THE GEEGM OF PHGTfiREACTI'v’lTY 1N EERTAIN KETONES POSSESSING LOWEST Dr, DI TRIPLE? STATES Wham for fire Deg?“ of DB. D. MICHISAE STAYE UNIVERSE" Dean Amaid Ersfeld I974 j HES U1 I‘AJChigan Sf. Universit y This is to certify that the thesis entitled RIGIN 0F PHOTOREACTIVITY IN CERTAIN £- ETONES POSSESSING LOWEST w,n* TRIPLET STATES I. presented by Dean Arnold Ersfeld has been accepted towards fulfillment of the requirements for Ph.D. degnmin Chemistry 1 - g D ‘ ‘« "'4 69W 1: a /ajor professor November 14 , J 974 Date ‘ 0-7639 LIBRAI'L'I ’ ABSTRACT THE ORIGIN OF PHOTOREACTIVITY IN CERTAIN KETONES POSSESSING LOWEST n,n* TRIPLET STATES By Dean Arnold Ersfeld The photochemistry of p-methoxy-Y-dimethylaminobutyrophenone (p-MDMAB), p—methoxy-a-methoxyacetophenone (p-MMAP), and p-methoxy-a- ethoxyacetophenone (p-MEAP) was studied to obtain further evidence in support of reaction from an equilibrium population of upper n,n* trip- lets. In addition, 4-dimethylamino-l-(B-naphthyl)-butanone (DMANB) was studied to determine from which excited states reaction would occur where equilibration of the lowest n,n* triplet state with upper n,n* triplets was not possible. p-MDMAB and DMANB were studied to also determine rates of intramolecular charge transfer from the y-amino group to n,n* triplet states as well as to obtain information concern- ing the requirements for type II elimination from a charge transfer complex. In benzene DMANB undergoes type II elimination only from its lowest excited singlet state, since greater than 3 M l,3-pentadiene does not quench the reaction. In methanol, because a long-lived (r = 0.8 x 10" sec) quenchable excited state is responsible for more than 90% of the type II elimination reaction, the n,n* triplet state as well as the Dean Arnold Ersfeld excited singlet state undergoes this reaction. However, the n,n* trip- let state undergoes this reaction via a charge transfer complex, which is undoubtedly formed in benzene (cpII = 0.01), acetonitrile (oII = 0.014), and methanol (qbII = 0.17), but only in the polar protic solvent does it go on to type II elimination products. The hydroxyl group probably under- goes O-H stretching in the solvation process and may even transfer a proton to the negatively charged oxygen of the complex, since a solvent deuterium isotope effect is observed in methanol-d1 (¢E¥30H/¢%¥30D = 1.4). The rate constant in methanol for charge transfer from the amino group to the n,n* triplet state is 1.2 x 106 sec". Based on the downward curvature in the Stern-Volmer plot for quench- ing type II elimination from p-MDMAB in benzene and the unlikelihood of excited singlet or n,n* triplet reaction in benzene, p—MDMAB probably photoreacts via an equilibrium population of upper n,w* triplets. The rate constant in benzene for charge transfer from the y-amino group to the n,n* triplet state in this case is detenmined to be 3 x 10° sec". Stern-Volmer plots for the quenching of type II elimination from p-MMAP and p-MEAP also curve downward. In these cases, however, a small amount of reaction occurs from the excited singlet states. When probable quantum yields for singlet reaction are subtracted from measured type II elimination quantum yields for both of these ketones, residual downward curvature in the Stern-Volmer plots is likely caused by quenching upset- ting the equilibrium between the reactive n,w* triplets and the unre- active n,n* triplets. THE ORIGIN OF PHOTOREACTIVITY IN CERTAIN KETONES POSSESSING LOWEST n,n* TRIPLET STATES By Dean Arnold Ersfeld A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1974 To Diana for her help and encouragement. ii ACKNOWLEDGEMENTS The author extends his sincere appreciation to Professor Peter J. Wagner for his guidance and support throughout the course of this research. The author thanks the National Science Foundation for research assistantships administered by Professor Wagner and the Department of Chemistry for financial assistance and for the use of its fine research facilities. The author thanks friends and acquaintances, whose fellowship has been invaluable. iii TABLE OF CONTENTS INTRODUCTION ........................... A. Photophysical Processes .................. B. Photochemical Processes .................. 1. Photoreduction .................... a. Early observations ................ b. The mechanism ................... c. Hydrogen sources ................. d. Evidence for the charge transfer interaction e. Summary of factors influencing the rate and quantum yield of photoreduction .............. 2 The Norrish Type II Photoprocess ........... a. Early observations ................ b. The 1,4-biradical ................. c. n,n* and n,n* triplets .............. d. Stern-Volmer kinetics ............... C. Research Objectives .................... RESULTS ............................. A. 4-Dimethylamino-l-(B-naphthy1)-l-butanone ......... 1. Quantum Yields .................... 2. Quenching of Z-Acetonaphthone Production ....... B. p-Methoxy-y-dimethylaminobutyrophenone .......... 1. Quantum Yields .................... iv Page 1 Dill-5&4:- 10 15 17 17 18 23 26 28 30 3O 3O 3O 35 35 TABLE OF CONTENTS (Continued) Page 2. Quenching of Excited p-MDMAB .............. 35 C. a-Alkoxy Ketones ...................... 38 1. p-Methoxy-a-ethoxyacetophenone ............. 33 a. Quantum yields ................... 33 b. Quenching of excited p-MEAP ............ 33 2. p-Methoxy-a-methoxyacetophenone ............ 41 a. Quantum yields ................... 41 b. Quenching of excited p-MMAP ............ 41 3. Methoxyacetone ..................... 46 a. Quantum yields ................... 46 b. Quenching of excited methoxyacetone ........ 46 D. Ben20phenone and 4,4'-Dimethoxybenzophenone Phosphorescence Quenching ......................... 48 DISCUSSION ............................ 51 A. Reactivity Where Lowest 3n,n* Is Far from 3n,n* ...... 51 B. The Nature of the Charge Transfer Complex in DMANB ..... 56 C. Reactivity Where Lowest 3n,n* Is Energetically Close to 3n,n* 58 1. Photoreactivity of p-Methoxy-v-dimethylaminobutyrophenone 58 2. a-Alkoxy Ketones .................... 53 0. Summary .......................... 68 E. Suggestions for Further Research .............. 70 EXPERIMENTAL ........................... 71 A. TABLE OF CONTENTS (Continued) Chemicals ......................... 1. Ketones ........................ a. p-Methoxy-y-dimethylaminobutyrophenone ...... b. p-Methoxy-a-methoxyacetophenone .......... c. p-Methoxy-a—ethoxyacetophenone .......... d. 4-Dimethylamino-l-(B-naphthyl)-1-butanone ..... e. Valerophenone ................... f. p-Methoxyacetophenone ............... g. 2-Acetonaphthone ................. h. Methoxyacetone .................. i. Acetone ...................... j. Ben20phenone ................... k. Acetophenone ................... 2. Quenchers ....................... a. 1,3-Pentadiene .................. b. l,3-Cyclohexadiene ................ c. Triethylamine ................... d. 2,5-Dimethyl-2,4-hexadiene ............ e. trans-Stilbene .................. 3. Solvents ....................... a. Acetonitrile ................... b. Benzene ...................... vi Page 71 71 71 72 73 74 75 75 76 76 76 76 76 76 76 77 77 77 77 77 77 77 TABLE OF CONTENTS (Continued) Page c. Heptane ...................... 78 d. Methanol ..................... 78 e. Pyridine ..................... 78 4. Internal Standards .................. 78 a. Octadecane and Tetradecane ............ 78 b. Cycloheptane ................... 79 c. 2-Methy1decane .................. 79 d. Pentadecylbenzene ................. 79 8. Methods .......................... 79 l. Readying Samples for Irradiation ........... 79 2. Irradiation ...................... 81 3. Photolysate Analysis ................. 82 a. Gas chromatography ................ 82 b. Identification of photoproducts .......... 83 c. Internal standard-product response ratios ..... 84 4. Actinometry and Quantum Yields ............ 85 C. Photokinetic Data ..................... 86 BIBLIOGRAPHY ........................... 116 TABLE 10 11 12 13 14 LIST OF TABLES Quantum yields for 2-AN production from DMANB ....... Molar ratios of 3-oxetane to p-MAP from the photolysis of p-MMAP and p-MEAP in various solvents ........... Quenching constants obtained from quenching the phospho- rescence emission of benzophenone and 4,4'-dimethoxybenzo- phenone with 2,5-dimethyl-2,4-hexadiene and triethylamine . An example of the preparation of solutions to be photolyzed cis-l,3-Pentadiene quenching of 2-AN formation from 0.04057 M DMANB in benzene irradiated at 313 nm .......... 1,3-Pentadiene quenching of 2-AN formation from 0.03062 M DMANB in benzene irradiated at 313 nm ........... 1,3-Pentadiene quenching of 2~AN fbrmation from 0.03314 M DMANB in methanol irradiated at 313 nm .......... 1,3-Pentadiene quenching of 2-AN formation from 0.02932 M DMANB in methanol irradiated at 313 nm .......... trans-Stilbene quenching of 2-AN formation from 0.0206 M DMANB in benzene irradiated at 313 nm ........... Photolysis of 0.05 M DMANB at 313 nm in benzene, 0.557 M pyridine in benzene, acetonitrile, and methanol ...... Page 31 43 50 80 88 89 90 91 92 93 Photolysis of 0.02 M DMANB at 313 nm in methanol, methanol-d1, various percentages of methanol in benzene, and in benZene A comparison of the cis-to-trans isomerization of cis-1,3- pentadiene sensitized by 0.05033 M benzophenone, 0.04998 M 2-AN, and 0.05116 M DMANB photolyzed at 313 nm ...... 1,3-Pentadiene quenching of p-MAP formation from 0.0399 M p-MDMAB in benzene irradiated at 313 nm .......... 1,3-Pentadiene quenching of p-MAP formation from 0.07388 M p-MDMAB in benzene irradiated at 313 nm .......... viii 94 95 96 97 TABLE 15 16 17 18 19 20 21 22 23 24 25 26 27 LIST OF TABLES (Continued) Page l,3-Cyclohexadiene quenching of p-MAP formation from 0.0434 M p-MDMAB in benzene irradiated at 313 nm ......... 98 l,3-Cyclohexadiene quenching of p-MAP formation from 0.04002 M p-MDMAB in benzene irradiated at 313 nm ......... 99 Determination of the disappearance quantum yield of 0.0300 M p-MDMAB in benzene irradiated at 313 nm .......... 100 Dependence of p-MAP quantum yield on initial p-MDMAB concen- tration and on pyridine concentration in benzene irradiated at 313 nm ......................... 10] The cis-to-trans isomerization cis-l,3-pentadiene sensitized by 0.05 M p-MDMAB in benzene irradiated at 313 nm ..... 102 l,3-Pentadiene quenching of p-MAP formation from 0.08693 M p-MMAP in benzene irradiated at 313 nm .......... 103 1,3-Pentadiene quenching of p-MAP formation from 0.1060 M p-MMAP in benzene irradiated at 313 nm .......... 104 The l-(p-methoxyphenyl)-l-hydroxy-3-oxetane to p-MAP ratio produced from the photolysis at 313 nm of 0.05 M p-MMAP in benzene, 1,3-pentadiene, cyclohexene, cyclopentene, and cyclohexane ........................ 105 The cis-to-trans isomerization of cis-l,3-pentadiene sensi- tized by 0.07 M p-MMAP in benzene irradiated at 313 nm . . 105 The 1-(p-methoxyphenyl)-1-hydroxy-2-methy1-3-oxetane to p-MAP ratio produced from the photolysis at 313 nm of 0.05 M p-MEAP in benzene and in l,3-pentadiene ............. 107 l,3-Pentadiene quenching of p-MAP formation from 0.05035 M p-MEAP in benzene irradiated at 313 nm .......... 108 The cis-to-trans isomerization of cis-l,3—pentadiene sensi- tized by 0.05 M p-MEAP in benzene irradiated at 313 nm . . 109 l,3-Pentadiene quenching of acetone formation from 0.2001 M methoxyacetone in benzene irradiated at 313 nm ...... 110 TABLE 28 29 30 31 32 LIST OF TABLES (Continued) Page Quantum yields for acetone formation from 0.201 M methoxy- acetone in benzene and in 1,3-pentadiene irradiated at 313 nm ........................... 111 Quantum yield for the disappearance of methoxyacetone from a solution 0.0521 M in methoxyacetone in benzene irradiated at 313 nm ........................... 112 The cis-to-trans isomerization of cis-1,3-pentadiene sensi- tized by 0.2001 M methoxyacetone in benzene irradiated at 313 nm ........................... 113 The 2,5-dimethy1-2,4-hexadiene and triethylamine quenching of phosphorescence from 0.0231 M benzophenone in benzene excited at 375 nm ......................... 114 The 2,5-dimethyl-2,4-hexadiene and triethylamine quenching of phosphorescence from 0.02306 M 4,4'-dimethoxybenzophenone in benzene excited at 375 nm ................. 115 FIGURE 1 10 11 12 LIST OF FIGURES Page Radiative and nonradiative transitions resulting from the absorption of light by a phenyl ketone .......... 1 Approximate relative energy levels of some electronically excited states of three types of ketones ......... 3 ¢2~AN as a function of the percent methanol in benzene solvent ......................... 32 Stern-Volmer plots for l,3-pentadiene quenching of 2-AN formation from DMANB in benzene (A), in methanol (0). and in methanol with the quantum yield for reaction from the short-lived state subtracted from ¢° and o GI) ...... 33 Stern-Volmer plot for t-stilbene quenching of 2-AN formation from DMANB in benzene .................. Dependence of quantum yields for p-MDMAB sensitized cis-to- trans isomerization of cis-l,3-pentadiene on diene concen- tration in benzene .................... 35 Stern-Volmer plot for 1,3-pentadiene quenching of p-MAP formation from p-MDMAB (0 0.03990 M, A 0.07388 M); 313 nm irradiation ....................... 37 Quenching of excited p-MDMAB by l,3-cyclohexadiene . . . . 39 Dependence of quantum yields for p-MEAP sensitized cis-to- trans isomerization of cis-l,3-pentadiene on diene concen- tration in benzene .................... 40 Stern-Volmer plot for 1,3-pentadiene quenching of p-MAP fonnation from p-MEAP in benzene ............. 42 Efficiencies of sensitization of the cis-to-trans isomeriza- tion of l,3-pentadiene by p-MMAP as a function of diene con- centration in benzene .................. 44 Stern-Volmer plot for l,3-pentadiene quenching of p-MAP formation from p-MMAP (0 0.0869 M,‘ 0.106 M) in benzene 45 Figure 13 14 15 16 LIST OF FIGURES (Continued) Page Dependence of quantum yield for methoxyacetone sensitized cis—to-trans isomerization of cis-l,3-pentadiene on diene concentration in benzene ................. 47 Stern-Volmer plots for the quenching of benzophenone phos- phorescence by 2,5-dimethyl-2,4-hexadiene (C3), and by triethylamine (AS), and of 4,4'-dimethoxybenzophenone phosphorescence by 2,5-dimethyl-2,4-hexadiene (Q), and by triethylamine (A) .................. 49 1,3-Pentadiene quenching of p-MAP formation from triplet p-MMAP .......................... 66 1,3-Pentadiene quenching of p-MAP formation from triplet p-MEAP .......................... 67 xii INTRODUCTION A. Photophysical Processes The creation of an electronically excited molecule results in the occurrence of a large number of photochemical and photophysical process- es. To fully appreciate any one of these processes it is imperative that the relative importance of those remaining be understood. In the absence of any photochemical reactions, the physical pro- cesses which occur upon the absorption of a photon by a ketone molecule are shown in Figure l, a scheme similar to that introduced by A. Jablonski in 1935.1 The lowest triplet, 1,, corresponds to H O ‘r‘ww “"’s, E $-- <1 F 13 PS 6' 9r it iL2_ Ground State Figure l. Radiative and nonradiative transitions resulting from the absorption of light by a phenyl ketone. 2 Jablonski's "metastable state, which he used in rationalizing the phenomena of delayed fluorescence and phosphorescence emission.2 The vertical wavy lines in the figure depict radiationless transitions. Internal conversion, 10, and intersystem crossing, 150, are represent- ed by horizontal wavy lines. The straight, slanted lines indicate absorption of light, while the straight, vertical lines represent fluorescence, F, and phosphorescence, P, emission. Vibrational and rotational levels are not shown for the sake of clarity. In the process of absorption of a photon, the spin angular momen- tum of the ground state must be conserved. Since ground-state ketones are singlets, photon absorption results almost exclusively in excited singlets. When upper singlets (52,53, etc.) become populated, vibra- tional relaxation and internal conversion bring about a very rapid (1012 seC") decay of these singlets to $1.3 Thus, fluorescence and inter- system crossing to the triplet manifold normally occur from the 5, state. One of the few established exceptions to this behavior is the fluorescence of azulene which occurs from $2.“’5 The rate of radiation- less crossing between states is inversely proportional to the energy gap between them.3 A large 51-52 energy gap is thought to allow the 52 state to fluoresce at a rate competitive with internal conversion to 51. Population of the triplet state is brought about by intersystem crossing from the 51 state. The rate of intersystem crossing is large- ly dependent on the amount of spin-orbit coupling which causes the sing— let to attain some triplet character and vice versa.6 The extent of this mixing is influenced by the energy difference between $1 and the triplet state,9 and the nature of the singlet and triplet states in- volved.6 Intersystem crossing from an n,n* singlet (a non-bonding 3 electron is promoted to an anti-bonding n orbital) to a n,n* triplet (a n electron is promoted to an anti-bonding n orbital) is about 103 times faster than from an n,n* singlet to an n,w* triplet? These considerations can serve to explain, at least in part, cer- tain behavior of aliphatic, phenyl, and naphthyl carbonyl compounds. Figure 2 shows the approximate energy levels of some of their electron- ically excited states.‘° Ben20phenone, benzaldehyde, and acetophenone Aliphatic Phenyl Naphthyl lflzfl* ($2) 3","* (T2) ‘N.Tr* ($2) l";“* (51) 1"9"* ($2) 3 * 1 * 3 * n,n (Tl) now (51) non (Ta) In N* (51’3n "* (T2) 3033:.(71) 31mr* (T1) Figure 2. Approximate relative energy levels of some electronically excited states of three types of ketones. intersystem-cross with a rate constant of about 101° - 10ll sec-1, phos- phoresce strongly in rigid media, and do not fluoresce."7 Aliphatic carbonyls, however, fluoresce (10s seC") as well as phosphoresce, and therefore intersystem-cross with a smaller rate constant (10' sec"‘).