KETO-ENOL TAUI‘OMERISM IN 3.D|CARBONYLS STUDIED BY NUCLEAR MAGNETIC RESONANCE $FECTROSCOPY Then: for i‘ho Degree of DH. D. MICHIGAN STATE UNIVERSETY Jane Louise Burdett 196.3 JHESbS w O Q- / 774:5 ,IF‘nlfrj,';.-~'~_g org-*7 gar EAST LRNSH‘JG, fviiCHlGAN m: N QTATVZ l'"-“"’PSIT v ABSTRACT KETO-ENOL TAUTOMERISM IN B-DICARBONYLS STUDIED BY NUCLEAR MAGNETIC RESONANCE SPECTROSCOPY by Jane Louise Burdett A study of keto-enol tautomerism in fi-dicarbonyls has been made by proton magnetic resonance techniques. Both fi-diketone and fi-ketoester molecules of the type RCOCH(R')COR" have been investi- gated with substitution of electron withdrawing and donating groups for R, R', and R". Chemical shift measurements have been made for the pure com- pounds, and the effect of substituents on chemical shifts has been noted. Separate signals may be seen for the acetyl methyl protons of the enol and keto tautomers and for the a-protons of each tautomer. Occasionally the methylene protons of the alkoxy group give different chemical shifts for the two tautomeric forms. Identification of resonance peaks has been made possible by varying the tautomeric equilibrium through change of solvent and, in some cases, by the effect of temperature on the equilibrium. The effect of solvent on chemical shifts has been studied and in most cases such measurements have been extended to give the chemical shifts at infinite dilution. In general the effect of the solvent is most pronounced on the keto and enolic a-protons and on the enolic OH proton which is involved in the intramolecular hydrogen bond. Specifically, dilution in various solvents has resulted in upfield shifts of the above- mentioned protons. Benzene represents the principal exception to this Jane Louis e Burdett result, and its behavior has been explained on the basis of complex formation. Equilibrium constants have been determined for the pure com- pounds by integration of the relative areas under the resonance peaks of enol versus keto tautomers. The effect of substituent on the equilibrium constant has been noted and discussed. Extensive solvent studies have been made on the position of equilibrium, and it may be noted that the nonpolar solvents tend to increase the enolic tautomer, whereas polar solvents tend to decrease the enolic content. Long-range spin-spin couplings in both enol and keto tautomers have been determined. These couplings are in general rather small. In the enol tautomer the spin-spin interaction between acetyl and a-protons increases over that in the keto as would be expected when the spin coupling is across a double bond. Variable temperature studies have been undertaken. Low and high temperature effects on the chemical shift are concentrated principally on those protons involved directly in the equilibrium--the keto and enol o-protons and the enolic OH proton. High temperatures tend to cause an upfield shift of these protons and result in negligible shift of other protons. The effect of temperature on the equilibrium constants has been observed. An increase in temperature results in a decrease in the enolic content. Low temperatures were particularly useful for studying compounds with low enolic content, but line broadening precluded extensive measurements. Where feasible, the enthalpy of tautomerization has been obtained from the slope of the graph of ln([enol]/[keto]) versus l/T. Finally, free energies and entropies of enolization have been calculated. KETO-ENOL TAUTOMERISM IN fi-DICARBONYLS STUDIED BY NUCLEAR MAGNETIC RESONANCE SPECTROSCOPY BY Jane Louise Burdett A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCT OR OF PHILOSOPHY Department of Chemistry 1963 To My Par ents ii ACKNOWLEDGMENTS I would like to thank Professor Max T. Rogers for his continual support and encouragement during the investigation of this research. Appreciation is expressed to Varian Associates for the use of their equipment and facilities at the Pittsburgh -Sa1es Applications Office. In this connection I would like to thank Mr. Dennis Benefield, Dr. Norman Bhacca, Dr. Jerry Holcomb, Dr. Donald Hollis, Mr. Cappy Joller, and Mr. Robert Williams of the Varian staff for their assistance. TABLE OF CONTENTS HISTORICAL REVIEW . . . . ..... . ........... Introduction - Nature of the Equilibrium ....... . Substituent Effects. . . . . . . . . ........... Structure........ ..... ..... Dipole Moments . . . . . . . . . . ........ Configuration of the Keto Tautomer ........ Configuration of the Enol Tautomer ........ Infrared Absorption ...... . . . ....... Ultraviolet Absorption ............... Summary of Structure Work on Keto-Enol Tautomers.............. Effect of Temperature and Pressure .......... Solvent Effects ...................... Bromine Titration ........... . . . . . . Dipole Moments ........ . . . . . . . . . . Raman Spectroscopy. . . . . . . . ........ Infrared Spectroscopy ..... . . . . . . . . . . Ultraviolet Spectroscopy ..... . . . . . . . . Nuclear Magnetic Resonance Spectroscopy. . . . Kinetics and Thermodynamics. . . . . . . . . . . . . . Equilibrium Constants and Thermodynamic Quantities ............ . . . . . . . . Rates of Enolization ..... . . . . . ...... Rates of Ionization and Ionization Constants . . . Nuclear Magnetic Resonance Studies ....... . . . Theoretical Studies ................... . Stabilities of Metal Chelates. . . . . . . . . . ..... THEORETICAL CONSIDERATIONS . . . . . . . ....... Nuclear Magnetic Moments. . . . . . . . . . . . . . . ChemicalShifts................ Chemical Shift due. to Bulk Susceptibility ..... Chemical Shift due. to Anisotropy. . . . ..... Chemical Shift due to van der Waals Forces . . . Chemical Shift due to Polar Effects ..... . . iv 22 24 27 27 27 3O 3O 33 36 36 37 40 41 48 55 59 63 63 63 64 65 70 72 TABLE OF CONTENTS - Continued 'Page Hydrogen Bonding ................ . . . . . 75 Relaxation Mechanisms ......... . ....... . 77 Longitudinal Relaxation .............. 77 Transverse Relaxation ........ . . ..... 78 Proton Exchange ..................... 78 Nuclear Spin-Spin Coupling. ............... 79 EXPERIMENTAL ...... . . . ........... . . . . 81 Instrumental ....................... 81 Compound Preparation .................. 82 Compound Purification .................. 83 Preparation of Solutions ................ . 83 Physical Properties ..... . ...... . . ..... 84 Solvent Purification ............ . . . . . . . 84 RESULTS AND DISCUSSION ...... . ........ . . . 87 Nuclear Magnetic Resonance Spectra ..... . . . . . 87 Chemical Shifts ..... . . . . . . . . . . . . . 87 Structure ......... . . . . . . . . ..... 124 Tautomeric Equilibrium ..... . . . ...... 125 SolventEffects...................... 129 Carbon Tetrachloride ........... . . . . 129 Benzene.......... ..... . ....... 136 Hexane...... ......... 143 Chloroform ..................... 147 Carbon Disulfide . ................. 150 Ether 150 Acetic Acid . ................... . 150 Acetonitrile. . . . . ......... . ...... 154 Dimethylsulfoxide .......... . . . . . . . 154 Alcoholic Solvents .............. . . . 155 Triethylamine ........ . . . . . . . . . . . . 155 Diethylamine . .......... . ........ 157 OtherBases........ ....... 158 Solvent Effect on Keto-Enol Equilibrium ..... . . . . 158 Long-Range Proton Spin-Spin Coupling .......... 166 Variable Temperature Nuclear Magnetic Resonance . . 170 TABLE OF CONTENTS - Continued Page Chemical Shift Measurements . . . . ....... 170 Equilibrium Constant Measurements ....... 173 SUMMARY ....... . ..... . ............ . . 178 REFERENCES . . . . . . . ................... 179 vi TABLE II. III. IV. VI. VII. VIII. IX. XI.- XII. XIII. LIST OF TA BLES . Percentage of Enol Tautomer in B—Dicarbonyls. . . . Effect of Substituents on Enolization in [3-Dicarbony1s Influence of Halogen Substitution on Enolization in Ethyl Acetoacetates. . . . . . . . . ’ Calculated Dipole Moments for Acetylacetone and Ethyl Acetoacetate . . . . . . . . ~Percentages of Enol Tautomer, Dielectric Constants, and Apparent Dipole Moments for Ethyl Acetoacetate in Various Solvents. . . . . . . . Infrared Frequencies for B-Diketones ....... . . Infrared Frequencies for fi-Ketoesters . . . . . . . . Ultraviolet Absorption for 8-Dicarbony1s ....... Temperature Dependence of Keto—Enol Equilibrium in some fi-charbonYIS ..... o o o o o o o o o o o o Percentages of Enol Tautomer and Equilibrium Con- stants for Enolization of Acetylacetone and Ethyl Acetoacetate in Various Solvents . . . . . . . . . . . Percentages Enol Tautomer for a-Substituted Ethyl Acetoacetates in Various Solvents ..... . . . . . . Free Energy of Tautomerization for B-Dicarbonyls in Various Solvents . . Enthalpy of Tautomerization for Acetylacetone and Ethyl Acetoacetate in Various Solvents . . . . . . . . 0 o o o o o o o o 0 9 Page 11 16 17 18 26 28 29 31 32 LIST OF TABLES - Continued TABLE XIV. XV. XVI. XVII. XVIII. XIX. XX. XXI. XXII. XXIII. XXIV. XXV. XXVI. XXVII. XXVIII. XXIX. XXX. Percentages of Enol Tautomer for B-Dicarbonyls in Various Solvents. Percentages of Enol Tautomer for Ethyl Aceto- acetate and Acetylacetone in Various Solvents . . . . Equilibrium Constants for B-Dicarbonyls. . . . . . . Free Energies of Tautomerization for 8-Dicarbonyls Rates of Enolization and Thermodynamic Constants for Tautomerization for Fluorinated [3—Dicarbonyls . Ionization Constants for B—Dicarbonyls ..... . . . Rates of Ionization of Pseudo Acids at 250C in Water Enthalpy and Entropy of Ionization of B-Dicarbonyls. Infinite Dilution Chemical Shifts of the Protons of Acetylacetone in Various Solvents. . . . Molecular Orbital Calculations for [3-Di- and 8-Tri- ketones. . Infrared Frequencies of Groups in 8-Diketones and Their Metal Chelates . . . . . . . Infrared Frequencies and Force Constants for Groups fer Copper Chelates of fi-Diketones. . . . . Physical Constants of 8-Diketones ..... . . . . Physical Constants of B-Ketoesters Molecular Formulas of 8—Diketones . . . . . . . . . Molecular'Formulas of 8-Ketoesters. . . . . . . . . Proton Chemical Shifts in Pure 8-Diketones ..... viii Page 34 35 38 39 42 44 45 45 49 57 60 62 85 86 88 99 116 LIST OF TABLES - Continued TABLE XXXI. X XXII . XXXIII. XXXIV. XXXV. XXXVI. XXXVII. XXXVIIIM ' XXXIX. XL. XLI. ’ XLII. XLIII. Proton Chemical Shifts in Pure 8-Ketoesters. . . . . Chemical Shifts of Enol OH for B-Dicarbonyls in Order of Increasing Interaction between OH and Cr-"D Percentages of Enol Tautomers and Equilibrium Constants for fit-Diketones as Determined by Nuclear MagneticResonance. . . . . . . . . . . . . . . . . Percentages of Enol Tautomers and Equilibrium Con- stants for [ES-Ketoesters as Determined by Nuclear MagneticResonance.................. Infinite Dilution Proton Chemical Shifts in [ES-Dike- tones in Carbon Tetrachloride. . . . . . . . . . . . . Infinite Dilution Proton Chemical Shifts in B—Keto- esters in Carbon Tetrachloride . . . . . . . . . . . . Infinite Dilution Proton Chemical Shifts in B-Dike- tonesinBenzene ..... Infinite Dilution Proton Chemical Shifts in B-Keto- estersinBenzene................... Chemical Shifts of Protons in B-Dicarbonyls at Infinite Dilution in’nU—Hexane . . . . . . . . . . . . . Chemical Shifts of Protons in 8-Dicarbonyls at Infinite Dilution in Chloroform . . . . . . . . . . . . Infinite Dilution Proton Chemical Shifts in Acetyl- acetone in Various Solvents. . . . . . . . . . . . . . Infinite Dilution Proton Chemical Shifts in Ethyl Acetoacetate in Various Solvents. . . . . . . . . . . Equilibrium Constants for Tautomerization of Acetyl- acetone in Various Solvents at 0. 1 Mole Fraction of selute ....... O O O O O O O O O ..... O O O O ix Page 117 121 126 127 130 131 137 138 144 148 152 153 159 LIST OF TABLES - Continued TABLE Page XLIV. Equilibrium Constants for Tautomerization of Ethyl Acetoacetate in Various Solvents at 0. l Mole Frac- ti on Of SOlute 000000 O O O 0 O O O O O ....... 160 XLV. Free Energies of Tautomerization of 8-Dicarbonyls at 0. 1 Mole Fraction in Various Solvents and Values of the Solvent Property ....... . . . . . . . . . . 165 XLVI. Long Range Spin-Spin Coupling in fi-Dicarbonyls . . . 169 XLVII. Chemical Shifts of the Enol OH and of Keto and Enol a-Hydrogen Protons in [3-Dicarbonyls at Various Temperatures..................... 172 XLVIII. Thermodynamic Quantities for the Tautomerization Equilibria in fi-Dicarbonyls . . . . . . . . . . . . . 176 FIGURE 10. 11. 12. 13. 14. 15a. LIST OF FIGURES . Configurations of keto and enol tautomer in acetyl- acetone. 0 O O O O O O O O O O I O O O O O O O ..... . Configurations of keto and enol tautomers in ethyl acetoacetate...................... . Configurations of keto tautomer in B-dicarbonyls Configurations of enol tautomer in 8-dicarbonyls . Possible structures for intramolecularly hydrogen- bondedmolecules................... . Conjugated structures for the enol tautomers in 8-dicarbonyls ............. . . . . . . . . . .Enolateionstructure................. . Structures of some fi-triketones . . . . . . . . . . . Keto and enol structures for hydroxymethylene com- Pounds ...... C O I O O O O C O O O O O O ..... Schematic representation of ring-current effect . . . Orientation of polar molecule relative to benzene ring in complex. 