was prepared by Dr. P. J. Wagner.63 r. p-Trifluoromethylvalerophenone was prepared by Dr. P. J. Wagner.63 63 s. o-Fluorovalerophenone was prepared by Dr. P. J. Wagner.63 t. m-Fluorovalerophenone was prepared by Dr. P. J. Wagner.63 u. p-Fluorovalerophenone (Pfaltz & Bauer) was puri- 63 fied by Dr. A. Kemppainen. v. Nonangphenone was prepared by Dr. A. Kemppainen.63 'w. Benzophenone (Eastman-White label) was purified 65 by Dr. B. Scheve. 2. Nitriles a. Benzonitrile (Eastman Organic Chemicals-aniline free) was used as received. b. o-Anisonitrile (Aldrich) was used as received. c. m-Anisonitrile (Aldrich) was used as received. d. p-Anisonitrile (Aldrich) was recrystallized. e. o-Tolunitrile (Aldrich) was used as received. f. m-Tolunitrile (Aldrich) was used as received. 9. p-Tolunitrile (Aldrich) was used as received. h. o-Chlorobenzonitrile (Aldrich) was recrystallized twice from petroleum ether. 1. m-Chlorobenzonitrile (Aldrich) was used as received. j. p-Chlorobenzonitrile (K.K. Labs) was used as received. k. o-Trifluoromethylbenzonitrile (Pierce Chemical Co.) was used as received. 64 1. m-Trifluoromethylbenzonitrile (Pierce Chemical Co.) was used as received. m. p:Trifluoromethylbenzonitrile (Pierce Chemical Co.) was used as received. n. o-Fluorobenzonitrile (Columbia Organic Chemical Co.) was used as received. 0. m-Fluorobenzonitrile (Columbia Organic Chemical Co.) was used as received. p. p—Fluorobenzonitrile (Columbia Organic Chemical Co.) was used as received. ‘q. o-Dicyanobenzene (Eastman) was recrystallized. r. m-Dicyanobenzene (Eastman) was recrystallized. ' s. p-Dicyanobenzene (Eastman) was recrystallized. 3. Solvents a. 'Ethanol.(Rossville-Gold Shield Alcohol) was used as received.‘ b. Isopentane (Matheson Coleman and Bell--spectral grade) was used as received. c. Methylcyclohexane (Eastman--spectral grade) was used as received. d. Heptane (J. T. Baker-~spectral grade) was used as received. 4. Methylmethacrylate (Aldrich) was purified according 6‘ Of the to the procedure of Unterleitner and Hormats. monomer, 125 ml. was washed twice each with 5 percent sodium nitrite, 5 percent sodium bisulfite, and 5 percent sodium hydroxide, and three times with water. After drying, it 65 was distilled under reduced pressure (aspirator). A center cut boiling at 68-70°C was collected and stored at 0-5°C until used. B. Methods 1. Phosphorescence Spectra and Lifetimes Two different phosphoroscopes were used in the course of these studies. a. All lifetime measurements and most of the phenyl ketone spectra were obtained on the equipment of Dr. A. Bang.67 The phosphoroscope was 20 cm. in diameter and, through reduction pulleys, could be operated at several speeds from about 30 to 1,775 rpm with less than 2 percent drift in frequency. The excitation and observation periods were of equal length and ranged from 0.9 sec. at the lowest speed to 15.9 msec. at the highest speed. For spectral measurements, the highest speed was used, whereas in the case of lifetime determinations, the speed was chosen so as to give excitation and observation intervals equal to more than the lifetime in all cases. The emission was focused onto the slit of a Jarrell-Ash Model 82-000, 0.5—m Ebert scanning spectrometer. This instrument was equipped with a grating blazed at 5,000 8 in first order, having a reciprocal linear dispersion of 16 g/mm.and effective aperture ratio of f/8.6. The slit- width was set at 0.4 mm. Light emerging from the exit slit of the spectrometer was collected on the cathode of a dry—ice cooled EMI 6256SA photomultiplier by means of appropriate 66 condensing optics. The photomultiplier output was fed into a Victoreen Model VTE-l electrometer where it was