“'8 The presence of the n,n* triplet between the n,n* singlet and triplet levels may enhance intersystem crossing in the phenyl carbonyl compounds? In the case of 2-acetonaphthone, the n,n* triplet is the lowest triplet, 4 but it is considerably lower in energy than the n,n* singlet.11 Thus, processes in competition with intersystem crossing have a better chance of occurring, since intersystem crossing may be slower in this situation. Upper excited triplet states undoubtedly cascade to T1 at a rate (10‘1- 1013 seC") similar to excited singlet states. Therefore, any chemical reaction from an upper excited triplet state must be very rapid to compete with this process. On the other hand, lowest triplets are relatively long-lived because intersystem crossing to S0 is relatively slow, and in most cases triplet reaction occurs almost exclusively from T1. The fact that some compounds react from upper excited states poses the interesting problem of what characteristics of an excited state cause it to react. To solve this problem it is necessary to identify the excited state or states responsible for the observed photochemical processes, as well as to elucidate the reaction mechanism. TWo kinds of photochemical processes have been the focal point of a great volume of research. These are photoreduction and the Norrish type II reaction, both of which deserve reviewing here. 8. Photochemical Processes l. Photoreduction a. Early observations As early as 1900, Ciamician and Silber12 observed the photoreduction of benzophenone and acetophenone. When an alcoholic solution of the ketone was placed in sunlight, benzopinacol and l,2-dimethyl-l,2-di- phenyl-1,2-ethanediol were produced respectively. Ciamician and Silber13 later showed that other ketones could be photoreduced. In 1920, in addition to corroborating this work, Cohenl“ found that a small amount 5 of sodium alcoholate present in the alcoholic solution of various sub- stituted ketones resulted in the decomposition of the pinacol into hydrol and ketone; so that the hydrol appeared to be the sole product. He also made the interesting observation that the following ketones in ethanol were not reduced to pinacols: 4,4'-bis(dimethylamino)benzo- phenone, phenyl a-naphthyl ketone, fluorenone, and p-phenylbenzophenone. Bachmannls reported that various para substituted benzophenones were converted to hydrols when solutions of the ketone in isopropyl alcohol containing small amounts of sodium isopropylate were irradiated by sun- light. Except for p-phenylbenzophenone, which he observed to photo- reduce nearly quantitatively to the corresponding hydrol, Bachmann's work corroborated that of Cohen. Bergmann and Hirshberg16 extended the list of ketones which did not pinacolize when irradiated in isopropyl alcohol to include among others a and B-acetonaphthone, di-a-naphthyl ketone, p-phenylbenzophenone (in conflict with Bachmann's report), and p-methoxypropiophenone. b. The mechanism In 1934, Backstroml7 postulated a mechanism for the first step of the photoreduction process, suggesting that a biradical formed from the light absorption process could abstract a hydrogen atom, and that the resulting radicals dimerized and underwent disproportionation. Soon after, Weizmann, Bergmann, and Hirshberglo suggested the same mechanism with one exception. They found that optically active phenyl methyl car- binol retained its optical activity when in solution with photolyzed acetophenone. This ruled out disproportionation. TWO decades later, Pitts and coworkers19 corroborated this finding by showing that the photolysis of benzophenone in optically active sec-butyl alcohol 6 resulted in no change in the optical activity of the alcohol. In addi- tion, in the absence of oxygen, benzopinacol was formed with a quantum yield of nearly unity while no pinacol, mixed pinacol, or benzhydrol was formed. Thus, the following set of reactions was proposed: thco=~ + (CH3)ZCH0H —-—> théow + (CH3)2C0H (l) (CH3)2C0H + thco ———-> thCOH + (CH3)2C0 (2) 2 thCOH —> PhZCOHCOHPhZ (3) Further, reaction (2) provides for observed quantum yields of benzo- phenone disappearance in excess of unity.2° The multiplicity of the excited state of benzophenone in reaction (1) has been shown to be triplet. Porter and Wilkinson21 and later Bell and Linschitz22 observed the triplet state of benzophenone directly by flash photolysis. Backstrom and Sandr0523’2“ showed that the sensi- tization of biacetyl phosphorescence by benzophenone took place by triplet energy transfer from benzophenone to biacetyl, thus, proving that electronic excitation of benzophenone results in population of the triplet manifold. 0n the basis of the lifetime of the reactive state (longer lived than the lowest singlet should be), Hammond, Moore, and Foss”,26 concluded that the reactive excited state must be the lowest triplet state. Porter and Wilkinson}1 and Moore and Ketchum27 showed that the reactive excited state was triplet by quenching the photore- duction of benzophenone with naphthalene, a quencher previously shown by Terenin and Ermolaev28 to accept triplet energy from a number of com- pounds including benzophenone. As it turns out, the nature of the lowest triplet has a major in- fluence on the ability of ketones to be photoreduced. Ben20phenone, which has a lowest n,n* triplet,29 is photoreduced with ease, while 7 ketones such as phenyl a-naphthyl ketone, p-phenylbenzophenone, fluore- none, and the acetonaphthones, which are believed to have lowest n,n* triplets3°»3"“, are photoreduced by alcohols only with difficulty. Further, acetophenone, with an n,n* triplet slightly lower than the n,n* triplet is about fifty times more reactive toward hydrogen abstraction than 3,4-dimethy1acetophenone, which has a n,n* triplet just below the n,n* triplet.32 These two types of excited triplets differ markedly in their elec- tron distribution. An n,n* triplet results from the excitation of a non- bonding, n, electron into a n antibonding, n*, orbital. This causes a net shift in electron density away from the oxygen atom, in the case of aldehydes and ketones, leaving an electron deficient oxygen. R20 = 0 --JP R C - O: G. S. 3n,n* This appears similar to an oxy radical, and indeed, benzophenone triplets have been shown to be qualitatively similar to the t-butoxy radical.3393“ The excitation of a bonding electron into an antibonding, n* orbital results in a n,n* excited state. In cases where the lowest triplet of an aldehyde or ketone is n,n* in nature, the carbonyl group is conjugated with a n system; so that the partial vacancy of the bonding n orbital is delocalized over the n system. Additionally, electron density is shifted :0: ll G. S. :0: :0- :0:- Qt“... 2...... 9.1)-. 3 * n,n toward the carbonyl oxygen.35 8 The n,n* and n,n* excited triplet states also differ spectroscopic- ally in several ways. The phosphorescence lifetime of n,n* triplets is of the order of 10'2 sec while that of n,n* triplets is at least five times longer. Polar solvents and electron donating substituents shift n,n* states to higher energy while shifting n,n* states to lower energy and to a greater extent.36’37»2° Kearns and Case have investigated singlet-triplet transitions by the phosphorescence excitation method and have found taht the addition of an external heavy atom (for example, adding ethyl iodide) enhances absorption to the n,n* triplet by a factor of two. No effect on the absorption to the n,n* triplet is observed.38 Pitts, et aZ.2° made one of the earliest and most enduring corr- elations between the nature of the lowest triplet level and reactivity. Substituted benzophenones with lowest n,n* triplets, such as p-phenyl- benzophenone, were said to be photoreduced inefficiently because the hydrogen abstracting ability of the carbonyl oxygen atom is greatly re- duced as compared to the case where the lowest triplet is n,n* in nature. This is not surprising given the difference in electron distribution and the rapid rate of decay from T2 to T1. In the same way Hammond and Leermakersll have explained the fact that l-naphthaldehyde and 2-aceto- naphthone are not photoreduced with secondary alcohols. In 1963, Becket and Porter39 explained the low efficiency of photoreduction of p-hydroxy- benzophenone in isopropyl alcohol also in this way. However, soon after this Porter and Suppan35»“° found that hydroxy— benzophenones were photoreduced by alcohols in a nonpolar solvent. Thus, they invoked a third discrete excited state, the charge transfer state. Charge transfer states, they said, were not reactive, n,n* states had kw reactivity, and n,n* states had high reactivity. Later, Suppan :4 ‘- ‘5 -' .2. 9 seemed to alter this view slightly when he said, "In aromatic ketones substituted with a strong donor function (such as NHZ) the dipole moment in the lowest states usually classified as n,n* is so large that this state is better called simply a charge transfer state"."1 It would seem, then, that the n,n* triplet has varying degrees of electron density shifted toward the oxygen atom of the carbonyl group depending on the nature of the substituents. In cases where the lowest n,n* and the upper n,n* triplets are energetically proximate, triplet reactivity is greater than when the triplets are well separated.32 More will be said about this in the section dealing with the type II reaction. c. Hydrogen sources In addition to alcohols, a wide variety of hydrogen sources have been used in the photoreduction of ketones. Among these are alkanes“2. 33, alkylbenzenes“2:33:“3, tributylstannane“, ethers““, and amines“5. Interestingly, compounds with lowest n,n* triplets, such as fluorenone, and 2-acetonaphthone, although not photoreduced by alcohols, abstract hydrogen atoms from tributylstannane with rate constants in the range of 105 to lOSM'lsec.“5»11 Hammond and Leermakers11 have suggested that hydrogen abstraction by n,n* states is an activated process, and that the low Sn-H bond energy and high polarizability of the tin atom lowers the activation energy sufficiently. Suppan"1 has measured the temperature effect on the very low rate constant for hydrogen abstraction from ethanol by 2-acetonaphthone. The pre-exponential factor from the Arrhenius plot was about the same as that for benzophenone, and Suppan considered this additional evidence that n,n* and n,n* states differ in their hydrogen abstracting ability because of a difference in activation energy. A i b ‘- .,p 10 Amines also demonstrate an ability to react rapidly with excited carbonyl compounds which have either n,n* or n,n* lowest triplets. For example, benzophenone interacts with triethylamine three orders of mag- nitude faster than with isopropyl alcohol.“7 Triethylamine"8 photo- reduces 2-acetonaphthone about 103 times faster than does ethanol."l Fluorenone abstracts hydrogen atoms from tributylstannane"6 about one tenth as fast as it interacts with triethylamine.“9 Cohen and Cohen5° found that the quantum yield of p-aminobenzphenone photoreduction in cyclohexane was fifty times greater in l M triethylamine than in l M 2-propanol. The Cohens suggested that the high reactivity of amines may be caused by rapid electron transfer from nitrogen to the excited trip- let followed by proton transfer rather than simple a-hydrogen atom abstraction. d. Evidence for the charge transfer interaction Cohen's electron-transfer hypothesis was not without precedent. By flash-spectroscopic investigations of perylene solutions containing amine, Leonhardt and Weller51 demonstrated that electron transfer, not possible between a ground state molecule and an appropriate donor or acceptor, could occur when the molecule was electronically excited. Mataga and Ezumi52 observed long-wavelength fluorescence from some aromatic hydrocarbons in N,N-dimethy1 aromatic amine solvents. This in conjunction with the electron affinities of the excited aromatic hydro- carbons and the ionization potentials of the amines led Mataga and Ezumi to conclude that the interaction between excited hydrocarbon and amine involved a charge transfer from amine to the excited hydrocarbon. Davidson and Lambeth52 proposed the same mechanism as Cohen's to explain some of their results. They compared the quantum yield of ll disappearance of benzophenone being photoreduced by several amines to that by diphenylmethanol. The resulting photoreduction quantum yield ratios were then converted to quantum yield ratios per a-hydrogen atom on the amine, and the following order was obtained: N,N-dimethylanaline > N-methylphenylamine > N-benzyldiphenylamine > N-methylcarbazole > N-benzylcarbazole. Since the N-benzylamines gave lower quantum yields per hydrogen atom than the N-methylamines, the hydrogens of which are less reactive, Davidson and Lambeth suggested the following mechanism: M = 0* + R'CHZNRi —-> ch-o' + R'CHZ +NR2—DR2C-0H + R'CHNR; More concrete evidence concerning the mechanism of photoreduction by amines was presented by Cohen and Chaos“ in 1968. By quenching the photoreduction of benzophenone by 2-butylamine with naphthalene, the rate of reaction was determined to be about 5 x 107 M"1 sec-1, which is more than an order of magnitude greater than that found when benzhydrol27 or 2-propanol55 are the hydrogen sources. Yet with 2-butylamine the highest quantum yield attaned for benZOphenone disappearance was appre- ciably less (e = 1.1) than that attainable when isopropyl alcohol is the hydrogen source (o = 2)2°. No light-absorbing transients were found which might lower the quantum yield. Optically active 2-butylamine was not racemized, thereby excluding reversible hydrogen abstraction as the cause for the inefficiency. An inverse isotope effect on the benzo- phenone disappearance quantum yield was found for 2-butylamine-N,N-d2 and cyclohexylamine-N,N-d2. Thus, initial abstraction of hydrogen from nitrogen was ruled out. A small isotope effect on the quantum yield was found with cyclohexylamine-l-d (oH/qbD = 1.6). In addition, photoreduction of p-aminobenzophenone by triethylenediamine was found by Cohen and Cohen56 to proceed with a respectable quantum yield (o = .21) even ,. 12 though the following type of resonance stabilization of hydrogen abstraction is much more difficult with this compound. . \ .. . _ {3. (bzc-O' H"/C‘XHH¢2C-0 H- +C-XH The reactivity of bridgehead nitrogen, the high rate of reaction, moderate quantum yields, the inverse N-deuterium effect, and the small a-deuterium isotope effect are consistent with a rapid charge-transfer interaction between the ketone triplet and the nonbonding electrons of nitrogen, followed by reverse charge-transfer and proton transfer.56 Further evidence has accumulated. Ware and RichterS7 used di- methylaniline to quench the fluorescence of perylene in various solvents of differing dielectric constants. As the dielectric constant of the media was increased, more nonradiative decay was observed. This led to the conclusion that a charge-transfer complex was being formed at a diffusion-controlled rate in solvents with high dielectric constants. Wagner and Kemppainen58 found that triethylamine and dimethyl-t-butyl- amine photoreduced triplet valerophenone with virtually identical rates, even though a significant difference existed between the reactivity of the C-H bonds in each. Cohen and Green59 reported that the rate of interaction of acetophenone triplets with a-methylbenzylamine was more than twenty times greater than that with a-methylbenzyl alcohol. Wagner and Kemppainen58 showed that the rate of interaction between valero- phenone and triethylamine was much higher in benzene and in aceto- nitrile than in methanol. The same effect was observed on the corre- sponding intramolecular interaction which occurs in v-dimethylaminobuty- rophenone. However, here it was also found that a five to ten fold increase in type II elimination quantum yield occurred in changing from acetonitrile and benzene to methanol, whereas no change of such a 13 magnitude was observed in the valerophenone - triethylamine system. Cohen and Litt"7 discovered the same phenomenon in the photoreduction of benzophenone by amines. Rates of interaction with the amines were higher in benzene than in aqueous medium, while the interaction with 2-propanol remained unchanged. Cohen and Parsonss° have correlated the rate of interaction between fluorenone and various ring-substituted N,N-dimethylanilines with 0+ values. Their correlation indicated the development of a positive charge at nitrogen. The ionization potentials of amines have been correlated with the rate of interaction between amine and electronically excited ketone. Davis, et aZ.,“‘ using amines of differing ionization potentials, quenched the fluorescence of fluorenone. The rate of quenching was found to be inversely proportional to the ionization potential of the amine and greater in acetonitrile than in benzene. Davidson and Lambeth61 reported that tri-p-tolylamine quenched the photoreduction of benzo- phenone by diphenylmethanol in benzene more than an order of magnitude faster than did triphenylamine. Cohen and Stein62 photoreduced 4-benzoyl- benzoate anion with amines of differing ionization potentials in a 1:1 water-pyridine solvent. Reactivity was observed to increase as the ionization potential of the amines decreased. Guttenplan and Cohen63 have recently observed a linear inverse relation between log kir (kir is the rate constant for interaction of the electron donor with, in this case, benzophenone triplets) and the ionization potential of the donor. Previously Weller, et aZ.°“ had related the rate of fluorescence quench- ing by electron donors of aromatic hydrocarbons to the free-energy change AGct, which is in turn related to the excitation energy of the acceptor, AEO 0, the ionization potential of the donor, I, the electron affinity 14 of the acceptor, A, and a constant, C. AG 2 -AEO 0 + I - A + C (4) ct , Guttenplan and Cohen used benzophenone as the acceptor with a variety of donors so that AGct 2 I + constant. The slope of log kir versus I was found to be much smaller than that observed where complete65 electron transfer took place, and only small solvent effects were observed com- pared to those observed when full electron transfer occurred.51 Gutten- plan and Cohen determined that the activation energies for full electron transfer from most of the donors to the benzophenone triplet would be too high to lead to the large values of kir which were observed. Thus, they made the following conclusions. The excited ben20phenone and donors interacted by both partial transfer of charge and a-hydrogen abstraction. While the initial contact between excited ketone and alcohols involved mainly the a-hydrogens, the initial interaction with amines involved mainly the n-electrons. At the transition state, however, resonance stabilization by the heteroatom of the transition for hydrogen abstrac- tion was considered about the same in both cases. Since the aromatic amines showed a greater sensitivity of kir to ionization potentials than did the aliphatic amines, a higher degree of charge transfer was occurring in the case of the aromatic amines. Other classes of compounds are thought to react with excited ketones via the charge-transfer process. Methyl phenyl sulfide and di-n-butyl sulfide have been shown to quench the photoreduction of benzophenone by isoborneol (¢ = 1.38) with a rate constant of about 108 M'1 sec'l while photoreducing benzophenone with only a very low quantum yield (o = 0.05).66 Caldwell67 hypothesized that simple olefins and benzo- phenone interact by substantial electron transfer from olefin to excited 15 ketone, since the rate of interaction was nearly two orders of magnitude faster with p-trifluormethylbenzophenone than with p,p'-di- methoxybenzophenone. Kochevar and Wagner68 have quenched triplet buty- rophenone with a large variety of olefins and have correlated the quench- ing rates with the ionization potentials of the olefins. They concluded that charge-transfer complex formation is a general process which pre- dominates when the alkene is electron rich. In corroberation, it has been found that enol ethers quench alkanone fluorescence via a charge- transfer complex.69 Furthermore, various dienes have been shown to interact with ketone singlets with rates which correlate best with diene ionization potentials."