0 O O C O O O O O O O O O O O O O O O O Proton NMR spectrum of acetylacetone. . . . . Proton NMR spectrum of o-bromoacetylacetone . Proton NMR spectrum of a-chloroacetylacetone . . . Proton NMR spectrum of cyclic isopropylidene malonate(inCHC13)............... xi 10 12 19 20 47 52 54 68 69 89 9O 91 92 LIST OF FIGURES — Continued FIGURE 15b. 16. 17. 18a. 18b. 19. 20a. 20b. 21. 22a. 22b. 23a. 23b. 24. 25. 26. 27. 28a. 28b.- Proton NMR spectrum Proton NMR spectrum Proton NMR spectrum Proton NMR spectrum Proton NMR spectrum Proton NMR spectrum (in CC14). . ...... Proton NMR spectrum (in acetone) . . . . . . Porton NMR spectrum Proton NMR spectrum (C82). 0 o o o o o o o o Proton NMR spectrum Proton NMR spectrum Proton NMR spectrum Proton NMR spectrum Proton NMR spectrum Proton NMR spectrum Proton NMR spectrum Proton NMR spectrum Proton NMR spectrum Proton NMR spectrum of hexafluoroacetylacetone. . . of dibenzoylmethane (in CCLl) . of trifluoroacetylacetone. . . . of 1, 3-indanedione (in CHC13) . of a-methylacetylacetone of l-phenyl- 1, 3-butanedione of 2-pheny1- 1, 3-indianedione of B-bromoethyl acetoacetate . of thenoyltr ifluor oac etone (in of butylacetoacetate ..... . of t-butyl acetoacetate ..... M of n-butyl o-chloroacetoacetate WV of/‘tJbutyl a-chloroacetoacetete of ethyl acetoacetate ..... . of ethyl a-allylacetoacetate . . of ethyl a-isoamylacetoacetate NW of ethyl benzoylacetate . . . . of ethyl a-bromoacetoacetate . of ethyl y-bromoacetoacetate . xii Page 92 93 94 95 95 96 97 97 98 100 100 101 101 102 103 104 105 106 106 LIST OF FIGURES - Continued FIGURE 29a. 29b. 30. 31. 32a. 32b. 33. 34a. 34b. 35a. 35b. 36. 37. 38. 39. 40. Proton NMR spectrum of ethyl a-iasvgbutylaceto- acetate. 0 O O O O O O O O O 0 O O O O O O ....... Proton NMR spectrum of ethyl a-BJbutylacetoacetate Proton NMR spectrum of ethyl a-chloroacetoacetate. Proton NMR spectrum of ethyl a-cyanoacetoacetate . Proton NMR spectrum of ethyl o-ethylacetoacetate. . Proton NMR spectrum of ethyl trifluoroacetoacetate. Proton NMR spectrum of ethyl a-fluoroacetoacetate . Proton NMR spectrum of ethyl a-methylaceto- acetate. 0 O I O C O O O C O O O O O O O C O O O O O O O Proton NMR spectrum of ethyl a-isopropylaceto- acetate. 0 O O O C O O O O O O O O O O O O O O O O O O O Proton NMR spectrum of ethyl a-JL-propylaceto- acetate. 0 O O O O O O O O O O O O O O O O O O O O O O C Proton NMR spectrum of methyl acetoacetate . . . . Proton NMR spectrum of ethyl a-chloroacetoacetate fellowing distillation. O O O O O O O O O O O C O O O O 0 Chemical shift of enol OH versus carbonyl stretching frequency for 8-dicarbonyls. . . . . . . . . . . . . . Possible enol tautomers for trifluoroacetylacetone . Chemical shift of protons of ethyl acetoacetate in benzene and carbon tetrachloride. . . . . . . . . . . Proton NMR spectrum of acetylacetone in carbon tetraChloride. O O C O 0 O O O O O O O O O O O O O O O O xiii Page 107 107 108 109 110 110 111 112 112 113 113 114 123 124 133 134 LIST OF FIGURES - Continued FIGURE 41. 42. 43. 44. 45. 46. 47. 48. 49. 50.- 51. 52. 53. 54. 55. Chemical shift relative to pure solute of OH proton in 8-dicarbonyls at infinite dilution in carbon tetra- chloride versus mole fraction solute ...... . . . . Proton NMR spectrum of ethyl acetoacetate in benzene........ ...... .. ........ . Chemical shift relative to pure solute of the OH pro— ton in 8-dicarbonyls at infinite dilution in benzene versus mole fraction solute. . . . . . . . . . . . . . . Proposed complex of acetylacetone with benzene. . Proton NMR spectrum of butyl acetoacetate in n-hexane. o o o o o ooooo o o o o o o o o o o o o M Proton NMR spectrum of t-butyl acetoacetate in M chloroform...................... Association of keto molecules in ethyl a-chloroaceto- acetate. 6 0' e e o o o ooooooo o o e o o o o o 0 Chemical shift of chloroform proton in 8-dicarbonyls asdonorsolvents ...... Proton NMR spectrum of acetylacetone in methanol . Percentage enol tautomer in acetylacetone in various 8°1ventSu O O O O I O O O O O O O O O O O O O O O O O O 0 Percentage enol tautomer in ethyl acetoacetate in varioussolvents.................... Free Energy of tautomerization for acetylacetone versus solvent property. . . . . . . . . . . . . . . . Free energy of tautomerization for ethyl aceto- acetate versus solvent property . . . . . . . . . . . Long range spin- spin coupling in ethyl acetoacetate . Graph of log ([enol]/[keto]) versus l/T for a-methyl- ac etYIacetone O O C O O O O O O O O O C O O O O O O ‘O i' O xiv Page 135 139 140 142 145 146 147 151 156 162 163 167 168 171 175 HIST OR ICAL REVIEW Introduction-~Nature of the Equilibrium Keto-enol tautomerism in B-dicarbonyls may be considered as an equilibrium between keto and enol forms through the anion, as pictured below: 0 =0 R-C= -CHR'-C R" 4—9 R-Cp =cauc=0 O :0 -R" 4—9 R-C’ H'-'-'CR‘-C R" . (1) As early as 1912 the keto-enol equilibrium in these compounds was studied by Meyer (1) using bromine titration. This method involves the reaction of the enolic tautomer with bromine, as seen in the following equations: Me-C(OH)=CH-COOEt 331» MeCO-CHBr-COOEt (2a) MeCO-CHBr-COOEt fit—1;; MeCOCHz-COOEt + Br" + 12 (2b) Iz+zszo3= ———> 21" + 5.06: . (2c) As stated by Meyer, one of the principal problems associated with this method involves the fact that during the reaction more enolic tautomer is formed and the end point is not sharp. For the molecule, R-CO-CHR'- CO-R", Meyer noted that the equilibrium is affected by substitution of various groups for R, R', and R" in Equation (1). The effects of both solvent and temperature on the equilibrium were observed. Polar sol- vents tended to decrease enolization, as did increased temperatures (2). Results are shown in Table I for pure compounds. Bromine titration has also been used as a measure of the tautomeric equilibrium by Park e_t a}. (3), Ratnakar (4), Malawski e_t a_1_. (5), Table I. Percentage of Enol Tautomer in [Bu-Dicarbonyls Percent Enol Compound Pure 6 Liquid Gas Method Reference Acetylac etone 76 Br; 1 60 IR 30 78 92 BI; 55 8 5 NMR 71 81 NMR 23 95 B172 51 Dibenzoyl'methane 100 Br; 1 Ethyl acetoac etate 7 Br; 1 8 46 Br; 55 7 NMR 58 63 Br; 52 Ethyl benzoylacetate 29 Br; 1 Ethyl a-bromoac etoacetate 29 Br; 50 Ethyl n-butylacetoacetate 6 14 Br; 55 "" 2 Br; 14 Ethyl a-chloroacetoacetate 46 Br; 50 Ethyl ethylacetoacetate 3 Br; 1 3 10 Br; 55 Ethyl methylacetoac etate 3 Br; 1 4 14 Br; 55 Ethyl igppropylacetoac etate 5 6 Br; 55 1 Br; 14 Ethyl n—propylac etoac etate 7 13 Br; 55 Ethyl trifluoroac etoacetate 89 NMR 18 Methyl acetoacetate 5 Br; 1 6 54 Br; 55 3-Methy1acety1acetone 30 Br; 1 30 44 Br; 55 l-Phenyl-l, 3-butanedione 100 Br; 1 Eistert and Reiss (6, 7), Kabachnik et al. (8-12), and Eistert and Geiss (13). Results from the literature for 8-dicarbonyls included in the present study are given in Table I. The keto-enol equilibrium is influenced by many factors, and several of these will now be considered. Substituent Effects Meyer (1) noted the effect of various substituents on the tauto- meric equilibrium. He found increased enolization in the order OMe < OEt < OH < NHPh < Me < PH < COOR" for R" in Equation (1). For substitution in the a—position, enolization was decreased with an alkyl group and increased with an acyl group. However, if the un- substituted compound is strongly enolized, an acyl group decreases enolization. He found that bromine in the a-position increased enolization. Henecka (14) also found a decrease in enolization with alkyl a-substitution and explained this on the basis of a disturbance of the inductive and conjugative effects. He noted further that branching of the R (or R") [see Equation (1)] alkyl group of a 8-dicarbonyl decreases enolization, which contrasts with the conclusions of Hammond (15). Hammond has stated that branching of the end group reduces the keto content compared to acetylacetone (15). With large groups the carbonyl groups can only be rotated about 900 from the parallel coplanar configuration seen in Figure 3(a). In cyclic enols there is no serious interference with even a pair of terminal tertiary butyl groups. - Intramolecular hydrogen bonding is not possible in cyclic fi-dicarbonyls with both carbonyl groups in the ring. The influence of branching of a hydrocarbon chain on enolization in derivatives of acetylacetone was studied by Rumpf and La Riviere (16), and the percentage enol was obtained by bromine titration. Results are given in Table II. As reported by earlier workers,” increased branching resulted in decreased enolization. The authors have stated that a- substitution prevents coplanarity not only of the enol, but also of the anion. Substitution of a more bulky group in the end position gave increased ketone form, in disagreement with Hammond e_t a_1_. (17). The high degree of enolization in dipivaloylmethane has been explained by Hammond e_t 841. , by the fact that the end groups force the carbonyls into a position in which electrostatic repulsion is a maximum. Therefore, the enolic tautomer becomes energetically more favorable. For a series of 2-pheny1acetoacetic acid nitriles substituted in the 4'-position, Malawski e_t a_L_l. (5), have determined the keto-enol equilibrium by bromine titration. An electron attracting group in the 4‘-position resulted in an increased enolic content, and an electron withdrawing group in decreased enolic form. This effect must be an electronic one, since a steric effect is not possible in this position, according to the authors. Filler and Naqvi (18) have found by indirect bromine titration and by nuclear magnetic resonance that fluorine substitution on the acetyl methyl of ethyl acetoacetate results in increased enolization. The presence of the fluorine atoms probably makes the enols more acidic, so that there is increased proton donor capacity and stronger hydrogen bonding to the ester carbonyl oxygen. These results are shown in Table III. Structure Dipole Moments. Wolf (19) has determined the dipole moment of acetylacetone in benzene at 16°C as 2.78 i O. 08 D. Wolf has also calcu- lated dipole moments for each tautomer in several possible configura- tions, and these values are reproduced in Table IV. Zahn (20) measured a: Table II. Effect of Substituents on Enolization in B-Dicarbonyls Compound Ke PKe CH3COCH;COCH3 0. 24 8 . 23 CH3COCH;COC;H5 0. 19 8. 57 ~ CH3COCH;COCH(CH3)2 0.09 8.33 CH3COCH;CO[CHCH(CH3);] 0. 09 8. 88 CH3COCH;CO(06H5) 0. 38 8. 12 CH(CH3);-COCH;CH(CH3)2 0. 08 8 . 76 CH3COCH(CH3)COCH3 0. 03 9.2 cragcocmczmsmocs3 0.01 g 9.5 CH3COCH[CH(CH3);]COCH3 5 0.003 _<_10.3 CH3COCH(C6H5)COCH3 0.09 _<_ 8. 3 cn3cocn(cn,)c0(c6H5) < 0.005 < 8.8 as From Rumpf and La Riviere (16) **Ke 3 [enol]/[keto] * Table III. Influence of Halogen Substitution on Enolization in Ethyl Acetoacetates Percent Enol Compound in Pure Liquid Method Ethyl acetoac etate 6. 0 'NMR Ethyl y-fluoroac etoacetate 7 . 2 NMR Ethyl ‘y, 'y-difluoroac etoac etate 53 :1: 4 NMR Ethyl 7, ‘y, y-trifluoroacetoac etate 89 NMR Ethyl 'y-chloroac etoac etate 1 1 Brio:c Ethyl 'y, 7, 'y-trichloroacetoac etate 40-50 Br; Ethyl d-bromoacetoacetate 4 Br; Ethyl 0., 7-dif1uoroac etoac etate 5 Br; as From Filler and Naqvi (18) *8: Reported by authors from other references >1: Table IV. Calculated Dipole Moments for Ac etylac etone and Ethyl A cetoac etate Dipole Moment (D) Configuration - _ . . Ethyl Ac etylac etone Ac etoac etate Free diketone: 3. 7 3. 2 Trans diketone 1. 8 1. 6 ** Diketone (a) 3. 2 2. 8 -----------------------.— --------------------------------------- Free enol; only CH3CO or COOEt firm (b)** 3.0-2.8 2.3 Free 3.3-3.6 2.6-3.0 Chelate (c)** 3.4-4.0 2.7-3.2 Dienol (d)** 2.2-2.6 - a): From Zahn (20) ** Letters in parentheses refer to configuration shown on page 8. the dipole moment of acetylacetone in the gas phase and has considered the structure of both the keto and the enol tautomers. , Among the , structures shown in Figure 1, Zahn has preferred (c) for the enol and (a) for the keto in which the carbonyl approaches the opposite methyl group. ‘72 I? / \ H\ /C\ O=C /C-Me O-C\/ /C-Me \Me 0/ Me 0’ —m \m C H / \ ‘\ 1 C\ / - Me /C = C -- C; O-H”'O Me CH2 (c ) (d) Me- Figure 1. Configurations of keto and enol tautomers in acetylacetone. These groups approach too closely to permit completely free rotation, according to. Zahn. The author discounted the presence of either free enol or di-enol forms. The observed dipole moment was 3. 00 D,- , which represents an average of enolic and ketone forms. Since the equilibrium in acetylacetone is very far in the direction of the enol form, one should compare the observed value with that calculated for the chelate form (c), 3.4-4.0 D. Temperature studies over the range SO—ZOOOC by Zahn indicated a dipole moment independent of temperature. This is perhaps surprising when one realizes that increased temperature favors the keto form (2). Lack of temperature variation may be due either tocompletely free rotation or to completely hindered rotation. 