amplified and the spectrum recorded on a strip chart recorder. Peak locations could be determined with an accuracy of :1 nm. Lifetimes were determined at the 0-0 band with the help of a Nuclear Data Model ND180 multichannel analyzer. All samples were contained in sealed 3 mm i.d. quartz tubing. Excitation was filtered through pyrex. Under these conditions, the solvents neither cracked nor emitted with the exception of heptane which "snowed." In cases of two- component emissions, decays at different phosphoroscope speeds were analyzed. No ketone was found to possess more than two exponential components to its emission. Quantum yields were not measured, but were comparable for all ketones studied (> .05). b. All nitrile spectra and some ketone spectra were obtained using an Aminco Bowman Spectrophotometer equipped with the phosphoroscope accessory. Peak locations could be determined with an accuracy of :1 nm. On both spectrometers, calibration was checked by comparison with known spectra. 3. EPR Spectra Two different EPR spectrometers were used in the course of these studies. 67 a. Varian 4502-15 The valerOphenone and substituted valerophenone spectra were obtained at 77°K on a Varian X-band 4502-15 spectrometer equipped with an optical waveguide, a low— temperature accessory, a Mark II magnetic field regulator, and a 100 KHz modulation unit. Light was supplied by a focused Engelhard-Hanoveria lKW mercury-xenon high-pressure lamp fitted with a water filter. The samples were contained in 3 mm i.d. quartz tubing. Ethanol was the solvent of choice and the optimum ketone concentration was found to be 5 X 10-2 M. Only the Am = 2 line was observed in each case. A con- certed effort was made to locate the Am =,l lines but the signals were too weak. The position of the Am = 2 signal was taken to be half-way between the maximum and minimum points. A field correction of +4G, determined by calibration with a Varian F-8 proton resonance fluxmeter, was then applied to this reading. The frequency was monitored with a Hewlett-Packard Model 5245 frequency counter. b. Varian E-4 The benzonitrile spectra were obtained at 77°K on a Varian E-4 spectrometer equipped with a low-temperature accessory. Light was supplied by a focused Engelhard Hano- veria lKW mercury-xenon high-pressure lamp sometimes fitted with a quartz water filter. The samples were contained in 4 mm thin-walled quartz tubing. Ethanol was the solvent of choice. The optimum nitrile concentration was found to be 5 X 10"2 M. The Am r 2 lines were easily observed. Due to 68 the rapid decay in signal strength upon irradiation, it was only possible to obtain worthwhile spectra for m30 see. This made the Am = 1 lines somewhat difficult to locate. Warming the sample to room temperature and then reusing it was found sufficient to restore signal strength. A "dark" run was always obtained with no light shining on the sample so that the triplet signals could be located unambiguously. The location of Hx and'HY was taken to be at their maximum (or minimum). The location of Hmi and Hz was taken n to be half-way between the maximum and minimum points. At a sweep width of 1,000 G, each square of the standard chart paper was equivalent to 52.8 G. At a 400-G sweep width, each square was covered by 22.5 G. By comparison with a carefully calibrated spectrum of biphenyl, a +3 G field correction was applied to the 1,420 G field setting. No significant correc- tion was found at higher field settings. Individual condi— tions (microwave frequency, etc.) under which each signal was observed are given on the respective spectra, as illustrated in Figures 13 through 17. c. Calculation of ZFS Parameters The Am = 2 signal gives Hm. from which D* is in. calculated, using: 13* = Y3/452 — 3(gB)2 ~112 min The Am = l signals give 3nz, 2H , 2H , 3H , 3Hy’ and 2H2 y x x from which the following calculations are made: I‘ll-Illil‘vi!" lull-'11? fit: 0!. T Lu X+Y+Z 31.. 