°’71 Triphenylphosphene and trimethylphosphite have been shown to quench excited ketones, and Davidson and Lambeth have suggested charge-transfer complex formation as the mode of interaction.72 The rates of interaction between a-trifluoroacetophenone reacts with alkylbenzenes by the formation of a charge-transfer complex which can either fall apart to the ground-state molecules or fall apart to radicals of the type normally found in photoreduction. e. Summary of factors influencing the rate and quantum yield ofjphotoreduction The rate of interaction between excited ketone and hydrogen source depends upon the nature of the interaction. The rate of hydrogen abstraction is largely influenced by the C-H bond strength of the hydrogen atom source, while the rate of charge-transfer is dependent on the electron affinity of the ketone and the ionization potential of the electron donor. The rates of both interactions are dependent, although to different extents, on the nature of the excited state, the energy of the excited state, and the solvent. 16 The quantum yield of photoreduction is really the probability that absorption of light will cause photoreduction to occur. In a typical case where the triplet ketone, such as that of benzophenone, abstracts hydrogen from a donor, such as hexane, the quantum yield for ketone disappearance will be determined by the relative rates of each of the process in the following scheme: W KO + hv —+Ksl I (rate of absorption of light) KSl ———>ko kd [KSJ KSl ——>Ko + th kF [K51] KS, —_—’ KT, kisc KS1] KT1 ———pko kd. [Kn] KT1 ——>ko + hvp kp [le] KTl + RH-—>KH + R k" [KTJ [RH] kH + 1.2—9 o products + (l-a)(Ko + RH) Under conditions of steady illumination the quantum yield for dis- appearance of ketone may be expressed as follows: ¢-K = kisc kr [RH] (5) 0. kisc + kd + kF kr [RH] + kd. + kp When the expression for the quantum yield of fonnation of a particular product is desired, a probability term, Pp, must be multiplied by the above terms. Pp is the probability that metastable intermediates pro- ceed on to the particular product rather than other by-products. In cases where R can selectively react with ground state ketone, another term must be added to the above expression, and the quantum yield will tend toward 2 rather than 1 as above. When charge—transfer becomes a possible mode of interaction between l7 excited ketone and donor, another rate constant and another term are introduced since the following processes may occur: Rate ._ +. ‘— KTl + RH——>K--- RH kct[KT1] [RH] ._ +. ._ +. K--- RH ——> k0 + RH k_ct [K--- RH] K:--+RH ——v KH + R kr. [KI-JRH‘) The reversion of the charge-transfer complex K:--+RH to the ground state is a significant cause of inefficiency in this photoreduction process. ¢-K = kisc fi[ k? [RH] + kisc + kd +kF Er [RH] +de. +kp + kCt [RHJT kct [RH] . kr' ] 'a (6) Er [RH]’+ k'. + kp +kct[RH]' kr' + k-ct 2. Norrish Type II Photoprocess a. Early observations In an effort to elucidate the machanism of photochemical decompo- sition of aldehydes in the gaseous state (RCHO ——-D RH + CO), Norrish and Appleyard’“ in 1934 photolyzed some ketones in the vapor state. Unexpectedly, methyl butyl ketone decomposed in the following way: CH3CH2CH2CH2C0CH3—§ CH3CH=CH2 + CH300CH3 (7) Further investigation showed that other carbonyl compounds with long hydrocarbon chains underwent a, 8 bond cleavage with the production of a smaller carbonyl compound and an olefin.75 This process was desig- nated type II as opposed to type I, which in the gas phase involved the elimination of carbon monoxide and the formation of free radicals.7'°"76 In 1947, Davis and Noyes77 suggested that a y-hydrogen in methyl butyl ketone might interact with the carbonyl oxygen in such a way that the excited compound could dissociate into the enol form of acetone and 18 propylene in a single step. Several years later Nicholson78 presented results consistent with the intramolecular rearrangement concept. A year later Martin and Pitts79 reported that the type II process was general for aliphatic ketones with v-hydrogen atoms. The involvement of a y- hydrogen atom in the process was fairly well established by Srinivasan”:o observation that the photolysis of 2-hexanone-5,5-d2 gave mainly C3H50 and a 50:50 mixture of deuterated and nondeuterated acetone. Srinivasan explained that propylene and the enolic form of acetone were formed in the primary process and that during the subsequent rearrangement of the enolic four to the ketone, hydrogen-deuterium exchange occurred on the walls of the reaction vessel. Several years later Coulson and Yang81 presented quite conclusive evidence for the involvement of a y-hydrogen in the primary process. They reported that y-deuteration caused a sig- nificant increase in the lifetime of excited 2-hexanone, a result con- sistent with the work of Borkowski and Ausloos82 which showed that a v- hydrogen was more reactive than a y-deuterium. This isotope effect pro- duced by y-deuteration has since been verified by a number of researchers. 8338.4,35 Cyclobutanol formation was first reported by Yang and Yang86 in 1958 to accompany a, 8 bond cleavage. The Yangs formulated a 1,4-bi- radical as an intermediate to explain both processes. b. The 1,4-biradica1 The intermediacy of a 1,4-biradica1 was intuitively readily accept- able because of its mechanistic simplicity and because it is merely the intramolecular analog of intermolecular hydrogen abstraction by excited ketones, the primary reaction which is known to occur in the photore- duction process. Yang, et aZ.°7 reported in 1963 that 6-hepten-2-one 19 photorearranges to methylcyclohexenol as well as methylvinylcyclobutanols. These results were thought to substantiate the step-wise mechanism, since the simultaneous appearance of both products indicated the intermediacy of a delocalized biradical. CHaCOHCHZCHzCHCH=CH2§——> CH3COHCH2CH2CH=CHCH2 However, the possibility of a concerted mechanism was brought into focus soon after. Schulte-Elte and Ohloff88 found that 2,6-dimethyloct- 7-en-3-one, optically active at the y-carbon, produced cyclization pro- ducts with some retention of configuration as well as a large amount of racemization of the recovered ketone. Orban, et al.39 photolyzed (5R)- 5,9,-dimethyldecan-2-one and found that the cyclization products retained some optical activity. Such results were thought to be explainable by a competition between the concerted and 1,4-biradical mechanisms or by the intermediacy of a short-lived biradical which could racemize and cyclize competitively. More support for the 1,4-biradical mechanism has since appeared. Wagner and Hammond9° postulated that reversible hydrogen transfer in the 1,4-biradical was the cause of low quantum yields in compounds where the excited state reactivity was high compared to the rate of radiationless decay. Very strong support for this idea was reported by Wagner,91 who found that the quantum yield for valerophenone disappearance increased to unity when photolyzed in solvents which could hydrogen bond with the biradical's hydroxy hydrogen and, thus, prevent it from disproportion- ating, while not inhibiting cyclization or cleavage. Rauh and Leermaker:2 reported a similar finding for butyrophenone. Various phenyl alkyl ketones, whose rates of excited state reaction varied by two orders of magnitude in benzene, were shown by Wagner and Kamppainen93 to possess 20 quantum yields which did not vary by more than a factor of three in ben- zene and which approached unity in polar solvents. The lack of correla- tion between quantum yield and reactivity was suggested to be caused by disproportionation of the biradical. Lewis°“ showed that the photolysis of the B,y-dideuterio derivative of y-hydroxy—y-phenylbutyrophenone pro- duced the deuterium-hydrogen exchanged ketone, a-deuterio-y-deuteroxy-y- phenylbutyrophenone, a finding which further suggested the presence of a biradical intermediate. Perhaps the most compelling evidence for a 1,4- biradical intermediate was presented by Wagner, Kelso, and Zepp.85 The photolysis of B,y-diphenylbutyrophenone produced no triplet stilbene from triplet ketone even though energy and spin considerations showed that such a process could have proceeded concertedly. In addition, the combined quantum yield for a,B cleavage, cyclization, and racemization of (4S)-(+)-methyl-1-phenyl-l-hexanone was very near unity. Alcohol sol~ vent caused the product quantum yield to be essentially unity while elim- inating racemization of starting material. Thus, racemization and product formation arose from the same biradical intermediate. Furthermore, the biradical of valerophenone was actually trapped using alkyl thiols as trapping agents. Using the photoprocesses of (4S)-(+)-4-methyl-l-phenyl- l-hexanone, MPH, as an example, the following scheme reveals the mechanism of the type II process in phenyl alkyl ketones.85 MPHO ——“L-> MPH” MPH‘ émampm” MPH 3*__:l®%_, BR 21 OH .,.CH2CH3 BR CH3 Ph ~3.8% 4" o H 4,,CH2CH3 Ph 0H [CHZCHg 0H .1! /J\v/F‘CH, /A§§ '1 \TFJ Ph \\\. ~1.2% P“ a 0H Ph ~95% Ph + H0 2R 0H 0H - ~25% 140% Ph/‘S +jf\ +Ph o (1;) .203 (i) The partitioning of the biradical between the various processes varies according to the structure of each specific ketone. The ratio of a,B-cleavage product to cyclization product is in- fluenced by the multiplicity of the reactive state(s) as well as by the structure of the compound. The a,B-cleavage reaction has been shown to occur from both singlets and triplets’“’9°, while cyclobutanol formation is thought to be mainly a triplet reaction“, since triplet quenchers almost completely suppress cyclobutanol fbrmation but only partially quench a,B-cleavage. Aliphatic ketones demonstrate singlet and triplet reactiongu’ao and have a corresponding low cyclization to a,B-cleavage ratio, while unsubstituted phenyl ketones, which intersystem cross much faster and, therefore, have an intersystem crossing quantum yield of 22 unity’s, react only from the triplet manifold and generally have a higher cyclization to a,B-cleavage ratio. By analogy to the photoreduction process, the l,4-biradica1 should be able to be produced through an initial charge-transfer process. Padwa and Eisenhardt" have reported the photochemical production of (B-t-butylamino)-trans-benzalacetophenone from trans-N-t-butyl-2-phenyl- 3-benzoy1aziridine. Because the reaction occurred from the triplet state and yet could not be quenched by triplet quenchers, it was pro- posed that rapid charge-transfer from the nitrogen to the n,w* triplet was occurring. Padwa, et al.97 reported also that a-dialkylamino ketones underwent type II elimination which could not be quenched with high con- centrations of dienes. These results have been explained by a mechanism involving intramolecular charge-transfer to the lowest n,n* singlet followed by proton transfer.9° The biradical thus fbrmed can then pro- ceed as before. Wagner and Kamppainen58 concluded that triplet y-di- methylaminobutyrophenone, DMAB, underwent charge-transfer complex for- mation and that in methanol this led to biradical, which in turn could proceed to product.9° The following scheme5° is illustrative: _ N(CH3)2 DMAB 0 0H ’ } Type II products Ph . £4. 23 c. n,n* and n,n* triplets Early attempts to correlate the type II reactivity of a given elec- tronically excited triplet state in aromatic ketones with the effects of ring substituents were complicated by a change in the relative energy levels of the lowest triplets.99 Ketones with n,n* lowest triplets had respectable quantum yields and were said to be reactive, while ketones with n,n* lowest triplets had type II quantum yields which were either low or near zero and were said to be unreactive. In the case of the photoreduction of benzophenone, substitution with an electron donating group caused a small change in the observed reactivity (~2 fold).100 In the case of acetophenone, substitution with an electron donating group was reported to cause a substantial decrease in the photoreduction quan- tum yield.”l Shortly thereafter, it was shown that for both the photo- reduction of acetophenones32 and the type II reaction of valerophenone;02 these quantum yield reductions were the result of large decreases in the apparent hydrogen abstraction rates. These large changes in reactivity were caused mainly by inversion of the lowest triplets present in the un- substituted molecule, and the small changes observed in the benzophenones were probably caused only by inductive changes in the individual excited states.35 This explanation presumes that the n,n* - n,n* triplet energy gap is smaller in unsubstituted phenyl alkyl ketones than in benzophenone. In- deed, the close energetic proximity of the lowest triplets in unsubsti- tuted phenyl alkyl ketones has been sufficiently demonstrated‘°’, and it is widely accepted that electron donating ring substituents such as bromo, methyl, and methoxy groups cause the lowest triplet to be n,n* 1T1 nature.29’32al°1’1°2 24 From the valence bond description of n,n* and n,n* triplets already presented, it is not surprising that when the triplet manifold is popu- lated mainly by n,n* triplets to the virtual exclusion of n,n* triplets, observed rates of hydrogen abstraction are low. However, that ketones with lowest n,n* triplets do react presents the very interesting problem of from where do such ketones draw their reactivity. There are three a priori sources of triplet reactivity: (1) both n,n* and n,n* triplets react; (2) only the n,n* triplets react; (3) only the n,n* triplets re- act. Two approaches to this problem have appeared. One approach assumes that the rapid rate of internal conversion precludes reaction from upper n,n* triplets and that only n,n* triplets react. Further, n,n* triplet reactivity is derived from n,n* character which has been mixed in.32’9°’ 1°2 Such mixing could occur in several ways.1°"’1°5 If the carbonyl group and the aromatic part of the molecule are not parallel, the n,n* triplet might configurationally mix with the n,n* triplet. Spin-orbit coupling between the n,u* singlet and the n,n* triplet could put some n,n* character into the n,n* triplet, and vibronic interactions between the nearby triplets could also cause a mixing of the n,n* and the n,n* states. It has been suggested that configurational mixing might well explain the low observed triplet reactivity”6 of compounds with n,n* triplets much below their n,n* triplets, while spin-orbit coupling and vibronic mixing would best account for the mixing of states where they were energetically close.35 However, ketones with similar phosphorescence lifetimes should have lowest triplets possessing similar amounts of n,n* character. Such ketones would be expected to have about the same reac- tivity, but have, at least in one instance, been shown to react with vastly different observed rates.35 25 The other approach is that where n,n* and n,n* triplets are ener- getically very close to each other, photoreactivity comes from an equilibrium concentration of the reactive n,n* triplet.”5 Wagner, Kemppainen, and Schott1°5’35 have presented evidence indicating that the reactive state in some p-methoxyphenyl ketones might be the upper n,n* triplet. They showed that rates of reaction were affected by y-C substituents in precisely the same way in both the phenyl and the ani- soyl ketones, the former possessing a lowest n,n* triplet, the latter a lowest n,n* triplet. Other than the work of Wagner, Kamppainen, and Schott, very little research concerning equilibration of excited states or reaction from upper triplets has appeared in the literature. Saltiel, et aZ.1°7 have shown that the lowest triplet in benzophenone can equilibrate with the lowest singlet, the two states being about five kilocalories apart. Yang, et al.1°° have suggested that 9-anthraldehyde undergoes the Paterno-Buchi reaction with 2-methy1-2-butene from its upper n,n* trip- let state, the large energy difference (~30 Kcal) between the lowest n,n* triplet and the upper n,n* triplet making internal conversion slow enough. Wagner and Nakahira”9 have very recently shown that the lowest triplet states in the benzoyl and anisoyl chromophores of l-benzoyl-4-p- anisoylbutane thermally equilibrate before undergoing any chemical reaction. Further evidence concerning the equilibration theory and reaction from an upper n,n* triplet comprises a major part of the research pre- sented herein. Since Stern-Volmer quenching plots are heavily relied upon, the kinetics involved are pertinent. 26 d. Stern-Volmer kinetics Under the steady state conditions commonly employed in the labora- tory, the quantum yield is the only kinetic parameter which can be directly measured. Thus, it must be related to rate constants which are being sought. In the situation where reaction occurs only from the lowest n,n* triplet, such as in a phenyl ketone like valerophenone, the following scheme describes the energetic disposition of the important photo- processes. ln,1r* k. A IN 3n,n* lSC / hv k. \Rkfl knn 1 / ‘(///kd \\\~rrn E— G. . Products From this, the Stern-Volmer expression (equation 8) may be derived.110 2 2 o° = 1 + K r [A] + kq Te [A] (a) q 9 k k 4? no + M where¢f’is the quantum yield of reaction in the absence of an energy acceptor, A, and o is that in the presence of A, kq is the rate constant for energy transfer from either excited triplet state to A, and T = kn. k. + knn (k + k ) '1 (9) e k—____—-__ l -——————-—— rn d n + + n knn knn knn Acceptor concentrations which have normally been used are in the range of 0-1 M. Under this condition the Stern-Volmer plot is essentially linear. (I) = 1 + que [A] («10) However, at higher acceptor concentrations the Stern-Volmer plot will curve upward as predicted by equation (8). In this situation it must be noted that if km is not significantly smaller than kn“, equilibration 27 of the n,n* and n,n* triplets will not be achieved, and the Stern-Volmer plot will remain linear even at high concentrations of quencher. The following scheme describes the case where the n,n* triplet is a few kilocalories below the n,n* triplet, and reaction may occur from both triplets. 1n,n* k Tk\ 'sc 3 . . n,n* 1 SC I( k h\) k nn rm rn d 3n,n* (5.. /k1 E— G. . X' \ Products The resulting Stern-Volmer expression is o 2 2 -l g = l + kgre [A] + k9 Te [A] (krm + knn) 1 + on kq[A] ¢n + ¢n knn (11) where on and on are quantum yields of reaction from the n,n* triplet and the n,n* triplet respectively. In this case, at low acceptor concentra- tionS(<.02 M) the Stern-Volmer plot may be simplified, and the plot is linear. o _SL= ¢ 1 + que [A] (12) At intermediate concentrations of acceptor the plot curves downward, and at high quencher concentration the plot may again be simplified to a linear form. ¢° = ¢n + Qn T k -+ ¢n + ¢n k Te Run [A] (13) o e e nn o q k + k n n on nn A Stern-Volmer plot which curves downward is entirely consistent with reaction occurring from an equilibrium concentration of n,n* triplets, and such a plot yields values for km and kw". 28 C. Research Objectives The type II reaction was chosen for the present study because cer- tain problems inherent in other systems can be avoided.99 The C-H bond strength at the y-carbon can be varied without significantly changing the environment of the excited states of the ketone. Such successful variations are more difficult in the case of the photoreduction reaction which by contrast is a bimolecular process, since unwanted solvent effects are difficult to avoid. In addition, the products of the type II reaction are well characterized and can, in most cases, be easily analyzed by VPC. Photoreduction often produces more products which are more difficult to analyze. Several ketones with lowest n,n* triplets were studied to obtain further evidence in support of reaction from an equilibrium population of upper triplets using Stern-Volmer kinetics. The ketones chosen were reactive enough to allow the use of high enough concentrations of quencher to upset the equilibrium without quenching so much of the reaction that product quantitation would be impossible. Thus, Stern- Volmer plots of the following ketones, which undergo the type II reac- tion, were obtained. 