RETURNING MATERIALS: )V1531_] RTace in book drop to LIBRARJES remove this checkout from 4-sngglg... your record. FINES will be charged if book is returned after the date stamped below. PART I DEACTIVATION OF TRIPLET PHENYL ALKYL KETONES BY CONJUGATIVELY ELECTRON-HITHDRAWING SUBSTITUENTS PART II REGIOELECTRONIC CONTROL OF INTRAMOLECULAR CHARGE-TRANSFER QUENCHING IN VARIOUS EXCITED TRIPLET PHENYL KETONES By Elizabeth Jane Siebert A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1981 ABSTRACT PART I DEACTIVATION OF TRIPLET PHENYL ALKYL KETONES BY CONJUGATIVELY ELECTRON-NITHDRAHING SUBSTITUENTS PART II REGIOELECTRONIC CONTROL OF INTRAMOLECULAR CHARGE-TRANSFER QUENCHING IN VARIOUS EXCITED TRIPLET PHENYL KETONES By Elizabeth Jane Siebert PART I Substituent effects on both photoreactivity (y-hydrogen abstrac- tion) and triplet energies (n,n* and n,n*) of phenyl alkyl ketones were' determined for conjugatively electron-withdrawing substituents (cyano, carbomethoxy, and acyl). Spectroscopic results indicate that para {-R) substituents stabilize the n,n* triplet much more than the n,n* triplet, resulting in n,n* lowest triplets. Conversely, meta (-R) substituents do not stabilize n.n* triplets sufficiently to invert the triplet levels. Ortho (-R) substituents are dominated by steric factors: o-cyano stabilizes the n,n* triplet, whereas o-carbomethoxy stabilizes the n,n* triplet. These results support a largely Elizabeth Jane Siebert l,4-biradical structure for the n,n* triplet and suggest resonance stabilization by the para (-R) substituents. Since only the n,n* triplet is reactive toward hydrogen abstrac- tion, substituent effects on triplet energies can also be determined from triplet photoreactivities. In agreement with the spectroscopic results, para (-R) substituents and ortho-carbomethoxy (n,n* lowest triplets) decrease triplet reactivity, while meta (-R) substituents and ortho-cyano (n,n* lowest triplets) increase reactivity in both benzene and acetonitrile. All (-R) substituted ketones undergo charge transfer quenching at nearly diffusion controlled rates. Various correlations involving Hammett parameters, reduction potentials, and n,n* triplet energies for various substituted valerophenones are presented. PART II Intramolecular charge transfer (CT) quenching rates were determined for amino-ketones in which the amine moiety was permitted only limited access to the site of triplet excitation. Rapid internal quenching 8 s") was found in para-valeryl B-dimethylaminoethyl “(mm 5 x IO benzoate (p-ZVB), which has a n.n* lowest triplet. The conformation- ally similar para-benzoyl B-dimethylaminoethyl benzoate, p-ZBB, (n,n* lowest triplet) displayed slower internal quenching (kCT «.105 5") as determined from its phosphorescence lifetime and quantum yield. Para-valeryl y-dimethylaminopropyl benzoate (n,n* lowest triplet) also undergoes rapid CT quenching whereas both m-ZVB and m-ZBB (n,n* lowest triplets) undergo little internal quenching. Since the amine Elizabeth Jane Siebert moiety cannot possibly come into contact with the carbonyl in any of these molecules, the results suggest that excitation in n,n* triplets is centered on the phenyl ring and excitation in n,n* triplets is localized on the carbonyl. Rapid CT quenching was observed for p-methoxy-y-dimethylamino- butyrophenone in both benzene and acetonitrile (kCT «.108 - 109 s"), while significant internal CT quenching (ch «:107 5") was observed for p-(3-dimethylaminapropoxy)valerophenone only in acetonitrile. These results suggest charge separation in the n,n* triplet, stabilized by polar solvents. To Mom and Dad and Barney ii ACKNOWLEDGMENTS The author wishes to thank Professor Peter J. Wagner for his guidance and support throughout the course of this endeavor. His advice to emulate his chemistry and not his desk hopefully will be heeded in the future. Also, the memos "signed" by Peter J. Wagner will always be remembered. The author would like to thank the members of the Wagner Group for all the memorable times shared at MSU. Special thanks to (M. A. Meador‘)2 for all the late night discussions, to the Kondili for their lasting friendship, and to R. Young for all the nickels. ' Also, hearty thanks to F. D. for the new beginning. The author would also like to thank both relatives and friends for their encouragement and Mrs. Peri-Anne Narstler for the excellent typing of this dissertation. Finally, the author would like to thank the Chemistry Department at MSU for the financial support and use of its facilitites and the National Science Foundation for the research assistantships ad- ministered by Dr. Wagner. TABLE OF CONTENTS Chapter Page LIST OF TABLES ....................... viii LIST OF FIGURES ...................... xviii INTRODUCTION ........................ l A. Photophysical Processes .............. l B Norrish Type II Photochemical Reaction ....... 3 C. Triplet States of Phenyl Alkyl Ketones ....... 4 D Energy Transfer and Charge Transfer Processes ..................... 9 1. Energy Transfer ................ 9 2. Charge Transfer Processes ........... l2 E. Stern-Volmer Kinetics ............... 25 F. Research Objectives ................ 26 l. Part I: Deactivation of Triplet Phenyl Alkyl Ketones by Conjugatively Electron-Withdrawing Substituents ....... 26 2. Part II. Regioelectronic Control of Intramolecular Charge-Transfer Quench- ing in Various Excited Triplet Phenyl Ketones ......................................... 27 PART I - DEACTIVATION OF TRIPLET PHENYL ALKYL KETONES BY CONJUGATIVELY ELECTRON- NITHDRAHING SUBSTITUENTS .............. 30 RESULTS ........................... 31 A. Photokinetic Studies ................ 3l l. Stern-Volmer Quenching Studies ......... 31 iv Chapter 2. Quantum Yields ................. 3. Maximum Quantum Yields ............. 4. Disappearance Quantum Yields .......... 5. Intersystem Crossing Yields .......... 6. Photoreduction Experiments ........... B. Spectroscopy .................... l. Phosphorescence Emission Spectra ........ DISCUSSION ......................... A. Models for Substituent Effects on Triplet Energies .................. B. Triplet Energies of Substituted Valerophenones ................... C. Electronic Absorption Spectra and Cyclic Voltammetry . . . .- ............. D. Rationalization of Substituent Effects on Both n,n* and n,n* Triplet States ........ E. Type II Quantum Yields ............... F. Determination of kd Values by Various Methods ....................... G. Ketone Photoreactivity and Calculation of Relative Triplet Energies ............ H. Attempted Correlation of Reactivities with Hammett a Values ................ 1. Charge Transfer Quenching Studies .......... J. .Correlation of Reduction Potentials with Hammett a Values .................. K. Correlation of n,n* Triplet Energies for Substituted Phenyl Ketones with Hammett a Values ...................... L. Correlation of n,n* Triplet Energies with Reduction Potentials ................ Page 40 4O 50 50 51 51 51 69 69 7O 73 75 BI 82 84 86 88 94 97 Chapter Page ”PART II - REGIOELECTRONIC CONTROL OF INTRA- MOLECULAR CHARGE TRANSFER QUENCHING IN VARIOUS PHENYL KETONES ............. 102 RESULTS .......................... 103 l. Stern-Volmer Quenching Studies .......... 103 2. Quantum Yields .................. 113 3. Disappearance Quantum Yields ........... 113 DISCUSSION ........................ 120 A. Bifunctional Molecules .............. 120 B Amino-Esters ................... 121 C. Amino-Ethers ................... 128 D Summary of Charge Transfer Quenching Results ...................... 131 E. Suggestions for Further Research ......... 135 EXPERIMENTAL ....................... 138 A. Preparation and Purification of Chemicals ..................... 138 l. Solvents and Additives ............ 138 2. Internal Standards .............. 141 3. Quenchers ................... 142 4. Ketones .................... 144 B. Photokinetic Techniques .............. 161 1. General Procedure ............... 161 2. Stern-Volmer Quenching Studies and Quantum Yields ................ 165 3. Disappearance Yields ............. 166 4. Intersystem Crossing Yields .......... 167 5. Photoreduction ................ 168 C. Spectra ...................... 168 vi Chapter Page 1. Phosphorescence Emission Spectra ....... 168 2. Electronic Absorption Spectra ......... 169 3. Infrared Spectra ............... 169 4. 1H NMR Spectra ................ 169 5. Mass Spectra ................. 170 6. Cyclic Voltammetry .............. 170 APPENDIX ......................... 172 REFERENCES ........................ 262 vii Table LIST OF TABLES Photokinetic Parameters for Ring- Substituted Valerophenones .......... Photokinetic Parameters for Phenyl Alkyl Ketones as a Function of y-Substi- tution .................... Spectroscopic Data fer Ring-Substituted Valerophenones ................ Spectrosopic Data fer Other Ring-Sub- stituted Valerophenones ........... Substituent Effects on Triplet Energies of Model Compounds .............. UV Spectroscopic Data ............ Predicted n,n* - n,n* Separations in Substituted Valerophenones in Benzene at Room Temperature ............. Relative Substituent Effects on Reduc- tion Potentials of Substituted Valero- phenones ................... Charge Transfer Quenching Data for Substituted Valerophenones .......... viii Page 32 34 56 57 59 64 71 76 90 Table 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Page Charge Transfer Quenching Parameters in Benzene ................... 92 Correlation of Hammett 0 Values with Reduction Potentials for Valerophenones ..... 95 Photokinetic Data for Valerophenone Models ..................... 107 Photokinetic Data for Amino-Esters ....... 108 Photokinetic Data for Amino-Ethers ....... 109 Phosphorescence Data for Substituted Benzophenones .................. 114 Charge Transfer Quenching Data for Valerophenone Models .............. 125 Charge Transfer Quenching Data for Amino-Esters .................. 127 Charge Transfer Quenching Data for Amino-Ethers .................. 130 Summary of Charge Transfer Quenching Results ..................... 132 Quenching of 0.039 M Valerophenone by Hexadiene in Benzene .............. 173 Quenching of Valerophenone by Hexa- diene in Acetonitrile .............. 174 Quenching of Valerophenone by Tri- ethylamine in Benzene .............. 175 . Quenching of Valerophenone by Tri- ethylamine in Acetonitrile ........... 176 ix Table 24 25 26 27 28 29 3O 31 32 33 34 35 Quenching of 0.040 M Butyrophenone by Hexadiene in Benzene ............. Quenching of o-Cyanovalerophenone by Hexa- diene in Benzene ............... Quenching of 0.040 M o-Cyanovalerophenone by Pentadiene in Acetonitrile ......... Quenching of m-Cyanovalerophenone by Hexa- diene in Benzene ............... Quenching of 0.039 M m-Cyanovalero- phenone by Hexadiene in Acetonitrile ..... Quenching of m-Cyanovalerophenone by Tri- ethylamine in Benzene ............. Quenching of m-Cyanovalerophenone by Tri- ethylamine-n1Acetonitrile ........... Quenching of p-Cyanovalerophenone by Hexadiene in Benzene ............. Quenching of 0.040 M p-Cyanovalerophenone by Hexadiene in Acetonitrile ......... Quenching of p-Cyanovalerophenone by Tri- ethylamine in Benzene ............. Quenching of 0.042 M p-Cyanovalerophenone by Low Concentrations of Triethylamine in Benzene .................. Quenching of 0.037 M p-Cyanovalerophenone by Triethylamine in Acetonitrile ....... Page 177 178 179 180 181 182 183 184 185 186 188 189 Table 36 37 38 39 4O 41 42 43 44 45 46 47 Quenching of 0.041 M p-Cyanobutyro- phenone by Hexadiene in Benzene ........ Quenching of 0.040 M p-Cyanobutyro- phenone by Hexadiene in Benzene ........ Quenching of 0.037 M p-Cyano- yMethyl- valerophenone by Hexadiene in Benzene ..... Quenching of o-Carbomethoxyvalerophenone by Hexadiene in Benzene ............ Quenching of 0.041 M o-Carbomethoxyvalero-» phenone by Hexadiene in Acetonitrile ..... Quenching of o-Carbomethoxyvalerophenone by Triethylamine in Benzene .......... Quenching of o-Carbomethoxyvalerophenone by Triethylamine in Acetonitrile ....... Quenching of m-Carbomethoxyvalerophenone by Hexadiene in Benzene ............ Quenching of m-Carbomethoxyvalerophenone by Pentadiene in Acetonitrile ......... Quenching of 0.040 M m-Carbomethoxyvalero- phenone by Triethylamine in Benzene ...... Quenching of p-Carbomethoxyvalerophenone by Hexadiene in Benzene ............ Quenching of p-Carbomethoxyvalerophenone by Hexadiene in Acetonitrile ......... xi Page 190 191 192 193 195 196 197 198 199 200 201 202 Table 48 49 50 51 52 53 54 55 56 57 58 59 Quenthing of p-Carbomethoxyvalerophenone by Triethylamine in Benzene .......... Quenching of p-Carbomethoxyvalerophenone by Triethylamine in Acetonitrile ........ Quenching of m-Divalerylbenzene by Hexa- diene in Benzene ................ Quenching of 0.032 M m-Divalerylbenzene by Hexadiene in Acetonitrile .......... Quenching of 0.029 M m-Divalerylbenzene by Triethylamine in Benzene ............ Quenching of m-Divalerylbenzene by Tri- ethylamine in Acetonitrile ........... Quenching of p-Divalerylbenzene by Hexa- diene in Benzene ................ Quenching of p-Acetylvalerophenone by Hexadiene in Benzene .............. Quantum Yields for 0.040 M Trifluoro- methylvalerophenones in Benzene ........ Disappearance Quantum Yields for Cyano- valerophenones in Benzene ........... Disappearance Quantum Yields for Cyano- valerophenones in Benzene ........... ‘Disappearance Quantum Yields for Cyano- valerophenones in Acetonitrile ......... xii Page 203 204 206 207 208 209 210 211 212 213 214 215 Table 60 61 62 63 64 65 66 67 68 69 70 Disappearance Quantum Yields for Carbo- methoxyvalerophenones in Acetonitrile ...... Disappearance Quantum Yields for Carbo- methoxyvalerophenones in Acetonitrile ...... Maximization of ¢II for 0.040 M p—Cyano- valerophenone by Pyridine in Benzene ...... Maximization of °II for 0.041 M p-Cayno- valerophenone by Dioxane in Benzene ....... Maximization of III for 0.040 M p-Cyano- valerophenone by t-Butyl Alcohol in Benzene ..................... Maximization of III for 0.040 M p-Cyano- butyrophenone by Pyridine in Benzene ...... Effect of Ketone Concentration on ¢II for p-Carbomethoxyvalerophenone in Acetonitrile .................. Intersystem Crossing Yield for 0.040 M p-Cyanoacetophenone in Benzene ......... Intersystem Crossing Yield for 0.050 M p- Carbomethoxyvalerophenone in Benzene ...... Quenching of 0.050 M p-Cyanoacetophenone by Naphthalene in Acetonitrile (1.0 M Toluene) .................... Photoreduction of 0.051 M p-Cyanoaceto- phenone by Toluene in Acetonitrile ....... xiii Page 216 217 218 219 220 221 222 223 224 225 226 Table 71 72 73 74 75 76 77 78 79 Page Quenching of 0.050 M p-Cyanoacetophenone by Naphthalene in Dry Acetonitrile (1.0 M p-Xylene). .' .............. 