'Similar calculations for ethyl acetoacetate gave dipole moments of 3. 2 D for the ketone form (a) and 2.7-3. 2 for enol form (c). The observed dipole moment in the gas phase was 2. 9 D, and no temperature dependence of this moment was observed in the interval 394-4310K. It has been found that the tautomeric equilibrium of ethyl acetoacetate is independent of temperature (2). Beyaert (21) attempted to determine the dipole moment of each tautomer of ethyl acetoacetate. The dipole moment of the ketone form as measured in benzene at room temperature was 3. 22 D. The benzene solution contained 16.8 percent enol. The moment of the enolic form, as measured in carbon disulfide at -80°C, was 2.04 D. The amount of ketone under these conditions was considered negligible. These results do not show good agreement with the calculations of Zahn. (20). The square of the apparent dipole moment has been determined by Le Fevre and Welsh (22) for ethyl acetoacetate in several solvents. These workers have assumed the following relationship, “2 = XeHeZ + Xka2 : (3) where p. is the apparent dipole moment, Xe and Xk are mole fractions of enol and ketone tautomer, respectively, and pe and pk are dipole moments of enol and keto tautomers, respectively. Le Fevre and Welsh criticized Zahn's gas phase moment measurements, since the percentage of enol tautomer was not accurately known. They have assumed the structures in Figure 2 for enol and keto tautomers. Me CH 0 Me CH; 0 \ \(ll-Jé \ Ci/M Et \ "/ \fi/M\Et e 9 o\ [,0 o o H/ (a) (b) Figure 2. . Configurations of keto and enol tautomers in ethyl acetoacetate. 10 Structure (a) gave a calculated moment of 2. 2 to 1.4 D, and structure (b) a moment of 3. 5 to 2. 7 D, for values of the angle 9 between 600_ and 120°. Apparent moments are listed in Table V. Dielectric con- stant measurements in Le Fevre and Welsh's work have been con- verted from weight fractions to mole fractions for the purposes of this thesis. These values have been plotted graphically for several solvents. The graphs deviate from linearity as would be expected for- a‘ decrease of percentage keto tautomer on dilution. Configuration of the Keto Tautomer. Possible structures for this tautomer are seen in Figure 3, Although Le Fevre and Welsh (22) have R! | R 7 CH u R" O R" \ C/ \C/ . I n O O O CHR O R CHR' O (a) (b) (C) Figure 3. Configurations of the keto tautomer of [3-dicarbonyls. pictured structure (a) for the keto tautomer, molecular models show that the carbonyls in this orientation have maximum electrostatic repulsion. Reeves (23) has pictured acetylacetone as structure (a) since the methyl groups are equivalent in the proton NMR spectrum. Hammond e_t 11. (17), have suggested that acetylacetone exists as either (b), (c), or both. Bulky groups on the end positions tend to force the molecule into the enol form rather than into the less favorable keto configuration (a). For some compounds with bulky end alkyl groups the carbonyl groups may only be rotated about 90° from their parallel coplanar configuration in structure (b). Zahn has suggested structure (c) as the most important for the diketone form. 11 Table V.* Percentages of Enol Tautomer, Dielectric Constants, and Apparent Dipole Moments for Ethyl Ac etoacetate in Various Solvents = i m Percentage Enol Solvent Observed Calculated** u E app solvent Chloroform 0 - 3. 36 - Ether 14 19 2.42 4.27 Carbon tetrachloride 25 53 3. 07 2. 23 Benzene 25 52 2.93 2.27 Toluene 39 49 2. 75 2. 37 Carbon disulfide 43 44 2 . 60 2. 63 Hexane 63 63 1. 70 1. 89 * From Le Fevre and Welsh (22), using literature values for observed enolic perc entage s . ** Calculated from Equation (6). 12 Configuration of the Enol Tautomer. The enol tautomer might exist in any of the forms seen in Figure 4. R! R! 1 Me CR' R" H c H c H \ \ \ \c/ \c/ \o-c/ /C-R" o-c/ \ -o/ ‘ " \Me o/ \\CH R" o ,o ’- \H/ (a) _ci_s-enol (b) trans-enol (c) di-enol Figure 4. Configuration of the enol tautomer in B-dicarbonyls. The 29. or intramolecularly hydrogen bonded species, (a), has been suggested by Sidgwick (24) and many workers since that time have assumed that this was the preferred configuration. Referring to Meyer's work, Sidgwick observed that the preference for the enolic tautomer in nonpolar solvents indicates the presence of a chelated enolic structure. According to Sidgwick's calculations, the sum of the angles in the chelate ring structure is 7200 or that of a plane hexagon, which indicates a ring entirely free of strain. Sidgwick noted that the chelate enol tautomer is less polar than the diketone form and will have a lower boiling point than it or the free enolic tautomer. Hammond (15) stated that enols have a c0planar ring of six atoms which gives minimum electrostatic repulsion and maximmn resonance stabilization. He has further sug- gested that e-methylac etylac etone may exist to some extent in the trans enol form, since infrared measurements show a weaker hydrogen bond than in acetylacetone. Henecka (14) suggested the presence of_tr_a_n_s_ enol in compounds with large groups in the u-position. Bromine titration indicated the presence of enol. However, the colorimetric reaction of ferric chloride, which involves only the Eli enolic tautomer was negative. The fact that bromine titration indicates the presence of enol, but the ferric chloride 13 reaction negates the possibility of cis enol, lends support to the presence of a trans enol tautomer. In this reaction the iron atom replaces the proton in the intramolecular bond. FeC13 + C6H902° OH '———9 FCC12C6H903 + HCI (4) Eistert e_t a}. (25), have called attention to the possibility of two £i_s_- and two m- enolic forms for the molecule, R'COCH;COR, where R and R' differ. In c_is enols the chelate structure essentially eliminates the difference between the two forms. For B-ketoesters the keto carbonyl is predominantly enolized. These authors suggest the presence of a m enol-solvent complex in a solution of acetyl- acetone in methanol. Kabachnik ‘3} al. (8), hypothesized that the keto-enol equilibrium of Bins-fixed B-dicarbonyl compounds is independent of solvent. Eistert and Geiss (13) found infrared and ultraviolet evidence indicating that these equilibria were actually dependent upon the solvent. In hydrophobic solvents enolic trans—fixed compounds were ketonized, whereas in hydrophilic solvents they were not. Kabachnik, gt a}. (9), extended this study to a series of compounds by ultraviolet and infrared spectroscopy and attempted to show the applicability of the relationship K6: EL + E'L', (5) where Ke is the equilibrium constant, E and E' the £2 and trans enolizability of the solute, and L and L' the tendency of the solvent to enhance gig and E13 enolizability in the given solute. They have demonstrated that for fi-dicarbonyls with bulky substituents in the a-position, the equilibrium is independent of the solvent. They have argued that a large percentage ofmenol exists in compounds with bulky a-substituents. In another work Kabachnik e} 31. (10), found that 14 enolization was essentially independent of the solvent for alkyl tetronic acids and for cyclic acetals of malonic acid. Only dimedone seemed to agree with Eistert and Geiss' hypothesis that the trans enol form pre- dominated in hydrophilic solvents and the keto in hydrophobic solvents. They further examined (10) enolization in a number of compounds and showed that the equilibrium depended to a considerable extent upon the type of compound. The keto-enol equilibrium of c-alkylacetylacetones has been investigated by Kabachnik e_t ail.( 11). Ultraviolet absorption spectra appear to indicate the presence of both c_i_s_ and m enolic forms, and the relative amounts of £13 and 132? forms are influenced by the solvent. For a-ethyl, u-prOpyl, a-butyl, and a-iAs/‘qbutyl acetylacetone, both £i_s_ and tra___n_s forms appear to be present. For a-methyl acetylacetone, only _<_:_i_§ seems to exist, whereas in a-i/s’o’propyl and a-sgcl-butyl acetylacetone,- only m may be present. The amount ofm enol has been considered constant, and a graph of l/L versus l/I has been made, where L is the solvent enolizability toward _c__i_§_ enol and I the intensity of the absorption peak. This graph gives an approximately linear relationship for solutions in hexane, ethanol, ether, and in 67, 75, and 85 percent aqueous methanol. Values for xmax and the extinction coefficients are given in Table VIII for ethyl:, igbutyln and i’g’gpropylacetoacetates. Bratoz e_t a_._l. (26), found infrared evidence in the form of a weak absorption band at 3570-3600 cm'1 which suggested the possibility of a small amount of mend in acetylacetone, but this has not been con- firmed by others. Rasmussen e_:_t 11- (27), have recognized that not only the formation of an intramolecular hydrogen bond but also the separation in space of the oxygen atoms may be important in influencing enolization of B-dicarbonyls. Certain gins-fixed 8-diketones which cannot form intramolecular hydrogen bonds may form dimers involving the enolic tautomer . 15 Shigorin (28) has measured the frequencies and intensities of the infrared bands of the enol and keto forms of several (3-dicarbonyls. Evidence was presented in support of the existence of dienol forms in equilibrium with enol forms. In a later work Shigorin (29) has intro- duced deuterium into 8-dicarbony1s and has tabulated Raman frequencies for the C-D bond and for the C=CH and C=CD double bond. The presence of dienol was confirmed, according to the author. Infrared Absorption. Infrared data are available for B-diketones in work by Mecke and Funck (30), Rasmussen e_t a_1. (27), Park (it a_1. (3), Angell and Werner (31), Bellamy and Beecher (32), Bratoz e_t a_1. (26), Holtzclaw (_e_t £11. (33), Hammond e_t a_1. (17), Tamm and Albrecht (34), Murthy e_t 31. (35), Powling and Bernstein (36), Kuratani (37), Belford g1: 11. (38), Shigorin (29), and in Sadtler Standard Spectra Nos. 3505, 5437, 5774, and 11982 (39). Spectra for B-ketoesters may be found in Le Fevre and Welsh (22), Rasmussen and Brattain (40), Hunsberger it 2:1. (41), Bratoz gt a_1. (26), Murthy gt a_1. (35), Powling and Bernstein (36), Kuratani (42), Belford e_t a_1. (3 8), Shigorin (28), Shigorin (29), Bar’ikowska (43), and in Sadtler Standard Spectra Nos. 65, 101, 8013, 8017, 14640, 17508, and 18941 (39). Data from these references are summarized in Tables V1 for B-diketones and in Table VII for B-ketoesters. In general the following frequencies may be listed for groups of principal interest in the tautomeric equilibrium. .1 Group Frequency, cm Reference Free OH intermolecular 3650-3590 sharp Bellamy (44) Intramolecular hydrogen bond 3570-3450 sharp Bellamy (44) Chelated OH 3200—2500 broad Bellamy (44) Normal conjugated carbonyl 1695-1672 Rasmussen (it a_1. (27) Carboxyl carbonyl 1733 Bellamy and Beecher (32) Carbonyl not chelated 1709 Mecke and Funck (30) Conjugated chelate carbonyl 1639-1538 Bellamy (44) C=C 1650 Mecke and Funck (30) 16 5.83.830 Q 6583 dowospeg smfidoofioasouozi. .oflo>o< 3 Hofimbcvucoo Go mucomofl s 3:030m Go mpsomoflvws “opiogombou 233.30% m ooe: w... ow: met 5.: 32 22866256666833; em mom: 6N Moe: oeem N... 8.: «NS 6:666:33 .m J 4.33m: mm m:6: woe:.oms: m: 8.: 383 6.806 6336 6161562- m em owe: ::e: mes: mm mm: .8 em: 3: do: 8:... 6.286683666863885 mm eeos:.mms: :m dc:e:.moee: emcs:q.oms: ssemmem eeocoeossn: NM ooe: em OHe: emcee em moo: seem 2. mm m: .8: 2856816235 om woe: woe: ooem 6N o:e: omen mm 2.: mm: 22866236088304 em eem: mme: mus: mm woe: mme: em mme: mas: om Mme: emcee em o:e: comm mm mmm: wHe: ooe:.ome: ommm.oeem om eHQ: eoe:.s~s: oosm em *mmm:-eme: mosm em o:e: ems: mm mm: on: we: .3: ommm 88663384 eeeoeoeem 0.6 9.6 Sam one 36: mo 3.30:0 6559500 A. .83 psmm coflmuowwoke ,,l|l!| 17 .5928 odd Begum? 94. Ha is omn: :0 Nmf mmn: omhfi Emma om mmoa pmoso. opmuoomouoom 3:32 om mmn: ow: em: mm mvmauowoma wood omf comm opmsmumoumomoaosdwfl {Hum om emf e895 1.. mm mNoH ma: mafia oumuoomoumomgmosmofl THU—H we. emf ow om: 3: we: 3883686625588 15m om . omofi pmofin. mumuoomoEomffflm Trim 2. m1: «0on mmnanoofifi cm can: 336 owOHoomosoEo 15mm om men: pecan. i mm mm: 3: we: 3369638233-: 15m om UMOHQ. mumpoomHIAONcoQ Twainm om Nmofi 3nd 03L ow omfi 3: NE; mm omnowofi of; mm: ommm mm 31.0on own: om own: pmoun om name: A» ova: omvfi mm men: 00: mm: Nv om: 0v: 33696303... 35H am 3.09: 38368on :3st vocouvmom 000 On0 Ou0 MOOU $0 33030 30M 30M UGSOQEOD A153 beam GoflmwOng. u it i)” E msvumGOuOMnn Hem mofiocofiwosh possum-cm .HH> ":an I rJ-dHUfimUF~A-Tu\rC h a .200 a . a H .. m: OO:M..OOm .Ma DEM sneaks. 30:0 fiCOuOUQTSQUdH her-rvvurwe iv: U Aa.:.vusai...)n \Aum-~flw.~AU-\/— dn~rv>~amfl ~V--00-rhunvsv ...... 