0 @k c... CH30 R = OCH, (1), OCHZCHa (II), CHZCH2N(CH3)2 (III) The 2-naphthyl ketone corresponding to compound III, 4-dimethylamino-l- (B-naphthyl)-butanone, was studied to see from which excited state(s) reaction would occur where return from the lower triplet is slow, and the y-hydrogen is quite reactive. 29 In addition, both amino ketones can undergo charge-transfer from the amino group to the excited carbonyl. Thus, the rate of intra- molecular charge transfer to a n,n* triplet can be estimated. It is also reasonable to expect that reaction from the n,n* triplet should occur, since 2-acetonaphthone has been shown to be photoreduced by triethylamine in a respectable quantum yield (o = .7 in acetoni- trile)“°, and n,n* singlets may be quenched by amines via charge- transfer interactions.51952:S7 RESULTS A. 4-Dimethy1amino-l-(B-naphthyl)-l-butanone 1. Quantum Yields Photolysis of 4-dimethylamino-l-(B-naphthyl)-l-butanone (DMANB) in degassed solution with 313 nm light results in the production of 2-aceto- naphthone. No other products are observed by glpc analysis in any of the solvents used in photolysis. Table 1 lists the quantum yields for 2- acetonaphthone formation determined with various solvents. Figure 3 shows the fluctuation of ¢2-AN when the percent methanol in benzene is varied. A comparison of the efficiencies with which 2-acetonaphthone (2-AN) and DMANB sensitize the cis-to-trans isomerization of cis-l,3- pentadiene in benzene gives a value of 0.76 :_.04 for the intersystem crossing quantum yield (¢isc) of DMANB, since the ¢isc for 2-acetonaph- thone is unity as determined by Vesley and Prichard111 as well as by this author (Table 12). In methanol, o is found to be 0.86 :_.02. isc 2. Quenching_of 2-Acetonaphthone Production In benzene, 2-acetonaphthone production is not quenched, even in the presence of 3 M 1,3-pentadiene. However, in methanol more than 95% of this type II photoelimination is readily quenched. By subtracting the quantum yield of unquenchable reaction from the total quantum yield at each concentration of 1,3-pentadiene, a triplet state quenching plot is obtained. The slope (qu) of this plot is found to be 790 :_12 M". Figure 4 contains these Stern-Volmer plots for the quenching of 2-aceto- naphthone production in benzene and in methanol. Diffusion controlled quenching with low concentrations of t-stilbene in benzene is shown in Figure 5. A slope of 5744 i 128 M'1 is obtained. 30 31 Table 1. Quantum yields for 2-AN production from DMANB. Solvent [DMANB], M ¢2-AN Benzene 0.0201 0.0098:.0026 * 0.0505 0007430005 0.56 M Pyridine 0.0503 0.0078_.0004 in benzene Acetonitrile 0.0501 0.0105:,0005 Methanol 0.0206 0.17 1.02 * 0.0331 0.12 1.01 Methanol-d1 0.0206 0.12 +.01 * These quantum yields are accurate. Quantum yields without an asterisk are shown here to demonstrate changes in quantum yields with changes in solvent. 32 II II 0.15 F .1 Cl 0.10 - ¢ 2-AN . I. 0.05 T '1 I. fi 1 l 50 100 percent methanol Figure 3. ¢2-AN as a function of the percent methanol in benzene solvent. 33 30 .. 25 ‘+ 0 . C1 20 — ID ¢°/¢ II 15 ‘. 10 5 1 AL lMIAl L l 1' u l 7 I 0.5 l 2 3 [Quencher], M Figure 4. Stern-Volmer plots for l,3-pentadiene quenching of 2-AN for- mation from DMANB in benzene (A), in methanol (0), and in methanol with the quantum yield for reaction from the short- lived state subtracted from 0° and 0 (II). 34 5 .. 4 .. 3 .. ¢°/¢ 2L 1 1 I, I 1 ‘ L, .4 1 2 3 4 5 [t-Stilbene], x 10“ M Figure 5. Stern-Volmer plot for t-stilbene quenching of 2-AN formation from DMANB in benzene. 35 B. p-Methoxy-y-dimethylaminobutyrophenone 1. Quantum Yields p-Methoxy-y-dimethylaminobutyrophenone (p-MDMAB), 0.03 M in degassed benzene solutions, undergoes type II elimination and cyclization when irradiated with 313 nm light. The quantum yield for the former process is 0.028 :_.002 when conversion to p-methoxyacetophenone (p-MAP) is taken to 10%. At 20% conversion to p-MAP the quantum yield for p-MAP appearance (¢p-MAP) is 0.025 i.-002 while the quantum yield for dis- appearance of p-MDMAB is 0.055 1 .005. In the presence of 0.5 M and l M pyridine, ¢p-MAP is 0.043 :_.003 and 0.046 :_.004, respectively. The intersystem crossing quantum yield (¢isc) 0f p-MDMAB is deter- nfined from the quantum efficiencies with which this ketone sensitizes the cis-to-trans isomerization (¢c+t) of various concentrations of cis-l,3-pentadiene in benzene. From the reciprocal of the intercept obtained by plotting 0‘555/¢c+t) versus l/[c-P]O (Figure 6; [c-P]o 5 initial concentration of l,3-pentadiene), the intersystem crossing quantum yield, 0.85 :_.02 is obtained. By contrast, ¢isc for p-methoxy- acetophenone is within experimental error of unity (o. = 0.99 :_.03) 15C based on repeated comparisons of the ¢c+t of cis-l,3-pentadiene sensi- tized by p-MAP with that sensitized by acetophenone, and benzophenone. 2. Quenching of Excited p-MDMAB l,3-Pentadiene and l,3-cyclohexadiene quench the production of p—MAP ‘ and 1-(p-methoxyphenyl)-l-hydroxy-2-dimethylaminocyclobutane from p-MDMAB. Both of these products are produced to a measurable extent in 3 M 1,3- cyclohexadiene. Using 1,3-pentadiene at concentrations ranging from 0 to 2.6 M, a plot of relative quantum yield of p-MAP production (4° -MAP/¢p-MAP) versus quencher concentration (Figure 7) curves downward p 36 10 20 30 [C'PJ-ls M-1 Figure 6. Dependence of quantum yields f0r p-MDMAB sensitized cis-to- trans isomerization of cis-l,3-pentadiene on diene concentra- tion in benzene. 37 6 _. AL AA AL 5 _. O ‘ O O 4 _. AL ¢°/¢ 0 3r— AL O 2 ._ AA 1 IL, I 1 2 [Quencher], M Figure 7. Stern-Volmer plot for l,3-pentadiene quenching of p-MAP forma- tion from p-MDMAB (0 0.03990 M, A 0.07388 M); 313 nm irra- diation. 38 with an initial slope, k T, of 9.3 :_.8 M". The intercept/slope of the q plot in Figure 6 gives a k r of 9.0 :_.4 M", in good agreement with the q initial slope of the Stern-Volmer plot. The final slope and intercept of this Stern-Volmer plot are not considered reliable because of a significant change in solvent charac- teristics at high concentrations of l,3-pentadiene. This change is evidenced by the fact that ¢S-MAP/¢p-MAP for p-methoxy-v-diethylamino- butyrophenone actually decreases as the 1,3-pentadiene concentration is raised above 4 M. l,3-Cyclohexadiene is not expected to produce sig- nificant solvent changes in benzene at high concentrations. Using l,3- cyclohexadiene at concentrations ranging from 0 to 5 M results in the Stern-Volmer plot shown in Figure 8. The initial slope is 9.4 :_1.1 M“; the final slope and intercpet are 0.19 :_.01 M'1 and 8.3 :_.1 respectively. C. a-Alkoxy Ketones l. p-Methoxy-a-ethoxyacetophenone a. Quantum yields The irradiation of p-methoxy-a-ethoxyacetophenone (p-MEAP) in de- gassed benzene solution with 313 nm light results in the production of p-methoxyacetophenone and l-(p-methoxyphenyl)-l-hydr0xy-2-methyl-3- oxetane. The quantum yield for p-MAP appearance is 0.46 :_.04. The intersystem crossing quantum yield as obtained by taking the reciprocal of the intercept of a plot of 0.555/0c+t versus 1/[c-P]0 (Figure 9) is 0.90 i .03. b. Quenching of excited p-MEAP The formation of p-MAP and l-(p-methoxyphenyl)-l-hydroxy-2-methyl-3- oxetane is quenched in the presence of l,3-pentadiene. The Stern-Volmer plot for quenching p-MAP production bends downward as can be seen in 39 ¢°/¢ l J .J l 1 2 3 4 [Quencher], M Figure 8. Quenching of excited p-MDMAB by l,3-cyclohexadiene. 4O 10 20 [c-P]", M'l Figure 9. Dependence of quantum yields f0r p-MEAP sensitized cis-to- trans isomerization of cis-l,3-pentadiene on diene concen- tration in benzene. 41 Figure 10. From the initial slope a k r of 12.1 :_.5 M'1 is obtained. This is verified by the sensitization :lot in Figure 9 where the inter- cept/slope gives a value of 11.1 :_.6 M'”1 for qu. Furthermore, as the concentration of l,3-pentadiene is increased to its maximum, the ratio of the 3-oxetane to p-MAP decreases from 0.97 at 0 M 1,3—pentadiene to 0.27 in neat l,3-pentadiene (Table 2). 2. prethoxy-a-methoxyacetophenone a. Quantum yields p-Methoxy-a-methoxyaceotphenone (p-MMAP) produces p-MAP and 1-(p- methoxyphenyl)-1-hydroxy-3-oxetane when irradiated in degassed benzene solution with 313 nm light. The quantum yield for p-MAP production is 0.71 i .07. Values of 0.72 for p-MAP production and 0.21 for oxetane production have previously been determined by Lewis and Turro.112 The ¢isc for p-MMAP, determined from a sensitization plot (Figure 11) in the same way as for p-MEAP, is 0.99 :_.01. b. Quenching_0f excited p-MMAP As with p-MEAP, the appearance of both p-MAP and l-(p-methoxy- phenyl)-l-hydroxy-3-0xetane is quenched by 1,3-pentadiene. From the initial slope of the Stern-Volmer plot for quenching p-MAP production (Figure 12), a qu of 63 :_3 M'1 is found. From the sensitization plot shown in Figure 11, qu is 56 :_3 M". Lewis and Turro112 have determined the qu value at low concentrations of 1,3-pentadiene (.01 - .l M) to be 42 M". However, in the present study the Stern- Volmer plot is found, as expected, to curve downward at higher concen- trations of quencher as shown in Figure 12. Table 2 lists the ratio of 3-oxetane to p-MAP in various solvents. The ratio diminishes signifi- cantly only in 1,3-pentadiene. 42 10 ¢°/¢ l l l l 2 3 [Quencher], M Figure 10. Stern-Volmer plot for l,3-pentadiene quenching of p-MAP for- mation from p-MEAP in benzene. 43 Table 2. Molar ratios of 3-0xetane to p-MAP from the photolysis of p-MMAP and p-MEAP in various solvents. * p-MAP, mole p-MMAP benzene 0.31 l,3-pentadiene 0.13 cyclohexene 0.39 cyclopentene 0.32 cyclohexane 0.38 p-MEAP benzene 0.97 1,3-pentadiene 0.29 The estimated flame ionization detector response ratios of p-MMAP to l-(p-methoxyphenyl)-1-hydroxy-3-oxetane and p-MEAP to 1-(p-methoxyphenyl)- l-hydroxy-Z-methyl-3-0xetane are 0.74 and 0.58, respectively. Multiplying the peak area ratios in tables 10 and 13 by the appropriate response ratio gives the above numbers. 44 0.555 ¢c+t 1 l I I l 1 1 LE 1 l l 2 3 4 5 6 7 8 9 10 [C'PJ-ls "-1 Figure 11. Efficiencies of sensitization of the cis-to-trans isomeriza- tion of l,3-pentadiene by p-MMAP as a function of diene con- centration in benzene. 45 A 2 . A O A 30- O A 20.- ¢°/¢ 1o . L. 1 C II ‘ . L 1 1 1 2 3 [Quencher], M Figure 12. Stern-Volmer plot for l,3-pentadiene quenching of p-MAP formation from p-WIAP (0 0.0869 M, A 0.106 M) in benzene. ‘I 46 3. Methoxyacetone a. Quantum yields Irradiation of methoxyacetone in degassed benzene solution with 313 nm light produces acetone with a quantum yield of 0.41:,02. Using a variety of temperatures and nitrogen flow rates, glpc analysis on a column containing 19.4% (w/w) FFAP on chromosorb P shows no sign of any products other than acetone. The same is true using a 6' x 1/8" aluminum column packed with 4% QF-l, 1.2% carbowax 20 M on 60/80 chromosorb G (the same type of column used in analyzing for the oxetanes 0f p-MMAP and p-MEAP). l-Methyl-1-hydroxy-3-oxetane may undergo an acid catalyzed reaction on the FFAP column, but it should not do this on the QF-l, carbowax column. The quantum yield of disappearance of methoxy- acetone in degassed benzene solution is found to be 0.47 i.-02- b. Quenching of exgjtgd methQXXAQQIQne. Up to l M 1,3-pentadiene causes no effect on the production of acetone from methoxyacetone. In 6 M or more l,3-pentadiene the quantum yield for acetone formation is reduced to 0.37 :_.02. Thus, the quantum yield for acetone appearance via a triplet reaction is on the order of 0.04. By sensitizing the cis-to-trans isomerization of various concen- trations of cis-l,3-pentadiene with methoxyacetone, a plot of 0’555/¢c+t versus [c-P];1 is constructed (Figure 13), from which the reciprocal of the intercept gives a value of 0.091 :_.001 for ¢isc and the intercept/ slope gives a value of 2.41 :_.03 M'1 for qu. In making this plot, the amount of trans-1,3-pentadiene produced from sensitization by the acetone photoproduct is subtracted from the total trans-l,3-pentadiene measured. 47 40- 30‘- 20'— 0.555 ¢c+t 10" 1 1 1 1L 1 1 l 2 3 4 5 6 [C'P]-13 ”-1 Figure 13. Dependence of quantum yield for methoxyacetone sensitized cis-to-trans isomerization of cis-1,3-pentadiene on diene concentration in benzene. 48 D. Ben20phenone and 4,4'-Dimethoxybenzophenone Phosphorescence Quenching Quenching plots for the triethylamine and 2,5-dimethyl-2,4-hexadiene quenching of benzophenone and 4,4'-dimethoxybenzophenone phosphorescence are shown in Figure 14. Resulting k r values are listed in Table 3. 0 Assuming kq = 5 x 109 sec'1 for the diene, the kq values for triethyl- amine quenching of benzophenone and 4,4'-dimethoxybenzophenone phospho- rescence are calculated to be 1.39 x 109 sec'1 and 1.40 x 109 sec“, respectively. 49 I 6- 51- 4— O ¢°/¢ A 3" A O O, 2 r . A .— O ‘ A 1 I l l l l l 1 2 3 4 5 6 [Quencher], M x 105 Figure 14. Stern-Volmer plots for the quenching of benzophenone phos- phorescence by 2,5-dimethyl-2,4-hexadiene ((3), and by tri- ethylamine (25), and of 4,4'-dimethoxybenzophenone phospho- rescence by 2,5-dimethyl-2,4-hexadiene (OD), and by tri- ethylami ne (A). 50 Table 3. Quenching constants obtained from quenching the phosphorescence emission of benzophenone and 4,4'-dimethoxybenzophenone with 2,5-dimethyl-2,4-hexadiene and triethylamine. ketone quencher k r x 10‘5, q M" benzophenone 2,5-dimethyl-2,4—hexadiene 0.54:,02 triethylamine 0.15:,02 4,4'-dimethoxybenzophen0ne 2,5-dimethyl-2,4-hexadiene 1.25:,01 triethylamine 0.35:,04 DISCUSSION A. Reactivity Where Lowest 3n,n* Is Far from 3n,n* The photoinduced type II elimination of DMANB in benzene (¢2-AN = 0.0098) cannot be quenched, even by 3 M l,3-pentadiene (Figure 4), in spite of the fact that triplet energy is being transferred to the diene as manifested by diene isomerization. The intersystem crossing quantum yield of DMANB in benzene is 0.76, whereas that for 2-acetonaphthone is unity. Since the n,n* triplet state of DMANB is considerably lower in energy than the n,n* singlet (AEln,n*-3 ~ 15 kcal’s), the rate con- N,N- stant for intersystem crossing (kisc) could be reduced relative to its value in a simple phenyl ketone. This may well contribute to the ability of DMANB to undergo singlet y-H abstraction. The fact that product formation is not quenched by l,3-pentadiene (a triplet quencher) shows that the n,n* triplet state of DMANB is not the origin of any 2-AN, either via y-H abstraction or via a charge transfer interaction. It must be concluded that the y C-H bond in DMANB is not reactive enough to allow n,n* triplets to y-H abstract at a rate competitive with other decay processes of n,n* triplets, the most notable of which is the charge transfer interaction. This information is useful in analyzing the behavior of other ketones which have lowest n,n* triplets. Furthermore, it is likely that the n,n* triplets of DMANB do not pro- duce a significant amount of type II product, since type II elimina- tion from the lowest n,n* triplet state of y-dimethylaminobutyrophenone (DMAB) is readily quenched by l,3-pentadiene, and there is no reason to believe that substituting a naphthyl group for the phenyl group in DMAB should cause the lifetime of the n,n* triplet state to decrease. 51 52 Yang and Shani113 have reported that B-valeronaphthone undergoes type II elimination with an approximate quantum yield of 0.002. Even 8 M 1,3-pentadiene did not attenuate the 2-AN yield. This along with the apparent lack of any cyclobutanols indicates that 2-AN is produced by B-valeronaphthone only via its n,n* singlets. Interestingly, kisc in B-valeronaphthone can be estimated from the above quantum yield along with values for k: and pS (k: is the rate constant for excited singlet y-H abstraction, and pS is the probability that any metastable inter- mediates which result from singlet y-H abstraction go on to type II k: s elimination product), since 0:1 = . p . Using the values k: + kisc 8.6 x 10° sec"1 for k: and 0.14 for p5 calculated from data obtained on 2-hexanone by Yang, et al.11“, kisc is calculated to be 1011 sec". If this number is accurate, k: and/0r pS is large enough in DMANB that a reduction in kisc is not required to account for singlet reaction. Recently, Coyle115 has reported that a-alkoxy-a-acetonaphthones, which have virtually the same energetic distribution of their excited states as 2-AN, undergo type II elimination in benzene with quantum yields comparable to that for DMANB (0II = 0.0016 for o-methoxy-a-aceto- naphthone; 011 = 0.010 for a-ethoxy-o-acetonaphthone; 011 = 0.015 for a-propoxy-a-acetonaphthone). These quantum yields are not affected by l M l,3-pentadiene or cyclopentadiene even though triplet energy is transfered to the dienes. This along with the fact that no oxetanols (type II cyclization products commonly produced in high yields by trip- let a-alkoxyacetophenones‘12) are observed, indicates that these o-alkoxyacetonaphthones react via their lowest excited singlet state. 53 It is interesting to note that the reciprocals of the triplet lifetimes of DMAB, armethoxyacetophenone, and urethoxyacetophenone are 8.4 x 109, 3.2 x 10°, and 8.4 x 10° sec", respectively, and the reciprocals of the excited singlet lifetimes should follow the same trend. Further- more, type II singlet quantum yields for the respective naphthyl com- pounds are 0.0098, 0.0016, and 0.010. Since the a-amino ketones do not undergo y-H abstraction as rapidly as the a-alkoxyketones, some of the excited singlets of DMANB may be producing 2-AN via a charge transfer interaction. An alternative explanation for the high ¢II in DMANB as compared with the more reactive a-alkoxy-a-naphthyl ketones is that ornaphthyl substitution in some way causes more singlet inefficiency in reaction from the excited singlet state than B-substitution. In methanol, DMANB produces 2-AN (o = 0.17) from two excited states, one quenchable using l,3-pentadiene and one not, as shown by the curved Stern—Volmer quenching plot in Figure 4. This plot becomes linear when the quantum yield for unquenchable reaction is subtracted from 0° and 0, indicating that 2-AN is coming from only one unquenchable upper excited state. Since the lifetime of the quenchable excited state, as determined by the t-stilbene quenching shown in Figure 5, is 0.8 x 10'5 sec (qu = 5,744; kq = 7.5 x 10° sec1 in methanollls), this excited state must be the lowest n,n* triplet state. There is no reason for a n,n* state to y-H abstract in methanol but not in benzene, and it follows that this reaction must proceed via a charge transfer process as is the case for the photoreduction of 2-AN by triethylamine."