227 Photoreduction of 0.050 M p-Cyanoaceto- phenone by p-Xylene in Dry Acetonitrile . . . . 228 Quenching of 0.040 M p-Carbomethoxyvalero- phenone by B-Dimethylaminoethyl Benzoate in Benzene ................... 229 Quenching of p-Carbomethoxyvalerophenone by B-Dimethylaminoethyl Benzoate in Aceto- nitrile .................... 230 Quenching of 0.040 M p-Carbomethoxyvalero- phenone by y-Dimethylaminopropyl Benzoate in Benzene ................... 231 Quenching of 0.040 M p-Carbomethoxyvalero- phenone by v-Dimethylaminopropyl Ben- zoate in Acetonitrile ............. 232 Quenching of 0.040 M m-Carbomethoxyvalero- phenone by B-Dimethylaminoethyl Benzoate in Acetonitrile ................ 233 Quenching of p-Methoxyvalerophenone by Pentadiene in Benzene ............. 234 Quenching of p-Methoxyvalerophenone by Pentadiene in Acetonitrile ........... 235 xiv Table 81 82 83 84 85 86 87 88 89 90 91 Quenching of 0.040 M p-Methoxyvalerophenone by N,N-Dimethy1aminopropy1 Phenyl Ether in Acetonitrile ................ Quenching of 0.020 M p-2VB by Pentadiene in Benzene ................... Quenching of 0.030 M p-2VB by Pentadiene in Benzene ................... Quenching of 0.040 M p-ZVB by Pentadiene in Benzene ................... Quenching of 0.010 M p-ZVB with Pentadiene in Acetonitrile . . °- ............. Quenching of 0.020 M p-ZVB by Pentadiene in Acetonitrile ................ Quenching of 0.040 M p-2VB by Pentadiene in Acetonitrile .................. Quenching of 0.060 M p-2VB of Pentadiene in Acetonitrile ................ Quenching of 0.080 M p-2VB of Pentadiene in Acetonitrile ................. Quenching of 0.10 M p—2VB by Pentadiene in Acetonitrile ................ Quenching of 0.020 M p-3VB by Pentadiene in Acetonitrile ............... Quenching of 0.040 M p-3VB by Pentadiene in Acetonitrile ............... XV Page 236 237 238 239 240 241 242 243 244 245 246 247 Table 92 93 94 95 96 97 98 99 100 '101 102 103 Quenching of 0.060 M p-3VB by Pentadiene in Acetonitrile ................ Quenching of 0.10 M p-3VB by Pentadiene in ACetonitrile ................ Quenching of 0.020 M m-2VB by Pentadiene in Benzene ................... Quenching of m-2VB by Pentadiene in Acetonitrile .................. Effect of Ketone Concentration on II for p-2VB and p-3VB in Acetonitrile ...... Disappearance Quantum Yields for p-ZVB and p-ZAB in Acetonitrile ........... Quenching of 0.020 M p-Me03E by Penta- diene in Acetonitrile ............. Quenching of p-3VE by Pentadiene in Benzene .................... Quenching of Ketone Disappearance for p- 3VE and p-2AE by Pentadiene in Acetonitrile . . Phosphorescence Lifetime for 1.0 x 10'4 M m-Carbomethoxybenzophenone and Irel 4 for 1.0 x 10- M m-ZBB in CCl4 ......... Phosphorescence Lifetime for 1.0 x 10"4 M p-Carbomethoxybenzophenone and Irel for 4 1.0 x 10' M p-ZBB in CCl 4 00000000000 Phosphorescence Lifetime for 1.0 x 10'4 M p-ZBB in CC14 ................ xvi Page 248 249 250 251 252 253 255 256 257 258 259 260 Table Page 104 Quenching of 0.040 M Butyrophenone in CCl4 ...................... 261 xvii Figure LIST OF FIGURES Modified Jablonski Diagram for Phenyl Alkyl Ketones ................ Stern Volmer Plots for Substituted Valerophenones in Benzene with Diene Quencher (O H;.o-CN;Am-CN;Ip-CN) . . . . Stern Volmer Plots for Substituted Valerophenones in Acetonitrile with Diene Quencher ( O H; O o-CN;A m-CN; II p-CN) .................. Stern Volmer Plots fer Substituted Valerophenones in Benzene with Diene Quencher (O o-CDZCH3; Um-c02ca3; A p-COZCH3) ................ Stern Volmer Plots for Substituted Valerophenones in Acetonitrile with Diene Quencher ( O o-COZCH3; E] m- cozcn3;Ap-cozcn3) ............. Stern Volmer Plots fbr Substituted Valerophenones in Benzene with Diene Quencher (An-cocH3; g p-COC4H9; O m-C0C4H9; .m-C0C4H9 in acetonitrile) . . . xviii Page 35 36 37 39 Figure Page 7 Stern Volmer Plots for Substituted Valerophenones in Benzene (0.5 M Pyri- dine) with Triethylamine Quencher ((3 H; 1:] n-cu;Ap-CN) ................ 41 8 Stern Volmer Plots for Substituted Val- erophenones in Acetonitrile with Tri- ethylamine Quencher (0 H; E] m-CN; A p-CN) ................... 42 9 Stern Volmer Plots for Substituted Val- erophenones in Benzene (0.5 M Pyridine) with Triethylamine Quencher (C) o-COZCH3; L'J m-cozcn3;A p-COZCH3) ........... 43 10 Stern Volmer Plots fOr Substituted Val- erophenones in Acetonitrile with Tri- ethylamine Quencher (C) o-COZCH3; A p-cozcu3) ................. 44 11 Stern Volmer Plots for m-Divalerylbenzene in Benzene (0.5 M Pyridine)() and Aceto- nitrileEl with Triethylamine Quencher ..... 45 12 Stern Volmer Plot for p-Cyanovalero- phenone in Benzene (0.5 M Pyridine) and Low Triethylamine Quencher Concentrations. . . 46 13 Maximization of II for p-Cyanovalerophenone in Benzene with Added t-Butyl Alcohol 0 and Added PyridineA ............... 47 xix Figure Page 14 Maximization of QII for p-Cyanovalero- phenone in Benzene with Added Dioxane ..... 48 15 Stern Volmer Plots for p-Cyanophenyl Alkyl Ketones in Benzene with Diene Quencher (E3 p-Cyanobutyrophenone; Ap-Cyanovalerophenone; O p-Cyano- - methylvalerOphenone) ............. 49 16 Results from Reaction of p—Cyanoaceto- phenone and Toluene in Acetonitrile (Formation of Bibenzyl Monitored). . ... . . . 52 17 Stern Volmer Plot for p-Cyanoacetophenone in Acetonitrile (1.0 M Toluene) with Naphthalene Quencher (Formation of Bi- benzyl Monitored) ............... 53 18 Results from Reaction of p-Cyanoaceto- phenone and p-Xylene in Acetonitrile (Formation of p-Ditolylethane Monitored) . . . 54 19 Stern Volmer Plot for p-Cyanoacetophenone in Acetonitrile (1.0 M p-Xylene) with Naphthalene Quencher (Formation of p- Ditolylethane Monitored) ........... 55 20 Phosphorescence Spectra of 10-4 M Valero- phones in 5:1 Methylcyclohexane/Iso- pentane at 77°K. Top, Valerophenone; Second, m-CyanovalerOphenone; Third, XX Figure 20 cont. 21 22 23 24 p-Cyanovalerophenone; and Bottom, 0- Cyanovalerophenone in Isopentane Glass ................... Phosphorescence Spectra of 10'4 M Valero- phenones in Isopentane at 77°K. Top, o-Carbomethoxyvalerophenone; Middle, me Carbomethoxyvalerophenone; Bottom, p- Carbomethoxyvalerophenone ......... Phosphorescence Spectra of 10'4 M Valero- phenones in Isopentane at 77°K. Top, Valerophenone (MCH/IP glass); Middle, m-Divalerylbenzene; Bottom, p-Divaleryl- benzene .................. Cyclic Voltammagrams for 10'4 M Valero- phenones in Acetonitrile (0.1 M TEAP) versus SCE (200 mV/sec). Top, Valero- phenone: Second. o-CyanovalerOphenone; Third, m~Cyanovalerophenone; Bottom, p- Cyanovalerophenone ............. 4 M Valero- Cyclic Voltammagrams for 10' phenones in Acetonitrile (0.1 M TEAP) versus SCE (200 mV/sec). Top, o-Carbo- methoxyvalerophenone; Middle, m-Carbo- methoxyvalerophenone; Bottom, p-Carbo- methoxyvalerophenone ............ xxi Page 60 61 62 65 66 Figure 25 26 27 28 29 Cyclic Voltammagrams for 10'4 Valero- phenones in Acetonitrile (0.1 M TEAP) versus SCE (200 mV/sec). pr, m-Di- valerylbenzene; Middle, p-Divaleryl- benzene; Bottom, p-Acetylvalerophenone. . . . Cyclic Voltammagrams for 10'4 M Valero- phenones in Acetonitrile (0.1 M TEAP) versus SCE (500 mV/sec). Top, o-Tri- f1uoromethylvalerophenone; Middle, p- Trifluoromethylvalerophenone; Bottom, m-Trifluoromethylvalerophenone ........ Plot of Relative kobs Values as a Function of Ground State Hammett 0 Values. 0 , -I Substituents;G, Elec- tron-donating Substituents;€9. para-R Substituents;.. p-c1;D . p-scr3 ...... Correlation of Charge Transfer Quench- ing Rates fbr Substituted Valerophenones in Benzene (0.5 M Pyridine) with Ex- cited State Reduction Potentials: O, n,:r* Lowest Triplet;El 1r,n* Lowest Triplet ................... Correlation of Half-wave Reduction Po- tentials for Substituted Valerophenones with Hamnett om and o ' Values: 0 meta- P xxii Page 67 68 87 91 Figure 29 cont. 30 31 32 33 34 Substitution; E1 para-Subs ti tuti on ...... Correlation of n,n* Triplet Energies (4:1 MCH/IP, 77°K)of Substituted Benzo- phenones with Hamett Parameters.89a U . I (-R) Substituents;., I Values Added From This Study ............ Correlation of n,n* Triplet Energies (4:1 MCH/IP, 77°K) of Substituted Benzo- phenones with Haumett Parameters: D , I (-R) Substituents;o. Cl Values Taken From Reference 89a ............. Attempted Correlation of n,n* Triplet Energies (Isopentane, 77°K) with Half- wave Reduction Potentials for Substituted Valerophenones: O n,1r* Lowest Triplets; [J'n,n* Lowest Triplets ........... Attempted Correlation of n,n* Triplet Energies (Ethanol, 77°K) with Half-wave Reduction Potentials for Substituted Valerophenones: C) n,n* Lowest Triplets; I 1r,1r* Lowest Triplets ........... Stern Volmer Plots for Diene Quenching of 0.020 M Ketones: C) p2VB in Benzene; O p2VB in Acetonitrile;A p3VB in Aceto- nitrile;1:] m2VB in Benzene; and-m2VB in Acetonitrile ................ xxiii Page 96 98 98 100 101 104 Figure Page 35 Stern Volmer Plot for Diene QuenChing of 0.020 M p3VE in Benzene ............ 105 36 Stern Volmer Plot for Diene Quenching of 0.020 M p-Me03E in Acetonitrile ...... 106 37 Extrapolation of Triplet Lifetimes of p2VB C) and p3VB.(D in Acetonitrile ...... 110 38 "Stern Volmer Plot for Diene Quenching of Room Temperature Phosphorescence of m- Carbanethoxybenzophenone O and p-Carbomethoxy- benzophenone 1| in CCl4 ............ 111 39 Stern Volmer Plot for Diene Quenching of Room Temperature Phosphorescence of p288 in CC14 ................. 112 40 Stern Volmer Plot for 0.040 M Butyrophenone in CCl4 with Hexadiene Quencher'C) or Triethylamine Quencher O .......... 115 41 Extrapolation of III for p2VBO and p3VB O in Acetonitrile to Infinite Dilution . . . 116 42 Extrapolation of ¢II for m2VB in Aceto- nitrile to Infinite Dilution ......... 117 43 Extrapolation of é-K for p2VB C) and p2AB . in Acetonitrile to Infinite Dilution ................... 118 xxiv INTRODUCTION A. Photophysical Processes Molecules can undergo a wide variety of photophysical and photo- chemical processes upon the absorption of light. In order to under- stand better the photochemical reactions that phenyl alkyl ketones undergo, the many available photophysical processes will be discussed briefly. A modified Jablonski diagram] for phenyl alkyl ketones shows the photophysical processes which can occur in the absence of photo- chemical reactions (Figure 1). In this diagram the absorption and Sn : I __J III k 51—— . '0' T‘ . Th «‘2 oil in: k} Nil kill $o- Figure 1. Modified Jablonski Diagram for Phenyl Alkyl Ketones. emission of light are depicted by straight lines whereas radiation- less transitions are shown as wavy lines. A molecule absorbs a photon and is excited to an upper singlet state which rapidly under- 012 s") to the lowest excited singlet 6 goes internal conversion (kic’b 1 5'1), radiationless 3 51.2 51 can then undergo ffiuorescence (kfvt 10 decay (kd N 107 5'1), or intersystem crossing to the triplet state. The rate of intersystem crossing depends on the amount of spin- orbit coupling which allows the singlet to attain some triplet character.4 The extent of spin-orbit coupling depends both on the energy gap between S1 and the triplet state3 and the nature of the 4a singlet and triplet states. Both experimental observations and selection rules derived from electronic overlap integrals suggest that intersystem crossing from an n,n* singlet to a n.n* triplet is approximately 103 times faster than from an n,n* singlet to an n,n* triplet.4b’c For phenyl alkyl ketones, which possess n,n* lowest singlets, the rate of intersystem crossing is extremely fast (k 'M 1010 s").5 ISC Thus, fluorescence and radiationless decay are too slow to be com- petitive singlet processes. Essentially all the singlets undergo spin inversion to populate the upper vibrational states of the triplet(s). Again, rapid internal conversion (kic’b 1012 5") leads to population of the lowest vibrational level of the triplet(s).2 In many phenyl alkyl ketones, two triplets (n,h* and n,n*) are in 5 close proximity and undergo vibronic coupling or thermal equilibra- 6 tion before undergoing reaction or decay. The triplet(s) can then undergo phosphorescence (kp m 10.1 to 103 5") or radiationless decay if other photochemical processes are unavailable.3 For many phenyl alkyl ketones, hydrogen abstraction is very fast compared to both radiationless decay and phosphorescence, and thus photochemi- cal reaction leading to product formation is the major pathway fbr deactivation of the triplet state. B. Norrish Type II Photochemical Reaction The Norrish Type II photochemical reaction is the major decay process for most phenyl alkyl ketones possessing y-hydrogens. This reaction was first discovered in 1934 by Norrish and Appleyard7 when they feund that methyl butyl ketone photodecomposes to give acetone and propene in the gas phase. Further investigation of other ketones showed a characteristic a,B-bond cleavage to give a smaller carbonyl compound and an olefin.8 In 1958 Yang and Yang9 reported the formation of cyclobutanols as minor products in addition to cleavage products and suggested the formation of a 1,4-biradica1 intermediate to explain the results. Further evidence supporting the 10 and others.11 intermediacy of a 1,4-biradical was provided by Wagner Thus. y-hydrogen abstraction by the triplet state to fbrm a 1,4-bi- radical, followed by cleavage or cyclization is the generally ac-: cepted mechanism for the Norrish Type II photochemical reaction (Scheme 1). The 1,4-biradica1 also can undergo reverse hydrogen transfer to give ground state ketone.10a The addition of a Lewis base suppresses the reversion of the biradical to starting ketone by hydrogen bonding the hydroxyl hydrogen, thus slowing down reverse hydrogen transfer 1 1 «it - 3 at X X X . k'sc W kr h, \cvc knev “all )( )( ’99>‘JF"J\I‘.