18 .Gofldufinoocou mo «newcomovafi .noflsfiv mfidflg ou Uvugommnuxm Jordon—um» HOG» mcmufixr 50535.3 Hos» 303 .303on 5?? mofium> 3% * * "w mm ooofiood J a wwm muodxo . w. #030 vacuvomfifivomouosflfinfi mm cow . w m hm m..o on . w. £030 0:80 0.350 umonodzmmxvm mv 33.059” wow and . o mcmumofl vuduoomgo Odo Hodgoud gaum mm .3..ooo..£ .. 330 umoum um 35mm .: .. mm m .. 28va S - NE ..- o H Us . N zonmu mamfiofiaoucoflm NH omom room. $032 @530 0.23m omamoummmmud NH foowd -.3oovw room .3oom . $032 vacumomfnuooaffianoflw .6 pm ooowofi cpm mm *iooonmfi - .5 ...ooH -. H a ... .. ocmxvgofio>0 mm oomuoomf 03 “ads n5T5 83032334 vodouflmfl u EEVNMEA 33.302 pavZOm UndomEoD mTwconumuflQun new :ofiumHOmokw ucfiogmufib 43> 3nt 19 In B-diketones Rasmussen (27) has found the OH frequency at about 2700 cm'l. The OH peak is often very broad due to its chelated nature and may be masked by the strong CH stretch band near 3000 cm"!. The - strength of the hydrogen bond is given in some measure by the frequency,‘ since a lower stretching frequency indicates a more nearly symmetrical bond, according to Rundle and Parasol (45). Rundle and Parasol have indicated that very broad bands which are characteristic of strong bonds become sharper as the bond becomes symmetrical. For a symmetrical bond, the OH bond length is longer, with a resulting lowering of the OH frequency. In compounds which appear symmetrical, such as nickel or ‘1 and palladium dimethylglyoxime, the OH frequencies are 1775 cm about 1600 cm’l, respectively, and the peaks are sharp. With increased hydrogen bonding, the peaks become first broadened and finally sharp once again. Mecke and Funck (30) showed that for the cis- enol form of acetyl- acetone the observed frequencies for the C=C and C=C bonds are between those for the normal CO and CC single and double bonds. Several struc- tures have been written for the hydrogen-bonded species in Figure 5. l l C C C C \ § \ 9” \fi 33K C a? 9 CK $ | o 0 +0 0- o o- o\ o \ /’ \ \ / H/ H H+ \H (a) (b) (C) (d) Figure 5. Possible structures for the intramolecularly hydrogen- bonded molecule. Structures (a) and ((1) together would lead to a symmetrical hydrogen bond. Forms (a) and (d) then could explain the lowering of the OH stretching frequency. Structures (b) and (c) might explain the displace- ment of the double bond frequency even though the hydrogen atom were 20 not symmetrically located according to these workers. Rasmussen (_e_t a_1. (27)have also suggested that structure (b) is stabilized by the hydrogen bond of (a). Bellamy and Beecher (32) have pointed out that the frequencies of the carbonyl stretch are essentially the same for acetylacetone,- benzoylacetone, dibenzoylmethane, a-thenoyltrifluoroacetone, and di- medone. These workers have assumed that the contribution of structure (b) [Figure 5] is the same in all these compounds. They have further calculated that form (b) contributes about 20-25 percent in aliphatic = (B-dicarbonyls. The authors have concluded that, since the equilibrium constants of enols vary widely, the proportion of resonant form (b) is not a determining factor in their ionization. Consequently, no relation- ship would be expected between the carbonyl frequencies and the equi- O librium constants. The chelated carbonyl stretching frequency occurs about 100 cm‘1 below that for the normal conjugated carbonyl group and is a much more intense peak (27). Hunsberger e_t 2:1. (41), have suggested that the dif- ference in carbonyl stretching frequencies may depend on the relative importance of structures such as (e) (0 Figure 6. Conjugated structures for the enol tautomers in fi-dicarbonyls. The relationship between the carbonyl frequency and double-bond Character has been investigated by these workers. A linear relation- Ship has"been found between the difference in chelated and unchelated stretching frequency and the double bond character, using Pauling's values for percentage double-bond character and neglecting ionic Character. 21 Rasmussen and Brattain (40) have assigned the 1650 cm'1 band to the chelated carbonyl of ethyl acetoacetate, in disagreement with earlier Raman results. Rasmussen (_e_t a_1.: (27), also noted a chelated carbonyl absorption band both intense and much shifted from the normal region. It was suggested that the greater participation of ionic struc- tures leads to greater decrease in the double-bond character of the carbonyl bond than would occur for simple conjugation. The increased charge on the carbonyl oxygen results in the observed high intensity. It was assumed that the C=C stretching frequency is hidden under the carbonyl absorption. Bellamy and Beecher (32) have stated that the carbonyl frequencies are determined almost entirely by the character of the double bond involved since its length effectively determines the O. . .0 distance. Unlike the results of Hunsberger 3133.1. (41), a linear relationship between double bond character and the difference in carbonyl stretching frequency for a chelate and non-chelate bond was not found to hold for aliphatic compounds with unconjugated double bonds. A graph of double bond character of fi-diketones versus the actual stretching frequency, however, shows a linear relationship. In assigning double- bond character, Pauling's values were used, and the contribution of ionic and Dewar forms to the resonance hybrid was ignored. For B-keto- esters such a plot is less satisfactory, probably because the second oxygen of the ester group has greater effect on the carbonyl to which it is attached. Stretching frequencies have been measured by Raman spectroscopy by Shigorin (46) for several B-dicarbonyls. The results are seen in Tables VI and VII. Ultraviolet Absorption. Ultraviolet absorption of (S-dicarbonyls has been studied by Belford at 6:1. (38), Hammond El; a_1. (l7), Murthy 2t 311. (35), Rasmussen e_t a_._l. (27), Bankowska (43), and Kabachnik e_t 11. (11). 22 Values for xmax and of extinction coefficients are given in Table VIII. Rasmussen, e_t a_1. (27), have pointed out that a simple conjugated system will have two maxima at 2350 and 3200 X, with extinction co- efficients of 10, 000 and 40, respectively. For the conjugated chelate systems studied here, however, only one maximum of considerable intensity was observed. Summary of Structural Work on Keto-Enol Tautomers. (1) The Keto Tautomer. Dipole moment measurements for ethyl acetoacetate and acetylacetone appear to indicate the presence of structure (b), Figure 3. Structure (a) has been discounted because of strong electro- static repulsion. Structure (c) would lead to a dipole moment sub- stantially lower than the experimental values and lower than the enolic tautomer. However, the boiling point data would indicate that the enol form is less polar since it distils at a lower temperature. Also in favor of structure (b) is the fact that little rearrangement would be necessary in order to form the coplanar anion which has been proposed as an intermediate in the keto-enol equilibrium reaction. This writer's calcu- lations for structure (b) give a dipole moment of 3. 0 D, in good agree- ment with Zahn's calculations (20). The presence of a small amount of the _t_r_a_r_13 structure (c) cannot be discounted as a possibility. However, compounds with large bulky groups on one or both ends show an in- crease in percentage of enol tautomer, presumably because of steric hindrance in structure (b) (Figure 3), whereas in structure (c) the non bonded interactions are negligible. (2) The Enol Tautomer. The presence of the cis-enol form has been clearly established by both infrared and nuclear magnetic resonance ‘ measurements. Kabachnik's evidence for the existence of trans enol in e—alkyl substituted B-ketoesters depends on the absence of the chelated cfilrbonyl absorption peak in infrared measurements (9). It is possible, 23 however, that the C=C absorption peak masks the chelate carbonyl absorption band. His assumption that the percentage of 323.23 enol is independent of the solvent appears unsubstantiated since bromine titration is neither precise nor accurate as an analytical tool, and since the ultraviolet spectra and graphs of equilibrium constant versus enolizability'of‘solvent toward _c_i_s_enol both suggest the oppOsite con- clusion. The present writer would prefer to have positive evidence in the form of free OH stretching bands in dilute nonpolar solvents. Ultraviolet evidence given by Kabachnik e_t 2:1. (1 1), for i’aqbutylacetyl- acetone does, however, appear to indicate two conjugated forms each of which depends upon solvent in both intensity and maximum wave- length. Shigorin's suggestion of a tra___n_s enol form for ethyl acetoacetate in inert solvents from infrared and Raman data does not appear to be proven by the facts (28). The present writer feels that additional solvent studies of the OH stretching frequency would be needed, and a clear distinction should be made between the intramolecular and intermolecular hydrogen bond in terms of stretching frequency and solvent dependency. Further studies of Shigorin (47, 48) do not appear to provide new evi- dence for the existence Mm enol in other fi-dicarbonyl compounds. As has been mentioned, Bankowska (43) was unable to reproduce Shigorin's infrared evidence for free or intermolecularly linked OH bands for ethyl c-chloroacetoacetate. Shigorin's Raman spectra of ethyl acetoacetate and acetylacetone ‘ do not appear to support the presence of dienol forms (28). The substi- tution of deuterium in B-dicarbonyls has provided evidence which may suggest the presence of a small amount of dienol structure (29). Shigorin has concluded that the low intensity of the O—D band in certain compounds indicates the presence of the dienol form. He also notes that the intensity Of the C=C-H intensity increases in deuterated over that in the undeuterated Compound. In a compound with mobile protons on the acetyl methyl and 24 in the a-position, these two facts together would substantiate Shigorin's claim. In this case, the data could only be explained by a structure such as the following: OD OH I I CHZ=C'CH=C'R However, there is no evidence from NMR that the acetyl methyl protons are labile. Dipole moment measurements indicate that the enol form of both“ acetylacetone and ethyl acetoacetate is an intramolecularly bonded species. A free hydroxyl group would result in a larger dipole moment than that for the keto tautomer, contrary to the evidence. The present writer has attempted to calculate the dipole moment for the enolic tautomer. If one considers the proton symmetrically placed between the oxygen atoms with an OHO angle of 1500, and takes the OH dipole moment from water, the calculated dipole moment is 3. 5 D, higher than the observed value. If the dipole moment for the OH bond is taken as 1.7 D as in free hydroxyl groups, the calculated moment is 2.9 D. Effects of Temperature and Pressure Meyer's observation of an increase in keto tautomer in acetyl- acetone with an increase in temperature and the independence of the keto-enol equilibrium in ethyl acetoacetate from temperature changes have been noted. The latter observation does not agree with expected results or with the work of Briegleb (52). In a study of acetylacetone and ethyl acetoacetate in various solvents by bromine titration, Gross- mann (49) has found that the keto concentration increases with a rise in temperature, although the rate of achieving equilibrium varied from 1'2 hours for alcohol solutions to 1-2 days for inert solvents. Briegleb e_t a_1. (50), have studied the temperature dependence of the keto-enol equilibrium of a-chloro- and a-bromoacetoacetic ester in 25 the gas state. The enolic content of the d-chloro- and a-bromoaceto— acetic ester over the range 0-1600 and 0-1000, respectively, is less than that of ethyl acetoacetate itself at the same temperatures. Earlier work provides temperature studies on acetylacetone (51) and ethyl acetoacetate (52). Equilibrium constants and thermodynamic quantities for these compounds are given in Table IX. The effect of pressure on the tautomeric equilibrium of aceto- acetic ester was studied first by Kabachnik e_t a_1. (53). The equilibrium constants were determined in two percent solutions in various solvents. The observed shift of the equilibrium to the enol form with increased pressure in water and methanol has been interpreted as a solvation effect. The authors have assumed that the 2323 enol form, although low in concentration, is in greatest concentration in aqueous solvents and in methanol, and is most readily solvated. In ethanol and chloroform, increased pressure gave smaller equilibrium constants, and in the pure compound and in inert solvents, no effect was observed. Le Noble (54) has questioned whether the systems in Kabachnik's study were in equilibrium. In the pure liquid, Le Noble found for ethyl acetoacetate a decrease in the equilibrium constant from 0. 