° The only other reported example of type II photoelimination from a n,n* triplet is the photochemical reaction of esters of aromatic carboxylic acids.117 Another example of intramolecular hydrogen abstraction by a n,n* triplet 54 involves the following interesting reaction:11° .1 HQ The rate constant for abstraction of a hydrogen atom 0 to the carbonyl is reported to be 9.3 x 10“ sec". This respectable rate constant for n,n* 0-H abstraction is the result of a very favorable geometry for hydrogen abstraction by the carbonyl group as well as very reactive C-H bonds at the 6 carbon atom. The behavior of DMANB in benzene and in methanol is illustrated in the following scheme, where the percentages shown are relative to the originally formed n,n* singlets: 1n ,11* 76768 ’ 311 ,11* 76%8 872M, 87%M 3CT Some of the biradicals return to the ground state when the reaction is run in benzene, because 0.6 M pyridine increases the quantum yield by ~5%. In the absence of any amine, the quantum yield in benzene is estimated to be about 0.009. The reduced amount of unquenchable reaction 55 in methanol (°2-AN = 0.007 in the presence of l M l,3-pentadiene) as compared with that in benzene is presumably the result of a reduced rate of y-hydrogen abstraction by the excited singlet state caused by a decrease in the ability of the lone pair of electrons on nitrogen to stabilize a radical center. The fact that all triplet reaction in DMANB proceeds via a charge transfer complex of some sort is useful. It has been suggested that DMAB, which has a lowest n,n* triplet, undergoes a good deal of its type II reaction via a charge transfer complex in meth- anol.116 The partitioning of the DMAB triplet state between charge transfer and y-H abstraction, however, has not been determined. This can now be estimated if the assumption is made that the same fraction of charge transfer triplets in DMAB go to biradical as in DMANB (0.19 of °CT DMANB goes to biradical in methanol). While the n,n* and n,n* triplets have different electronic distributions", after charge trans- fer the resulting complexes may well be very similar so that the above assumption could be reasonably valid from that standpoint. In methanol, DMAB intersystem crosses and type II eliminates with quantum yields of 0.72 and 0.25, respectively.116 The quantum yield in methanol for type II elimination from the triplet of DMAB is described in terms of ky_ H and kct in equation 14. ¢II = 0.35 = ky-H + 0.19 kct (14) ky-H + FCi: ky-H + kct Letting ky-H =1; and kct = 1-0.a is calculated to be 0.20. ky-H +kct ky-H +kct The following scheme approximates the behavior of DMAB in methanol: 1 3 n,1r* 72% $ n,1r* / 28/ 47 X N / 0H - \ DMAB 0% . T Q 125% type II elimination B. The Nature of the Charge Transfer Complex in DMANB The large discrepancy between ¢2~AN in benzene and in methanol cannot be explained by an increased rate of charge transfer in meth- anol, since methanol hydrogen bonds with the lone pair electrons on nitrogen and thereby reduces kct as well as kv-H‘ Rather, charge trans- fer complexation occurs in both solvents, but in methanol some of the complex proceeds to biradical which then goes on to product. In acetonitrile ¢2-AN is only a factor of 1.4 greater than in benzene even though electron transfer is known to proceed considerably faster in this solvent than in benzenei7’llglhe low quantum yields for product forma- tion in acetonitrile and in benzene indicate that the complex is too tight to open up sufficiently for y-proton transfer to occur. In methanol, the ketyl radical anion may be solvated sufficiently to allow the positive nitrogen to move away and a y-hydrogen to approach the carbonyl. This explanation has been used to rationalize the large methanol quantum yield enhancement in the type II reaction from DMABI16 A small reduction in the quantum yield for 2-AN fonnation is observed in methanol-d1 relative to the quantum yield in methanol (ogfiRgH/ogfiigo= 57 1.4). Actual proton transfer from solvent to the complex may be occur- ring although there are very few systems with known solvent deuterium isotope effects which are mechanistically similar. It may be possible to compare the above deuterium isotope effect with that in the following system where ¢H/¢D = 1.6 (o = the quantum yield for benzophenone dis- appearance), and the concentration of cyclohexylamine used was in the range of 0.7-1 M.5“ H(D) 0- Ph 00* + 21-» PhCPh +° /H 2 K . N\ ——>Pn,c0 + H w H H(D) l 0H(D) ”(NH PhCPh+ <:>‘—N<: OAK: Here the probability of product formation is dependent on the ease of proton abstraction from the a-carbon and does not change greatly with deuterium abstraction. Another system which may be used for comparison is the following: ( ) CH, 03 0 l PhCOCF: + PhCH.(Da) -——> PhCCFa @ 0H(D)\ PhCOCFa + PhCH3(03) PhCCF, + PhCH2(Dz) In this instance the isotope effect on the quantum yield (for bibenzyl formation from coupling of the benzyl radicals) at toluene concentrations greater than 2 M is 3.5 (oHloD = 3.5).73 Whether complete protonation of the complex by the solvent occurs or hydrogen bonding between the ketyl radical anion and the solvent occurs cannot be distinguished from the data present here. In fact, something between these extremes may be happening. At all events, it is reasonable to conclude that the 58 hydroxyl group of methanol is involved in the quantum yield enhancement found in methanol. The fOllowing scheme depicts the probable mechanism for type II product formation from DMANB in methanol. N/ DMANB —“—"—> DMANB” 0MANBl*———> DMANB°* a 0MANB3*———> 301 5 31:1 ———> DMANB 301 + CH,0H «NET- - --H00H3 s 301- - -H00H3->0MANB H ;( 31:1- - -H00H,—>BR 2 0O BR——§ 2-AN It is interesting to note that the plot of ¢2-AN versus concentra- tion of methanol in benzene (Figure 3) never levels off; so that what- ever effect methanol has on the mechanism, it has not reached its 1naximum even in neat methanol. Quite possibly, the rate of proton transfer or partial proton transfer to the ketyl radical anion is not 1naximized. If this is the case, further quantum yield increases should occuriin more acidic solvents which are not so acidic that they proto- ruate the amino group. 2,2,2-Trifluoroethanol is apparently too acidic, since the quantum yield in this solvent is diminished relative to that CF3CH20H/¢CH30H = 0.3). C. Reactivity Where Lowest ’n,n* Is Energetically Close to 3n,n* in methanol (0 l. Photoreactivitygof,p-Methoxy-yrdimethylaminobutyrophenone In the case of p-methoxyphenyl alkyl ketones, it is known that the 'n,n*’triplet state is 3 kcal lower in energy than n,n* triplet state?5 59 Furthermore, by analogy with DMANB, the n,n* triplet state of p-MDMAB in benzene is not expected to give any perceptible type II products either via charge transfer or via y-H abstraction. Any possible but improbable increase in the rate of'y-H abstraction by the n.n* triplet state brought on by an increase in n,n* character in the n,n* state is undoubtedly more than matched by a much enhanced rate of charge transfer to the n,n* triplet state. The result must be that no type II products are formed via the n,n* triplet state of p-MDMAB in benzene. The downward curvature in the Stern-Volmer plots (Figures 7 and 8) for 1,3-pentadiene and 1,3-cyclohexadiene quenching of type II elimina- tion from p-MDMAB shows that an excited state shorter lived than the lowest triplet state is reactive. The observed curvature is not caused by the high concentration of l,3-pentadiene used in these quenching experiments, f0r it has previously been shown that significant devia- tions from linearity do not occur until concentrations of l,3-pentadiene in excess of 4 M are used.‘°° The n,n* triplet and singlet are the only excited states that could be responsible for this reactivity. DMAB does not undergo singlet reaction in spite of the fact that only fifty-eight per cent of its singlets become triplets in benzene.116 By contrast, eighty-five per cent of the singlets in p-MDMAB become triplets in benzene. If the p-methoxy group reduced kisc’ one might expect an increase in the lifetime of the n,n* state, which could then undergo some reaction. A decrease in kisc would, however, cause a significant drop in ¢isc if the rates of hydrogen abstraction (ki-H) and charge transfer (kit) did not drop significantly. Substitution of p-methoxy groups in benzophenone causes little change in the rate of charge trans- fer to the n,n* triplet (kEt = 1.39 x 10° for triethylamine reacting 60 with benzophenone; kit = 1.40 x 10’ sec‘1 for triethylamine reacting with 4,4'-dimethoxybenzophenone). The same should be true for the n,n* singlet state in p-MDMAB. By comparing changes in the rate constants for triplet y-H abstraction (kt-H) in going from valerophenone to y-methylvalerophenone and DMABIIS’12 with changes in the rate constants for singlet y-H abstraction in going from 2-hexanone to 5—methyl-2-hexanone,11“ a probable k:_H value of ~ 3 x 109 in 5-dimethylamino-Z-pentanone is arrived at. The n,n* singlet of DMAB should y-H abstract with much the same rate constant. This is only 3% of kisc (assumed to be 1011 sec'1 as it is in benzophenone’). so that even if ki-H dropped to zero in DMAB with p-methoxy substitution, ¢isc would not increase perceptibly. It follows, then, that since ¢isc is higher in p-MDMAB than in DMAB, kisc must not decrease much in going from DMAB to p—MDMAB, for a large decrease (~10 x) in kisc would require a very large decrease in kit (~50 x). In the absence of a large decrease in kisc’ it is very unlikely that p-methoxy substitution in DMAB would result in singlet reaction. It must be concluded then, that p-MDMAB does not undergo significant reaction from its singlet. Furthermore, given the strong probability that the n,n* triplet state does not y-H abstract fast enough to give any measurable type II products as is the case with DMANB, downward curvature in the Stern- ‘Volmer plots cannot be caused by reaction from two nonequilibrated trip- lets. In addition, the lack of n,n* singlet reaction as is apparently the case with DMAB precludes the possibility that the downward curvature is the result of the n,n* singlet and triplet states producing type II product, triplets being quenchable and the singlets being unquenchable. 61 The only remaining explanation for the downward curvature is reaction from the equilibrated upper n,n* triplet state. Since the upper reactive excited state in p-MDMAB is quite probably the n,n* triplet, the downward curvature in the Stern-Volmer plots in Figures 7 and 8 may be interpreted as further evidence that the type II elimination reaction in some p-methoxyphenyl ketones is derived from an equilibrium concentration of upper n,n* triplets.105 Recently, Wagner and Nakahira1°° have shown that the lowest triplet states of the two chromophores in l-benzoyl-p-anisoylbutane equilibrate before any reaction can occur. This lends support to the conclusion that the n.n* and n,n* triplets in p-MDMAB are equilibrated. From the Stern-Volmer quenching plot in Figure 8, it is possible to calculate the rate constants for intrachromophore energy transfer, k and knn’ using equations 12 and 13, since the initial slope = k 1 1111’ q e’ the final slope = xnflqueF-l, and the extrapolated intercept of the = -1 final slope kflnTeF where X = k1m ’ and F = °n . on no n n If it is assumed that the o“ = 0, the following quantities result: Te = 1.9 x 10'° sec; X = 0.02; k = 4 x 10° sec"; k = 2 x 1011 sec'l. n nn nn n Thus, kTm and knn are likely fast enough to allow the n,n* and n,n* triplets to equilibrate before either can undergo charge transfer, the fastest mode of decay available to either excited state. The following equation describes the triplet lifetime under equilibrium conditions: 14. -- 4.42:. + (<20 + 4.42. + 40 1... Substituting the values of 0.8 x 10° 'sec'1 for k$fiH and 7.2 x 10° sec‘1 for knTT ct’ as found for DMAB,116 in this equation, and assuming k$fH << 62 kgg, a value of 4 x 10° sec"1 for k2: is calculated from the above num- bers found for To and Xnn (Xfiw = 1-Xn"). As with DMAB, the type II reaction of p-MDMAB in benzene most prob- ably proceeds by direct y-H abstraction.11° The following scheme depicts the behavior of p-MDMAB in benzene: ’ ln.11* 8512 3n.n* 10% 15% I \ 75% I OH / 301 hp 15% 75 p-MDMAB 5.5% (0%) 4.6% (10%) type II products All of the percentages in the above scheme are relative to the originally formed n,fl* singlet. The value of 4.6% is arrived at by deducting the enhancement in the disappearance quantum yield (¢_K = 0.055) caused by the presence of 0.03 M amine in the form of the aminoketone.122 The value of 5.5% is determined from the fact that the quantum yield for p-methoxyacetophenone (¢p-MAP) increases to 0.046 in the presence of l M pyridine. This is an increase of a factor of 2.2, since in the absence of any amine ¢p-MAP would equal 0.021. Thus, the quantum yield for the total loss of ketone in the absence of any amine (¢_K = 0.046) is increased by a factor of 2.2 in the presence of 1 M pyridine. Values in parenthesis indicate precentages in the presence of l M pyridine. In DMAB, about 10% of the n,n* singlets become biradicals via the n,n* triplet state11°, and the same is true for p—MDMAB. 0n the other hand, the percentage of the triplets of p-MDMAB which undergo charge transfer is 88% as compared with 83% for DMAB. This is reasonable 63 because of the very large population of n,n* triplets in p-MDMAB which do not hydrogen abstract but can undergo charge transfer. 2. a-Alkoxy Ketones In addition to the p-MDMAB ketone, p-methoxy-a-methoxyacetophenone (p-MMAP) and p-methoxy-a-ethoxyacetophenone (p-MEAP) demonstrate non- linear Stern-Volmer quenching behavior (Figures 10 and 12). Here again it is possible to use large enough concentrations of quencher to upset the equilibrium between the n,n* and n.n* triplets. p-MMAP and p-MEAP have intersystem crossing quantum yields of 0.99 and 0.90, respectively. Maximum rate constants for excited singlet reaction in p-MMAP and p-MEAP may be calculated from these values of o. , if the above intersystem lSC crossing quantum yields are reliable and if kisc can be assumed to be 1011 sec"1 for these ketones. For p-MMAP, k: is then 10° sec“, and k: is 11 x 109 sec-1 for p-MEAP. For methoxyacetone ¢isc = 0.09, and assuming kisc = 4 x 10° (as for other aliphatic ketones‘l“), the rate constant for singlet reaction is calculated to be 4 x 10° sec". From the qut obtained from Figure 13 (qut = 2.4 M“), the rate constant for triplet reaction is determined to be in the range of 2 x 10° sec". (kq is taken to be 5 x 10° sec'l.) t Extrapolating the rate differences between kr and k: in methoxyacetone to a-methOXyacetophenone (k: = 3 x 10° sec")112 and a-ethoxyaceto- phenone (k: = 8 x 10° sec")‘1°, k: should be on the order of 6 x 10° sec"1 and 2 x 1010 sec'1 for a-methoxyacetophenone and a-ethoxyaceto- phenone respectively. Since the p-methoxy group reduces the electro- philicity of the n,n* singlet and thereby reduces the rate of'y-H abstraction, the values for singlet reaction calculated above for p-MMAP and p-MEAP are reasonable, and it may therefore be concluded that at _—., .s'~_- s 64 least p-MEAP may undergo some type 11 product formation from its excited singlet. Indeed, type II elimination product is coming from the n,n* singlet of p-MEAP as manifested by a decrease in the ratio of the cyclization to elimination product from 0.97 when no triplet quencher is present to 0.29 when the reaction is run in neat l,3-pentadiene. p-MMAP exhibits similar behavior. In benzene the cyclization to elimination ratio is 0.31 while in neat l,3-pentadiene the ratio drops to 0.13. This is not a solvent effect, since various solvents such as cyclopentene produce no significant change in the peak ratio as compared with that in benzene (Table 2). The quantum yield of type II elimination from the n,n* singlet (¢II) of methoxyacetone is 0.37. Since 91% of the singlets do not cross over to the triplet manifold, 40% of the singlets which do not become triplets form type II elimination product. The same partitioning should occur in p-MMAP. Here, however, only ~1% of the singlets do not become triplets; so that ¢II for p-MMAP is estimated to be about 0.004. When this number is subtracted from each of the measured quantum yields used in preparing the Stern-Volmer quenching plot in Figure 12, the plot still curves downward. In fact, to make the plot linear, ¢II would have to be 0.02. This would require the intersystem crossing quantum yield to be 0.95, a number which lies outside experimental error. In the case of p-MEAP, the partitioning of the n,n* singlet between product formation and reversion to the ground state is not known. There is further evidence that p-MMAP and p-MEAP react from the n,n* triplet state as well as from the singlet state. In neat l,3- pentadiene there is still a significant amount of cyclization product; the quantum yield for cyclization is about 0.01 for p-MEAP and 65 approximately 0.002 for p-MMAP. This means that if all of the unquenched reaction is singlet reaction, at least 10% of the p-MEAP singlets that do not become triplets and 17% of the p-MMAP singlets that do not be- come triplets produce cyclization product. However, even 10% of these singlets cyclizing is more than would be expected from an n,n* singlet state.°1 It is of interest that if the assumption is made that no cycliza- tion products come from the excited singlet states of p-MMAP and p-MEAP, values for ¢II can be calculated from quantum yields of type II elimina- tion and cyclization in the absence of triplet quencher along with type II elimination quantum yields in neat triplet quencher. The results are shown in the following table. 0 q t s ¢elim ¢cyc ¢elim ¢cyc ¢elim ¢elim p-MMAP 0.71 0.22 0.020 0.0026 0.008 0.01 p-MEAP 0.46 .45 0.045 0.013 0.013 0.03 s elim various quencher concentrations, the Stern-Volmer plots for both p-MMAP When the respective 0 are subtracted from measured quantum yields at and p-MEAP still curve downward although less markedly as shown in Figures 15 and 16. In the case of p—MEAP the plot does not appear to have curved enough to have reached a final slope, but with p-MMAP it looks as though a final slope may have been achieved. Assuming this to be the case, the following values are obtained for p-MMAP in the sane way as f0r p-MDMAB: r = 14 x 10'9 sec; X n = 1010 sec'1 e n p-Methoxy substitution in a-methoxyacetophenone causes a twentyfold = . = 9 -1. n 0.2, knn 2 x 10 sec , kn ‘reduction in the triplet reactivity}12 whereas a two hundredfold reduction 105 is normally caused by p-methoxy substitution. In p-methoxyphenyl alkyl 66 ¢°/¢ 20“ 10'. 1' . , . Figure 15. 1,3-Pentadiene quenching of p-MAP formation from triplet p-MMAP. 67 30 T’ O O O O 20 - O ¢°/¢ 10 T O O ‘T. 1 1 l l 2 3 Figure 16. l,3-Pentadiene quenching of p-MAP formation from triplet p-MEAP 68 ketones where the alkyl side chain does not interact with the carbonyl group in such a way as to change the excited state energy levels, about 1% of the excited triplets are in the n,n* configuration.35 For this reason, the above mentioned order of magnitude difference between the effect of p-methoxy substitution on the observed triplet reactivities in a-alkoxy phenyl ketones and alkyl phenyl ketones indicates that about 10% of the excited triplets in p-MMAP are n,n*. The number cal- culated for XMT from the final slope in Figure 15 is too large by this reasoning. However, even a 10% population of upper n,n* triplets is high compared with other p-methoxyphenyl ketones, and it indicates that the n,n* and n,n* triplet states are closer together in the a-alkoxy ketones. It is possible that the nonbonding electrons of the ether oxygen interact with the carbonyl group to stabilize the n,n* triplet state. Although the 0-0 bands must be approximated, the phosphorescence spectra of p-MAP and p-MMAP do indicate that the lowest triplets have nearly the same energy. Furthermore, the lowest triplet lifetimes of p-MAP and p-MMAP are essentially the same at 77°K in 4:1 methylcyclo- hexanezisopentane; so that the lowest triplet is n,n* in both compounds. 