} . -+ gees. tui Scheme 1. Norrish Type II Photochemical Reaction. and enabling cleavage or cyclization to occur exclusively.mb’12 The quantum yield for cyclization products averages only approxi- 13 mately 10-20% of the quantum yield for cleavage products. Also, the rate of yehydrogen abstraction is dependent on both the stability of the incipient radical center,”’15 and the reactivity of the triplet state toward hydrogen abstraction. Both the electronic nature of the triplet state and the effects of substituents determine the reactivity of phenyl alkyl ketones.1]b’]6'26 C. Triplet States of Phenyl Alkyl Ketones Phenyl alkylketones possess two triplet states which differ significantly in their electronic distribution. An n,n* triplet results from excitation of a non-bonding electron of the carbonyl to a n-antibonding orbital, thus producing electron deficiency at the oxygen atom. The n,n* triplet resembles an alkoxy radical and under- goes typical radical reactions such as hydrogen abstraction.16’17 Analogously, a n,n* triplet results from excitation of a n electron to a n-antibonding orbital, resulting in increased electron density 18.19 at the oxygen atom. Consequently, the n,n* triplet is not re- active in typical electrophilic radical reactions. Valence bond representations of the two triplet states are given below. a. I 6 3m!" I “ I : Z: 3 :1 2 ’ . 311-,1'r‘ Many examples in the literature indicate that phenyl ketones with n,n* lowest triplets are reactive towards hydrogen abstraction re- actions; whereas, ketones with n,h* lowest triplets are either un- reactive or display substantially decreased reactivities. Yang and 20 Dusenbery reported that substituted benzophenones (4-methy1 or 4-trifluoromethyl) that retain an n,n* lowest triplet are easily photoreduced by 24propanol. Conversely, 4-phenylben20phenone, which possesses a n,n* lowest triplet,2] displays a 103-fold decrease in reactivity compared to benzophenone itself.22 Similarly, the photo- reduction of 4-trifluoromethylacetophenone (n,n* lowest triplet) dis- plays a six-fold increase in the rate of hydrogen abstraction, whereas 4-methylacetophenone (n,n* lowest triplet) shows a tenfold decrease in 23’24 Likewise, the Norrish reactivity, both compared to acetophenone. Type II photochemical reactions of substituted butyrophenones and valerophenones follow a similar trend: ketones with n,n* lowest triplets undergo intramolecular y-hydrogen abstraction readily, while ketones with definite n,n* lowest triplets are essentially unreac- tive.]]b’25'29 In the past two decades the determination of substituent effects on the photochemistry of phenyl ketones has received considerable attention. The effects of ring-substituents on both the triplet energies and triplet-state photoreactivity have been investigated extensively.11b’21’23424a27:28.30 In general, the effects of ring- substituents on the energies of both triplets, thus determining the nature of the lowest triplet, overshadow their electronic effects on the individual triplet states. In benzophenone the n,n* triplet is far enough below the n,n* triplet such that most substituents do not 31 cause an inversion of the two triplet states. In unsubstituted phenyl alkyl ketones the n,n* triplet lies only approximately 2.8 30b Polar solvents kcal below the n,n* triplet in nonpolar solvents. stabilize the n,n* triplet and destabilize the n,n* triplet, since the n.n* triplet has a lower dipole moment and the n,n* triplet has a higher dipole moment than the ground state."b’30d Electron-donating substituents stabilize the n,n* triplet and destabilize the n,n*. triplet, resulting in inversion of the two triplets and decreased reactivity.28 Conversely, electron-withdrawing substituents stabilize the n,«* triplet relative to the n,n* triplet and also make the n,n* triplet more electrophilic, thus making such triplets more reactive than those of unsubstituted ketones (Scheme 2). Q . 315‘" e 3 * ‘ I ~‘~__ “3“ e :BDRW' I! \)I‘ 31w“ I ‘\ 3nar" \ ‘x 34w“ «0"DKJNKIR '-tl {f'VVfTFNDHUUNlflR Scheme 2. Substituent Effects on Triplet Energies for Phenyl Alkyl Ketones. It is generally accepted that n,n* triplets are much less re- active than n,n* triplets towards hydrogen abstraction, and ketones with n.n* lowest triplets derive their reactivity primarily from equilibrium populations of upper n,n* triplets.28’32 Wagner28 has provided evidence for photoreactivity from upper n,n* triplets for p-methoxyphenyl alkyl ketones, since their rates for hydrogen ab- . straction exactly parallel the rates for unsubstituted phenyl alkyl ketones (n,n* lowest triplets) with similar y- or d-substitution. Also, the variation of photoproduct distribution with temperature for l-benzoyl-4-p-anisqy1butane is consistent with thermal equilibration of the two nonconjugated chromophores, and reaction of the anisoyl moiety from an equilibrium population of its n,n* triplet.32b At a certain temperature, T, the equilibrium population of both triplets is dependent upon the energy separation between the two triplets. AET, and fbllows a Boltzmann distribution (Equation 1) -AE RT 52.11“, T, (1) Xn,n where X is the equilibrium fractional population of a given triplet state. At room temperature the two triplets must be in close prox- imity (AET g_3 kcal) in order for thermal equilibration to occur. If the observed rate of hydrogen abstraction reflects reactivity from both triplets, then Equation (2) describes the reaction kinetics. obs n h "H ’3 XnnrkH I ankh (2) When AET §_3 kcal, reaction occurs primarily from population of the upper n,n* triplet.28 obs _ n kr ' xn,irkr (3) If the two triplets are in close proximity then vibronic coupling can induce some n,n* character into the n,n* triplet and vice versa.5 Therefore, the wave function W of the lowest triplet is a linear combination of the two unperturbed wave functions o, with the mixing coefficients a and b dependent upon both the energy gap between the triplets and symmetry considerations.5 RT] = aW(n.n*) + bw(n,n*) (4) Thus, some reactivity possibly can occur from the vibronically induced n.n* character of the n,n* triplet.23b 0. Energy Transfer and Charge Transfer Processes 1. Energy Transfer Processes a. Intermolegglar - Energy transfer is an important photo- chemical process involving a radiationless transfer of electronic excitation from a donor molecule to a suitable acceptor molecule. Triplet-triplet energy transfer (Equation 5) is the most common type of energy transfer, since the lowest triplet state of a molecule is longer-lived than the corresponding lowest singlet state and therefore has a greater probability of transferring its electronic excitation energy to a suitable acceptor molecule.32 0*(T1) + A(So)-+ 0(50) + A*(T1) (5) Triplet energy transfer can occur when the acceptor has a lower 1O 33 triplet energy than the donor molecule. Usually, the efficiency of an acceptor or quencher is determined solely by the position of its lowest triplet level and not by its molecular structure.34 In general, triplet energy transfer in solution occurs via an electron- exchange mechanism which requires diffusion of the donor and acceptor molecules to within collisional separation as a rate-limiting step.33 Thus, in relatively viscous solvents is triplet energy transfer diffusion controlled and described by a modified Debye equation.35’36 k = k = 8RT/2000 n 1 nor1 5" (6) et dif Only in solvents of low viscosity, the exothermic transfer of triplet excitation energy is not completely efficient, and an excited donor and an acceptor molecule can diffuse apart before energy transfer can occur.33’37 Since every collision between donor and acceptor does not result in energy transfer, then a conformational requirement for effective orbital overlap may be necessary in order for energy trans- 38 Wagner and coworkers39 have indicated that efficient fer to occur. energy transfer occurs at van der Waals separation (approximately 4 A for n-system overlap with a carbonyl moiety) between donor and acceptor molecules. In order to gain more insight into the conforma- tional requirements involved in energy transfer, several researchers have studied intramolecular energy transfer in systems containing isolated chromophores having known spatial dispositions relative to each other.40'43 11 b. Intrgmolecglar - There are many examples of intramolecular triplet-triplet energy transfer reported in the literature. Hammond gt_gl,4o found that triplet energy transfer from a benzophenone chromophore to a naphthalene moiety separated by one to three methylene units (the distance between the carbonyl carbon and the center of the naphthalene group is at most 10 A in all three cases) occurs with 100% efficiency with a rate constant greater than 1010 5']. Likewise. Cowan and Baum41 investigated a series of compounds in which aceto- phenone and trans-B-styryl chromophores were separated by one to four methylene units and found efficient isomerization of the styryl group upon irradiation of the benzoyl chromophore. However, the reported 20-fold decrease in the rate of triplet energy transfer in going from two to four methylene groups (ket = 7.2 x 1010 s'1 for n = 2 and ket = 3.3 x 109 s'1 for n = 4) suggests that an increase in the number of methylene units results in a decrease in the number of desirable con- fbrmations (donor and acceptor in close proximity) relative to the number of available conformations. Thus, it appears that the two chromophores have to be in close contact in order for energy transfer to occur. Also, there are a few examples of intramolecular triplet-triplet energy transfer occurring in compounds containing two isolated chromo- phores held at fixed distances in rigid molecules. Zimmerman and 42 investigated the phosphorescence emission of 1-benzoyl-4- McKelvey «z-naphthyl)-bicyclo[2.2.2]octane, in which the benzoyl chromophore (donor) and naphthyl moiety (acceptor) are separated by approximately 7 A, and found that intramolecular triplet energy transfer occurs’ 12 with 100% efficiency. In contrast, Keller and Dolby43 found that the rate constant for triplet energy transfer from benzophenone to a naphthalene group held approximately 14 A apart by a rigid steroid bridge was only 25 5'1. 43 0f even greater interest is the fact that these authors also discovered that energy transfer from a carbazole donor (n,n* lowest triplet) to a naphthalene acceptor held at a com- parable separation (approximately 15 A) was 103 times slower (ket = 0.04 5"). Thus, the conclusion was reached that triplet-triplet energy transfer is not only distance dependent but also may be struc- ture dependent, with n,n* +-n,h* triplet-triplet energy transfer being slower by three orders of magnitude than n,n* +-n,n* triplet- triplet energy transfer for the same donor-acceptor separation. 2. Charge Transfer Processes a. Intermolecular - Carbonyl compounds can be photoreduced 44 45 alkanes, 49 by a wide variety of hydrogen sources including alcohols, 45’46 47 48 and amines. alkyl benzenes, tributylstannane, ethers, It is well established that hydrogen abstraction from alcohols and alkanes by triplet carbonyls involves the simultaneous cleavage of the R-H bond and the formation of the 0-H bond to give two radical species.50 3ch = 0* + R'-H + RzC-OH + R'. .However, there are many differences in the photoreduction of ketones 13 by amines as compared to alcohols, alkanes, and many other hydrogen donors. For example, the rate of hydrogen abstraction for triplet benzophenone with alkanes depends upon the hydrocarbon C-H bond strength,51 whereas the photoreduction of benzophenone by N-alkylated diphenylamines displays little dependence on donor bond strength.52 53 reported that the photoreduction of benzophenone by Cohen and Litt triethylamine is 103 times faster than with isopropanol as hydrogen donor. Also, Cohen and Green54 found that the rate of photoreduction of acetophenone with a-methylbenzylamine was twenty times greater than the rate of photoreduction with a-methylbenzyl alcohol. Both fluor- enone and p-aminobenzophenone (n,n* lowest triplets) are not reduced by alcohols, but are readily photoreduced by tertiary amines.55’56 In order to account for the greater reactivity of amines, Cohen and 56 proposed the following mechanism involving the initial trans- Cohen fer of an electron from the nitrogen atom to the excited carbonyl. followed by proton transfer and electron redistribution. 3R c=o* + R' NCH R" :ch [R to R' 'N+CH W 2 2 2 2 2k 2 + 4 RZCOH + R'ZNCHR" The charge transfer complex can also undergo spin and charge destruc- 57 tion to give ground state ketone. Thus, charge transfer interaction between the excited carbonyl moiety and the amine can result in + k ° ' I ° u q .. l u [RZC-O R ZNCHZR ] -> RZC-O + R ZNCHZR 14 either radical fermation or deactivation to ground state ketone, depending on the nature of both the donor and acceptor molecules in- volved in the charge transfer complex.58 Rates of charge transfer interaction between donor and electron- ically excited acceptor have been fbund to be dependent on their thermodynamic properties. In 1968 Weller59 related the rate of fluores- cence quenching of aromatic hydrocarbons by amines to the change in free energy involved in electron transfer, AGCT’ which was found to be dependent on the oxidation potential of the donor, the reduction potential of the acceptor, the singlet excitation energy of the ac- ceptor, and a Coulombic term related to the free energy gained by bringing the ions to the encounter distance (Equation 7). AGCT . sin/0*) - E(A‘/A) - A‘E(A) - efi/ea (7) Weller6o determined that in solvents with large dielectric constants, the rate of charge transfer quenching of singlets produces free radical ions, and is diffusion controlled when AGCT < -10 kcal/mole. The rate constant of charge transfer quenching decreases with increas- ing AGCT and becomes proportional to exp (-AGCT/RT) when AGCT > 5 kcal/mole. Thus, when AGCT is endothermic, a plot of log kCT versus . AGcT has a slope of -16.5 eV for processes involving full electron transfer. Mataga gt_gl,6] observed the fluorescence from pyrene and N,N-dimethylaniline in various solvents. From the fluorescence quantum yields and lifetimes they concluded that in polar solvents the emiSsion was due to a strong charge transfer complex, while the 15 emission in nonpolar solvents was due to a weak CT complex (only partial electron transfer). Davis and coworkers62 showed that the rate of fluorescence quenching of fluorenone by various amines was inversely proportional to the ionization potential of the amine. Thus, Heller's equation appeared to be applicable to a large number of systems. A similar relationship was sought for the reaction of excited- 63 reported that for the state triplets with amines. Cohen and Stein photoreduction of sodium 4-benzoylbenzoate by tertiary aliphatic amines, the rate of interaction between ketone triplets and amines increased as the ionization potential of the amine decreased. In 64 related the rate constant for donor- 1972 Guttenplan and Cohen acceptor charge transfer interaction, kCT’ to the change in free energy, AGCT, fer carbonyl excited triplet states and various donors. A modified "Heller's Equation" was proposed which related AGCT to the ionization potential of the donor, the reduction potential of the acceptor, and a constant (Equation 8). 