0834 to 0.0685 from 1 atm to 3700 atm. For an equilibrium constant which is the ratio of enol to ketone tautomer, one would expect an increase in the equilibrium constant with increasing pressure, which was observed. Le Noble has pointed out that the molar volume of [-C(OH)=CH] is less than that of COCHZ, and that this difference is even greater for the intramolecularly hydrogen-bonded species. According to him the molar volume of the enol is greater than that of the diketone by about 1. 0-1. 5 ml/mole. 26 ”soéroaa ..... om... NJL... ©w.mu mo.N Mw.mn «$0.0 w Nimdw om 3.309... «.m- 8:? om .o S. .7 Bed om N .mZ 639880.36 1.3m om.m.. NmJn 3‘.“ o~.o.. moH.o A: mmv om oumuood w.2- 3.? god 3.? $3.0 3. ~12... -838830é 1.3m NHKV: mHJqI MQJ... mm.Nu moa.o u: N.mmv mm m.N~u awn: NON .o: H>.mu on; no N.MNN oumuoomouoom .Tfthm mmé»: :..NI ooé: swan .... ow N.MH¢ Am 3.0-. 2 .o- 3.7 3.2- of mo ~55 2.882384 .mp 305 308 308 305 0020 \Ho 0% \Q ox \aoox \3 o& 0M 3cm MOB vasomaoo 383m m4 m3. .3 ma .. 208$ maconumgflun 050m GM 8.939%:de HOdMuOqu mo oocovcomofl ousumhomnflofi .XH £nt 27 Solvent Effects Bromine Titration. Meyer (1) has studied the effect of solvent on enolization in ac etylac etone and ethyl ac etoacetate by bromine titra- tion in 3—5 percent solutions in various solvents. The percentage of enol tautomer and equilibrium constants have been determined, and results are shown in Table X. A marked increase in the percentage of enol may be noted in the nonpolar solvent, hexane, over that in more polar solvents. Bromine titration has been used by Kabachnik (9) in a study of the effect of solvent on the tautomeric equilibrium in ethyl acetoacetate. Percentages of enols are given in Table XI, which indicates greatest enolization in nonpolar solvents. More recently Kabachnik et al. (10), followed the enolization of a- substituted B-keto- esters. These results are also shown in Table XI. For all compounds the greater enolization is in the direction 67 percent aqueous MeOH < MeOH < EtOH < benzene, but it should be noted that concentrations are not recorded. » Dipole Moments. The effect of the solvent on the apparent dipole moment of ethyl acetoacetate has been noted by Le Fevre and Welsh (22). Percentages of the enol form have been taken from literature values and extrapolated to infinite dilution. The moments have been rearranged_ in order of increasing percentage of enol tautomer, as shown in Table V. Calculated values are those obtained from assuming a simple relationship between the solvent dielectric constant and the equilibrium constant, as shown in the following equation, E = 100 e '“e (en/Pk , (6) Where E is the percentage enolic tautomer, P‘e and “k are dipole moments for the enol and keto tautomer, respectively, and e is the 28 * Table X. . Percentages of Enol Tautomer and Equilibrium Constants for Enolization of Acetylacetone and Ethyl Acetoacetate in Various Solvents o Ethyl ac etoac etate Acetylacetone Solvent T Percent Ke** Percent Ke** Enol Enol Water 0 0.4 0.004 19 0.24 Formic acid 20 1.1 0.011 48 0,9 Methanol 0 6.9 0.074 72 2.6 Melt 20 7.4 0.079 76 3.2 0.084*** Chloroform 20 8. 2 0.089 79 3.8 0.106*** Ethanol 0 12.7 0.15 84 5.3 Benzol 20 18.0 0.22 85 5.7 Heptane -- - — 0. 92=I<=i=>i< -- -- Hexane 20 48 0. 9 92 12 *Bromine titration by Meyer (1) in 3—5 percent solutions. **Ke F [enol]/[keto]. sgkakFrom Kabachnik (9). u n!!! .lelvva I.-.n.- ...:v a... - ~L- (lav. mun-und-uruv. ~n v-uqu. .' 9t Iii-,1 ill-L- 7,! 1. "III 29 £83953 0580.5 a3 3 .3 xflfiomnmvH «Gash... o.~m 93 0.3 cg; «a .36 m5 od mg as Ne :m.o.m as; 0.3 «0.2 mg: ... .3 fidum We To one m.m .36....“ ea in o..m ed .mnoum N6 dim m... «.m Eda 0.... NJ. o.m a; ammo».um N4. w.~ as To no.0 Nd mam m.~ a; .mo :3 5m ion mg: ...: is o.~ m OOUMEUOUoE Gd .m «Goduflmfidm 255: :20 .80 05.8 £6 moss $002 $002 «Guzman mucoxfiom moofinm> 5 moomuoomopoo< 123nm oopsoflmnfimud HOH ~053de Hocm mo mommudoouonm .HX 03oF * 3O dielectric constant of the solvent. There does not appear to be a good relationship between the solvent dielectric constant, as expressed in Equation (6), and the tautomeric equilibrium, although some agreement may be noted between observed and calculated values. Raman Spectroscopy. From the vibrational spectrum of OH, Shigorin (28) has noted that an intermolecular bond between keto and enol forms is preferred in pure ethyl acetoacetate. In inert solvents such as hexane, carbon tetrachloride, and carbon disulfide at low concen- trations, the enol tautomer forms an intramolecular hydrogen bond. The formation of this bond is then followed by displacement of the equilibrium toward the enol form (28). The present writer does not find adequate evidence for these conclusions. Infrared Spectrosc0py_. Infrared techniques were used by Powling and Bernstein (36) to show the effect of solvent on the keto-enol equi- librium in acetylacetone, ethyl acetoacetate, benzoylacetone, and other compounds. Free energies and heats of tautomerization for these sub- stances vary in a linear fashion with the solvent parameter [(6-1)/2€+l)] (p/M). Tables X11 and XIII for free energies and enthalpies of tauto- merization, respectively, incorporate not only the authors' results but also values taken from the literature by them. For the purpose of this thesis values at infinite dilution have been extrapolated from figures in reference (36). The trend for both free energies and enthalpies is in the same general direction for all compounds, and algebraically smaller values are found in the gas phase, or in a nonpolar solvent such as hexane, than in polar solvents. In an infrared study Kuratani has found ethyl acetoacetate more enolizable in the nonpolar solvents carbon disulfide, hexane, ether, and benzene than in methanol and pyridine (42). In a more recent work, 31 >5: Table XII. Free Energy of Tautomerization for fl-Dicarbonyls in Various Solvents Compound 25°C AF kcal/mole Solvent Ac etyl- Ethyl aceto- l-Phenyl- 1, 3- acetone acetate butanedione Gas phase -1.4 +0.2 - Ethanol -1.0 +1.2 +1.0 Methanol —0.6 +1.5 +1.5 Water +0.5,+1.0 +3.3 - Acetic acid -0.6 +1.7 +1.7 Formic acid -0.1 +2. 7 - nv-Hexanemk - 1. 4 +0. 2 - Toluene,“< -1.1 +0.8 +0.9 Benzene** -1.0 +0.9 +1.2 Decalinwg - 1. 2 — - Hexachloroethylenew< - l. l - - Chloroform“: -O. 8 +1. 4 +1. 5 Bromoform** -0. 7 - - Acetone:M -0. 5 +1. 5 - Values interpolated from figures in Powling and Bernstein (36). Solutions were 0. 1 M and were considered infinitely dilute. ** Solvent considered inert. 32 >:< Table XIII. Enthalpy of Tautomerization for Acetylac etone and Ethyl Acetoacetate in Various Solvents compound AH kcal/mole Solvent Acetylac etone Ethyl acetoacetate Gas phase -2.4 -2.0 Decalin - 2. 3 - Hexane - - 2. 0 Tetrachloroethylene - 2. 2 - Carbon tetrachloride - - l. 3 Bromoform - 2. 1 - Ether - - l. 0 ’13: Pentanol - - 2. 0 Ethanol - -1. 6 Methanol - -1. 3 * Values interpolated from figures in Powling and Bernstein (36). Solutions were 0. 1 M and were considered infinitely dilute. 33 Kuratani has concluded that acetylacetone is also more enolizable in nonpolar solvents than in polar solvents (37). Infrared studies of the effect of solvent on the tautomeric equilibrium in dimedone were made by Angell and Werner (31). For this compound the keto form is preferred in nonpolar solvents. The authors have concluded that, where intramolecular hydrogen bond- ing cannot stabilize the enol form, the keto form is preferred. In acetonitrile and dioxane both keto and enol forms are present. Mecke and Funck (30), in infrared studies of acetylacetone, have found that the percentage of enol increased from an acetonitrile solution through the pure liquid to hexane solutions. These results are in Table XIV. The percentage enol has been determined for a number of fl-dicarbonyls by Conant and Thompson (55) in absolute alcohol and in hexane, and these values are also in Table XIV. Ultraviolet SpectrosccHDy. In an ultraviolet study of a-acylphenyl- acetonitriles, Russell (56) has concluded that in hydroxylic solvents, the keto-enol equilibrium is determined by solvation, whereas in non- hydroxylic solvents, it is determined by chelation. The effect of solvent on the equilibrium constant for acetylacetone has been recorded by Hammond e_t a_1.(17). Where this constant is defined as Enofl/flmtdl, the values are 1.4 and 0. 15 in acetonitrile and water, respectively. Murthy e_t a_1. (35), have determined the tautomeric equilibrium of ethyl ac etoacetate and acetylacetone in several solvents from the equation, 6 O 6a where E is the percentage enol, £0 is the observed molar extinction coefficient, and Ea the assumed value for the enol tautomer. Results of these calculations are shown in Table XV. The authors have plotted 34 Table XIV. Percentages of Enol Tautomer for fi-Dicarbonyls in Various Solvents m Percent Refer- Compound Enol Solvent Molarity ence Ac etylac etone 9 0 hexane 1/10 (vol) 30 45 CH3CN 1/10(vol) 30 9 2 hexane 0 . l 5 5 8 3 ethanol* 0 . 1 5 5 Dibenzoylrnethane ~100 ether - - 1 Ethyl a-butylacetoac etate 10 hexane 0. l 55 6 ethanol 0 . 1 55 Ethyl a- ethyl ac etoac et ate 15 hexane 0 . l 5 5 4 ethanol 0 . 1 5 5 Ethyl a-methylac etoac etate 12 hexane 0. l 55 5 ethanol 0 . 1 5 5 Ethyl a-i’agpropylac etoac etate 6 hexane 0 . l 55 Ethyl a-nfl-propylac etoac etate l4 hexane 0 . l 55 Ethyl ac etoac etate 49 hexane 0 . 1 5 5 10 ethanol 0 . 1 5 5 13 ether - l l-Phenyl- 1, 3-butanedione 84 ether - 1 :4: Absolute ethanol in all cases. 35 :1: Table XV. Percentages of Enol Tautomer for Ethyl Acetoacetate and Ac etylacetone in Various Solvents Pe rc entage Enol Solvent Ethyl Acetoacetate Acetylacetone Cyclohexane 43 88 Dioxane 8 7 3 Di ethyl ether 25 85 Ether/CH3CN (50/50) 7 68 Ether/CH3CN (75/25) 7 71 Ac etonitrile 4 55 t: Butanol 10 80 iv- Propanol 1 1 85 Ethanol 8 73 Methanol 6 70 Water 0 . 5 16 >5From Murthy e_t a}. (35), calculated from extinction coefficients. 36 the percentage of enol tautomer against the carbonyl stretching fre- quency and the n ->1r* frequency in various solvents. The trends in both infrared and ultraviolet data indicate the stabilization of the carbonyl groups of the keto form by polar solvents. In the case of polar solvents which cannot donate protons, the solvent blue shifts are determined mainly by the dielectric constant of the solvent. In proton-donating solvents, however, the major contribution is hydro- gen bonding . Nuclear Magnetic Resonance Spectroscopy. Nuclear magnetic resonance (NMR) spectroscopy has been used by Reeves (23) for the determination of the keto-enol ratio for acetylacetone in various solvents. At infinite dilution the percentage enol was found to be 73 in acetic acid, 91 in cyclohexane, 58 in pyrrole, and 100 in triethylamine. In diethyl- amine only enol was observed, but keto may be present as an inter- mediate among rapidly interconverting forms, according to Reeves and Schneider.(57). Giessner-Prettre (58) has observed the effect of solvents on the equilibrium in ethyl acetoac etate. In this compound an increase in enol content was noted in carbon tetrachloride, dichloro- ethylene, and benzene. No change in enol content was observed in either chloroform or pyridine. In pyrrole a decrease in the percentage of enol was found. Kinetics and Thermodynamics In a study of equilibrium constants over a range of solvents of different polarities, Kosower (59) has plotted graphically log Keq versus Z, an empirical measure of solvent polarity. The source of equilibrium constant values is not given, but the figure shows that this basis of com- paring the effect of various solvents on the equilibrium has some merit. 37 Equilibrium Constants and Thermodynamic Quantities. Equi- librium constants for enolization have been determined by Meyer (1), Kabachnik (9), and Hammondet a_1.; .(17). The present writer has calculated constants from the data of Conant and Thompson (55). For the purposes of comparison, equilibrium constants have also been calculated from nuclear magnetic resonance data of Reeves (23), Jarrett _et 11. (69), Giessner-Prettre (58), and Filler and Naqvi (18). These data are given in Table XVI. Free energies of enolization have also been determined by Conant and Thompson (55) using indirect bromine titration. Measure- ments in the gas phase were made by noting the effect of a change of temperature on the composition of the distillate, but only over a short temperature range. Determinations were also made in the pure liquid and in hexane and alcohol solutions, and results are seen in Table XVII. Dewar (60) has stated that, other things being equal, a ketone is more stable than the corresponding enol because the sum of bond energies in H-C-C=O is greater than that in C=C-O-H. Powling and Bernstein (36) estimate that the open chain end is about 18 kcal less stable than the diketone from bond energies. The present author has used Pauling's (61) bond energies to obtain a difference of 15 kcal/mole. In (B-diketones where there is the possibility of intramolecular hydrogen bonding, one must consider the additional energy obtained from this hydrogen bond. Wheland (62) estimates that the intramolecular hydrogen bond of acetylacetone stabilizes the end tautomer by 5-10 kcal, and the conjugated system further stabilizes this tautomer by 2-3 Real. Powling and Bernstein (36) have determined the heats of tauto- merization of acetylacetone in dilute solutions in decalin, tetrachloro- ethylene and bromoform, and in the gas phase. The optical densities of the enol and keto carbonyl stretching frequency bands were measured to obtain the concentrations of the tautomers. Measurements were made 38 >:< Table XVI. Equilibrium Constants for B-Dicarbonyls ** Ke Pure 0. 