0. Summary In benzene, DMANB produces no type II elimination product via its excited triplet states, even though the triplet manifold is heavily populated. This is evidence that charge-transfer triplets do not continue on to product in nonpolar solvents. In methanol, the n,n* triplets of DMANB produce a large amount of type II elimination product. This is a unique example of type II elimination via a n,n* triplet 69 state in a ketone. n,n* Triplets have already been shown to abstract hydrogen atoms in several instances where very reactive hydrogens are available, but in this case reaction must occur via a charge-transfer complex. The methanol solvent effect is a function specifically of protic solvents, since acetonitrile improves the quantum yield only slightly above that in benzene. A tight complex is indicated for the benzene case. Methanol must solvate the complex in such a way that at least some solvent O-H stretching occurs. This solvation allows v-proton transfer to the ketyl anion or to a complexed solvent molecule to occur. The rate constant for charge transfer from the amino group to the n,n* triplet state is found to be 10° sec". p-MDMAB does not undergo excited singlet reaction by analogy with DMAB, since kisc is not reduced by p-methoxy substitution. p-MDMAB does not undergo n,n* triplet reaction by analogy with DMANB, since charge transfer to the n,n* triplet is much faster than in DMANB, while v-H abstraction may be facilitated only slightly in p-MDMAB by the expected small increase in n,n* character in the n,n* triplet. Downward curva- ture in the Stern-Volmer plots where type II elimination from p-MDMAB is quenched is interpreted as resulting from triplet quenching upsetting the equilibrium between the n,n* and n,n* triplet states. The rate con- stant for charge transfer to the n,n* triplet state in this case is determined to be 3 x 10° sec". Stern-Volmer plots for p-MMAP and p-MEAP also curve downward. Here, however, singlet reaction is responsible, in part, for this curvature. When probable quantum yields for singlet reaction are subtracted from measured type II elimination quantum yields for both of these ketones 70 residual downward curvature is interpreted as being caused by quenching upsetting the equilibrium of the reactive n,n* triplets with the unreactive n,n* triplets. E. Suggestions for Further Research It may be possible to distinguish between charge transfer triplets formed from n,n* and n,n* triplet states. The concentration of a series of polar protic solvents with different pKa values could be varied in benzene to determine the effect of solvent acidity on the type II quantum yields of DMANB and DMAB. A difference in basicity of the negatively charged carbonyl might be demonstrated. Another useful probe would be a comparison of the solvent deuterium isotope effect on type II elimina- tion quantum yields of DMANB and DMAB. For compounds, such as DMANB, which have a large energy separation between the n,n* and n,n* triplets, the determination of the relative rates of intersystem crossing from the n,n* singlet to the n,n* and n,n* trip- let states would be interesting. In compounds which have large Tl-Tz energy gaps it is possible to preferentially quench the upper triplet with an appropriately chosen acceptor.”s Hammond, et aZ.‘2° have shown that the isomerization of diethylmaleate (DEM) and diethylfumarate (DEF) is sensitized by 2-AN (ET1 = 59 kcal) in spite of the fact that the triplet (energy of DEM is in the range of 72-77 kcal and that for DEF is 61-67 kcal. Preliminary work by this author shows that ¢DEF+DEM sensitized by 2-AN changes nonlinearly with concentration of DEF but fairly linearly with 2-AN over the concentration range studied. Further work on this system may be profitable. EXPERIMENTAL A. Chemicals 1. Ketones a. p-Methoxy-y:dimethy1aminobutyrophenone p-Methoxy-y-dimethylaminobutyrophenone was prepared from the Grignard reagent of p-bromoanisole and y-dimethylaminobutyronitrile. An ether solution of y-dimethylaminobutyronitri1e (0.187 mole) was added to the Grignard reagent of p-bromoanisole (0.27 mole) under an atmosphere of nitrogen at room temperature with stirring and at a rate which maintained vigorous refluxing. The reactants were then refluxed for five hours. The mixture was cooled in an ice bath, and 75 ml of 5.2 M aqueous ammonium chloride was added. The ether was then distilled off, and the aqueous phase was heated for an additional hour. Following this, the mixture was extracted with ether, and the combined ether extracts were extracted with cold 2 N hydrochloric acid. The cold acid solution was neutralized with sodium hydroxide and extracted with ether. The combined ether extracts were washed with a saturated aqueous solu- tion of sodium chloride and dried. After the ether was removed, the residue was vacuum distilled and gave a 36% yield of p-MDMAB (b.p. 112- 114°C at .05 mm, uncorrected). F0r further purification, p-MDMAB was run through a short column of alumina and then vacuum distilled. GLPC analysis showed the p-MDMAB to be 99% pure. The 70eV mass spectrum of p-MDMAB shows a parent peak at m/e 221 in .addition to the following characteristic fragments: m/e 150 (loss of N,N-dimethylvinylamine), m/e 135 (anisoyl), m/e 107 (anisyl), m/e 71 (loss of p-methoxyphenacyl radical and a hydrogen atom), and the base 71 72 peak at m/e 58 (N,N-dimethyliminonium ion). The proton nmr chemical shifts in deuterated chloroform are: 6 1.93 (quintet, two protons B to the carbonyl group), 2.24 (singlet, six protons on amino methyl groups), 2.33 (triplet, two protons a to the amino group), 2.93 (triplet, two protons a to the carbonyl group), 3.84 (singlet, three protons on the methoxy group), 6.92 (doublet, two protons ortho to the carbonyl group), and 7.98 (doublet, two protons meta to the carbonyl group). The ir spec— trum of this compound has strong bands at 2769 cm‘1 and 2820 cm-1 result- ing from the symmetric C-H stretching of the methyl groups on nitrogen, a very strong band at 1678 cm'1 produced by the C=0 stretching vibra- tions, and medium to weak bands at 1601 cm", 1575 cm", 1510 cm", and 1415 cm'1 produced by benzene ring stretching vibrations. The ultra- violet spectrum of p-MDMAB in heptane exhibits a Amax at 214 nm (e = 11,900) and a Amax at 263.5 nm (e = 12,600). The mass, nmr, ir, and uv spectra were produced on a Hitachi Perkin- Elmer RMU-6 mass spectrometer operated by Mrs. Lorraine A. Guile; a Varian T-60 nmr spectrometer; a Perkin-Elmer model 237-8 infrared spectrometer; and a Cary 15 spectrometer, respectively. Unless otherwise indicated, the same is true for mass, nmr, ir, and uv spectra henceforth refered to. b. p-Methoxy-a-methoxyacetophenone p-Methoxy-a-methoxyacetophenone was prepared by reacting the Grignard reagent of p-bromoanisole with methoxyacetonitrile. After the addition of inethoxyacetonitrile (0.2 mole) to the Grignard reagent (0.2 mole) was com- plete, the mixture stood at room temperature for two hours. Three hundred inilliliters of water and cracked ice was added to the cooled mixture; and “then 70 ml of cold dilute sulfuric acid was added. After two hours of stirring, the ether layer was separated, and the aqueous layer was 73 extracted with ether. The combined ether layers were washed first with 5% aqueous sodium carbonate and then with water. The other solution was then dried and the ether removed. The residue was vacuum distilled, giving p-MMAP with a boiling point of 101 - 103°C at 0.25 mm. After f0ur recrystallizations from low boiling petroleum ether, the yield of 99.9% pure (determined by glpc) p-MMAP (m.p. 40 - 41.5°C, uncorrected) was 30%. The 70eV mass spectrum of p-MMAP includes the following important peaks: m/e 180 (molecular ion), m/e 150 (loss of formaldehyde), m/e 135 (base peak, anisoyl), m/e 107 (anisyl), m/e 92 (anisyl less a methyl radical), and m/e 77 (phenyl cation). The proton nmr spectrum of p-MMAP includes the following in carbon tetrachloride: 0 3.32 (singlet, three protons on the a-methoxy group), 3.78 (singlet, three protons on the p-methoxy group), 4.73 (singlet, two protons o to the carbonyl group), 6.78 (doublet, two protons ortho to the carbonyl group), 7.76 (doublet, ‘bwo protons meta to the carbonyl group). The ir spectrum includes a strong band at 1127 cm’1 from asymmetric stretching of CHz-O'CHg, a very strong band at 1687 cm'1 from C=0 stretching, and medium bands at 1599 cnf“, 1570 cm", 1505 cm", and 1415 cm'1 from benzene ring stretching ‘vibrations. The uv spectrum of p-MMAP in heptane has two Amax’ one at 217 nm (e = 11,500), and one at 267 nm (e = 15,500). c. p-Methoxy-o-ethoxyacetophenone p-Methoxy-o-ethoxyacetophenone was prepared by the Friedel-Crafts acylation of anisole. This was accomplished by very slowly adding cathoxyacetyl chloride (0.2 mole) to a cooled solution of anisole (0.2 nx)le) in carbon disulfide in the presence of aluminum chloride (0.21 nwale). When the addition was complete the mixture was stirred at room 74 temperature for seventy-five minutes. The carbon disulfide was decanted off, and the dark, redish residue poured into a stirred solution of ice water and hydrochloric acid. After standing f0r two hours, the mixture was extracted with several portions of ether. The combined ether extracts were washed with dilute aqueous sodium hydroxide, with water, and dried. The ether was removed, and the residue was vacuum distilled using a short path condenser with hot water. Adequate purity (m.p. 45.5 -47.5, uncorrected; 98.5% pure by glpc) was attained after several re- crystallizations from low boiling petroleum ether. Because of the great ease of y-H abstraction by the electron de- ficient carbonyl oxygen of the molecular ion, no parent peak is observed in the 70eV mass spectrum of p-MEAP. However, as with p-MMAP, fragments at m/e 150, 135 (base), 107, 92, and 77 are present. The proton nmr in CCl. reveals the following peaks: 0 1.20 (triplet, three protons on 0- carbon), 3.56 (quartet, two protons on y carbonyl), 3.88 (singlet, three protons in the methoxy group), 4.45 (singlet, two protons on the a car- bon), 6.89 (doublet, 17710 protons ortho to the carbonyl group), and 7.95 (doublet, two protons meta to the carbonyl group). The ir spectrum in- cludes the asymnetric stretching of CHz-O-CHz at 1133 cm", the C=0 istretching band at 1688 cm", and the benzene ring stretching bands at 1598 cm", 1570 cm", 1512 cm", and 1420 cm“. The uv spectrum of p-MEAP in heptane has three xmax: 216 (e = 10,700), 268 nm (e = 14,300), and 320 nm (e = 110). d. 4-Dimethylamino-l-(§-naphthyl):l-butanone 4-Dimethylamino-l-(B-naphthy1)-l-butanone (DMANB) was prepared from iflne Girgnard reagent of 2-bromonaphthalene and y-dimethylaminobutyroni- ‘trile in the same way that p-MDMAB was prepared. The residue distilled at: 128-135°C at 0.05 mm and gave an 86% yield of DMANB. Final 75 I purification of DMANB was attained by recrystallizing its hydrochloride salt several times from a methanol-butanone solution and carefully re- generating the free amine. Mass spectrometry (70eV) produced the following peaks: m/e 241 (molecular ion), m/e 155 (naphthoyl), m/e 127 (naphthyl), m/e 71 (loss of naphthacyl radical and a hydrogen atom), and m/e 58 (base, N,N-di- methyliminonium ion). The proton nmr of DMANB shows the following peaks: 6 1.94 (multiplet, two protons B to the carbonyl group), 2.18 (singlet, six protons in the dimethylamino group), 2.3 (multiplet, two protons o to the amino group), 3.06 (triplet, two protons a to the carbonyl), 7.77 (multiplet, six protons in the naphthyl ring), and 8.48 (singlet, the isolated proton in the naphthyl ring a to the carbonyl group. The ir spectrum has strong bands at 2810 cm‘1 and 2760 cm'1 resulting from symmetric C-H stretching of the methyl groups on nitrogen, a very strong band at 1677 cm"1 produced by the C=0 stretching vibrations, and medium bands at 860 cm", 820 cm", and 747 cm'1 caused by the C-H out of plane bending vibrations of the B-substituted naphthalene ring. The uv spec- trum of DMANB in heptane eXhibits the following A 238 nm (e = max’ 29,1000), 246 nm (e = 31,500), 281 nm (e = 5,140), 324 nm (e = 1,210), 339 nm (e = 1,410). e. Valerophenone The valerophenone used was either from Eastman Kodak Company or was iarepared by the acylation of benzene with valeryl chloride in the presence of aluminum chloride. In both cases the valerophenone was distilled, run through activated alimina, and distilled again before use. f. p:Methoxyacetophenone p-Methoxyacetophenone from Aldrich Chemical Company was twice 76 recrystallized from ligroin by Dr. Herbert N. Schott. The purified ketone melted at 39°C (uncorrected). g. 2-Acetonaphthone Z-Acetonaphthone (Eastman Kodak Company) was dissolved in ethanol, treated with decolorizing charcoal, and recrystallized several times from aqueous ethanol. It was further recrystallized several times from ligroin. The uncorrected melting point of the pure 2-acetonaphthone was found to be 53-53.5°C. h. Methoxyacetone Methoxyacetone was used as received from Chemical Procurement Laboratories, Inc., since glpc analysis showed it to be 98% pure. i. Acetone Spectrophotometric grade acetone from J. T. Baker Chemical Company was used as received. j. Ben20phenone Eastman White Label benzophenone was recrystallized from low boil- ing petroleum ether by Dr. B. J. Scheve. The uncorrected melting point was 47.5-48.7°C. k. Acetophenone Acetophenone from Matheson Coleman and Bell was distilled at re- duced pressure and a middle fraction taken for use. 2. Quenchers a. 1,3-Pentadiene When l,3-pentadiene was used in obtaining Stern-Volmer plots, a inixture of the cis and trans isomers was used. Before use, the 1,3- pentadiene (Aldrich Chemical Co.) was run through a 50 cm column of neutral alumina and then distilled. When the isomerization of 1,3- pentadiene was studied, cis-l,3-pentadiene, 99%, from Chemical Samples 77 Co. was used as received. Glpc analysis indicated that this cis-l,3- pentadiene was >99.7% cis. b. l,3-Cyc10hexadiene 1,3-Cyclohexadiene (Chemical Samples Co.) was distilled befbre use. c. Triethylamine Triethylamine (Eastman Organic Chemicals) was distilled through a 22 cm vigreaux column before use. d. 2,5-0imethyl-2,4-hexadiene 2,S-Dimethyl-2,4-hexadiene (Chemical Samples Co.) sublimed in the bottle at refrigerator temperature(~0°C)1and atmospheric pressure. The sublimed compound was used. e. trans-Stilbene trans-Stilbene suitable for photosensitizer use Baker grade from J. T. Baker Chemical Co. was used as received. 3. Solvents a. Acetonitrile Acetonitrile (J. T. Baker Chemical Co.) was distilled from potassium permanganate through a 45 on column packed with glass helices. A center cut of about 70% was taken for use. b. Benzene Nanograde benzene supplied by Mallinckrodt Chemical Works was stirred over successive portions (~5% by volume) of concentrated sulfuric acid until the acid layer was not yellow after stirring with the benzene for 24 hours. This usually required about four portions of concentrated Strlfuric acid, the previous portion of concentrated sulfuric acid having been removed before the next was added. The benzene was then stirred over 10% aqueous sodium hydroxide for 24 hours, washed with saturated 78 aqueous sodium chloride, and dried over anhydrous magnesium sulfate for 24 hours. The benzene was distilled from phosphorous pentoxide (~1009/ gallon of benzene) through a 95 cm column packed with glass helices. A reflux ratio of at least 10:1 was maintained, and after about 10% of the benzene had distilled over, approximately 90% of that remaining was distilled for use. c. Heptane Heptane (Matheson Coleman and Bell) was purified in the same way as benzene, but in smaller amounts. d. Methanol Absolute methanol from J. T. Baker Chemical Co. was distilled from magnesium turnings through a 45 cm column packed with glass helices. A reflux ratio of 10:1 was maintained, and a center cut of 70% was collected. e. Pyridine "Analyzed reagent grade" pyridine from J. T. Baker Chemical Co. was used as received from a bottle freshly opened. 4. Internal Standards a. Octadecane and Tetradecane Octadecane (Aldrich Chemical Co.), and tetradecane (Columbia Organic Chemical Co.) were stirred over concentrated sulfuric acid until freshly added acid would not discolor after stirring for a day with the alkane. The alkane was washed with dilute base, dried over calcium chloride, and distilled at reduced pressure. The octadecane was further purified by recrystallization from absolute ethanol. These purifications were done by Professor P. J. Wagner. 79 b. Cycloheptane Cycloheptane (Aldrich Chemical Co.) was further purified by distil- lation. c. 2-Methyldecane 2-Methyldecane (Aldrich Chemical Co.) was used as received. d. Pentadecylbenzene Pentadecylbenzene (Chemical Samples Co.) was used as received. 8. Methods 1. Readying Samples for Irradiation . All glassware involved in sample preparation and photolysis was scrupulously cleaned. In all cases class A volumetric flasks and pipettes were used. Before being filled, photolysis tubes were con- stricted to facilitate sealing. This was accomplished by rotating the area of the tube (10 x 1.3 cm) three to six centimeters from its top in an oxygen-house gas flame until this area was soft enough to allow pulling the tube out to a length of about 18 cm. Solutions to be photolyzed were prepared from stock solutions of desired compounds. In some cases, however, a compound(s) was (were) iweighed directly into the volumetric flask which was to contain the solution to be photolyzed. Table 4 is an example of the preparation of solutions to be photolyzed. Once prepared, solutions to be photolyzed were placed in the con- stricted tubes with a 5 ml capacity syringe attached via a luer lock fitting sealed with teflon tape to a 15 cm, blunt point hypodermic needle. In almost all cases, 2.8 ml of solution was syringed into the tubes. However, in cases where the solution to be photolyzed was pre- pared in a 5 m1 volumetric flask, and yet two tubes of the solution 80 Table 4. An example of the preparation of solutions to be photolyzed. Soln # of 1.021 M 0.021 M 0.739 M Compound Final Final conc. # tubes quencher C1. ketone weighed vol. in benzene (Q) 0.030 M out 510 l 3 1 ml 10 ml 0.0030M C18 0.0739M ketone 2 2 0.5 ml 1 ml 10 m1 # 1 plus 0.051 M Q 3 3 1 ml 1 ml 10 ml # 1 plus 0.102 M Q 4 l 1.5 ml 1 ml 10 ml # 1 plus 0.153 M Q 5 3 2 ml 1 ml 10 ml # 1 plus 0.204 M Q 6 2 5 ml 1 ml 10 m1 # 1 plus 0.510 M Q 7 3 1 ml 0.66109 10 ml # 1 plus Q 0.972 M 0 8 2 1 ml 1.04269 10 ml # 1 plus 0 1.533 M Q 9 l 0.5 ml 0.75529 5 m1 # 1 plus Q 2.221 M Q 10 l 0.5 ml 0.87589 5 ml # 1 plus Q 2.576 M Q 11 3 2 ml 0.15359 10 ml 0.0946 M Valero- Valerophe- phenone none 0.004172 M Cu 81 were desired, 2.4 ml of solution was syringed into tubes containing eleven 4 mm glass beads. Tubes thus prepared were attached to a vacuum line via one-hole 00 rubber stoppers (Fisher Scientific Co.). To achieve the best vacuum (10'3-10‘“ mm), tubes were fitted very snuggly onto new stoppers which had been checked to assure the absence of defects which could cause a leak. The tubes containing solutions to be photolyzed were then frozen in liquid nitrogen. If tubes required the presence of air, they were sealed at this point. Otherwise, the tubes immersed in the liquid nitrogen were opened to the vacuum for ten to fifteen minutes and then closed to the vacuum and thawed. By subjecting the tubes to several of these freeze-pump-thaw cycles, oxygen was effectively removed from the solutions to be photolyzed. In nearly all cases four freeze-pump-thaw cycles were performed, more or fewer having been performed as required. Tubes were sealed by heating the constricted portion of the tubes as evenly as possible with an oxygen-house gas torch. This was done while the tubes were open to the vacuum and while the portion of the tubes containing the solution to be photolyzed was immersed in liquid nitrogen and held with tongs. After sealing, tubes were thawed, mixed well, washed with acetone, and rinsed with distilled water. 