3 Eo,o + The use of the more readily available ionization potentials instead of the oxidation potentials of the donors only changes the value of the constant term. Thus, for a series of donors with a constant acceptor, log kCT «'IPD + C' and for a series of acceptors with a constant donor, log kCT m -E(A'/A) - A3Eo,o + C“. Guttenplan and Cohen64b proposed a linear correlation between log kCT and the excited 16 state reduction potential (-E(A'/A) - A3Eo o) for the quenching by triethylamine of various carbonyl acceptors, which possessed widely differing reduction potentials, triplet energies, and triplet con- 643 also reported a linear inverse rela- figurations. These authors tionship between log kCT and the donor ionization potential for the charge transfer quenching of benzophenone by various donors includ- ing amines, sulfides, mercaptans, ethers, alcohols, aromatic hydro- carbons, and olefins. TWo lines were obtained, one for aliphatic donors (slope = -0.067 mole/kcal) and the other for aromatic donors (slope 8 -0.105 mole/kcal), on a graph of log kCT versus IPD. The values of the slopes are much smaller than that observed in systems involving full electron transfer (slope = -0.74 mole/kcal).65 Also, only small polar solvent effects were observed for the charge trans- fer quenching rate; (a factor of 2 in acetonitrile compared to benzene). A 13-fbld increase in the fluorescence quenching rate for a correspond- ing process involving full electron transfer was reported.66 Many studies indicate that full electron transfer is not involved in charge transfer complex formation between carbonyl triplets and amines. The small polar solvent effects and the decreased slopes of log kCT versus IPD plots suggest that at most only approximately 20% electron transfer is involved in the charge transfer complex formation.64a’67’68 69 Bartholomew gt_g1, later conducted flash studies on the photoreduc- tion of benzophenone with tertiary amines and found that a charge transfer complex forms in acetonitrile solvent; an exiplex may be 70 fermed in less polar solvents. Shaefer and Peters used picosecond absorption techniques to confirm the initial formation of radical 17 ions from benzophenone and triethylamine in acetonitrile, followed by either proton transfer to form the benzhydrol and amine radicals or decay back down to ground state molecules. Also, Amouyal and 7] used laser flash absorption techniques to investigate Bensasson the photoreaction of triplet duroquinone with tertiary amines and proposed that an exiplex is initially formed in both polar and non- polar solvents, followed by the formation of a charge transfer com- plex in solvents of high dielectric constant. The rate of charge transfer interaction between tertiary amines and carbonyls is fast, since these donors have relatively low ioniza- tion potentials.72 Tertiary amines undergo charge transfer complex formation with n,n* triplets of ketones such as benzophenone and acetophenone at nearly diffusion-controlled rates, leading to both quenching and radical formation.57"73 Mueller74 quenched the type II photoelimination of many ring-substituted valerophenones with tri- ethylamine in both benzene and acetonitrile solvents and observed diffusion-controlled quenching rates for those ketones with rela- tively low reduction potentials. Also, Wagner and Kemppainen75 reported an order of magnitude decrease in the quenching rate of .valerophenone with triethylamine in methanol as compared to benzene and acetonitrile solvents. The decrease was attributed to the de- creased availability of the hydrogen-bonded lone-pair electrons on the nitrogen atom. In addition to amines, other types of compounds are suitable electron donors for excited ketones. Kochevar and Wagner76 have investigated the quenching of the type II photoelimination of 18 butyrophenone with a wide variety of olefins and determined that charge transfer quenching predominates for electron-rich olefins, while energy transfer quenching is dominant for electron-deficient 77 reported that in the photoreduction olefins. Wagner and Leavitt of a-trifluoroacetophenone with alkylbenzenes, both charge transfer complex formation and direct hydrogen abstraction are involved. Cohen and Guttenplan78 determined that thioethers quench the triplet benzo- phenone in benzene and acetonitrile (kCT m 107 - 109 M'1 5"), but at a slower rate than tertiary amines, since sulfides have higher ioniza- tion potentials than tertiary amines. Also, both aromatic and ali- phatic mercaptans quench the phosphorescence of benzophenone with rate constants between 107 - 109 M'1 5'], respectively. In addition, ‘phosphorus, antimony, arsenic, and bismuth are thought to quench the type II photoelimination of butyrophenone by a charge transfer mechan- ism.50 Thus, charge transfer complex formation is a common type of interaction between molecules capable of donating or accepting addi- tional charge in the excited state. b. Intramolecular - Due to the limited number of conformations possible in bifunctional molecules, the study of intramolecular charge transfer interactions in these compounds can provide insight into the structure of the CT complex and the differences in charge trans- fer interactions with n,h* and n,n* triplets. In the past decade a number of studies concerning intramolecular charge transfer complex fbrmation have been published.73’75’79'84 The photochemical reaction of dialkylphenylacylamines was used as 19 a synthetic route to 3-hydroxyazetidines in only modest yields (9-22%), since type II photoelimination products (30-55% yields) were also ob- tained.79 80 Later, Padwa and coworkers investigated two N,N-dibenzyl- phenacylamines having either an n,n* or n,n* lowest triplet and found 1 [cu-12m I \cwzph —'-‘V——p + PhCH=NCH2Ph n ‘wflwene~ ll==ti,F!i similar type II quantun yields (on = 0.14 and 0.12, respectively) in ethanol. The lack of quenching by 1,3-cyclohexadiene and piperylene and the low type II quantum yields suggested that a charge transfer complex was fonmed at a rate exceeding diffusional quenching followed 20 by either back electron transfer to regenerate starting ketone or . proton transfer to generate a 1,4-biradica1. Padwag_i_:_31_.81 found that 3-aroylazetidines, which possess either an n,n* or n,n* triplet, give only arylpyrroles as photoproducts in low quantum yields (o 4.0.1). The authors proposed the rapid fbrmation of a charge transfer complex, fellowed by proton transfer and electron reorganiza- tion to form a biradical, bond closure, and elimination of water to give the observed products. The corresponding hydrochloride salts were found to be photochemically inert, since the nitrogen lone pair is no longer available. Thus, it appears that both n,n* and n,n* triplets can undergo charge transfer interactions. Wagner and coworkers73’75 investigated the photoreactions of a,y, and 6-dia1kylaminoketones in a variety of solvents. The inter- system crossing yields for these ketones are less than unity, with n: 1,3,4 21 lower values for decreasing values of n (AISC m 0.01 for n = l in benzene or acetonitrile). When n = 1, an unquenchable singlet state is responsible for all the photoelimination in benzene and aceto- nitrile and two-thirds of the photoelimination in methanol. In contrast, the photoelimination in compounds where n = 3 or 4 results from direct triplet-state y-hydrogen abstraction. Rapid intra- molecular charge transfer quenching is a competitive triplet-state process which results in deactivation of the triplet state to form ground-state ketone. Charge transfer quenching is 3.5 times faster for n = 3 than for n = 4, and five times faster in benzene or aceto- nitrile than in methanol. Since charge transfer quenching of the triplet does not result in biradical formation, the nitrogen lone pair must be held near the carbonyl and the y-hydrogens away from the carbonyl moiety in the charge transfer complex. 67 also investigated 4-benzoyl-4,N-dimethyl- Wagner and Scheve piperidine (n,n* lowest triplet) in which the nitrogen lone pair is held approximately 4-6 A from the carbonyl group. The two possible conformers formed distinct, non-interconverting triplets: the triplet having the benzoyl group axial underwent rapid y-hydrogen abstraction, fbllowed by cyclization of the 1,4-biradical, while the triplet having the benzoyl group equatorial underwent only a-cleavage. When the nitrogen lone pair is axial, the relative orientation and overlap of orbitals is poor and charge transfer quenching does not occur. The rate constant for CT quenching of the other triplet cannot exceed 2x1061 5' . This rate is slower by a factor of 103 as compared to the rate of CT quenching in the acyclic y-amino ketones.77 From these 22 31.¢ 11* 3" 1* L ‘Me‘ .9" l l ., 4. i“ g .t results the authors concluded that CT quenching in acyclic y-amino- ketones occurs solely by through-space rather than through-bond inter- actions. Wagner and Ersfeld67’82 photolyzed B-naphthyl-y-dimethylamino- propyl ketone and found inefficient photoelimination (411 = 0.008) in both benzene and acetonitrile solvents due to a n,n* lowest triplet and efficient CT interaction between the triplet state and the amine. The authors proposed that the cyclic charge transfer complex formed must be restricted to conformations in which the hydrogens a to the nitrogen are kept away from the carbonyl such that 1,5-proton transfer cannot occur. In contrast, bimolecular photoreduction of triplet 23 naphthyl ketones by amines results in efficient radical formation.83 In methanol the photoelimination proceeds at moderate efficiency (all s 0.17) and it was suggested that protonation of the oxygen atom in the CT complex by methanol leads to diradical formation and type II photoproducts.82 84 Winnik and Hsiao studied a series of benzophenone derivatives in which the benzophenone chromophore and an ethenyl moiety were separated by one to nine methylene units and found significant intra- molecular quenching of benzophenone room temperature phosphorescence emission when n = 9. The authors concluded that charge transfer quenching occurs only when the double bond approaches a van der Waals distance from the benzophenone carbonyl moiety and that CT quenching involves electron transfer to the electron deficient n orbital loca- lized on the carbonyl oxygen. Later experiments perfbrmed by Mar and Winnik85 with an extension of the methylene chain showed an in- crease in the quenching rate until n = 12, and then a rapid decrease in quenching (n = 13 to 21) due to the rapid increase in the total number of chain conformations as compared to the only modest increase in the total number of chain conformations which lead to quenching. Thus, these experiments suggest that a specific geometry is required in order for efficient charge transfer to occur. 24 86 Masuhara t 1. used laser spectroscopy to study systems in which benzophenone and N,N-dimethylaniline are connected by a methylene n==1;2;3 chain. The transient absorption spectra observed in acetonitrile could be reproduced by superposition of those of the benz0phenone anion and DMA cation, and the spectra in benzene were similar to those of the benzophenone ketyl radical or triplet benzophenone. Thus, the authors proposed that intramolecular hydrogen abstraction resulting in the formation of benzophenone ketyl radical occurred in benzene whereas intramolecular electron transfer occurred in aceto- nitrile, and both processes were almost completely independent of the number of methylene units. Thus, they concluded that a definite, restricted structure was not necessary for electron transfer or hydrogen abstraction, since these processes occurred rapidly both in a loose structure (n = 1,2) or in a sandwich-type structure (n = 3). These results contradict those of previous researchers and seem suspect to error, since when n = l the lone pair on the amine cannot possibly come close to the carbonyl and the possibility for intramolecular hydrogen abstraction seems remote when n = 1. Thus, many unanswered questions still remain concerning the con- fbrmational requirements involved in charge transfer complex formation. 