1M 0. 1M Compound Gas Liquid Hexane Alcohol Acetylacetone 11 3.6 11.1 4.88 4. 423 5. 6669 Ethyl acetoacetate 0.93 0.082 0.96 0.111 0. 075358 Ethyl a-g-butylacetoacetate 0.163 0.0650 0.111 0.0650 Ethyl a- ethylacetoac etate 0. 11 O. 031 0. 170 0. 0363 Ethyl e-methylac etoacetate 0.16 0. 043 0.131 0. 0537 Ethyl a-i/sgpropylac eto— 0. 066 0. 051 0 . 0638 - acetate Ethyl a-g—propylaceto- 0.15 0. 075 0.168 - acetate Ethyl trifluoroac etoacetate — 8. 0918 - — Methyl acetoacetate 1. 19 0. 062 - - 3-Methy1ac etylac etone 0 . 79 0 . 44 l. 44 0 . 460 >5: Calculated for purposes of this thesis from data in Conant and Thompson (55), and calculated from other references given as super- scripts. >l<>t< Ke = enol/keto. 39 * Table XVII. Free Energies of Tautomerization for fi—Dicarbonyls 0 AF25 C kcallmole Pure 0. 1M 0. 1M Compound Gas Liquid Hexane Alcohol Acetylacetone -1.3 -0.7 -1.3 -1.0 -1.4, -1. 83" Ethyl ac etoacetate 0. 1 1. 5 0 1 . 2 +0. 2-036 Ethyl u-fnJ-butylaceto- 1. 1 1. 6 1. 3 l. 6 acetate Ethyl a-ethylacetoacetate 1. 3 2. 1 1. 1 2. 0 Ethyl a-methylaceto- 1. 1 1. 9 1. 1 1. 7 acetate Ethyl a-i/ggpropylac eto- 1. 6 1. 8 1. 6 - acetate Ethyl a-An-propylaceto- 1. 1 1. 5 l. 1 - acetate Methyl acetoacetate -0. 1 1. 6 — - 3—Methylac etylacetone 0. 1 0. 5 -0. 2 0. 4 1-Phenyl- 1, 3-butanedione 2. 13'6 — - __ From Conant and Thomson by bromine titration at 259C, unless specified by superscript, in which case values have been obtained from Powling and Bernstein (36) by extrapolation of figures to infinite dilution. 40 in the gas phase at 5 mm pressure, and in 0.1 M solution, which was assumed to be infinite dilution. Temperature ranges were 25--200O and 25-1000 for the gas and solution determinations, respectively. It may be shown that the heat of isomerization in the gas phase is related to that in dilute solution by the equation, - 1 “gas = “.01. +15???) "£78m (H1- H3" (81 where e is the solvent dielectric constant, p the density of the solvent, M the molecular weight of the solvent, [.11 and Hz the dipole moments of the less stable and more stable isomer of the solute, respectively. A plot of AHS against the solvent property will give a straight line with a slope ezliihl to "(HI - (.3) and the intercept AHgas' Data have been plotted graphically for acetylacetone from measurements made by Powling and Bernstein, and the extrapolated value for AHgas' -2.4 kcal/mole, agrees very well with that found experimentally. Literature values have supplied data for a plot of ethyl acetoacetate from which the extrapolated AHgas is -2. 0 kcal/mole. Literature values of AF were plotted versus the solvent property. AH and AF values, interpolated for the purposes of this thesis, are given in Tables XII and XIII, respectively, for 0.1M solutions in various solvents. From nuclear magnetic resonance measurements of acetylacetone at various temperatures, Reeves (23) has found that increasing the temperature favors the keto form. By plotting log(c1/cz) versus l/TOK, Reeves has determined the enthalpy difference between tautomers as -2. 7 :1: 0.1 kcal/mole. Rates of Enolization. Reid and Calvin (63) have studied the rate of enolization in water for compounds with a trifluoromethyl group. These workers have found that the rate of formation of the enolate ion is the rate determining step in the reaction, 41 d [ketol dt = - ke [keto] . (9) Rate constants for enolization and AH . , for enolization are activation shown in Table XVIII. Also included in the table are values of AFfs, A535, and AH0 for enolization. Bankowska (43) has studied the enolization of ethyl o. - and 'y- chloroacetoacetates. Following distillation, ethyl u-chloroacetoacetate reached the equilibrium mixture in alcohol and ether solution within a few days. In nonpolar solvents the enol concentration changed only slightly over several weeks, but pyridine catalyzed enolization in these solvents and also changed the position of equilibrium. Infrared spectra indicated the presence only of _c_13 enol, in contrast to the work of Shigorin (28). Infrared and ultraviolet data are given in Tables VI- VII and VIII, respectively. The enol content of ethyl a-chloroacetoacetate was measured bromometrically by Bankowska (64). A difference in the attainment of equilibrium following distillation was noted in various solvents. In polar solvents equilibrium was attained quickly, but in nonpolar solvents, the rate of approaching equilibrium was slow and increased with the addition of pyridine. Only Sis: enol appeared to be present both from infrared and ultraviolet spectra. The slow rate of tautomerization in nonpolar solvents has been explained by a weakening of the basicity of the a-chloroester from the inductive effect of the chlorine atom. Rates of Ionization and Ionization Constants. Pearson and Mills (65) have determined the rate of ionization of B-dicarbonyls in water by measuring bromine uptake conductimetrically. The reactions are H20 + HA 5L» H3O+ + A’ (10a) k' 2 Br; + A- 1(3—> ABr + Br-, (10b) 42 .mucmHmsoo GOSMUOmmflo no“ maoquHOm occuox msooscm mo cofimfifi EOmZ magma .33 :3de was 30% gosh .HouoxH—\:ocou_ u 6M . E coflmsvm semison ** .w v- n: wJ- ow a; 03 .o osouson. .: mm @000 .o odouson A... m: v.N+ mao.o H.®H+ omoo.o w; modo.o nouns? oac .o Hogo . 0 mm Nooo .o nouns? econoomososfifiuu 395:8 u .. u u «A wooo .o unencum— : mm N010 ocouson 5: 0m. m.N+ mmo.o 0.2+ Nwmod u: Nwood .3595. mac . o . mmwo . 0 mm moco . 0 Head? ocouoomonodzflnu aoucom Fm a; mmwo :0 0.8583 $.01 v... 1N- mm mm mmwod oconcon 53+ good o 38.0 333 m . H ... N. .0... m .1. :o .0 «file .0 mm wmwo .o .533 occuoomguoomononflfinfi £05 .mov 305 205 \Hmox \Hmox \fimox Hommox n..oem ... m». o o 034 “me. or: ......m mq is oh 3:202 Evian 828880 mfienconumowflun poumcfisOSHh new cofimufiuofiounmh new muamumcoD UMEMC>UOEH0£H was COSMNSOGH mo men—mm .H:>K “:an ......" 43 where Equation (10a) is the rate determining step, and the ionization constant is given by the following relationship: k1 K 3 — e Kion ( k2 )° (11) The rates of ionization were correlated with ionization constants, and these constants are shown in Table XIX. Rates of ionization are given in Table XX. The authors have calculated the changes in enthalpy and entropy from transition state theory for the forward and reverse ionization reactions for compounds where activation energies for ioniza- tion and heats of ionization are known (see Table XXI). Pearson and Dillon (66) have extended the study of ionization rates of pi-dicarbonyls in water. Ionization constants are shown in Table XIX and rates of ionization in Table XXI. A plot of log [rate constant] for ionization versus log [equilibrium constant] is approximately linear. Alkyl groups on the central carbon produce negative deviations, or have rates less than expected, and a fluorine atom on the central carbon has the opposite result. Gould (67) has noted that the rate of enolization seems to be inde- pendent of acidity. Those substances which show more electron- withdrawing power by induction are, in general, more effective in stabilizing the anion than the activated complex leading to it. The influence of branching of the chain on enolization and on the acidity of alkyl derivatives of acetylacetone has been studied by Rumpf and La Riviere (16). Acidity constants have been determined electro- metrically, spectrophotometrically, and iodometrically. The acidity constant has been defined as K=K - 1+E+ E 12 Fe Papp10g( )1og. () where pKe is the pK of the enol tautomer, pKapp is the observed pK, .N! 44 .o>oo.m amt/Mm mdofimdvo GM mucmumcou mdoflumkr mo Gofifiaflov 90w «use mom "w ooooom com: 0:30 03.30 UMOHOSGCH ooow occuoomonosHfiSEOCosfi 8. a 3 4 8233.3; J -135“; Am . o 2 ma . o macaw omguoomlfiozu m N . 2 oomooomofiwom T332 Hoooo .0 V oomuoomouvoomgmonmndeIa HELEN omoo .o momooomgoomasuoguo 15mm moooo . o 3.30 omega ongodnwmud TEKH m A 00 . o 3 oo . o oomuoomOooomTEHo .6 15M on: 330 omouo omogounud TCfiH mad 3 .o :20 mood oumooomouoom TEHH Hmo .o ocmsuoagouaonwfl .. - $06 3 mam 3 5: 3; 3333334 a: a: a: a: a: s: :3 $3 dam vasomfioo we : ma oo .3 co me on .mom A33 5 353930 coflmuficoH * mataconumowflua new mucmumcoU GOSMNMGOH .Xan 3an 45 Table XX. Rates of Ionization of Pseudo Acids at 250C in Water k1(min'l) Compound Pearson and Mills (65) Pearson and Dillon (66) Ethyl acetoacetate 1.29x10'3 7.2x10'z Ethyl -a-bromoacetoac etate 1. 59x10'2 3 . 6x10"l Ethyl a- ethylacetoacetate 4. 53x10“ 4. 5x10" Ac etylac etone 6 . 8x10'1 1. 0 3-Methy1acety1ac etone 4. 97x10'3 5:»:10'3 Benzoylac etone - 6 . 6x10'l Trifluoroac etylac etone - 9 . OxlO'l Thenoyl trifluoroac etone - 6 . 0x10"l 3' Table XXI. F Enthalpy and Entropy of Ionization of B-Dicarbonyls in Water a: 3.. 3*. 1? AR; AS 2 Compound Kc a1 / mole e. u. Ethyl ac etoac etate l 3 . 6 - 34 . 4 Ethyl a- ethylacetoac etate .14 . 9 - 31 . 6 a-Methyl ac etylac etone 17 . 4 - 16 * From Pearson and Mills (65). 46 and E is the ratio of enol to keto tautomer. ~ For this thesis, the pKe values have been converted to Ke and are given in Table XIX. From the data it is seen that a-substitution results in an increase in pKe, a decrease in E, and apparently prevents coplanarity not only of the enol, but also of the anion. This is an example of the steric inhibition of conjugation. In a study of the acidity of fi-dicarbonyls in aqueous solution, Rumpf and Reynaud (68) have concluded that the influence of alkyl sub- stituents on ionization constants is comparable in acetylacetone and ethyl acetoacetate. Ionization constants for some B-ketoesters have been determined, and results are given in Table XIX. PKe Values have again been converted for the purposes of this thesis to K,3 values. The potentiometric determination of dissociation constants for both acetylacetone and ethyl acetoacetate has been reported by Walisch e_t a_._l. (9). The empirical dissociation constant has been defined as [H+][A'] Kb = [HA] (13) where [H] and [A] are activities, and [HA] is the total [B-dicarbonyl content. Values for the constants are given in Table XIX. Hammond (15, 17) has considered concentration acidity constants, Q , defined the same as K; in Equation (13), but using concentrations instead of activities. It may be noted that values for PKa may be obtained by adding 1.55 to -log QA' For the purposes of this thesis the reported value 3 of -log QA have been converted to values of DAL-K6 and are given in Table XIX. Hammond observed that acetylacetone is a stronger acid than dibenzoylmethane, although the latter is more enolized. Hammond has defined acidity as the ability toform anions. Steric strain can be important in producing variations in acidity. For the enolate ion structure shown in Figure 7, a large 3 group introduces 47 Figure 7. Enolate ion structure. strain in the planar trans configuration which should ordinarily be the most stable form. For dibenzoylmethane the tra___n_s configuration of the ion can only be constructed with models if the benzene group is perpendicular to the plane of the dicarbonyl system. Calvin and Wilson (70) have determined the constant, KD, for several fl-dicarbonyl compounds. This constant is defined as _ [H+][K;] .. 