34H :24... 2 3 2 2 2 (98) ( Hz - Hz ) 2 3 2 2 2 H - H (98) ( y Y ) 2 3 2 2 2 (98) ( Hx - Hx ) 0 (a check also useffil for locating missing signals) The ZFS parameters are then given by: D E 0* In all the above calculations: h 9 B 6 l erg 3/2 Z 1/3 D - Y = X - 1/3 D VD? + 332 6.625 X 10'27 erg-sec. 2.0023 Mc/gauss 9.273 X 10-21 erg/gauss hv 5.0348 x 1015 cm“1 3. Preparation of Polymethylmethacrylate Samples Lucite samples were prepared by dissolving the ketone (5 X lO-3M) in purified methyl methacrylate. The solutions were pipeted into 4 mm thin-walled quartz tubing which had been sealed at one end. The samples were then degassed by three freeze/thaw cycles under vacuum and sealed. They were heated in an oil bath at 40°C for 3 days and then at 70°C for 1 day. age. Slow cooling in the oil bath helped to prevent break— Some bubbles were invariably present, but they did not 70 adversely affect the spectra which were taken. The samples were always rotated to give maximum emission intensity. A LIST OF REFERENCES LIST OF REFERENCES l. A. Jablonski, Z. Physik, 94, 38 (1935). 2. E. F. Ullman, Accts. Chem. Res., 4, 353 (1968). 3. N. J. Turro, "Molecular Photochemistry," W. A. Benjamin, Inc., New York, 1967, pp. 58-59. 4. M. A. El-Sayed, Accts. Chem. Res., 4, 8 (1962). 5. P. M. Rentzepis and C. J. Mitschele, Anal. Chem., 42, 20A (1970). 6. M. A. El-Sayed, Accts. Chem. Res., 4, 23 (1971). 7. A. Padwa, Tetrahedron Letters, 3465 (1974). 8. C. Walling and M. J. Gibian, J. Amer. Chem. Soc., 41, 3361 (1965). 9. P. J. Wagner, Accts. Chem. Res., 4, 168 (1971). 10. A. A. Lamola, J. Chem. Phys., 47, 4810 (1967). 11. J. N. Murrell, "The Theory of the Electronic Spectra of- Organic Molecules," Methuen and Co., Ltd., London, 1963, pp. 206 and 227. 12. S. Nagakura and J. Tanaka, J. Chem. Phys., 23, 236 (1954). 13. D. Bryce-Smith, Pure Appl. Chem.,}fi, 47 (1968). 14. R. D. Rauh and P. A. Leermakers, J. Amer. Chem. Soc., 22, 2246 (1968). 15. Y. H. Li and En C. Lim, Chem. Phys. Letters, 1, 15 (1970). 16. R. M. Hochstrasser and C. A. Marzzacco, in "Molecular Luminescence," E. C. Lim, Ed., W. A. Benjamin, New York, p. 631. 17. J. W. Diamn, Chem. Rev., 58, 689 (1958). 18. J. N. Pitts, Jr., H. W. Johnson, Jr., and T. Kuwana, J. Phys. Chem., 44, 2456 (1962). 71 72 19. D. R. Arnold, Adv. Photochem., 5, 301 (1968). 20. D. R. Kearns and W. A. Case, J. Amer. Chem. Soc., 55, 5087 (1966). 21. W. A. Case and D. R. Kearns, J. Chem. Phys., 52, 2175 (1970). “— 22. R. J. Zwarick and L. Goodman, Chem. Phys. Letters, 1, 609 (1970). 23. M. Koyanagi, R. Zwarick, and L. Goodman, ibid., 5, 73 (1971). 24. M. Koyanagi, L. Goodman, J. Chem. Phys., 55, 2959 (1971). 25. S. P. McGlynn, T. Azumi, and M. Kinoshita, "Molecular Spectrosc0py of the Triplet State," Prentice Hall, New Jersey, 1969, p. 40. 26. s. K. Lower and M. A. El-Sayed, Chem. Rev., 55, 199 (1969). 27. N. C. Yang and S. L. Murov, J. Chem.'Phys., 45, 4358 (1966). 28. M. Kaplan, J. R. Bolton, and G. K. Fraenkel, ibid., 43, 955 (1965). 29. J. C. M. Henning, ibid., 4, 2139 (1966). 30. J. Mispelter, J.-Ph. Grivet, and J.-M. Lhoste, Mel. Phys., 24, 1015 (1971). 31. J. R. Lombardi, J. W. Raymonda, and A. C. Albrecht, J. Chem. Phys., 44, 1148 (1964). 32. N. C. Yang, D. S. McClure, S. L. Murov, J. J. Houser, and R. Dusenbery, J. Amer. Chem. Soc., 55, 5466 (1967). 33. N. C. Yang and R. Dusenbery, Mol. Photochem., l, 159 (1969). 34. Y. Kanda, J. Stanislaus, and E. C. Lim, J. Amer. Chem. Soc., 55, 5085 (1969). 