25 E. Stern-Volmer Kinetics Phenyl alkyl ketones undergo Norrish Type II photoelimination by y-hydrogen abstraction to give a 1,4-biradica1 which then undergoes cleavage, cyclization, or disproportionation back to ground-state ketone (Scheme 1). The triplet lifetime, 1, is dependent on both the rate constant for y-hydrogen abstraction, kr’ and the rate constant for radiationless decay, kd (Equation 9). _ -1 T ' (kr + kd) (9) Quantum yields for type II photoelimination are determined by competitive excited state reactions and by partitioning of the bi- radical: (10) where AISC is the quantum yield for intersystem crossing, kr is the rate constant fbr y-hydrogen abstraction, and P is the probability p that the biradical will go on to form product. The addition of a Lewis base usually results in solvation of the biradical, thus in- hibiting disproportionation back to ground-state ketone, and conse- quently maximizing the type II quantum yield (Equation 11). ' k ' T (11) 87 Stern-Volmer quenching kinetics can be used to determine 26 triplet lifetimes: o°l¢ = kq - T ° [0] + l (12) where 4° and a are the quantum yields for acetophenone formation in the absence and presence of external quencher, respectively and kq is the rate constant for bimolecular quenching by external quencher, Q. Intersystem crossing yields can be determined by using the ketone to sensitize a well-known triplet reaction, such as the cis + trans isomerization of 1,3-pentadiene: -1 g -1 1 sens ¢ISC (1 + kq - T - {03’ (13) oi°Pem eeeee.ene=m z.e x =.e 3 PI PI m .mmcocmgaogmpm> umpzuvumnzmimcva Low msmumEugnm ovumcpxouo;m .p mpnmh 33 .gmppmaz .m conga: an mapamme um;m_—n=a:= : a .uamuemu:.\maopm n p gm .mu oucmemmwm ease mwspa> a .nnn mo copm_umgq wmugm>a ;a_3 was; ououwpazv we mango>m appaamzm .mcm~:wn :p ago?» Ezucmzc moccsamaammpo u .eeeeeese : o._ a» z m.c ee.3 unneeexee ape.» eeeeeecd .cowumsgom mcocmnqoumua com upowx Espcmacn .copmgm>:ou Nopim on E: mpm an vagueness? macaw; z wo.ow emm em. ---- em.c eom.o eedneem mau-e emnem ee_ ---- e~.o em~.o eednedm «20-5 eo_nmm can ---- cm.o eo~.o eeeneem mdo-e ---- omnomm ---- ---- Po.cnmo.o dennede m:uou-a dAszwv eAeeenwv x-e xeee HHe eed>_em eee:e.eme=m 2;; Eex e a PI .edaeeeeeu ._ deeee Table 2. Photokinetic Parameters for Phenyl Alkyl Ketones Function of y-Substitution.a ‘scwzcuzaimnz x R1 R2 311 -dmax~ de,M-] 1/t,107s"1 H H H 0.23 ---- 550c 1.3 H H CH3 0.33d 0.85 47:1 11. H CH3 CH3 0.25d 0.90 10c 50. CN H H 0.12 0.31 400240 2.5 CN H CH3 0.17 0.47 74:1 5.7 cu CH3 CH3 0.13 0.47 18 23. a0.04 M ketone in benzene irradiated at 313 nm to ~10% conversion. b formation. cValue in Reference 15 is 625. dValues taken from Reference 15. Values determined from Stern Volmer runs monitoring acetophenone 35 6.0 - 5.0 " 4.0 r ‘0 O O \ \o 1.0 1 l l l l 0.02 0.04 0.06 0.08 0.1 0 101, M Figure 2. Stern Volmer Plots for Substituted Valer0phenones in Benzene with Diene Quencher (OH;.o-CN;Am-CN;Ip-CN). 6.0 5.0 4.0 6- o 3.0 Figure 3. 36 0.02 0.04 0-06 0.08 101ml Stern Volmer Plots for Substituted Valerophenones in Aceto- nitrile with Diene Quencher (OH;.0-CN ;Am-CN;Ip-CN). 37 ' I I. 'I (I A 2.0 ' 0 I A 1° ¢ ‘ V D 1;) a 1.0 0.02 0.04 0.06 0.08 101,M Figure 4. Stern Volmer Plots for Substituted Valerophenones in Benzene with Diene Quencher (Oo-COZCH3;D m-CDZCH3; 38 A I 3.0 - ‘3 IA 0 1:1 qbo ¢ I A I 2.0 (- IA ’1 (D A . ‘ A I, ' 1.0 I 4 1 a l 0.0 0.02 0.04 0.06 0.08 0.1 0 [OLM Figure 5. Stern Volmer Plots for Substituted Valerophenones in Acetonitrile with Diene Quencher ( O o-COZCH3;E] m-CDZCH3; 39 D 4.0 " ’l . d>° 3.0 - cp O H 0 A 0 2.0 - 1‘ ’ A' I A. I y 1.0 1 L 4 L 1 l 0.0 0.02 0.04 0.06 lQl.M Figure 6. Stern Volmer Plots for Substituted Valerophenones in Benzene with Diene Quencher (Ap-CDCH ;E]p-C0C4H9; o m-COC4H9;. m-C0C4H9 in acetonitrilg). (Figures 7-12). In those cases where the intercept was below 1.0, qu was determined by dividing the lepe by the intercept. The values obtained are also included in Table l. 2. anntum Yields Quantum yields for acet0phenone formation were determined by parallel irradiation at 313 nm of 0.04 M ketone solutions and usually benzene solutions of 0.1 M valerophenone as actinometer in a merry- go-round apparatus.9] All samples were degassed prior to irradiation and conversion was usually kept below 10%. In all cases both the ketone solution and the actinometer absorbed all the incident irradia- tion. The values are included in Tables 1 and 2. 3. Maxim Quantun Yields Maximum quantum yields for acetophenone formation were obtained by adding various amounts of Lewis bases to 0.04 M ketone solutions in benzene. For most ketones a maximum quantum yield of unity f0r total product formation could not be attained, since the addition of Lewis base up to a certain value maximized the quantum yield and further Lewis base either had no effect or slightly decreased the quantum yield. The values are given in Tables 1 and 2 and Figures 13 and 14 show the effects of added pyridine, t-butyl alcohol, and dioxane on the Type II quantum yield for p-cyanovalerophenone. 41 a 5L1) ' A El /A qbo» -— 2.0 F o 9’ ' o 0.04 0.08 0.12 101." Figure 7. Stern Volmer Plots for Substituted Valerophenones in Benzene (0.5 M Pyridine) with Triethylamine Quencher (O H;Um—CN;Ap-CN). 9. 3.0 2.0 1.0 ' Figure 8. 42 Q Q 7 l '1 r ‘ f5- 0.04 0.08 0.12 101. M Stern Volmer Plots for Substituted Valerophenones in Acetonitrile with Triethylamine Quencher (O H;E] m-CN; Ap-CN). ~e. Figure 9. 3.0 2.0 1.() 43 ’D >\ '0 ‘.> 0 \J> I l l l 0.04 . 0.08 101, M Stern Volmer Plots for Substituted Valerophenones in Benzene (0.5 M Pyridine) with Triethylamine Quencher (o o-0020H3; |'_'_] m-0020H3;Ap-0020H ) 3 O 0.12 44 6.0 - A 5.0 - 4.0 -
——-1T
“~‘ ,,,,
‘—
7 . n-br’
. ------ n
* _ --------- ~ ~“
fl ‘ ~
8 ” s~~ ‘~‘§
. s ‘\
~ ~
— ‘\~~ ”——’-‘-—_"'
7 §--
1 —_
Scheme 4. Mixing of h-Orbitals in Benzoyl System.
79
transition negligible (Scheme 4).
Another (-R) substituent reinfbrces this orbital splitting,
since inductive effects on transition dipoles greatly affect the
3La energy level. The vector addition of individual substituent
dipoles for 1,4-acceptor, acceptor substituents results in greater
stabilization than does a similar 1,3 pair. The M0 orbital diagram
indicates that a para substituent should lower the n,n* triplet by
the same amount as the first substituent; whereas, a meta substituent
should lower the triplet energy by only 1/4 the amount, if the spin
density at-the site affects the amount of stabilization affbrded by
the substituent. The n,n* triplet energies measured for substituted
benzonitriles and methyl benzoates seem to corroborate this simple
M0 representation of the h,n* triplet.
The valence bond representations below describe the La n,n*
triplet of phenyl ketones.
x 'i
The most important resonance f0rm is the 1.4-diradical “quinoidal”
ferm of the triplet.98 EPR studies99 showed that very little charge
density lies at the meta position, as is depicted in the middle
valence bond structure. The third structure represents a high-energy
charge-transfer state which mixes somewhat with the low-lying 3La
80
1 103
state but to a greater extent with the high-energy La state.
For phenyl alkyl ketones containing substituents capable of conjuga-
tion with the ring, the following valence bond structure depicts a
folly conjugated h-system.
Thus, significant contribution of this structure, or the analogous
ortho-structure, would account for the significantly larger stabiliza-
tion of the n,n* triplet afforded by 0- and p-CN, p-COZCH3, and p-COR
substituents as compared to the model CF3 substituents.
As long as the n-orbital remains orthogonal to the n-system, an
n,n* transition can be regarded simply as a one-electron reduction
‘04 EPR studies105
of the n-system. on benzonitrile radical anions
indicated that the additional electron density in the benzene ring
occupies primarily the l- and 4-positions with negligible spin
density at the meta position. Thus, substituents have diminished
effects in the meta position. Valence bond representations of the
n,n* triplet are shown below.
81
Electron-withdrawing substituents inductively stabilize the n,n*
triplet. The substituent effects determined for benzophenone n,n*
triplets confirm the larger stabilization provided by para- versus
meta-substitution by electron-withdrawing substituents.
E. Type II anptgm Yields
The Type II quantum yields in benzene, wet acetonitrile, and in
benzene with added Lewis base for these substituted valer0phenones
are given in Table l. The quantum yields for most ring-substituted
valerophenones can be maximized to unity with added Lewis base as
long as there are no competing triplet processes. The Lewis base
hydrogen bonds to the hydroxy proton and prevents disproportionation
from occurring.10b
The quantum yields for these conjugatively electron-withdrawing
substituted valerophenones rises to maximum values considerably less
than unity with added Lewis base (i.e., pyridine, t-butyl alcohol,
or dioxane) as shown in Figures 13 and 14. The quantum yields in
wet acetonitrile are generally higher than with added Lewis base.
The inability to sufficiently maximize the quantum yield is dif-
ficult to explain, since electron-withdrawing substituents increase
the acidity of the hydroxy proton. Thus, these biradicals must be
completely solvated by the added base, but for some reason do not
always go on to form product. One plausable explanation involves
resonance stabilization of the solvated biradical which would perhaps
make disproportionation competitive with product formation.
82
Polar solvents or Lewis bases would stabilize such charge separation.
Since the possibility remains that some other triplet decay
process, kd, could be competitive with y-hydrogen abstraction, kr’
28 suitable experiments
as in the case of p-methoxyvalerophenone,
were conducted in order to determine kd values. Only if y-hydrogen
abstraction is found to be the only triplet process can one equate
1/1 with kr'
F. Determination of k, Values by Various Methods
The rate of y-hydrogen abstraction is dependent on changes in
the y-C-H bond strength.‘5:23"°“ Table 2 lists quantum yields in
benzene, maximum quantum yields with added pyridine, and qu values
for both unsubstituted and p-cyanosubstituted butyrophenones (1°y-H),
valerophenones (2°y-H), and y-methylvalerophenones (3°y-H). In both
series the y-methylvalerophenone is approximately f0ur times more
reactive than the corresponding valerophenone. In contrast, butyro-
phenone is much less reactive (factor of 6) than valerophenone,
while p-cyanobutyrophenone is only 3 times less reactive than the
corresponding valerophenone. The reason for the greater reactivity
of p-cyanobutyrophenone than predicted by the unsubstituted models
is not apparent, but appears to be real.
The Type II quantum yield for acetophenone formation is determined
83
by competitive excited state reactions:
“11 = “150 ' kr ' Pp ' T (1°)
where °ISC is the quantum yield for intersystem crossing, kr is the
rate for y-hydrogen abstraction, PD is the probability that the bi-
radical will go on to form product, and T = (kr + kd)']. The inter-
system crossing yields for both p-cyano and p-carbomethoxyvalero-
phenones were measured and found to equal unity. Within each series
of compounds both P and kd remain constant, since increased y-sub-
stitution should no: affect these parameters. Since the quantum
yields remain nearly constant within each series, whereas, the trip-
let lifetimes decrease appreciably with increased y-carbon substi-
tution, then an increase in the rate of y-H abstraction with increased
v-carbon substitution must be responsible. Therefore, kd is not
competitive with kr’ and kd must be less than 106 s'].
In order to insure that kd is indeed smaller than kr’ bimolecular
photoreduction experiments were conducted.. Photoreduction experi-
ments on p-cyanoacetophenone using either toluene or p-xylene as
donor and naphthalene as quencher in acetonitrile solvent indicate
that kd = 4 x 105 5']. Thus, kd << kr and I/T can be equated to kr'
Consequently, the exceptional inability to maximize the total quantum
yield to unity for these (-R) substituted valerophenones must be due
to the reduced ability of the intermediate biradical to go on to form
product.
84
G. Ketone Photoreactivity and Calculation of Relative Triplet
Energies
It has been shown that the observed rate constant, kobs’ for
ketones with n,n* lowest triplets is described as follows:
k - kg” (3)
obs = Xn,n
where x",1r is the thermal equilibrium fractional population of the
n,n* state and k3’" is the intrinsic rate of hydrogen abstraction
for the appropriate n,n* triplet. The population of the n,n* trip-
let at a certain temperature can be calculated from the Boltzmann
equation as follows:
AE/RT)-1
x =(l+e (l4)
n,h
I where AE is the energy separation between the two triplets and T is
the temperature in degrees Kelvin. Thus, if one can estimate kg’",
then calculation of the energy separation between the two triplets
becomes possible.
Trifluoromethyl groups increase the energy gap between the two
triplets by stabilizing the n,n* triplet while having little effect
on the 1r,1r* triplet such that population of the upper 11,1r* triplet
is essentially zero. Since CN, 002R, and COR all have Hammett a
values similar to that of CF3,107
.their intrinsic kg’" values must
nearly equal kr for the analogous o-, m-, or p-trifluoromethylvalero-
phenone. Since the rate of hydrogen abstraction is enhanced only
85
by a factor of 2-3 by the inductive effect of m- and p-CF3 on the
electrophilic n,n* triplet, the error involved in estimating the
other ka’" values must be negligible.
From the Stern Volmer quenching studies and the other experi-
ments verifying that y-hydrogen abstraction is the only process
responsible for triplet decay, I/T = kgbs. The data from Table 1
indicate that p-C0C4H9, p-CN, and p-COZCH3 substituents decrease
triplet reactivity to 10%, 25%, and 40%, respectively, as observed
“for p-CF3 valer0phenone. From these xn,h values, the AET separations
calculated are 1.3, 0.67, and 0.3 kcal/mole, respectively, in benzene
at room temperature. Since I/T increases by an average of 50% at
most in acetonitrile, then these values cannot be very different in
acetonitrile.
It is interesting that p-divalerylbenzene is twice as reactive
as p-acetylvalerophenone. Although each compound has two chromo-
phores capable of absorbing light, only excitation of the valeryl
carbonyl can lead to y-hydrogen abstraction. Since both carbonyls
have an equal probability of becoming excited, then the decreased
reactivity of p-acetylvalerophenone reaffirms the idea that excitation
in the n,n* triplet is localized primarily on the carbonyl.