2. Irradiation Sealed tubes containing the solutions to be photolyzed were placed in the one inch thick walls of an aluminum cylinder that had been bored (out to accommodate thirty tubes positioned vertically. Windows provided on the inside walls of the cylinder were all tooled to precisely the same dimensions so that each tube received the same amount of light 82 from the light source, which was mounted in the center of the cylinder. This "merry-go-round" apparatus123 was rotated about the light source to insure that any lack of homogeneity between the tubes and the lamp was experienced in the same way by each tube. The light source was a 450 watt medium pressure Hanovia mercury arc lamp, positioned in a water- cooled quartz emersion well. In nearly all cases, tubes were irradiated with light from the 300 - 330 nm region, which was isolated by placing the emersion well in a cylindrical pyrex jar containing a solution 0.002 M in potassium chromate and 1% in aqueous potassium carbonate. The emersion well was positioned so that the light traveled through 1 cm of this solution. This provided for the maximum transmittance of light at 313 nm (40%) and less than four percent transmittance at 300 nm and at 330 nm. However, light output from the lamp in this region is essen- tially confined to 3000-3026 A, 3090-3200 A, and 3310-3341 A regions, the maximum intensity being in the 3090-3200 A region.‘2“ When needed, the 3660 A region was isolated with a set of Corning no. 7083 filter combinations. 3. Photolysate Analysis a. Gas Chromatography The analysis of photoproducts or of the disappearance of photo- reactants in photolyzed solutions was carried out in all cases by gas- liquid partition chromatography. For this purpose two Vari an Aerograph Hy-Fi III Model 1200 gas chromatographs and an Aerograph Hy-Fi Model 6000 gas chromatograph were used. In some cases an Aerograph Hy-Fi Model 550 oven was used with the electrometer of the Model 6000. Flame ionization detectors were used exclusively. Signals were recorded with Leeds and Northrup Speedomax recorders equipped with Disc integrators. 83 In some cases an Infotronics Model CRS-208 Automatic Digital Integrator was used to integrate peaks. In all cases 1/8" 0. 0. aluminum or stain- less steel columns were set up for on-column injection. Nitrogen (about 25 ml/min) was used as the carrier gas, and hydrogen (about 25 ml/min) and air (about 250 ml/min) were used for the flame. The particular column packing and conditions used for specific runs may be found in the tables of photokinetic data. b. Identification of photoproducts The photoproducts, p-methoxyacetophenone, acetophenone, acetone, 2-acetonaphthone, and trans-piperylene were identified by comparing their retention times with authentic compounds on analytical glpc columns. Peaks in the glpc traces of the photolysates of o-methoxy- p-methoxyacetophenone and o-ethoxy-p-methoxyacetophenone not assigned to the parent ketone or to p-methoxyacetophenone were assumed to be the corresponding 3-oxetanols on the basis of their expected and observed proximity to the parent ketone peak, the fact that they were the only photoproducts observed other than p-methoxyacetophenone, and Lewis and Turro's112 isolation and identification of the 3-oxetanols of a-methoxy- p-methoxyacetophenone and a-ethoxyacetophenone photolysis. A photoproduct of the photolysis of p-MDMAB believed to be the corresponding cyclobutanol was subjected to mass spectrometric analysis on an LKB 9000 Gas Chromatograph-Mass Spectrometer with a Digital Equip- inent Corporation PDP 8/I on-line digital computer, an incremental plotter, and a KSR 35 teletypewriter operated by Mr. Jack E. Harten. A 6' x 2 mm 1.0. glass column containing 3% SE-30 liquid phase and programmed from 130° to 200°C at 5°/min was used for the separation of the components of the photolyzed solution, 0.02 M p-MDMAB in benzene. The following 84 important peaks indicating the structure, l-(p-methoxyphenyl)-1-hydroxy- 2-dimethylaminocyclobutane, are present in the mass spectrum: m/e 221 (molecular ion; relative intensity 0.5), 220 (M+ - H; 0.5), 219 (M+ - 2H; 2.8), 203 (M+ - H20; 1.7), 202 (M+ - H - H20; 10.3), 150 (loss of N,N-di- methylvinylamine; 3.6), 135 (anisoyl; 13.7), 107 (anisyl; 3.0), 84 (H2C=CH-CH=N(CH3)2; 100), and 71 (loss of p-methoxyphenacyl radical and a hydrogen atom; 41.0). Furthermore, the ir spectrum of photolyzed p-MDMAB contains a broad O-H stretching band at 3480 cm". c. Internal standard-product response ratios To allow for variations in the size of the sample injected onto the gas chromatograph and to permit the accurate determination of the concen- tration of photoproducts, a known concentration of a photochemically inert compound was placed in solutions to be photolyzed. This internal standard was chosen on the basis of its ability to separate from all other components under the analytical conditions employed and was present in a concentration sufficient to produce a peak comparable in size to the photoproduct peak. To determine the concentration of a photoproduct, P, the response of the QC detector to a given concentration of P was com- pared to the response to a similar concentration of internal standard, S. Thus, a standardization factor, SF, was calculated. SF = ER] x area S peak (15) [S] area P Peak When analysis of P and S in a photolyzed sample was carried out under the same conditions that the SF was determined, the concentration of P could be calculated as follows: [P] = SF area P peak [S] (16) area S peak 85 Standardization factors for specific runs are given in the tables of kinetic data. 4. Actinometry and Quantum Yields As previously stated, the quantum yield is the only directly mea- surable parameter under the steady-state conditions employed in the laboratory. The quantum yield of formation of a given product is cal- culated by dividing the number of photoproduct molecules produced by the number of photons needed to produce this number of molecules. The actinometer provides the number of photons when it is irradiated in parallel with samples for which quantum yields are desired. TWo chemical systems were used for actinometers in this work. One system contained approximately 0.1 M valerophenone with a known concen- tration (usually in the range of 0.004 M) of tetradecane (0,.) as the internal standard. The quantum yield for acetophenone formation in this system in benzene solvent is known to be 0.33”2 and is constant out to 75% conversion122 of valerophenone into photoproducts when benzene is the solvent. Thus, the photon count in mole 1‘1 is as fol- lows: number of photons = [acetophenone] (l7) ' 0.33 Acetophenone was quantitated on various columns fitting the following general description: 6' - 9' x 1/8" aluminum or stainless steel con- taining 4-5% QF-l, 1% carbowax 20 M usually on chromosorb G. Column temperatures ranged from 110 to 140°C. The standardization factor for tetradecane and acetophenone peaks was usually checked for each analysis and was nearly always 2.05:3%. The other system used consisted of some laiown concentration of cis-l,3-pentadiene and an amount of sensitizer such that the same amount of light was absorbed by the sensitizer as by 86 the samples for which quantum yields were desired. The requirements for the sensitizer in this actinometer were: 1.) that its intersystem crossing quantum yield be known, and 2.) that its lowest triplet state be long enough lived and properly positioned energetically to transfer all of its energy to the ground state cis-l,3-pentadiene. Under these conditions, the photon count in mole 1"1 is as follows:95 0.555 /¢. (18) number of photons = [Cls'innitial]n 1sc 0.5554% trans-p The % trans-l,3-pentadiene (% trans-p) was measured on a 25' x 1/8' aluminum column containing 25% 1,2,3-tris(2-cyanoethoxy)propane on 60/80 chromosorb P held at 50-55°C. C. Photokinetic Data There follow tabulations of data obtained from experiments involv- ing quenching of photoproducts or phosphorescence, quantum yield deter; minations, photoproduct ratio determinations, and sensitization of cis- to-trans isomerization of cis-l,3-pentadiene. Tables include the glpc peak areas of the photoproduct of a given kinetic run relative to the peak area of an internal standard. Where actinometers were employed the concentration of photoproduct, the amount of light absorbed in mole 1", and the quantum yield of photoproduct formation are tabulated. Quenching data include relative quantum yields (9°), which are obtained by dividing the peak area ratio found in the absgnce of quencher by that found in the presence of quencher. Tabulations of phosphorescence quenching data include the relative phosphorescence emission at each concentration of quencher along with the corresponding relative quantum .yields. In tables of sensitization experiments, trans-l,3-pentadiene peak areas are tabulated as the percent of the sum of the cis- and 87 trans-1,3-pentadiene peak areas. Also tabulated are the concentrations of trans-l,3-pentadiene corrected for back reaction ([t-p] ), the corr amount of light absorbed in mole 1", and the quantum yields for cis- to-trans isomerization, ¢c+t' The concentrations of each compound in the photolyzed solutions are given in the tables as well as the analytical conditions employed. Average deviations are given in the tables for relative peak areas, relative quantum yields, quantum yields, and standardization factors. When qu and ¢isc values reported in the results section are determined from slopes and intercepts of lines containing three or more points, these values are reported along with standard deviations. All other k T and ¢. q 15c values are reported with their average deviations. 88 Table 5. cis-l,3-Pentadiene quenching of 2-AN formation from 0.04057 M DMANB in benzene irradiated at 313 nm. 2-Angeak area °2-AN [quencher], M C1. peak area 5;:Xi' [Z-AN], M °2~AN 0 0.971:,022 1 0.00162 0.0080:,0004 0.0052 1.05:,02 0.92:,04 0.00175 0.0087:,0004 0.0104 1.06:,02 0.923904 0.00177 0.0088:,0004 0.0208 1.07:,02 0.911904 0.00178 0.0088:,0004 0.0312 1.061302 0.92:,04 0.00177 0.0088:,0004 0.0416 1.051902 0.92:,04 0.00175 0.0087:,0004 0.520 1.071902 0.91:,04 0.00178 0.0088:,0004 0.0832 1.07:. 03 0.91:.04 0.00178 0.0088:.0004 Photon count: 0.202:,004 mole 1'1 Internal standard: 0.001004 M C13; SF = 1.661501 Analytical conditions: 5' x 1/8" stainless steel column packed with 5% SE-30 on 60/80 aw dmcs chromosorb W at 200°C. 89 Table 6. l,3-Pentadiene quenching of 2-AN formation from 0.03062 M DMANB in benzene irradiated at 313 nm. - 4° 2 AN peak area 2-AN [quencher], M C19 peaE'area -———- [2-AN], M ¢2-AN °2-AN 0 0.743:.018 1 0.00128 0.0083:,00060 1.14 0.776:,004 0.96:,03 0.00134 0.0086:,0004 2.03 0.762:,004 0.98:,03 0.00132 0.0085:,0004 2.48 0.756:,012 0.98:,04 0.00131 0.0085:,0005 3.40 0.722:,018 1.03::05 0.00125 0.0081:}0006 Photon count: 0.155:,004 mole 1'1 Internal standard: 0.0009980 M CID; SF = 1.73:,03 Analytical conditions: As in table 5. 90 Table 7. l,3-Pentadiene quenching of 2-AN formation from 0.03314 M DMANB in methanol irradiated at 313 nm. [quencher], M 2-AN peak area ¢2-AN [2-AN], M ¢2-AN Cl. peak area ¢2-AN 0 6.97306 1 0.0148 0.12301 0.0101 1.07303 6.51325 0.00227 0.017830014 0.0201 0.6883008 10.132 0.00146 0.011430007 0.0302 0.5543006 12.633 0.00118 0.009230006 0.0402 0.4813005 14.533 0.00102 0.008030005 0.0604 0.4253009 16.435 0.000903 0.007130005 0.101 0.3703004 18.834 0.000787 0.006130004 0.302 0.3003004 23.235 0.000638 0.005030003 0.503 0.2823005 24.737 0.000599 0.004730003 1.01 0.2793003 25.035 0.000593 0.004630003 Photon count: 0.128:.0001 Internal standard: 0.0009980 M C1,; SF = 2.13:,09 Analytical conditions: 5' x 1/8" stainless steel column packed with 5% SE-30 on 60/80 aw dmcs chromosorb W at 189°C. 91 Table 8. l,3-Pentadiene quenching of 2-AN formation from 0.02932 M DMANB in methanol irradiated at 313 nm. 2-Angeak area ¢2~AN _____. _ 9 _ [quencher], M C1. peak area ¢2-AN [2 AN]’ M 2 AN 0 1.44305 1 0'00283 0.13301 0.00102 0.8183014 L751: '0 0'00"“ 0.0773006 0.00508 0.2803002 5° '43-'23 °°°°°5°I 0.0263002 Photon count: 0.0214190002 mole 1" Internal standard: 0.0009406 M C10; SF = 2.13:,09 Analytical conditions: 5' x 1/8" stainless steel column packed with 5% SE-30 on 60/80 aw dmcs chromosorb W at 195°C. 92 Table 9. trans-Stilbene quenching of 2-AN formation from 0.0206 M DMANB in benzene irradiation at 313 nm. * 2-AN peak area [t-Stilbene], M C'° peak area ¢2-AN/¢2-AN 0 0. 5883 022 1 0. 0000696 0. 4103 003 1.433 06 0.000111 0.34513003 1.70:,08 0.000174 0.29513007 1.99:,12 0.000348 0.201i,004 2.93::17 0.000522 0.14519001 4.061317 Internal standard: 0.0008216 M 0,. Analytical conditions: 2' x 1/8" stainless steel column packed with 5% SE-30 on HP chromosorb W at 160°C. Corrected for t-stilbene absorbing a fraction of the light. 93 Table 10. Photolysis of 0.05 M DMANB at 313 nm in benzene, 0.557 M pyridine in benzene, acetonitrile, and methanol. 3 2-AN peak areas photons ¢ solvent C13 peak areas [Z-AN], M mole 1-1 2-AN benzene 0.59119016 0.000967 0.130:,003 0.0074:,0005 0.557 M pyridine in benzene 0.616:,014 0.00101 0.130:,003 0.0078¢,0004 acetonitrile 0.758:,019 0.00136 0.130:.003 0.0105:,0005 methanol 6.973061 0.01482 0.128:,001 0.12:,01 Internal standard: 0.0009980 M C1. in benzene; SF = l.641,02 0.0009996 M C1. in acetonitrile; SF = 1.79:,01 0.0009980 M C1. in methanol; SF = 2.133909 Analytical conditions: 5' x 1/8" stainless steel column packed with 5% SE-30 on 60/80 aw dmcs chromosorb W at 200°C. 1 Photolysis of 0.03314 M DMANB; the sum of P/S from three sets of tubes irradiated sequentially; column temperature 189°C. 2 The actual concentration of 2-AN in each tube was 0.0049 M. 3 The cis-l,3-pentadiene - benzophenone actinometry used here indicated a higher lamp output than valerophenone-Cl. actinometry. 94 Table 11. Photolysis of 0.02 M DMANB at 313 nm in methanol, methanol-d1, various percentages of methanol in benzene, and in benzene. 2-AN peak area * Solvent 018 peak area [2-AN], M Photons ¢2-AN mole l'l Methanol 0.969:,027 0.00183 0.0106130005 0.17:,02 Methanol-d1 0.74413001 0.00128 0.0106:,0005 0.12:,01 90% Methanol 0.909:,007 0.00172 0.0106:,0005 0.16:,01 in benzene 70% Methanol 0.782:,002 0.00148 0.0106130005 0.14:,01 in benzene 50% Methanol 0.609:,010 0.00115 0.0106:,0005 0.11:,01 in benzene 30% Methanol 0.470:,017 0.000889 0.0106:,0005 0.084:,009 in benzene 20% Methanol 0.870:,028 0.00164 0.0279:,0017 0.059:,007 in benzene 10% Methanol 0.610:,020 0.000915 0.0279:,0017 0.033:.002 in benzene 5% Methanol 0.602:,016 0.000903 0.0438:,003l 0.021:,002 in benzene 1% Methanol 0.306;,023 0.000459 0.0438:,0031 0.010:,002 in benzene Benzene 0.375:,034 0.000562 0.0517150031 0.0098:,0026 Internal standard: Analytical conditions: benzene; SF = 1.461303 0.0009406 M C15 in methanol and 3 20% methanol in benzene; SF = 2.01:,05 0.0009588 M C19 in methanol-01; SF = l.79+.01 SE-30 on HP chromosorb W at 160°C. 0.001027 M C1. in benzene and 5 20% methanol in 2' x 1/8" stainless steel column packEd with 5% * Valerophenone-C1. actinometry was used in this case, since the relia- bility of the cis-l,3-pentadiene actinometer used in obtaining the data in table 10 was questionable. 95 Table 12. A comparison of the cis-to-trans isomerization of cis-l,3- pentadiene sensitized by 0.05033 M benzophenone, 0.04998 M 2-AN, and 0.05116 M DMANB photolyzed at 313 nm. ketone [c-PJO. M % 1'” [t'P1corr’ M [t'PJuncorr’ M °isc benzophenone in benzene 0.2008 9.081307 0.0199j30002 1.00l 2-AN in 1.04: benzene 0.2008 10.3133 0.0207130006 .042 0.1004 19.134 0.0803 20.6133 DMANB in benzene 0.2008 7.84314 0'32;- 0.74: 0.1004 14.2:31 .023 0.78: 0.0803 16.035 .04“I 0.03079 M benzophenone in methanol 0.2221 4.451303 0.0103130001 0.03042 M DMANB in 0.86: methanol 0.2076 4.26:306 0.008841300012 .02“ Analytical conditions: 25' x 1/8" aluminum column packed with 25% l,2,3-tris(2-cyanoethoxy)propane on 60/80 chromosorb P at 55°C. 1 Ref. 95 2 Calculated by ¢. 3 Calculated by ¢. 1SC 15C for both ketones. “ Calculated by o. lSC [t-P] (2-AN)/[t-P] (benzophenone). UHCOFP COFF [t-P] (DMANB)/[t—P] (benzophenone) UHCOFP COT? % t-P (DMANB)/% t-P (2-AN), [c-P] being the same 96 Table 13. 1.3-Pentadiene quenching of p-MAP formation from 0.0399 M p-MDMAB in benzene irradiated at 313 nm. [1.3-pentadiene], M peMgPpggikaggga izg53ig' [g-HSEIM ¢p-MAP 0.000 0.9023016 1 3.802 0.0353003 0.049 0.619:3006 1.463304 2.609 0.024j3002 0.104 0.490:3002 1.841304 2.065 0.01913001 0.148 0.44413015 2.031310 1.875 0.01713002 0.207 0.4003021 2.25316 1.686 0.0153002 0.518 0.2813018 3.20326 1.190 0.0113001 1.025 0.2103023 4.30356 0.885 0.00823 0014 1.482 0. 1933 011 4.67337 0. 813 0. 00743 0009 2.069 0.186i3018 4.851358 0.784 0.0072j30012 Photon count: 0.1093004 mole 1’1 Internal standard: Analytical conditions: 0.001686 M octadecane (013); SF = 2.50:306 9' x 1/8" aluminum column packed with 4% QF-l, 1% carbowax 20 M on 60/80 chromosorb P at 145°C. Table 14. 97 p-MDMAB in benzene irradiated at 3l3 nm. l,3-Pentadiene quenching of p-MAP f0rmation from 0.07388 M pyMAP3peak area Q°E-MAP [p-MAP] x [1.3-pentadiene], M Cl. peak area 0 p- l03, M ¢p-MAP 0 0.89513027 1 6.739 0.028:,003 0.051 0.63l:,006 l.42:,06 4.753 0.02019001 0.l02 0.5l0:,004 l.75:,08 3.84 0.016:,00l 0.l53 0.43413002 2.06:,08 3.27 0.0l41300l 0.204 O.378:,009 2.37:}l2 2.85 0.0l2:,00l 0.5l0 0.25713006 3.48:3l7 l.93 0.0080:_ .0006 0.972 0.200:,005 4.47:,27 l.5l 0.0063:_ .0006 1.53 0.l7l:,007 5.23:,37 l.29 0.0054:_ .0005 2.22 0.l60:,004 5.59:,34 l.2l 0.0050:- .0005 2.58 0.1543006 5.81:.41 l.l6 0.0048: .0005 Photon count: Internal standard: Analytical conditions: 0.24l:,0l0 mole 1"1 0.003012 M 0103 SF = 2.50:,06 9' x l/8“ aluminum column packed with 4% QF-l, l% carbowax 20 M on 60/80 chromosorb P at l45°C 98 Table 15. 1,3-Cyc10hexadiene quenching of p-MAP formation from 0.0434 M p-MDMAB in benzene irradiated at 313 nm. * [l,3-cyc10hexadiene], M 9-??Z £22: 2:2: gzgffi§§' 0 0.90513008 l 0.020 0.80113036 1.131306 0.051 0.66113077 1.371319 0.108 0.44313008 2.041306 0 1.641302 1 0.540 0.40113006 4.081311 1.00 0.27213011 6.021332 2.02 0.20513003 7.991320 3.01 0.18513005 8.851334 3.71 0.18213008 9.001353 4.40 0.18013005 9.101336 5.01 0.17713008 9.251354 Internal standard: 0.001993 M C13 Analytical conditions: 9' x 1/8" aluminum column packed with old 5% QF-l, 1% Carbowax 20 M chromosorb G at 120°C. * Corrected for the reduction of p-MAP formation caused by the failure of p-MDMAB to absorb all of the light. 99 Table 16. 1.3-Cyclohexadiene quenching of p-MAP formation from 0.04002 M p-MDMAB in benzene irradiated at 313 nm. * [l,3-cyclohexadiene], M FETflpiggzkaigga %?%E$%¥; 0 0.7323046 1 0.020 0.5573004 1.31309 0.050 0.47913007 1.53:.11 0 1.50306 1 0. 