35. S. Dyn, R. M. Hochstrasser, and M. Schafer, J. Chem. Phys., 45, 646 (1968). 36. T. Takemura and H. Baba, Bull. Chem. Soc. Jap., 42, 2756 (1969). 37. J. N. Pitts, Jr., D. R. Burley, J. C. Mani, and A. D. Broadbcnt, g. Amer. Chem. Soc., 59, 5052 (1968). 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 73 A. A. Lamola, ibid., 55, 5045 (1970). Unpublished results of Dr. P. J. Wagner. N. C. Yang and R. Dusenbery, ibid., 55, 5899 (1968). W. D. K. Clark, A. D. Litt, and C. Steel, ibid., 55, 5413 (1969). J. Saltiel, H. Curtis, L. Metts, J. Miley, J. Winterle, and M. Wrighton, ibid., 55, 410 (1970). M. E. Long, Y. H. Li, and E. C. Lim, Chem. Phys. Letters, 55, 445 (1972). R. Castonguay, and A. C. Albrecht, ibid., 5, 89 (1970). G. J. Karabatsos and D. J. Fenoglio, in "Topics in Stereo- chemistry, Vol. 5" eds. E. L. Eliel and N. L. Allinger, Wiley Interscience, New York, 1970, p. 167. H. Suzuki, Bull. Chem. Soc. Jap., 55, 613 (1960). R. Hoffman and J. R. Swenson, J. Phys. Chem., 54, 415 (1976). G. Porter and P. Suppan, Trans. Faraday Soc., 55, 1664 (1965). P. J. Wagner, M. May, A. Haug, and D. Graber, J. Amer. Chem. Soc., 55, 5269 (1970). P. J. Wagner, M. May, and A. Haug, Chem. Phys. Letters, 55, 545 (1972). J. B. Gallivan and J. S. Brinen, ibid., 55, 455 (1971). A. M. Nishimura and J. S. Vincent, ibid., 55, 89 (1972). S. W. Mao, T. C. Wong, and N. Hirota, ibid., 55, 199 (1972). D. R. Arnold,J. R. Bolton, and J. A. Pederson, J. Amer. Chem. Soc., 54, 2872 (1972). J. S. Brinen and M. K. Orloff, Chem. Phys. Letters, 5, 276 (1967). R. F. Riley and J. Rosenthal, J. Chem. Phys., 55, 1034 (1969). P. H. H. Fischer and K. Hausser, Chem. Phys. Letters, 5, 665 (1968). 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 7o. 71. 72. 73. 74. 75. 76. 77. 78. 74 P. H. H. Fischer, H. Zimmermann, and J. Behnke, J. Magn. Resonance, 5, 250 (1970). E. G. Rodgers and J. S. Vincent, J. Chem. Phys., 55, 4627 (1970). E. Harris, Master's thesis, Mich. State Univ. H. Schott, Ph.D. thesis, Mich. State Univ. I. Kochevar, Ph.D. thesis, Mich. State Univ. A. Kemppainen, Ph.D. thesis, Mich. State Univ. J. McGrath, Ph.D. thesis, Mich. State Univ. B. Scheve, Ph.D. thesis, Mich. State Univ. F. C. Unterleitner and E. I. Hormats, J. Phys. Chem., 55, 2516 (1965). E. B. Priestly and A. Haug, ibid., 44, 622 (1968).. P. F. Jones and S. Siegel, ibid., 50, 1134 (1969). -_ S. W. Mao and N. Hirota, Mol. Phys., 51, 309 (1974). N. Hirota, C. A. Hutchinson, and P. Palmer, J. Chem. Phys., 45, 3717 (1964). A. Hudson and K. D. J. Root, Advan. Magn. Reson., 5, l (1971) . A. D. McLachlan, Mol. Phys., 5, 511 (1962). P. H. Rieger and G. K. Fraenkel, J. Chem. Phys., 55, 2795 (1962). J. Petruska, J. Chem. Phys., 54, 1120 (1961). J. N. Murrell, ”The Theory of the Electronic Spectra of Organic Molecules," Methuen and Co. Ltd., London, 1963, p. 126. ' M. A. El-Sayed, W. R. Moomau, and J. B. Chodak, J. Chem. Phys., _5__7_, 4061 (1972). N. Hirota, T. C. Wong, E. T. Harrigan, and K. Nishimoto, Mol. Phys., 55, 903 (1975). J. Kordas and M. A. El-Bayoumi, J. Amer. Chem. Soc., 55, 3043 (1974). 75 79. J. Tournon, M. Abu-Elgheit, Ph. Avouris, and M. A. El-Bayoumi, Chem. ngs. Letters, 55, 430 (1974). 80. N. Chu and D. R. Kearns, J. Amer. Chem. Soc., 94, 2619 (1972). “ 81. M. E. Long, Y. H. Li, and E. C. Lim, Mol. Photochem., 5, 221 (1971). 82. N. Kanamaru, M. E. Long, and E. C. Lim, Chem. ngs. Letters, 55, (1974). 83. H. J. Pownall, R. E. Connors, J. R. Huber, ibid., 55, 403 (1973). 84. M. A. Souto and C. T. Lin, Chem. Phys., 55, 129 (1976). 85. P. J. Wagner and M. J. May, Chem. Phys. Letters, 55, 350 (1976). 86. B. Smaller, E. C. Avery, and J. R. Remko, J. Chem. Phys., 45, 3976 (1967). .5 0311,1110 WI H1 T llHlllmlllll 3