Both m-CN and m-CDZCH3 valerophenones show reactivities comparable
to that of the model m-trifluoromethylvalerophenone. All three
ketones have n,n* lowest triplets. Conversely, m-divalerylbenzene
is only half as reactive as the model ketone, thus implying that the
n,n* and n,n* triplets are isoenergetic and equally populated.
The spectroscopic data seem to be in good agreement with this
86
assignment.
The observed photoreactivity fer the ortho-substituted valero-
phenones is more difficult to interpret, since steric factors are
also involved. o-Cyanovalerophenone is twice as reactive and o-
carbomethoxyvalerophenone is only 1/4 as reactive as the o-trifluoro-
methylvalerophenone model. Unfortunately, o-CF3 valerophenone is not
an ideal model since it displays decreased reactivity as compared to
its meta and para isomers, presumably due to steric factors. Thus,
calculation of the energy separation for the two ketones from the
kinetics is impossible. From the spectroscopy o-cyanovalerophenone
has nearly isoenergetic triplet levels; whereas, o-carbomethoxy-
valerophenone has a definite n,h* lowest triplet. At least a quali-
tative correlation exists between the observed reactivities and the
energy separations predicted from the spectroscopic data.
Thus, the photoreactivity of these ketones prevides a more sensi-
tive method for determining the energy separation between the two
triplets. In general, both photochemical and spectroscopic data
agree that p- (-R) substituents stabilize the n,n* triplet to such
an extent that these ketones possess n,n* lowest triplets.
H. Attempted Correlation of Reactivities with Hammett 6 Values
Wagner et 31.29 have correlated the relative rates of y-hydrogen
abstraction of ring-substituted phenyl ketones with Hammett o param-
eters (Figure 27). Both meta and para inductively electron-with-
drawing (-I) substituents (n,n* lowest triplets)(dashed line) and
weakly electron-donating substituents (m-, or p-CH3, -0CH3) (n,n*
87
00 9 /
O/
/
/ 90e
0
9
El
0
9
'ogkobs-1.0
-42JO
4 A L
Figure 27. Plot of Relative k b5 Values as a Function of Ground
State Hamett o VaTues.o -,I Substituents;0, Electron-
donating Substituents;e, para-R Substituents;., p- -Cl;
U , p-SCF3o
88
lowest triplets) (solid line) gave linear correlations with dif-
ferent slopes. p-Cl did not fit on either line, nor does our p-CN,
108 29 I
p-COZCH p-COR, and m-COR, or p-SCF3. As already pointed out,
3.
a Hammett plot is inappropriate f0r ketones with n,n* lowest triplets
where kobs values are dominated by Boltzmann factors rather than by
substituent effects on actual kc’" values. Also, since Hammett
constants do not adequately describe substituent effects on n,n*
triplet energies, then “ET values cannot Correlate with Hammett
substituent constants. Ground state a values decrease in the order
CN > COR > CDZCH3, whereas, conjugative effects on n,n* triplets dis-
play a different order: COR > CN > COZCH3. The coplanarity of the
excited benzoyl groups compared to the twisting of the ground states94
would explain the greater conjugating ability of acyl groups in the
excited state.
I. Chppge Transfer Quenchipg Studies
Substituted phenyl alkyl ketones can undergo charge transfer
interactions with tertiary amines.
61 65
The equation developed by Weller and later modified by Cohen
for carbonyl compounds describes the relationship between the rate
89
constant of charge transfer quenching and the various parameters
affecting the change in free energy involved.
log kcT a. AGCT n. IPD - E(A'/A) - ABEO’O + c (8)
Since in all the experiments triethylamine was used as donor, then
IPD remains constant, as does C. Thus, if one can determine both
the reduction potential and the energy of the triplet involved for
the acceptor, then one can calculate relative AGCT values.
From comparison of diene and triethylamine quenching rates under
similar conditions, one can determine the rates of charge transfer
quenching from the following equation:
k T(Et N)
k=“3-k (15)
CT quIdiené) q
9 M"1 s'1 in benzene
where kq is the rate of diene quenching (5 x 10
and 1 x 1010 M’1 s'1 in acetonitrile). Table 9 contains the values
of kCT in both benzene and acetonitrile solvents. The parameters neces-
sary to plot log kCT versus excited state reduction potential (Figure
28) are summarized in Table 10. Assuming that the rates of charge
transfer quenching of the n,n* and n,n* triplets are not significantly
different,83 then the energy of the lowest triplet was used in cal-
culating the excited state reduction potentials for the graph. Even
if this assumption proves to be erroneous, then only small errors
are involved since most of the ketones studied possess n,n* and
n,n* triplets in close proximity.
90
Table 9. Charge Transfer Quenching Data for Substituted Valerophenones.
Substituent Solvent k", 107s" kCT, 109s"
H Benzene 11. 1.2
H Acetonitrile 16. 2.1
o-CN Benzene 23. ---
o-CN Acetonitrile 30. --
m-CN Benzene 26. 5.8
m—CN Acetonitrile 50. 5.0
p-CN Benzene 6.8 6.1
p-CN Acetonitrile 11. ll.
o-COZCH3 Benzene 3.6 2.2
o-COZCH3 Acetonitrile 3.0 2.5
m-COZCH3 Benzene 28. 6.9
m-COZCH3 Acetonitrile 48. ---
p-COZCH3 Benzene 12. 5.5
p-COZCH3 Acetonitrile 17. 16.
m-COC4H9 Benzene 14. 3.8
m-COC4H9 Acetonitrile 22. 7.2
p-COC4H9 Benzene 2.7 ---
p-COC4H9 Acetonitrile 1.4 ---
o-CF3 Benzene 13.a 7.2
m-CF3 Benzene 32.a 8.4
p-CF3 Benzene 28.a 15.
p-CF3 Acetonitrile 42.a 21.
aUnpublished results by Warren 8. Mueller.
91
10.4 "
10.0 -
IOQ kc-r T
9.2 -
L 1 l l 1
1.0 1.2 1.4
' *
'Ered,eV
Figure 28. Correlation of Charge Transfer Quenching Rates for Sub-
stituted Valer0phenones in Benzene (0.5 M Pyridine) with
Excited State Reduction Potentials: O n,ii* lowest Triplet;
Un,n* Lowest Triplet.
92
Table 10. Charge Transfer Quenching Parameters in Benzene.
k
ET ET CT 109
Substituent (n,n*) (n,n*) 'Ereda '*Ered(ev)b 109s'] kCT
H 73.4“ 75 5“ 47.7 1.11 (1.21) 1.2 9.1
o-CN ~7l.6 71.0 37.8 1.47 (1.44) --- ---
MPCN 73.3 73.8 40.9 1.41 (1.43) 5.8 9.8
p-CN 71.2 59.3 35.3 1.51 (1 43) 5.1 9.8
o-0020H3 73.4 71.5 44.5 1.25 (1.17) 2.2 9.3
n-coch3 73.6 74.5 43.8 1.29 (1.33) 5.9 9.8
p-0020H3 71.5 70.3 37.5 1.47 (1.42) 5.5 9.7
m-COC4H9 73 4573 42.8 1.31 (1.31) 3.8 9.6
p-C0C4H9 71 57.7 35.7 1.53 (1.39) --- ---
o-CF3 m 73 73.0 43.8 1.27 (1.27) 7.2“ 9.9
m-CF3 72.8 74.7 42.9 1.30 (1.38) 8.4“ 9.9
p-CF3 72.4 74.3 40.6 1.38 (1.46) 15.“ 10.2
aVal ues in kcal/mole.
bValues for excited state reduction potentials f0r n,n* triplets;
values in parentheses for n,n* triplets.
cValues taken from Reference 29.
dUnpublished results by Warren 8. Mueller.
93
Weller's equation seems to hold for these (-R) substituted
valerophenones. In all cases these compounds possess excited state
reduction potentials lower than valerophenone itself and exhibit
increased rates of charge transfer quenching. The rate of charge
transfer quenching generally increased at most by a factor of 2 in
acetonitrile. As previously determined, the separation between
the two triplets cannot be significantly different in acetonitrile
than in benzene. The rate of charge transfer quenching with tri-
ethylamine exceeds the rate of diene quenching for many of these
(-R) substituted valerophenones.
In summary, these electron-withdrawing substituted valerophenones
readily undergo charge-transfer interactions with tertiary amines,
many at rates which approach diffusion control. Since the n,n*
and n,n* triplets are close in energy and the triplet energies are
estimated values, then correlation of the rates of charge—transfer
quenching with the nature of the triplet undergoing the interaction
is impossible. All Stern Volmer quenching experiments involving
triethylamine in benzene solvent contained 0.5 M pyridine to solvate
the biradical. For many ketones it is apparent that small amounts
of triethylamine (pK; solvate the biradical to a greater
extent than does pyridine (pKa = 5.2)]09, since slightly more photo-
product was produced in the lowest concentration quenched tubes
than in the unquenched tubes.
94
J. Correlation of Reduction Potentials with Hammett a Values
Nadjo and Saveant96 have related the reduction potentials of
substituted acet0phenones, fluorenones, and benzophenones to Hammett
oIn and op parameters. A similar correlation was sought fer the wide
variety of substituted valerophenones studied. The reduction po-
tentials and Hammett 0 values for both meta and para substituted
ketones are presented in Table 11. A plot of "Ered versus am or
0D values (Figure 29) gave a linear correlation which suggests that
resonance effects are indeed important in stabilizing the radical
anion formed by the one-electron reduction of phenyl ketones. When
Hammett op values were used instead of Op- values, the reduction po-
tentials of p (-R) ketones fell well below the line. The only
anomalies found were the COZH and C02" substituted valerophenones
which fell symmetrically about the line. Perhaps acid or base
catalyzed reactions with the solvent or reduction of the carboxy
functionality are responsible. Whatever the explanation for these
discrepancies, the other ketones seem to correlate very well with
Hammett parameters.
K. Correlation ofippn* Triplet Energies for Substituted Phenyl
Ketones with Hammett 0 Values
d89 has published several correlations of n,n* triplet
Arnol
energies for both substituted acet0phenones and benzophenones with
Hammettom and op values. The acet0phenones gave a good linear
correlation (no -R substituents were included). The benzophenones
95
Table 11. Correlation of Hammett a Values with Reduction Potentials
for Valerophenones.
Substituent cm or op“ op‘ “ -Eredb
H 0 47.7
mPCN l +.56 40.9
p-CN ~ +.66 +.90 36.3
1500ch3 +.37 43.8
p-COZCH3 +.45 +.68 37.5
m-COC4H9 +.38 42.8
p-000H3 +.50 +.87 35.7
m-CF3 42.9
p-CF3 +.54 40.6
m—CH3 -.07 48.0
p-CH3 -.17 49.7
m-OCH3 +.12 47.3
p-OCH3 -.27 -.2 51.0
m-F +.34 44.5
p-F +.06 ‘ -.02 48.1
m-Cl +.37 43.9
m—COZ' -.10 42.9
p-COZ' 0 40.9
m-COZH +.37 49.7
p-COZH +.45 +.73 47.0
aValues taken from Reference 107.
bValues given in kcal/mole.
96
52 -
48 :-
'Erod
(koal/ 44 '
1110101
40 r
3.6 D (:1
-0.1 0.1 0.3 0.5 0.7 0.9
amid-p-
Figure 29. Correlation of Half-wave Reduction Potentials for Sub-
stituted Valerophenones with Hanmett “m and op" Values:
0 meta-Substi tuti on; C1 para-Substi tuti on .
97
(both mono- and symmetrically substituted compounds) did not correlate
well with the additive Hammett parameters for disubstitution (Figure 30),
but a somewhat better correlation can be obtained when Hamett am and
a; values are used and symmetrically substituted benzophenones are
treated as monosubstituted compounds (Figure 31). Arnold89b later
studied a wide variety of both mono- and symmetrically and unsymmetri-
cally substituted benzophenones. They found that all the symmetric
derivatives correlate fairly well with a Hanmett two-parameter treat-
ment employing both a and 0' values, indicating that both polar and
radical-stabilizing effects influence the n,n* triplet state. For un-
symmetrically substituted benzophenones, the triplet energies were
110
lower than predicted, presumably from merostabilization from substi-
tuents that can stabilize charge separation in the free spin system.
L. Correlation of nin* Triplet Energies with Reductioanotentials
In 1973 Loutfy and Loutfy92
published a correlation between n,n*
triplet energies and half-wave reduction potentials for substituted
acet0phenones. Also, similar correlations were published for substi-
89b and thiophenes.97 Since the n-orbital is orthog-
tuted benzophenones
onal to the n-system, substitution of a group into the phenyl ring
would not affect the energy of the n-orbital but would affect the
energies of both the n and 5* levels.104 Loutfy maintains that p-
cyanoacetophenone has a lowest n,n* triplet, since it lies on the line
98
(kcal/
mole)
-O.4 O 0.4 0.8 1 .2
(Tm 9 Up
Figure 30. Correlation of n,n* Triplet Energies (4:1 MCH/IP, 77°K
of Substituted Benzophenones with Hammett Parameters.89a U ,
I (-R) Substituents... Values Added From This Study.
69"
ET 68
(kcal/
mole)
67-
l 1 1 I L _1_
-0.2 0 0.2 0.4 0.6 0.8
cm 9 Up-
Figure 31. Correlation of n,n* Triplet Energies (4:1 MCH/IP, 77°K)
of Substituted Benzophenones with Hanmett Parameters:
U ,I (-R) Substituents;o, [JValues Taken From Ref. 89a.
99
formed by -H, p-Cl, p-Br, and p-CF3 substituted acet0phenones. Since
the triplet energy for p-cyanoacetOphenone was measured in ethanol
glass, while the other ketones were measured in 4:1 methylcyclohexane/
isopentane glass, we measured the triplet energy for p-cyanoacetophenone
in a comparable glass and found it to be only slightly lower (0.5 kcal)
than the reported value in ethanol. In contrast, our study involving
both spectroscopic and photokinetic data suggests that p-cyanovalero-
phenone (therefore p-cyanoacetophenone also) has a n,n* lowest triplet.