500 0.3853007 3.89323 3.00 0.1763003 8.52351 Internal standard: 0.002050 M C1. Analytical conditions: 9' x 1/8" aluminum column packed with old 5% QF-l, 1% carbowax 20 M on chromosorb G at 120°C. * Corrected for the reduction of p-MAP formation caused by the failure of p-MDMAB to absorb all of the light. 100 Table 17. Determination of the disappearance quantum yield of 0.0300 M p-MDMAB in benzene irradiated at 313 nm. p-MDMAB peak area PDB peak area p:MAngeak area 0]. peak area Before irradiation: 0.92113004 After irradiation: 0.51813013 Concentration of p-MDMAP removed: (0.0300) (0.921 - 0.518) = 0.0131 M 0.921 ¢-p-MDMAP = 0.055: 005 Photon count: 0.23713014 mole 1'1 Internal standard for p-MDMAB analysis: 0. Analytical conditions for p-MDMAB analysis: 1.181303 Concentration of p-MAP which appeared: 0.00602 M ¢p_MAP = 0.02533002 00600 M pentadecylbenzene (PDB) 6' x 1/8" stainless steel column packed with 4% QF-l, 1% carbowax 20 M on chromosorb G at 180°C. Internal standard for p-MAP analysis: 0.001993 M C18; SF = 2.561301 Analytical conditions for p-MAP analysis: 9' x 1/8" aluminum column packed with old 5% QF-l, 1% carbowax 20 M on chromosorb G at 144°C. In. “A .- 101 Table 18. Dependence of p-MAP quantum yield on initial p-MDMAB concen- tration and on pyridine concentration in benzene irradiated at 313 nm. [p-MDMAB] , [pyridine], p-MAP peak [p-MAP] photons ¢ 0 p-MAP "ea x 103 M mole 1--1 M M 16 Pea ’ area 0.0281 0.0300 0 0. 3.06 0.11013007 .002 6003008 0.0433 0.0300 0.500 0.64513006 3.29 0.07713004 0.822+ 0.0300 1.003 0.69613014 3.55 0.07713004 0.3g3+ 0.0507 0 4.39 0.147+.009 .003_' 0.8613013 — 0.029: 0.0635 0 0.8275013 4.22 0.1473009 .003 Interna1 standard: 0.001993 M Clo; SF = 2.561301 Analytical conditions: 9' x 1/8" aluminum column packed with old 5% QF-l. 1% carbowax 20 M on chromosorb G at 144°C. 102 Table 19. The cis-to-trans isomerization cis-1,3-pentadiene sensitized by 0.05 M p-MDMAB in benzene irradiated at 313 nm. [c-P];‘, M'1 % t-P [t'P1corr’M agggoq§l 2.555 c’t 0.194 2.38312 0.125 0.2683011 1.19311 0.969 6.86311 0.0755 0.1753002 1.29304 1.94 11.131 0.0637 0.1753002 1.52303 4.99 4.72310 0. 00989 0.03073 0005 1.72307 7.96 5.83321 0. 00774 0. 030730005 2.20313 15.6 7.50316 0.00516 0.030730005 3.30313 15.9 7.63318 0.00515 0.030730005 3.31313 31.2 9.171323 0.00321 0.0307130005 5.311327 Analytical conditions: 25' x 1/8" aluminum column packed with 25% 1,2,3-tris(2-cyanoethoxy)propane on 60/80 chromosorb P at 58°C. til:- u- .l‘ I A 103 Table 20. l,3-Pentadiene quenching of p-MAP formation from 0.08693 M p-MMAP in benzene irradiated at 313 nm. P-MAngeak area 2:2;ugg_ [p-MAP] ¢ [quencher], M C1. peak area ¢ p-MAP x 102," p-MAP 0 1.363. 02 1 1.02 0- 62.:- 06 0.0480 0.3403013 4.00315 0.256 0- "51-02 0.0961 0. 2033.005 6. 703. 17 0.153 °-°93i-°‘° 0.144 0.1503005 9.07330 0.113 905812008 0.192 0.1253002 10.932 0.0941 “571-005 0.480 0.0743002 18.435 0.0557 0'0341-004 0.987 0.0513004 26.732.1 0.0384 0'0231-00" 1.65 0.0433003 31 .632 .2 0.0324 0- 0201303 2.08 0.0403003 34.032.5 0.0301 “0181-003 3.10 0.0373003 36.8-13.0 0.0279 0.0173003 1'? Photon count: 0.0165130011 mole 1“ Internal standard: 0.003012 M C18; SF = 2.501306 Analytical conditions: 9' x 1/8" aluminum column packed with 4% QF-l, 1% carbowax 20 M on 60/80 chromosorb P at 145°C. 104 Table 21. 1,3-Pentadiene quenching of p—MAP formation from 0.1060 M p-MMAP in benzene irradiated at 313 nm. p-MAP peak area ¢ E-MAP [P-MAP] ¢p_MAp [quencher], M C1. peak area ¢p~MAP x 102,M 0 2.0763019 1 1.56 0.80307 0.0501 0.4983003 4.17303 0.375 0.19302 0.150 0.23713007 8.761326 0.178 0.09113008 0.501 0.11113008 18.71].3 0.0836 0.04313007 1.61 0.07713006 27.01?.1 0.0580 0.03013005 2.05 0.0653004 31 .932.0 0.0489 0.0253004 2.52 0.0623003 33.531.6 0.0467 0.0243003 3. 09 0. 0573. 002 36 .431 .3 0. 0429 0.0223003 3.58 0.05513003 37.812.1 0.0414 0.02113003 Photon count: 0.0195130012 mole 1'1 Internal standard: 0.003012 M C16; SF = 2.501306 Analytical conditions: 9' x 1/8" aluminum column packed with 4% QF-l, 1% carbowax 20 M on 60/80 chromosorb P at 146°C. 105 Table 22. The 1-(p-methoxypheny1)-1-hydroxy-3-oxetane to p-MAP ratio produced from the photolysis at 313 nm of 0.05 M p-MMAP in benzene, 1.3-pentadiene, cyclohexene, cyclopentene, and cyclohexane. 3-oxetanolpeak area solvent p-MAP peak area benzene 0.42013011 1.3-pentadiene 0.17013005 cyclohexene 0.53113006 cyclopentene 0.42713012 cyclohexane 0.50913014 Analytical conditions: 8 2/3' x 1/8“ aluminum column packed with 4% QF-l, 1% carbowax on 60/80 chromosorb P at 180°C. 106 Table 23. The cis-to-trans isomerization of cis-l,3-pentadiene sensi- tized by 0.07 M p-MMAP in benzene irradiated at 313 nm. photons ' 0.555 [c-P];1,M'l % t-P [t'P1corr’M mole 1'1 ¢c+t 0.67 4.911306 0.07713001 0.14213006 1.021305 1.0 7.161336 0.07713004 0.14213006 1.021309 1.88 11.4132 0.06813001 0.12913001 1.051303 4.86 8.271329 0.0185130006 0.0355130001 1.071306 4.95 4.161306 0.0088130002 0.0174130003 1.091304 9.90 7.451314 0.0081130002 0.0174130003 1.191305 Analytical conditions: 25' x 1/8" aluminum column packed with 25% 1,2,3-tris(2-cyanoethoxy)propane on 60/80 chromosorb P at 58°C .1} 107 Table 24. The 1-(p-methoxypheny1)-l-hydroxy-2-methy1-3-oxetane to p-MAP ratio produced from the photolysis at 313 nm of 0.05 M p-MEAP in benzene and in 1.3-pentadiene. 3-oxetanol peak area solvent p-MAP peak area benzene 1.671303 l,3-pentadiene 0.49913016 Analytical conditions: As in table 22. 108 Table 25. l,3-Pentadiene quenching of p-MAP formation from 0.05035 M p-MEAP in benzene irradiated at 313 nm. p-MAPJeak area Mg [p-MAP] ¢p—MAP [quencher], M C1. peak area ¢p-MAP x 102,M 0 2.43307 1 1.20 0.46304 0.0511 1.503. 06 1.62306 0.738 0.28303 0.102 1.15306 2.11311 0.566 0.22302 0.511 0.5383018 4.52314 0.265 0.1023008 1.53 0. 3003. 006 8. 103. 16 0.148 0.0573004 2.22 0.2653027 9.173.92 0.130 0.0503008 2.75 0.25513007 9.531329 0.125 0.04813004 3.04 0.24313007 10.0133 0.120 0.04613004 3.64 0.23613007 10.3133 0.116 0.04513004 Photon count: Internal standard: Analytical conditions: 0.026030008 mole 1-1 0.002008 M C18; SF = 2.451306 9' x 1/8" aluminum column packed with 5% QF-l, 1.2% carbowax 20 M on chromosorb G at 155°C. 109 Table 26. The cis-to-trans isomerization of cis-1,3-pentadiene sensi- tized by 0.05 M p-MEAP in benzene irradiated at 313 nm. [c-P];' , n-I % t-P [t-chorrm $333351 gig c~t 0.194 2.52314 0.132 0.2683011 1.133.11 0. 999 6. 103. 10 0.0647 0.1353003 1.16305 4.86 5.75315 0.0126 0.03553. 0009 1.57308 9.72 8.44325 0.00937 0.03553. 0009 2.10312 14.0 9.82321 0.00769 0.035530009 2.563.12 19.5 10.937 0.00627 0.035530009 3.14329 28.1 12.731.4 0.00513 0.03553. 0009 3.84360 Analytical conditions: 25' x 1/8" aluminum column packed with 25% 1,2,3-tris(2-cyanoethoxy)propane on 60/80 chromosorb P at 58°C. 110 Table 27. 1,3-Pentadiene quenching of acetone formation from 0.2001 M methoxyacetone in benzene irradiated at 313 nm. [quencher],M aci;g?ghg:::ngrea ¢°acetone [acetone],M ¢acetone peak area 4)acetone 0 3.871307 1 0.0563 0.421303 0.105 3.801312 1.021305 0.0553 0.411303 0.210 3.9013041 0.991303 0.0568 0.421303 0.421 3.901312 0.991305 0.0568 0.421303 0.631 3.831309 1.011304 0.0558 0.421303 1.05 3.781312 1.021305 0.0550 0.411303 6.0 3.501307 1.111304 0.0510 0.381302 Photon count: 0.13413003 mole 1"* Internal standard: 0.004208 M cycloheptane: SF = 3.461307 Analytical conditions: 9'6" x 1/8" aluminum column packed with 19.4% FFAP on 60/80 chromosorb P at 68"C. * Corrected for the actinometer absorbing more light than the methoxy- acetone in the samples. 111 Table 28. Quantum yields for acetone formation from 0.201 M methoxy- acetone in benzene and in 1.3-pentadiene irradiated at 313 nm. acetone pgak area solvent cycloheptane peak area [acetone],M ¢acetone benzene 1.341303 0.0146 0.401302 1.3-pentadiene 1.201302 0.0131 0.361302 Photon count: 0.0361130007 mole 1'1 Internal standard: 0.003156 M cycloheptane; SF = 3.461307 Analytical conditions: As in table 27. 112 Table 29. Quantum yield for the disappearance of methoxyacetone from a solution 0.0521 M in methoxyacetone in benzene irradiated at 313 nm. methoxyacetone peak area 2Lmethyldecane peak area °-MA Before irradiation: 0.75213008 After irradiation: 0.45113006 Concentration of methoxyacetone removed: (0.0521) (0.752-0.451) = 0.0209130004 M 0. Photon count: 0.0446130013 mole 1"1 Internal standard: 0.01259 M 2-methy1decane 0.473302 Analytical conditions: 9'6" x 1/8" aluminum column packed with 19.4% FFAP on 60/80 chromosorb P at 100°C. 113 Table 30. The cis-to-trans isomerization of cis-l,3-pentadiene sensi- tized by 0.2001 M methoxyacetone in benzene irradiated at 313 nm. - - 0.555 [c-P] ',M 1 % t-P [t-P] ,M1 o corr ¢c+t 0.950 1.05303 0.00485 15.338 1.58 1.64310 0.00411 13131.4 2.38 2.31306 0.00341 21.831.1 Photon count: 0.13413003 mole 1‘1 2 Analytical conditions: 25' x 1/8" aluminum column packed with 25% 1,2,3-tris(2-cyanoethoxy)propane on 60/80 chromosorb P at 55°C. Corrected for t-P produced from acetone sensitization as well as for back reaction. 2 Corrected for the actinometer absorbing more light than the methoxy- acetone in the sample. 114 Table 31. The 2,5-dimethy1-2,4-hexadiene and triethylamine quenching of phosphorescence from 0.0231 M benzophenone in benzene excited at 375 nm. [2,5-Dimethyl-2,4-hexadiene], [triethylamine], RPE* 9:_ x 105 M x 105 M o 0 0 4813 1 1.10 2913 1.71.3 2.19 2212 2.2133 3.29 1812 2.7135 4.38 1111 3.4132 3.06 3211 1.51.1 4.08 3213 1.5132 5.10 2613 1.8133 6.13 2513 1.91.3 * The relative phosphorescence emission at 450 nm measured on an Aminco-Bowman Spectrophotofluorometer. 115 Table 32. The 2,5-dimethyl-2,4-hexadiene and triethylamine quenching of phosphorescence from 0.02306 M 4,4'-dimethoxybenzophenone in benzene excited at 375 nm. [2,5-dimethyl-2,4-hexadiene], [triethylamine], ' RPE* 9:_ x 105 M x 105 M ¢ 0 0 7137 1 1.10 3135 2.33.2 2.19 1932 3.71.8 3.29 1432 5.13.7 4.33 1132 6.411.2 3.06 3335 2.13.5 4. 08 2813 2.53. 5 5.10 2332 3.11.6 6.13 2432 3.03.5 * The relative phosphorescence emission at 440 nm measured on an Aminco-Bowman spectrophotofluorometer. BIBLIOGRAPHY mummpw 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. BIBLIOGRAPHY A. Jab1onski, Z. Physik, 2&3 38(1935). J. G. Calvert and J. N. Pitts, Jr., "Photochemistry“, John Wiley and Sons, Inc., New York, 1966, p. 286. Edwin F. U11man, Accts. Chem. Res., 13 353(1968). M. Beer and H. C. Longuet-Higgins, J. Chem. Phys., 23, 1390(1955). G. Viswanath and M. Kasha, J. Chem. Phys., 213 574(1955). M. A. El-Sayed, Accts. Chem. Res., 1, 8(1968). P. M. Rentzepis, C. J. Mitschele, Anal. Chem., 333 No. 14, 20A(1970). N. C. Yang, E. D. Feit, M. N. Hui, N. J. Turro, J. C. Dalton, J. Am. Chem. 800., 223 9674(1970). N. J. Turro, "Molecular Photochemistry", w. A. Benjamin, Inc., N. Y., 1967. pp. 58-59. ?. J.)Nagner and G. S. Hammond, Advances in Photochemistry, §3 21 1968 . G. S. Hammond and P. A. Leermakers, J. Am. Chem. Sbe., 313 207(1962). G. Ciamician and P. Si1ber, Berichte der.Deutschen Chemischen GeseZZ- schaft, 333 2911(1900). G. Ciamician and P. Silber, ibid., 34, 1530519011. G. Ciamician and P. Silber, ibid., 313 1280 1911 N. 0. Cohen, Rec. TTav. Chim,, 323 243(1920). N. E. Bachmann, J. Am. Chem. 580., §§3 391(1933). F. Bergmann and Y. Hirshberg, ibid., §§3 1429(1943). H. L. J. Backstrfim, Z. physik. Chem., §§§3 99(1934). Ch. Weizmann, E. Bergmann, and Y. Hirshberg, J. Am. Chem. Soc., §Q3 1530(1938). J. N. Pitts, Jr., R. L. Letsinger, R. P. Taylor, J. M. G. Recktenwald, and R. B. Martin, J. Am. Chem. Soc., 313 1068(1959). 116 20. 21. 22. 23. 24. 25. 26. 27. 28. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 41. 42. 117 J. N. Pitts, Jr. , H. W. Johnson, Jr. , and T. Kuwana, J. Phys. Chem., 66, 2456(1962). G. Porter and F. Nilkinson, Trans. Faraday Soc., 323 1686(1961). . A. Bell and H. Linschitz, J. Am. Chem. 500., 33, 528(1962). . L. J. Backstrhm and K. Sandros, Acta Chem. Scand., 123 823(1958). J H H. L. J. BHckstrhm and K. Sandros, ibidL 153 48(1960). G. 5. Hammond and N. M. Moore, J. Am. Chem. Soc., 313 6334(1959). W . M. Moore, G. 5. Hammond, and R. P. Foss, J. Am. Chem. Soc., 333 2789(1961). w. M. Moore and M. Ketchum, J. Am. Chem. Soc., 333 1368(1962). A. Terenin and V. Ermolaev, Trans. Faraday Sbc., 333 1042(1956). ?. ngbomura, N. Yamamoto, and S. Tanaka, Chem. Phys. Let., 1, 309 1967 C. Parkanye, E. J. Baun, J. Hyatt, and J. N. Pitts, Jr. , J. Phys. Chem. Z3_,1132(1969). K. Yoshihara, D. R. Kearns, J. Chem. Phys., 333 1991(1966). C. Yang, R. L. Dusenbery, J. .Am. Chem. Sbc. ., 90, 5899(1968). . Padwa, Tetrahedron Letters, 3465(1964). N. A C. Walling, M. J. Gibian, J'. Am. Chem. Soc. ., 86, 3902(1964). C. Wa11ing, M. J. Gibian, ibid, 87, 3361(1965]—. p 9.5. J. 5Wa?ner,)A. E. Kemppainen, and H. N. Schott, J. Am. Chem. 300., 1973 . 5604 G. Porter and P. Suppan, Trans. Faraday Soc., 313 1664(1965). J. W. Sidman, Chem. Rev., §§3 689(1958). D. R. Kearns, N. A. Case, J. Am. Chem. Soc., 33, 5087(1966). A. Beckett and G. Porter, Trans. Faraday Soc., 333 2051(1963). G. Porter and P. Suppan, Pure and’AppZied Chem., 33 499(1964). P. Suppan, Ber. Bunsenges. Physik. Chem., 23, 321(1968). E. J. Bowen and E. L. A. E. De La Praudiere, J. Chem. Soc., 1503(1934). 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. (i S. 2795(1 s. 0. 33, 86 s. 0. s. 0. 118 Ham?ond, W. P. Baker, and W. M. Moore, J. Am. Chem. Sbc., 33, 961 . . Cohen and S. Aktipis, Tetrahedron Letters, 579 (1965). . Cohen and R. J. Baumgarten, J. Am. Chem. Soc., 31, 2996(1965). . Davidson, Chem. Comm., 575(1966). . Cohen and R. J. Baumgarten, J. Am. Chem. Soc., 33, 3471(1967). . Davis, P. A. Carapellucci, K. Szoc, and J. D. Gresser, J. Am. . Sbc., 31, 2264(1969). . Cohen and A. D. Litt, Tetrahedron Letters, 837(1970). Cohen, G. A. Davis, and W. D. K. Clark, J. Am. Chem. Sbe., 9(1972). Cohen and J. B. Guttenplan, Tetrahedron Letters, 5353(1968). Cohen and J. I. Cohen, J. Am. Chem. Soc., 32, 164(1967). Von Horst Leonhardt and Albert Weller, Ber. Bunsenges. Physik. Chem., 91. 79 1(1963). N. Mataga and K. Ezumi, Bull. Chem. Soc. Jap., 33, 1355(1967). :MMM S G G G. R J G . Reh . Kni LIULMWMM‘U G5 . Gut Davidson and P. F. Lambeth, Chem. Com, 1265(1967). . Cohen and H. M. Chao, J. Am. Chem. 580., 33, 165(1968). . Cohen and Sherman, J. Am. Chem. Soc., 33, 1642(1963). Cohen and J. I. Cohen, J. Phys. Chem., 13, 3782(1968). . Ware and H. P. Richter, J. Chem. Phys., 33, 1595(1968). . Wagner and A. E. Kemppainen, J. Am. Chem. Soc., 31, 3085(1969). . Cohen and 8. Green, ibid., 31, 6824(1969). . Cohen and G. Parsons, ibid., 33, 7603(1970). . Davidson and P. F. Lambeth, Chem. Comm., 511(1968). . Cohen and Nina Stein, J. Am. Chem. Soc., 31, 3690(1969). . Guttenplan and S. G. Cohen, J. Am. Chem. 500., 33, 4040(1972). m and A. We11er, Ber. Bunsenges. Physik. Chem., Z3, 834(1969). bbe, D. Rehm, and A. Weller, ibid., 13, 839(1969). tenplan and S. G. Cohen, Chem. Comm., 247(1969). 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 91. 92. 119 R. A. Caldwell, .7. Am. Chem. 565., 32, 1439(1970). I. N. E. Kochevar and P. J. Wagner, J. Am. Chem. Soc., 33, 3859(1972). J. Turro, C. Lee, N. Schore, J. Barltrop, and H. A. J. Carless, ibid., 23, 3079(1971). R. N. R. H1>JDCJJO—I>ZXJO;U'U .0 X22 I. (19 P. P. R. R. Hautala and N. J. Turro, ibid., 33, 5595(1971). C. Yang, M. H. Hui, and S. A. Be11ard, ibid., 33, 4056(1971). . Davidson and P. F. Lambeth, Chem. Cbmm., 133311969). . Wagner and R. A. Leavitt, J. Am. Chem. Sba., 33, 3669(1973). . Bamford and R. G. W. Norrish, ibid., 1504(1935). S J . G. W. Norrish and M. E. S. Appleyard, J. Chem. Soc., 874(1934). N G . w. Norrish, Trans. Faraday 300., 31, 1521(1937). Davis, Jr., and W. A. Noyes, Jr., J. Am. Chem. 880., 33, 2153(1947). . J. C. Nicholson, Trans. Faraday Sbc., 3Q, 1067(1954). N. Martin and J. N. Pitts, Jr., J. Am. Chem. Sbc., ZZ, 5465(1955). Srinivasan, ibid., 31, 5061(1959). R. Coulson and N. C. Yang, ibid., 33, 4511(1966). P. Borkowski and P. Ausloos, J. Phys. Chem., 33, 2257(1961). Padwa and W. Bergman, letrahedron Letters, 5795(1968). . D. Lewis, J. Am. Chem. Sbc., 33, 5602(1970). . J. Wagner, P. A. Kelso, and R. G. Zepp, ibid., 33, 7480(1972). . C. Yang and D. H. Yang, J. Am. Chem. Soc., 33, 2913(1958). . C. Yang, A. Morduchowitz, and D. H. Yang, ibid., 33, 1017(1963). . H. Schulte-Elte and G. Dhloff, Tetrahedron Letters, 1143(1964). Organ, K. Schaffner, and 0. Jeger, J. Am. Chem. Soc., 33, 3033 63 . J. Wagner and G. 5. Hammond, ibid., 33, 1245(1966). J. Wagner, ibid., 33, 5898(1967). D. Rauh and P. A. Leermakers, ibid., 33, 2246(1968). 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 120 . Wagner and A. E. Kemppainen, ibid., 33, 5896(1968). . Wagner and G. 5. Hammond, ibid. 87, 4009(1965). . Dougherty, ibid., 31, 4011(1965].— CJCJC... Padwa and W. Eisenhardt, J'. Am. Chem. 800., 90, 2442(1968). Padwa and W. Eisenhardt, ibid., 33, 1400(19717 . Padwa, W. Eisenhardt, R. Gruber, and D. Pashayan, ibid., 31, 857(1969). P. J. Wagner and T. Jellinek, ibid., 33, 7328(1971). p p T A. A. Lamola and G. S. Hamond, J. Chem. Phys., 13, 2129(1965). A. A. A. 1 E. J. .)Baum, J. K. S. Wan, and J. N. Pitts, Jr. , ibid., 88, 2652 1966 J. N. Pitts, Jr., D. R. Burkey, J. C. Mani, and A. D. Broadbent, ibid., 33, 5902(1968). N. C. Yang and R. L. Dusenbery, Mel. Photochem., 1, 159(1969). N. C. Yang, D. S. Mclure, S. L. Murov, J. J. Hauser, and R. Dusen- bery, J. Am. Chem. 500., Q, 5466(1967). P. J. Wagner, A. E. Kemppainen, ibid., 90, 5898(1968). .A. Lamola, J. Chem. Phys. ., 31, 4810(1967). .Dym and R. M. Hochstrasser, J. Chem. Ph 3. 51, 2458(1969). .M. Hochstrasser and C. A. Marzzacco in 'MoleEUlar Luminescence", C. Lim, Ed, W. A. Benjamin, N. Y, N. Y., 1969, pp. 631-656. 0:0 “150(1) > J. Wa ner, A. E. Kemppainen, and H. N. Schott, J. Am. Chem. 500., 2, 528011970). .A.Bar1trop and J. D. Coyle, J. Chem. Soc. ., B, 251(1971). 3 C- C.. L. . Saltiel, H. C. Curtis, L. Metts, J. W. Miley, J. Winterle, and . Wrighton, J. Am. Chem. Soc., 32_. 410(1970). N. C. Yang, M. Nussim, M. J. Jorgenson, and S. Murov, Tetrahedron Letters, 3657(1964). P. J. Wagner and T. Nakahira, J. Am. CheM- See-3 2§, 8474 (1973). P. J. Wagner in "Creation and Detection of the Excited State", Vol. 1A, A. A. Lamola, Ed., Marcel Deker, New York, N. Y., 1971, p. 173. G. F. Vesley, B. A. Prichard, MoZ. Photochem., 3, 355(1973). F. D. Lewsi and N. J. Turro, J. Am. Chem. Soc., 33, 311(1970). 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 121 . Yang, A. Shani, Chem. Comm., 815(1971). . Yang, S. P. E11iot, 8. Kim, J; Am. Chem. Sbc., 21, 7551(1969). . Coy1e, J. Chem. 580., Trans. Perkin 11, 233(1973). “CL-UGO .Wagner, A. E. Kemppainen, T. Je11inek, J'. Am. Chem. Soc. , 7512(1972). A. Bathrop, J. D. Coy1e, J. Chem. Soc. ., B, 251(1971). .G.Schu1tz, C. D. DeBoer, W. G. Herkstoeter, R. H. Sch1essinger, Am. Chem. Soc. , gg_, 6086(1970). Kaneta, M. Koizumi, Bull. Chem. Sbc. Japan, 49, 2245(1967). P??«?$~|§P9.2.= J. Wagner, and I. Kochevar, J. Am. Chem. Sbc., 99, 2232(1968). U . J. Wagner, A. E. Kemppainen, J. Am. Chem. 500., 94, 7495(1972). P. J. Wagner, I. Kochevar, A. E. Kamppainen, J. Am. Chem. 800., 94, 7489(1972). F. G. Moses, R. S. J. Liu, and B. M. Monroe, Mel. Photochem., l, 245(1969). J. G. Ca1vert and J. N. Pitts, Jr., "Photochemistry", John Wi1ey and Sons, Inc., New York, N. Y., 1966, p. 688. ..H Liu andJ. R. Ednan, J. Am. Chem. 500., 90, 213(1968). H. Liu and J. R. Edman, ibid.,91,1492(196§7'. Ladwig and R. S. H. Liu, ibid.,_96, 6210(1974). . Hammond, J. Sa1tie1, A. A. Lamo1a, N. J. Turro, J. S. Bradshaw, .)Cowan, R. C. Counse11, V. Vogt, C. Da1ton, ibid., 86, 3197 19 4 . AUG) 070W mom OMU‘D . -.II-'¢. @mywmn gm 5 0 3 o 3 9 2 1 3 IIHIHEIIN