1 KB [HKé] ’ ( 4) where KD represents the acid strength of the enol for the reaction c - o\ /c - o' \c/ I} <——> \c + H+. (15) / \ / / \ C = O/ C = 0 Values forthe ionization constants are given in Table XIX. These values indicate that the anion of acetylac etone is more stable than that of ethyl acetoacetate. As noted earlier, Bellamy and Beecher (32) concluded from the work of Calvin and Wilson that the equilibrium constant is independent of the percentage double-bond character of the carbonyl bond. 48 Nuclear Magnetic Resonance Studies The initial application of nuclear magnetic resonance techniques to [S-dicarbonyls was by Jarrett, Sadler, and Shoolery (71). For both acetylacetone and 3-methylacetylacetone the 30 Mc proton spectra allowed integration of enol versus keto peaks for the calculation of the percentage of each tautomer. Jarrett e_t a_tl. , had assigned all acetyl methyl protons to one peak and CI-_-I_3C(OH)= methyls to another. Assign- ment to the methyl absorption peaks was corrected by Reeves (23) using a 40 Mc instrument at variable temperatures. Reeves showed that both methyls are equivalent in the enol tautomer of acetylacetone, and that the two methyl resonance peaks are due to keto and enol tautomers. Solvent studies of acetylacetone were performed by Reeves (23). Chemical shift measurements were made in pyrrole, cyclohexane, triethylamine, and acetic acid. Results are 'given in Table XXII. Their reported values have been converted to cps at a spectrometer frequency of 60 Mc and are referred to tetramethylsilane as a standard for purposes of this thesis. The use of a methyl group as an internal standard, as was done by Reeves, is advisable only with fairly inert solvents, such as cyclohexane. From chemical shift measurements both acetic acid and cyclohexane appear to be inert solvents toward acetylacetone in the sense that they do not change the chemical shifts of the various protons. In acetic acid the OH signal of the solute was broadened, but remained a separate resonance from that of the carboxyl OH group throughout the dilution range at room temperature. In pyrrole the OH peak was once again broadened. Here the keto u-methylene protons shifted upfield by 28 cps on dilution, while the enol OH protons shifted downfield by 33 cps. Triethylamine showed only the enol tautomer and gave a very large upfield shift of the enol 49 2: Table XXII. Infinite Dilution Chemical Shifts of the Protons of Ac etyl- acetone in Various Solvents Proton Resonance Peak Solvent Enol Keto Keto Enol Enol CH3** CH3 CH; CH: OH Pure 119 129 218 334 933 Pyrrole 119 124 190 333 966 Cyclohexane 119 127 212 331 935 Triethylamine 119 - - 334 455 Acetic acid 119 130 218 334 919 a: From Reeves (23), changed to cps at 60 Mc relative to tetramethyl- silane. ** Used as an internal standard. 50 OH by 478 cps. This large shift was explained by Reeves as due to both breaking of the intramolecular hydrogen bond of acetylacetone and the formation of a hydrogen bond with the solvent. Both the enol OH and the enol (it-proton resonances were broadened in this solvent. Proton transfer in acetylacetone in diethylamine and triethyl- amine has been observed by Reeves and Schneider (57). Solutions in triethylamine are completely enolic. Amines not only change the keto- enol equilibrium, but also increase the keto—enol conversion rate. With diethylamine, the rate of exchange is sufficiently rapid to result in an averaged spectrum. Assignment of the OH, NH, and enol a-proton peaks was made from temperature studies. Broadening of the above peaks has indicated that there is exchange among all of these protons. The following exchange processes are proposed to explain the behavior in diethylamine: OH...N ('9 O”°H-N (16a) H ' C E g H " c-H° 0 ON H "H-N l (16b) C C enol keto H-CHI {-9- H-Cll-H”°O (16c) c "'H§o c enol keto Proton exchange in hydrogen-bonded systems has been reviewed by Schneider and Reeves (72). A study of exchange in acetylacetone and acetic acid mixtures was made. The subject of proton transfer in such systems is covered in the theoretical section of this thesis (page 78). Solvent studies of acetylacetone in pyridine, water, and mixtures of pyridine and water have been conducted by Balta Calleja (73). 51 Measurements were made at 25 Mc relative to the enolic methyl protons as an internal standard. Only the enol OH proton showed a change in chemical shift in these solvents. With pyridine alone, no change was observed, but both with water and in mixtures, the enol OH proton showed a high-field shift. This shift has been explained on the basis of dissociation, but solubility limited the study of mole fractions above 0.6 M. Enolized B-triketones have been extensively investigated by nuclear magnetic resonance by Forsen and Nilsson. In a study of triw acetylmethane, 2-acy1cyclohexane-1, 3-diones, and ceroptene (see Figure 8a), chemical shifts for the enol OH protons varied from 1040 cps (downfield from TMS) in triacetylmethane to 1130 cps in ceroptene. Two enolic tautomers are possible, and the NMR spectrum indicated a ratio of 1/4, but no attempt was made to distinguish between these forms. The presence of more than one enolic tautomer was detected in several compounds. Equilibrium constants were dependent on the solvent and are in the range 3-8, corresponding to a free energy change on enolization of less than 1 kcal/mole. The tautomeric forms are evidently similar, with hydrogen bonds of about ‘equal strength. Infrared carbonyl stretching frequencies in the chelated forms were measured, and a linear correlation was observed between these frequencies and the chemical shift of the enolic proton. A lower stretching frequency was accompanied by a proton resonance at lower applied fields. Forsen and Nilsson (75) further found complete enolization in methyl and ethyl diacetoacetate. The enol proton chemical shifts in these compounds occurred at lower applied fields than in triacetyl- methane. Both ethyl cinnamoylacetoacetate (see Figure 8b), and 3-cinnamoy1pentane-2, 4-dione indicated the presence of two enolic tautomers. The latter exhibited two separate signals for the enol 52 If! \ 5 _ I / " I) // CH3. Ph-CHaCH-C-C\ g {—9 Ph-CH8CH-C—C (f: . = 0 g) CO-CHzCH-Ph 67% CH; 33% Me' (a) ceroptene (b) ethyl cinnamoylacetoacetate (c) dehydroacetic acid (d) usnic acid H ,H /” \ l’ \ O O O O 3 a --> g; a; \, Me/ \C/ -Me Me \CIS/ \H 80% CHO 20% COMe (e) di ac etoac etaldehyde /H H\ o’ \o o/ ‘ II I ......> I II c c c c / \ / \ / \ / \ CH, (I: H CH cl: OEt 90% COOEt 10% ‘ CHO (f) ethyl 2 - ac etyl - 2 - for mylac etate Figure 8. Structures of some p-triketones. 53 OH protons of these two tautomers, whereas in other B-triketones the lifetime in the two sites has been too short for detection by nuclear magnetic resonance. This separation of signals may also be explained by the ring current effect on the protons in different locations. Both infrared and NMR evidence indicate that dehydroacetic acid (see Figure 8c) exists entirely in one enolic form (76). The intramole- cular hydrogen bond in this compound was found to be weaker than those in other B-triketones studied, both from the chelate carbonyl frequency and the enol proton chemical shift measurements. The authors have suggested that the chelate system may be affected by the nature of the 4-5 double bond in the compound. Additional conjugation appears to give a stronger intramolecular hydrogen bond. In usnic acid compounds (see Figure 8d), the enolic hydrogen bond appeared to be comparable in strength to that in ceroptene (77). Hydrogen bonding is possible at positions x, y, and z in the figure. Only one enolic tautomer is seen in the NMR spectrum. Neither infra- red nor nuclear magnetic resonance were conclusive in determining which tautomer was present. Nilsson (78) has extended the original correlation figure (74) between enol proton chemical shift and chelate carbonyl stretching fre- quency. Nils son pointed out that, in a comparison of triacetylmethane and diacetoacetic esters and their conjugated analogues, the esters show stronger hydrogen bonds. Corrections Were not made for magnetic anistropy effects. Whether this strengthening of the hydrogen bond is mostly due to inductive rather than resonance effects is not known. The effect of conjugation on such systems varies with the compound. The presence of an adjacent aromatic group appears to weaken the intramolecular hydrogen bond, whereas a non-aromatic ring seems to strengthen this bond. However, there are discrepancies between infra- red and nuclear magnetic resonance results among such compounds. 54 Spin coupling effects have been observed by Forsen and Nilsson (79) in hydroxymethylene- and anilinomethylene compounds, which have possible structures as shown in Figure 9. ,He H\ o 0’ \o o/ “0 £1: II II I II c c c c \ R/ \c/ \H — R/ c/ \H — R/ \c/ \R / \ I a I H R' R' R' Figure 9. Keto and enol structures for hydroymethylene compounds. These compounds are almost entirely enolized, and in many cases more than one enolic form is observed by nuclear magnetic resonance. Spin coupling was seen between the enol OH proton and the aldehydic proton. This coupling varied from about 6 cps in hydroxymethylene compounds to 12.5 cps in ethyl 2-formyl-2-pheny1acetate. Spin coupling generally decreased with increased temperature and varied with concentration. For diacetoacetaldehyde two of the possible chelated enolic forms were seen in the nuclear magnetic resonance spectrum. The dominant form was the chelated 3-formyl-4-hydroxypent-3-en-2-one seen in Figure Be. In ethyl 2-acetyl-Z—formylacetate, two of six possible chelated enols was present. Spin- spin coupling constants and infrared spectra make possible the identification of these enols, shown in Figure 8f. A review of nuclear magnetic resonance studies of hydrogen bond- ing in B-triketone systems has been written by Forsen (80). In ethyl acetoacetate, Giessner-Prettre has shown by nuclear magnetic resonance studies at 25 Mc that carbon tetrachloride, benzene, and tetrachloroethylene increased enolization, whereas pyrrole de- creased enolization on dilution. In pyridine and chloroform, no change 55 in the amount of enol tautomer was observed. In carbon tetrachloride,- benzene, and tetrachloroethylene, the keto CH2 resonance peak was displaced toward higher fields on dilution. This was interpreted as due to a breaking up of intermolecular associations between CH2 and C=C. Basic solvents such as triethylamine and piperidine did not favor enolization, in contrast to the effect on acetylacetone observed by Reeves. Giessner-Prettre was unable to observe the OH resonance in binary solutions of ethyl acetoacetate with various solvents. There was approximately 50 percent enol at infinite dilution in carbon tetrachloride and tetrachloroethylene. Dudek and Holm (81) have noted a decreased shielding of the enolic OH proton of acetylacetone in benzene and have suggested that there is some hydrogen bonding of the carbonyl group of the solute with the solvent. For k£s(acetylacetone)-ethylenediimine there was a shift to high field of certain protons. The authors have explained this shift on the basis of the orientation of these protons above the plane of the benzene ring. 'y-Fluoro-B-ketoesters have been studied by Filler and Naqvi (18). The percentage of enol was determined, and an attempt was made to explain these percentages on the basis of the electron-withdrawing power of fluorine. Theor etic a1 Studie s Forsen (82) has used a simple molecular orbital approach to calculate the effect of substituents on the carbonyl stretching frequency. He has calculated bond orders for the carbonyl bond and has neglected resonance and overlap integrals. A linear relationship was found between bond order and stretching frequency, and it was concluded that the inductive effects determine the stretching frequency in these com- pounds for the most part. A correlation was also found between the 56 charge on the oxygen atom (and the dipole moment) and the stretching frequency. Molecular orbital calculations of the Hiickel type have been carried out by Forsen (83) for some enolized di- and tri—ketones. _ For the keto and the various enol tautomers the total Tr-electron energy and the delocalization energy have been calculated. There appears to be a general correlation between the calculated difference in delocali- . zation energy and the position of the keto-enol equilibrium, since smaller differences in delocalization energy accompany a higher enol content. Some results have been reproduced in Table XXIII. However, Forsen has pointed out that the molecular orbital calculations involve many uncertainties, and that entropy differences, energies of intra- molecular hydrogen bonds, and solvent effects have been neglected. Forsen has extended the molecular orbital calculations to an estimation of charge distributions and bond orders in these compounds (84) (see Table XXIII). No definite correlation was found between the enol proton chemical shift and the calculated charge density on the carbonyl oxygen. Neither was good correlation discovered between the calculated bond order of the C=C bond in the chelate ring and the stretching frequency of the chelated carbonyl group. For sen has suggested that the lack of correlation may be due to the fact that the Hiickel method may not give an adequate description of relative charges in the electronic distribution among molecules. In a-formyl ketones such calculations were used to predict which of the enolic tautomers pictures below is most stable (85). a“? $3 ‘B‘H'fi’ t1“? H-c-c|:-c-CH3 H-czc—c-CH, H-C-(llzc-CI-I3 R, R, R, (a) (b) (C) 57 >k Table XXIII. Molecular Orbital Calculations for 8~Di and 8oTri-Ketones Parameter Set I a; c J, H '0 8 c, o 3% 33 g .5 *3 '9. SM 0 g s: x- 2 *7 . $3 8 O 11] E H £11 3 8% 54 ‘5‘ "o 9' v ' :1 9' m c: as ‘1’ o O E t3 >\ m >\ Q) N QB > c: 3'} o ... a” o :0 3;: o H o ,_, u m 0 rd 3H 0 u q) 5 +4 Q) .—4 Q) :1: ...: O 0'} C .M m o c: o c: ... 8 a, c .0 0 U) E“I E-* M Q £11 £1 ... .oI-IJ O U 0 0 ll *3 II 1 -C-C-C k 40. + 9.0608 6.2608 0.5548 76.4 e 60. + 13.1068 5.7068 0 O n Iii II 2 -C-C-C- k 6a + 11. 