In order to probe the matter further, a similar correlation was attempt-
ed on a wide variety of substituted valerophenones in both a fluid
glass (isopentane) and a more rigid and polar glass (ethanol) at 77°K
(Figures 32 and 33). It appears that in both isopentane and ethanol
glasses, most of the ketones with n,n* lowest triplets fall on a line
and those ketones with n,n* lowest triplets fall below the line. The
most noticeable exceptions are o-CN and m-COC4H9 in isopentane glass,
since they are predicted to have nearly isoenergetic energy levels by
the other methods. In general the energy separations between the n,n*
and n,n* triplets predicted by this correlation are much larger (iso-
pentane glass) than those previously estimated in benzene at room tem-
perature. Fortunately, the Boltzmann equation (Equation 1) would pre-
dict a somewhat larger energy separation at lower temperatures, and
also, n,n* triplets in solution are generally lower by approximately 2
kcal than in rigid glasses due to differences in conformation between
the ground state and the excited state.93 It appears that Loutfy's
correlation is at least qualitatively correct in most cases, and gives
a rough estimation at best of n,n* triplet energies.
100
761-
I
ET’ II
(kcaU' ‘58 '
flung) II
- In I
“4 I
l l 1 1 L l l
1.6 1.8 2.0 2.2
Figure 32. Attempted Correlation of n,n* Triplet Energies (Iso-
pentane, 77°K) with Halfewave Reduction Potentials for
Substituted Valerophenones: C) n,n* Lowest Triplets;
[j'n,n* Lowest Triplets.
101
68 F '
II
64 .'.
1
.1
I i I I I L l
1.6 1.8 2.0 2.2
‘Erod,ov
Figure 33. Attempted Correlation of n,n* Triplet Energies (Ethanol,
77°K) with Half-wave Reduction Potentials for Substituted
Valerophenones: O n,ir* Lowest Triplets; I11,1r* Lowest
Triplets.
PART II
REGIOELECTRONIC CONTROL OF INTRAMOLECULAR CHARGE TRANSFER
QUENCHING IN VARIOUS PHENYL KETONES
102
RESULTS
A. Photokinetic Studies
1. Stern-Volmer Quenching Studies
In general 0.01 M ketone solutions containing varying amounts
of quencher in benzene or wet acetonitrile (2% H20) were irradiated
at 313 nm to 5 - 10% conversion (acetophenone photoproduct was
measured). A linear correlation was obtained for Stern Volmer
quenching plots of ¢°/¢ versus quencher concentration (Figures 34,
35 and 35). The slope is equal to k T, where T is the triplet life-
q
time of the ketone. The triplet lifetimes can be calculated, since
91
s' in benzene and l x 10 M- s' in aceto-
nitrile. Duplicate runs usually agreed to within 10% of each
other. Table 12 contains the photokinetic data for the model com-
pounds and Tables 13 and 14 contain the data for the amino-ketones
studied. Also, Figure 37 depicts the extrapolation of triplet life-
times to infinite dilution in order to eliminate intermolecular CT
processes (both lines must have similar slopes since intermolecular
CT quenching rates are approximately the same).
The phosphorescence lifetimes for the model m- and p-carbo-
methoxybenzophenones and p-2BB (1 x 10'4 M) were measured by diene
quenching of the phosphorescence emission in carbon tetrachloride
solution at room temperature (Figures 38 and 39). Also, emission
103
104
I
3.0 L-
I
.4
2.0 L C]
¢O
<5
1.0
0.02 V 0.06 0.10 0.14
[OLM
Figure 34. Stern Volmer Plots for Diene Quenching of 0.020 M
Ketones: 0 p2VB in Benzene;. p2VB in Acetonitrile;
‘p3VB in Acetonitrile; DmZVB in Benzene; and
I m2VB in Acetonitrile.
105
4.0 r
3.0 -
. (pa
Cb
2.0 L
1.0
4.0 8,0 ‘ 12.0
[01,10‘4111
Figure 35. Stern Volmer Plot for Diene Quenching of 0.020 M p3VE
in Benzene.
106
I L 1 I
1.0 2.0
[01."
Figure 36. Stern Volmer Plot for Diene Quenching of 0.020 M p-
Me03E in Acetonitrile.
107
Table 12. Photokinetic Data for Valerophenone Models.a
"Q—(Lr
-X Solvent Q IIIg Amax qu
p-COZCH3 .Benzene Dieneb 0.19 0.33 42:2
p-coch3 Benzene“ Amine“ ---- ---- 32
p-0020H3 Benzene“ Amine: ---- ---- 25
p-COZCH3 Acetonitrile Diened 0.60 ---- 5825
p-COZCH3 Acetonitrile Amine ---- ---- 4020.5
p-0020H3 Acetonitrile Amine“ ---- ---- 47
m-coch3 Benzene Dieneb 0.25 ---+ 18:1
m-COZCH3 Acetonitrile Dieneb 0.90 ---- 21:1
m-COZCH3 Acetonitrile Amine“ ---- ---- 8.0
p-OCH3 Benzene Dieneb 0.11 ---- 4,300:500
p-OCH3 Benzenec Aminef -—-- ---- ~501
p-00H3 Acetonitrile Dieneb 0.21 0.25h 6,400:400
p-OCH3 Acetonitrile Aminef ---- ---- 202
a0.040 M ketone solutions irradiated at 313 nm to 5_10% conversion.
b2,5-Dimethylhexa-2,4-diene or 1,3-pentadiene.
c0.50 M Pyridine added.
dB-Dimethylaminoethyl benzoate.
ey-Dimethylaminopropyl benzoate.
fN,N-Dimethylaminopropyl phenyl ether.
gUsually 0.10 M solutions of valerophenone in benzene used as acti-
nometer; formation of acetophenone monitored.
h
Value from Reference 28.
1Value estimated from solvent effects on quenching p-methoxyvalero-
phenone with triethylamine obtained by Warren 8. Mueller.
108
Table 13. Photokinetic Data for Amino-Esters.a
rib-U
Name Position R Solvent bub quc
p-2VB para -(CH2)2N(CH3)2 Benzene“ 0.07 10.
p-ZVB para -(CH2)2N(CH3)2 Acetonitri1ee 0.12 14.9
p-3VB para -(CH2)3N(CH3)2 Acetonitriiee 0.11 11.8
m-ZVB meta -(CH2)2N(CH3)2 Benzene 0.15 16.
m-2VB meta -(CH2)2N(CH3)2 Acetonitrilee 0.31 24.
a0.01 - 0.10 M Ketone in degassed solutions at room temperature.
bQuantum yield for acetophenone formation, extrapolated to infinite
dilution.
cDetermined by SV quenching with 1,3-pentadiene, extrapolated to
infinite dilution.
d0.40 M Pyridine added.
“2% H20 added.
109
.mcmvumacwa-m.— sue: mucmemmnammvc ocoumx acvgucosc >5 emceeLmawv mapm>
.mn «unmemwmm soc; mm=~m>
y
.PPP mucmemems son» mmape>m
u
.mcmvuoucmaim.— saw: mcwgozoaa >m x: necessmuoau
.covumsgom mcocasaoumuo cow upm_a ssacazon
.»m an Aomz “NV mppeuvcoumum am: so mcm~=mn vmmmmomu cw acoumx z no.0 . ~o.co
.em.F eomo.o o..eeeeaeoo< Nfimzuvz- I- <-m
aoe.o emmo.o oeoneom ~2m=evz- z- <-m
e.m omo.o a__eoeeaeoo< «Amzevz- mzeo- mmoo:-a
om.a omuo.o oeomeom Nfimzovz- mzuo- ”moo=-a
aommom. .oe o_,eo_aaooo< mxu- Nflmzuvzmfimzuvo- m>m-a
oo_nooom eo.o oeoaeom mzu- Nfimzuvzmfimzuvo- m>m-a
oeax ea peo>_om . .a a 52a:
a
a.meoeam-ae_e< toe apao oeeoe_eoeoea .4. opaae
110
15.0
13.0
.1.1141
T.
1033"1
9.0
7.0
I I I I I l
0.04 0.08 0.1 2
[Ketone] .14
Figure 37. Extrapolation of Triplet Lifetimes of p2VBO and p3VB.
in Acetonitrile.
111
6.0 +-
o
5.0 -
4.0 +-
IO
T o
3.0 F
o
2.0: F
o
C
1.0
2.0 4.0 5.0 8.0 100
[01,10’5111
Figure 38. Stern Volmer Plot for Diene Quenching of Room Tempera-
ture PhoSphorescence of m-Carbomethoxybenzophenoneo
and p-Carbanethoxybenzophenone. in CC14.
112
I k I I L
0.40 0.80 1.2 1.6 2.0
[01,10'4M
Figure 39. Stern Volmer Plot for Diene Quenching of Room Tempera-
ture Phosphorescence of p2BB in CC14.
113
from 1:1 solutions of carbomethoxybenzophenone and the model N,N-
dimethylaminoethylbenzoate quencher were measured and compared to
the emission from an equimolar amount of the amino-ketones in order
to obtain an estimate of self-quenching by the amino-ketones. The
results are summarized in Table 15. As a comparison, the rate of
quenching of butyr0phenone by both 2,5-dimethylhexa-2,4-diene and
triethylamine in carbon tetrachloride was measured (Figure 40).
2. Quantum Yields
Quantum yields for acetophenone formation were determined by
parallel irradiation at 313 nm of 0.01 - 0.10 M ketone solutions
and 0.10 M valerophenone actinometer in a merry-go-round apparatus.13
All samples were adequately degassed prior to irradiation and con-
version was usually kept below 10%. In most cases both the ketone
solution and the actinometer absorbed all the incident irradiation;
in those cases (low ketone concentration) where some of the light
was transmitted, then suitable corrections were made. The values
are included in Tables 12-14 and the extrapolation of quantum yields
to infinite dilution for p-2VB and p-3VB and for m-2VB in aceto-
nitrile are shown in Figures 41 and 42.
3. Dippppearance Qpantum Yields
Disappearance quantum yields determined for p-2VB and p-2AB
were extrapolated to infinite dilution (Figure 43) in order to
determine whether the Norrish Type II photoelimination is the major
114
.gmumm pmuos op coconsou mo upo_z saacmsa cw mmmmeowu soc» cmpms_um0 o=Pe>
u
.xnoumogauoam zmmpe x; vmezmmme mE—umewp 03m: oo_ a museum; mm mucogmwmmu
.4 a
eaoaeeoma Fee a, .-a .-z a
o_ x e.a xv nee—5-4
.AxupmoUmv> mu? sogw
wiuxw:~>:umspuim.~ saw: mcwcucmac >5 umcpsgmumu mmspm>
a
.ogaaoemnsmu soon as mcopuzpom «poo ummmmmmc a? macaw; 2 etc, x o.p so» cmcwsemumo mmspm>o
ee._a --- mo.o cue Nfimzovzmxumzu- apes mm~-e
~_.o a.~ .P ewe mzu- aeoe
m.P mm.o mo.o oee NAmzuvzmzomze- aeaa mmN-a
mo.o o~.m .2 see «:0. aeaa
m o_ .e\~ op .ecx no o.o a cowupmoe use:
_- m m Foe 25:3 2
.mococmcao~:om cmusapumaam go; mama mucmumogozamoga .mp mpnmh
m
115
6.0 -
5.0 - O
4.0 - 0
(to
3.0 (-
O
0
2.0 L
O
O
1.0 0'
2.0 . 4.0 . 6.0
[01,10'3M
Figure 40. Stern Volmer Plot for 0.040 M Butyrophenone in CC14
with Hexadiene-Quen’chero or Triethyl amine Quencher..
116
O O
14.0 ~
. a
O
12.0 -
. Q C
.1. -
0n .
D
10.0 "
3
8.0 -
0.04 0.08 0.12
[Ketone],M
Figure 41. Extrapolation of on for p2VBO and p3VB. in Aceto-
nitrile to Infinite Dilution.
117
13:1 -
l I 1 1 l l
0.02 0.04 0.06
[Ketone] .M
Figure 42. Extrapolation of 011 for m2VB in Acetonitrile to In-
finite Dilution.
118
0.5 - O
b o
0-x 0.3 t- .
0.1 -
1 L 1 J L
0 02 0.04 0.06 0.08 0.1 0
[Ketone] ,M
Figure 43. Extrapolation of A-K for p2VBO and p2AB. in Aceto-
nitrile to Infinite Dilution.
119
source of ketone disappearance or if charge transfer processes (both
intramolecular and intermolecular) efficiently lead to product forma-
tion. The errors involved are fairly large, but accurate enough for
this purpose. The ketones were usually photolyzed to greater than
20% conversion.
DISCUSSION
A. Bifunctional Molecules
The study of bifunctional molecules enables one to limit the
number of conformations in which the donor and acceptor orbitals
overlap. In order to investigate the effect of triplet state
electronic configuration on CT quenching, the following system
was chosen for study, since both m-ZVB and p-ZBB have n,n* lowest
23
I
- n-C4H9 (Z-VB)
:0
ll
Ph (2-BB)
triplets, while p-2VB has a n,n* lowest triplet. Identical confor-
mational restraints are present in both p-2VB and p-ZBB.
Another interesting system includes the following amino-ethers
OMe, R' = NMe2(p-MeO3E)
- 0(CH2)3NMe2, R'=CH p-3VB)
‘0
I
:0 :0
1 11
3(
in which the only significant difference involves the different ap-
proach of the donor to the acceptor orbital. Both compounds possess
identical energetic parameters, including a h,n* lowest triplet.
The following kinetic expression holds for these bifunctional
120
121
1/1 = kr + kd + k k ~ [Ketone] (16)
01 intra + 01 inter
molecules since the triplet lifetime, 1, is dependent on the rate
constant for hydrogen abstraction, kr’ the rate constant for radia-
tionless decay, kd, the rate constant for intramolecular charge
transfer quenching, kCT intra’ and the bimolecular rate constant for
charge transfer quenching, kCT inter' The sum of kd and kr can be
obtained from the triplet lifetimes of the model carbomethoxy ke-
tones and estimates for aminoiketone self-quenching rates can be
obtained by quenching the carbomethoxy models with suitable tertiary
amines. Extrapolation of triplet lifetimes to infinite dilution
eliminates any intermolecular CT quenching by the amino-ketone and
simplifies the kinetic expression.
l/r“ = kr + kd + kCT intra (17)
Thus, the rates of intramolecular CT quenching in bifunctional
molecules can readily be determined.
8. Amino-Esters
Winnik et_gl,84’85 have studied the intramolecular CT quenching
of benzophenone emission by alkenes in the following system
° CHzlnCH=CH2
122
and observed no intramolecular charge transfer quenching when
n < 9. For short methylene chains the double bond can come into
contact with the phenyl ring, but cannot approach the carbonyl
moiety. Thus, the lack of CT quenching in these compounds suggests
that the electron-deficiency in the n,n* excited benzophenone must
be located on the carbonyl n-orbital.
64a the
Since tertiary amines are better donors than alkenes,
analogously substituted benzophenones would provide better evidence
for the localization of excitation energy on the carbonyl in n,n*
triplets. Table 15 contains the phosphorescence data in carbon
© © 0(cH212NM02
tetrachloride for p-ZBB and m-ZBB, as well as the data for the model
p- and m-carbomethoxybenzophenones. The triplet lifetimes were
measured by diene quenching of phosphorescence emission. p-ZBB
has a shorter triplet lifetime than the model p-carbomethoxybenzo-
phenone (ID = 8 usec and 130 usec, respectively) and the decrease
in lifetime is paralleled by a decrease in emission intensity.
The triplet lifetime of m-carbomethoxybenzophenone measured was
0.6 usec and the triplet lifetime of m-ZBB, as determined by com-
parison of emission intensities of equimolar solutions of m-2BB
and model ester, is 0.05 usec. The rates of intermolecular CT
123
quenching in carbon tetrachloride were determined by comparing the
emission of the model compounds containing equimolar amounts of
y-dimethylaminoethyl benzoate with the emission from equimolar
amounts of p-ZBB or m-ZBB. The values for bimolecular CT quench-
1 1
ing for the para- and meta-amino-ketones are 3 x 107 M' s- and 2 x
108 1 1, respectively. These values are somewhat lower than the
M' 5'
value one would predict on the basis that triethylamine quenches
triplet butyrophenone at a slightly faster rate in carbon tetra-
chloride (l.6 x 109 M'ls'1) than in benzene (1.2 x 109 "-15-1).
In any case, the rate of intramolecular CT quenching is 04 x 105
s‘1 which is 102-103
times slower than the rate for intermolecular
CT quenching in carbon tetrachloride. Thus, the low rates of
intramolecular CT quenching by the amine moiety in these short chain
compounds support Winnik's results with the alkenyl moiety,84’85
suggesting that the excitation of the benzophenone n,n* triplet is
localized on the carbonyl moiety.
Unfortunately, since the aminobenzophenones react with trace
impurities in the carbon tetrachloride or with the solvent itself
to form quaternary ammonium salts (isolated and identified by NMR)
upon standing, it was necessary to establish the amount of free
amine in solution when the experiments were conducted. A FT-250
MHz NMR of 1.0 x 10'4 M p-ZBB in carbon tetrachloride pulsed for 4
hours showed only a 15% decrease in free amine as compared to a
tetrachloroethane standard. This result would imply no more than
a 30% decrease in free amine over a 4 to 4-1/2 hour period,
since a time-average value was obtained. Thus, within the time
124
frame of these experiments, only a small percentage of the free
amine became protonated. Also, the quaternary alimonium salt was
not sufficiently soluble in the carbon tetrachloride solution to be
detected in the NMR spectrum and a blank of the HCl salt of p-ZBB
in carbon tetrachloride showed no appreciable room temperature phos-
phorescence. The lifetime of the quarternary salt would be expected
to be similar to the carbomethoxybenzophenone models, since the lone
pair on the nitrogen atom is no longer available for CT quenching
interactions. Thus, it was concluded that the observed phosphor-
escence emission was indeed from the free amine. Other solvents
such as acetonitrile, benzene, heptane, methylene chloride, t-butyl
alcohol, distilled water, and chlorobenzene also had been tried, but
solvent emission prevented detection of ketone phosphorescence.
Further studies to determine the triplet lifetimes of the amino-
benzophenones in acetonitrile will be undertaken in this group.
Table 17 contains the Type II quantum yields, triplet lifetimes
extrapolated to infinite dilution, and the rate constants for y-
hydrogen abstraction and intramolecular CT quenching for the amino-
esters p-2VB, m-2VB, and p-3VB. Both p-ZVB and p-3VB show rapid
= 5.1 x 108 1 and 6.6 x 108
intramolecular CT quenching (kCT intra s'
1
s' , respectively, in acetonitrile), while m-2VB shows negligible
intramolecular CT quenching. Also, p-2VB undergoes rapid intra-
8 5") in benzene,
molecular CT quenching (kCT intra = 3.8 x 10
while the meta isomer does not. The low quantum yields for the para
amino-ketones reflect the intramolecular CT quenching, but the
reason for the low quantum yield for m-2VB is not fully understood.
125
Table 16. Charge Transfer Quenching Data for Valerophenone Models.a
WI
b CT inter’
7 -1 8 -1 -1
R Solvent kH,lO s 10 M as
p-coch3 Benzene 12. 38.“
p-COZCH3 Benzene 12. 31.e
p-c020H3 Acetonitrile 17. 69.“
p-0020H3 Acetonitrile 17. 81.e
m-COZCH3 Benzene 28. ---
m-c020H3 Acetonitrile 48. 38.“
p-OCH3 Benzene 0.12 0.6f
p-00H3 Acetonitrile 0.16 3.29
a0.040 M Ketone in degassed benzene or wet acetonitrile (2% H20)
solution.
bDetermined by sv quenching with 1,3-pentadiene.
cDetermined by SV quenching with various tertiary amines.
dB-Dimethylaminoethyl benzoate quencher.
'ev-Dimethylaminopropyl benzoate quencher.
fEstimated from solvent effects on quenching of p-methoxyvalero-
phenone with triethylamine obtained by Warren 8. Mueller.
gN,N-Dimethylaminopropyl phenyl ether quencher.
126
Long alkyl chains are known to lessen the efficiency of product
formation from the intermediate biradicals112
and could partially
explain the lower quantum yields. From the kinetic and spectros-
copic studies presented in Part I, it was concluded that p-carbo-
methoxyvalerophenone has nearly isoenergetic triplet states. Assum-
ing a Boltzmann population, the n,n* and n,n* triplets are nearly
equally populated in p-carbomethoxyvalerophenone, and analogously,
in p-ZVB. Thus, one would estimate the rate for intramolecular CT
quenching of the n,n* triplet in these amino-esters aS‘MI x 109 s'1
in acetonitrile.
Since both carbomethoxyvalerophenones and carbomethoxybenzo-
phenones have similar excited state reduction potentials (Table 19),
these results seem to support the idea that excitation of the n,n*
triplet is localized primarily on the carbonyl, while excitation of
the n,n* triplet lies primarily on the benzene ring (Scheme 5).
1r,1r triplet
Scheme 5. Valence Bond Representations of p-ZBB and p-2VB.
127
-4.2 :42: 442454444 >4 44 eoe2ELoeo4w
42422 2:244:44 .24 44 445224244 24
.co2ps22u mu2c2m=2 on noun—cameuxm .mcm2ueucmn
.co2ua22u ma2c2ec2 on umu42oameuxm .covumsgom ococogaoumue com
1 any ~222u2=ou504 no: go mco~coa 43444444 :2 mcouox.z o2.o-2o.o4
--- 4.4 4.4 44.4 422:22452454 4244422424442- oboe 4>4-e
-- 4.4 2.4 42.4 5:44:54 4244422424442- aooe 4>4-e
--- 4.4 4.4 44.4 422222454554 444- 4455
--- 4.4 4.4 44.4 4:44:54 4:4- 4245
4.4 2.2 4.4 22.4 422:22eaeooe 4244422424442- 4:44 424-4
2.4 2.2 2.4 42.4 422222442454 4244422424442- 4:44 4>4-a
4.4 4.2 4.4 24.4 oeoNeo4 4244422424442- aeaa 4>4-a
--- 2.2 2.2 44.4 422242442554 4:4- 4:44
--- 4.2 4.2 42.4 4:44:44 4:4- 4:44
2-4 442 2-4 442 52-4 442 224 244.254 4 44222454 oeaz
.42442 244 .44 .e\2 a
.mempmm-o:25< see 4449 mc2gocoao memcae» mmgegu .42 42444
a
128
Since the para-valeryl amines undergo rapid internal and external
CT quenching, it was necessary to determine whether CT quenching
leads to product formation. The disappearance quantum yields for
p-ZVB and p-ZAB (the acetophenone analogue) were measured and
extrapolated to infinite dilution (Figure 42). At 0.040 M ketone
Q-K = 0.39 and 0.43 for p-ZVB and p-ZAB, respectively. while at in-
finite dilution. °~K = 0.28 and 0.02, respectively. Since 011
= 0.l3 for p-ZVB at infinite dilution and relatively large errors
are involved in the measurement of disappearance quantum yields,
then one can assume that intramolecular CT quenching leads to little,
if any. nonvolatile photoproduct formation. This result is expected
since the hydrogens a to the nitrogen are unavailable for intra-
molecular abstraction by the radical anion.
, GS.
Thus, intermolecular CT interactions must lead to photoreduction
60
products by a-hydrogen abstraction and subsequent radical-coupling
to give nonvolatile photoproducts.
C. Amino-Ethers
In order to determine the generality of these results, the
following amino-ethers were investigated and the results are
129
O
R= OCHB, R'=N(CH3)2(p-MeO3E)
presented in Table l4. In both acetonitrile and benzene solvents,
p-Me03E undergoes rapid intramolecular CT quenching (kCT intra =
2.8 :< 109 s'1 and 5.4 x 108 s'], respectively) as compared to inter-
molecular CT quenching rates of 3 x l08 and 6 x l07 in acetonitrile
and benzene. Conversely, p-3VE undergoes CT quenching with a rate
constant of 6 x 107 s"1 in acetonitrile and only 5.] x l04 s"1
in benzene, compared to rate constants for intermolecular CT quen-
ching of 3 x 103 s“ and 5 x 107 s", respectively. The major dif-
ferences between the two compounds involves the approach of the
donor electron pair to the acceptor orbital, and the charge separa-
tion involved in the excited state.
One would expect a polar solvent such as acetonitrile to stabilize
these charge separated forms. In the case of p-MeO3E, such charge
separation would not be expected to significantly increase the rate
of intramolecular CT quenching, since the nitrogen lone pair does
not have access to the positively-polarized portion of the phenyl
130
.m4 wocmgmmmm Eng» 4444» mmzp4>m
.444 44:444444 see» 4444» 443—4>
.4444444444- um. 2 :44: muc4g4m444444 mcoumx mcvnucmsa 44 4444544444 4:24>4
.444442 co mucwspgmaxm 4:440:43: 44 4444544444 omx 44:4 covu4eucmucou 444444 5444 4444234444 4:24>
m
4
.4444444444-4.4 :44: mcpgucmzc >4 44 4444544444 44=F4>u
.copu4egom meocmsaoumU4 com 4444» 5444440“
.44 44 copuzpom 404: 444 :o_u=_o4 4444444444U4 no: go 4:44:44 44444444 44 mcoumx z mo. o-No. o4
.4444 .442 4.4444 4444.4 424444444444 4444442- 4- 4-4
.88 .22 W88 %84 2354 416;- 4- <4
.4444 4.4 4.2 .4444 444.4 424444444444 4444444- 4444- 44442-4
.444 4.44 4.2 4.444 4444.4 4444444 4444442- 4444- 44442-4
.44 4.4 4.2 4.44 .4e 424444444444 444- 44444424444444- 444-4
24.4v 4.2;. 4.2 4.2 44.4 4444444 444- 44444424444444- 444-4
---- -- 4.2 4.2 24.4 424444444444 444- 4444-
---- -- 4.2 4.2 22.4 4:44:44 444- 4:44-
442 442 442 42-4 442 4444 4e4>2e4 .4 4 4442
..Fox .hux .14— 4.4.).
GLPCF L3:_.
0 . mLQEHmIO—bwflz LCM. mHflG a: P802030 80% mcmsh Umhmnu .mp mpnmh
13]
ring. Thus, one would expect the normal polar solvent effects on
the rate of intramolecular charge transfer quenching, as indicated
by bimolecular CT quenching of p-methoxyvalerophenone (Table 16).
Both the intermolecular CT quenching of p-methoxyvalerophenone and
the intramolecular CT quenching of p-Me03E show a 5-fold increase
in acetonitrile as compared to benzene. 0n the other hand, the
nitrogen lone pair in p-3VE can come into contact with the electron-
deficient portion of the phenyl ring, and one would predict rapid
intramolecular CT quenching in acetonitrile. A 103-fold increase
in intramolecular CT quenching for p-3VE was observed in acetonitrile
as compared to benzene.
0. Summary of Charge Transfer_9uenching¥Results
Table l9 summarizes the rate constants for CT quenching of
amino-ketones. Since tertiary amines have oxidation potentials of
113 these CT interactions are either
approximately 34 kcal/mole
slightly endothermic or thermoneutral. The rapid intramolecular CT
quenching of both y-dimethylaminobutyrophenone (3-A)73 and p-Me03E111
demonstrates the ready access of the amine lone pair to the electron
"hole" in the n,n* and n,n* triplets, respectively. The slower rate
of intramolecular CT quenching in p-3VE as compared to p-Me03E-
in benzene is hard to explain, except on conformational factors,
while the large increase in the intramolecular CT quenching rate of.
p-3VE in acetonitrile as compared to benzene is probably due to charge
separation in the excited n,n* triplet state. Charge transfer quench-
ing is thermodynamically favorable for the amino-esters and both
132
4 442 x 4 4 442 x 4 444.444 4.44 4e.e
n mop x 4 4 442 x 4
4 442 x 4 4 442 444.444 2.24 44.4
n mop x m a map x m
5.; no— x m 4 mo— x m mam.mmv ~.—~ a: =
Eo—UQOFXN HQFGCF KN
4.4 442 x 2 4.4 442 x 4 444.444 4.44 44.4
4wwm44uz o 4-4 nmogm- ”whuhmw 444442
4 box mews? .hux
.4442444 mcpgucmac 4444:44» omg4zu mo >u4es=m
"mp mpn4h
133
mmNuE
x WOF A X moP x —.V $¢oGN *2}:
x New A x mop x —v $~.Nm k: c
4 me. x M). 4 co. v 444.4mv m.m~ *4 4
H mo— x w H mop x n on.¢mv m.Nm *=.=
-m -z -m uogw- um_4_44 «444m4
up — uF a; «$33
4ma4_ .494 44444 .494
.4m44.ucou .m_ «.444
134
.ppp mocmgmmmm sag» :mxmu mwzpo>e .u:m>_om Focmcpozp .u:m>pom uvvgopzu
-mgumu congoux .Aomz “NV ucm>pom mpvguwcoumu