0608 8. 2608 0.1268 95 C=C e 80. + 15.5348 8.1348 0 O n 171 u 3 -C-C-C- k 60 + 13.5908 9.39C8 0.1128 _ C=C e 80. + 18.0788 9.2788 0 O H I? u 4 >C=C -C-C-C- k 60 + 11.5168 8.7168 - (‘3-H~—O ll -C=C-C=C-C- e 8e 15.6128 8.2128 0.5048 7“"? —C=C-C-C=C- e 8d 15. 5268 8.1268 0. 590.8 >'tFrom Forsen and Nilsson (75, 76) o o ** II 1.4 u R,-—C-C-C-R3 R2 58 nowfinsmsfi : AI. : “Confirm mofiwuocm Goflmfioxfl 0.1.0, HopHO paom Ono 3225 HopnO pcom 38de couuooflw ‘5?wa USOGH RSBGMQ cokuovfiw OHU oumaonU $63 mo fizz ,/ 1.527 1.968 0.777 0. 863 1.5198 -5.4 1608 1.961 0.792 0.766 1.4808 -7.4 1.518 1580 59 Several sets of parameters were tried, but the results of these calcu- lations are inconclusive . Stabilities of Metal Chelates A brief discussion of metal chelates is included because of their interest in the separation of metal ions and because of possible corre- lations of stability with the 8 -dicarbony1s themselves. Calvin and Wilson (70) studied the stability of chelate compounds by titrating H+ in a solution containing Cu(II), 8-dicarbonyl, and excess H+. Factors of importance in determining the stability of the chelate include not only coulombic effects but also resonance in the enolate ion. For increased resonance in the chelated form of the ion, increased stability of the chelate is expected. Holtzclaw e_t .31. (33), have studied the infrared absorption of metal chelate compounds of 1, 3-diketones. Stability was established by polarographic reduction of copper chelates. In the chelate ring of the enolic form, the basic effect of oxygen donor atoms is responsible for bonding. Stability is also determined by resonance in the chelate ring and by steric hindrance. Trifluoromethyl groups make the oxygen less basic, and the trifluoroacetylacetone chelate is between acetyl- acetone and hexafluoroacetylacetone in stability. The Z-thenoyltrifluoro- acetone chelate is less stable than that of trifluoroacetylacetone since the thenoyl group is electron attracting and also interferes with resonance in the chelate ring. The order of increasing stability of metal. chelates is hexafluoroacetylacetone > 2-thenoyltrifluoroacetone > trifluoroacetylacetone > 1-pheny1-1, 3-butanedione > 2, 4-pentanedione > 3-methylpentanedione. Infrared frequencies of interest are given in Table XXIV. Nakamoto St a_._l. (86), have studied the infrared spectra of metal chelate compounds and have determined the effect of substituents on 60 >:< Table XXIV. Infrared Frequencies of Groups in 8-Diketones and Their Metal Chelates W Absorption Band (cm-l) Compound Hydroxyl Keto Chelate .- Perturbed OH C=O C=O C=O C=C Acetylacetone 3460w 1720s 16185 338 0w 1700s ** Cd(II)-acety1acetone 1613s 1518s 3-Chloroacety1acetone 1725m, b 16135, b Cu(II)-3-chloroacetyl- 15715 15205 acetone Dibenzoylmethane 1595 s , b Cu(II) - dibenzoylmethane 1552 s 15 3 5 s Cu(II)-hexafluoroacetyl- 16155 1541a acetone Cu(II)—methylacetylacetone 15805 1530w, 1515w 1-Phenyl-1, 3-butanedione 16045, b Cu(II)-l-phenyl-l, 3-butane- 15615 15255 dione Thenoyltrifluoroacetone 16405, b Cu(II)-thenoyltrifluoro- acetone 1570 s - Cu(II)-trifluoroacetyl- 15985 15385 acetone a): From Library of Congress, ADI 5129, referred to in Holtzclaw and Collman (33). am _ Dependent upon concentration. w Weak 8Strong b Broad 61 the spectra of metal acetylacetonates. Results for the copper chelates of hexafluoroacetylacetone, trifluoroacetylacetone, benzoyl- acetone, and dibenzoylmethane are given in Table XXV. Force con- stants have been calculated and are included in the table. The effect of the trifluoromethyl group is to increase the C=C and C=O bond distances and decrease the C-R and M-0 (M=metal) bonds. Thus the strong inductive effect of the trifluoromethyl group strengthens the C=C and C=O bonds and weakens the M-0 bonds. Phenyl substitution could lead to electron release and a strengthening of M-O bonds by an increase in the negative charge on oxygen. This effect strengthens the C-C and M-0 bonds more than the C-0. bond of the ring. These results differ from those of Holtzclaw, e_et a_LI. (33), who said that the phenyl group weakens the M-0 bond, since the C=O bond is weakened by conju- gation with the ring. Stability constants agree with those found by Holtzclaw e_t a_1. 62 Table XXV. Infrared Frequencies and Force Constants for Groups in COpper Chelates of 8-Diketone5 Frequency Force Constant Ligand cm" 105 cyne/cm C=C C=O C=C C=O Hexafluoroacetylacetone 1644 1614 5. 81 7. 90 Trifluoroac etylac etone 1611 1600 Acetylacetone 1580 1548 5. 35 6. 90 1524 Benzoylacetone 1590 1554 Dibenzoylmethane 1593 1544 5. 49 6. 82 a: From Nakamoto gt a_1. (86). THEORET ICAL C ONSIDERATIONS Nuclear Magnetic Moments All nuclei with non—zero nuclear spin may be studied by nuclear magnetic resonance techniques. For a nuclear spin of 1/2, as in protons, the nuclear magnetic quantum number _1‘2 may have the values _1; 1/2. In the absence of a magnetic field, there will be equalpopu- lations of nuclei with m = 1/2 and m = -1/2. In the presence of a magnetic field those nuclei with m = +1/2 are the more favored state, and the distribution of nuclei between states with m = .1; 1/2 may be expressed by the Boltzmann equation (87), N = Ape 'e/kT, (17) where E = -p H H (18) z 111. m H Z‘IT ’ and “H is the component of the (magnetic moment along the field axis, H is the magnetic field, and 'y is the gyromagnetic ratio. Chemic a1 Shift 8 In an applied magnetic field, the field at the nucleus is given by H = Ho(1- a“), (19) where H0 is the applied field and o— is the shielding constant. In liquids where molecular rotation is rapid, the shielding constant may be expressed as (88) 63 64 (011 + 022 + 033). (20) Q l tWIH where on, on," and 033 represent the principal components of the shielding tensor. An expression may be derived for the screening constant for a particular nucleus when the applied field is in the z—direction, as follows (88) Z Z Z Z -3 2 e x + e 9 = r _— °zz 2m62f r3 p d +mzczAE 0 irk 20 20 0 ° (21) j 'k In this equation p is the electron density, Oj is the azimuthal angle for rotation about the z-axis, and AE represents the average electronic excitation energy. The first term is positive and represents the local diamagnetic contribution, and the second is negative so is termed the contribution from local paramagnetic currents. From a practical standpoint, one is able to calculate the screening constant only for small molecules. The shielding constant receives contributions from several sources and has been arbitrarily divided by Buckingham (89), and Buckingham e_t .11. (90), into the following components, o=qb+oa+ow+oE, (22) where (This the contribution from the bulk magnetic susceptibility of the medium, 0a is due to anisotropy in the susceptibility of the solvent molecules, ow is due to van der Waals forces between solute and solvent, and GE is the polar effect. Chemical Shift due to Bulk Susceptibility. The magnetic field experienced by a proton in a cylindrical sample is given by 65 = H 1 - Z 3 2 proton O o( Tr7Cv/ )' ( 3) where H0 is the applied field, 0 is the shielding constant, and 7C v is the volume susceptibility. For the chemical shift between protons of two substances in a liquid mixture, the chemical shift becomes 6 = —J————1— x106, (24) where 03 and 0'i are proton shielding constants for the reference and observed proton, respectively. For the case of Equation (24) the chemical shift is independent of volume susceptibility. Bothner-By and Click. (91) found good agreement between experimental values and those calculated by this equation for mixtures of methylene- chloride and methylenebromide. Bothner-By and Glick (92) also studied the chemical shift of a given proton in media of different susceptibilities using the equation, x 106, (25) 217A = V 6 3 where Afiv is the difference in susceptibility of the media. Agree- ment was only fair when the shape factor 217/3 was used. The error may be explained not only by experimental'error, but also by unre- liability of literature susceptibility values. As a result of these studies, Bothner-By and Glick. suggested the use of the empirical equation for predicting the behavior of regular mixtures, as follows: H = O'.H.O (1 — 2.6 jCV). (26) 1 1 Chemical Shift due to Anisotropy. The volume magnetic suscepti- bility is given by 66 —-> —-> M = IVH, (27) —> where Wis the magnetic moment per unit volume, H the magnetic field, and 7Cv depends on the nature of the material. If )CV>0 or the induced moment is parallel to the magnetic field, then the sub- stance is paramagnetic, and if/Kfo, it is diamagnetic. The suscepti- bility of a mixture is given by (88) XM,mixt= x, X M, + X22942 ’ (28) where x, and x2 are mole fractions, and jCMl and 7CM; are suscept1- bilities of the pure compounds. The diamagnetic susceptibility of a molecule has been expressed by. Pascal (88) in the empirical equation, 7CM= 2794+ x, ‘ (29) where )CA are atomic susceptibilities and )\ are constitutive cor- rections. The susceptibility of a bond may be Written in terms of com— ponents as follows: )Cm =1/3OC1 + 9C2 + 7C3). (30) where 7C1.Z, 3 represent susceptibilities along the three axes of the bond. If the components are unequal, then the bond shows diamagnetic anisotropy. Similarly a molecule may show diamagnetic anisotropy if the components along principal axes differ. It may be shown that the anisotropic contribution to the shielding constant is given by 67 2 0-3 = _n%/(3 cos 0 - l), (31) where n is the number of molecules within the range R, fi/AJ- , are susceptibilities parallel and perpendicular to the molecular axis, 0 is the angle between the direction of principal susceptibility and the line from the molecular axis to the proton under observation, and R is the distance from the axis to the proton. For disc or rod- shaped solvent molecules close to the solute, where 9' is 00 and 900, respectively, the shielding contribution reduces to: °_a(disc) : '21“ 2C” ' 7C¢V3R3 (32) (Ta(rod) = n( 77/- )EV3R3 ° In solvents with disc-like molecules, there is a resultant high-field shift of the solute protons, whereas for rod-like molecules with the largest susceptibility along the molecular axis, there is a low-field shift of the solute protons. Calculations for methane in benzene give 0- a = 1. 3 ppm, and for methane in carbon disulfide, 0' a = -0. 5 ppm. Precise agreement with the observed values of 0. 33 and —0.42 ppm is not expected because of the assumptions made in deriving Equation (31). Bothner-By and (Slick (93) studied aliphatic solutes in aromatic solvents and noted irregular behavior in solvents with large magnetic polarizability anisotropy. For the benzene protons extrapolated to infinite dilution in several solvents, the resonance position is at lower applied fields by 0. 6-0. 7 ppm than that calculated using Equation (25). The aromatic compounds have effective volume susceptibilities smaller by about 0. 25-0. 30 x 10"6 than literature values. These effective susceptibilities when substituted into Equation (25), give a shape factor 217 3 obtained from electromagnetic theory. of 2. 6 instead of the value 68 *Chemical shifts of solutes in benzene move to higher applied fields as infinite dilution is approached, which indicates anomalous diamagnetic shielding by the aromatic ring (93). The size of the shift is irregular and depends on the specific solute. An approximate shift of 0. 12 ppm has been calculated for a proton in the vicinity of a benzene molecule and directly above the ring. This calculation disregards factors such as statistical variation in intramolecular distances, the occurrence of interaction with more than one solvent molecule at a time, and the possibility of a solute molecule lying in the plane of the ring. The large diamagnetic anisotropy of the benzene ring may be explained in terms of the ring current model, shown in Figure 10. Figure 10. Schematic representation of ring—current effect. Protons situated above or below the plane of the ring experience a magnetic field less than that elsewhere, and as a consequence require a greater applied magnetic field for resonance. Protons lying in the plane of the ring will experience a downfield shift by similar reasoning. In solutions concentrated in benzene, there is a greater chance for other benzene molecules to be above and below the plane of the ring than in the plane, with a resulting high field shift of the solute protons. Johnson and Bovey (94) have calculated the magnetic field in the vicinity of the benzene ring and have determined the chemical shift for 69 a proton in any position relative to an aromatic ring. A plot is given for the diamagnetic and paramagnetic shift as a function of distance from the ring. Reeves and Schneider (95) have calculated the distance of chloroform molecules from the plane of the benzene ring for assumed complexes of the chloroform molecule with benzene. For an observed shift of 1. 25 ppm in benzene, the chloroform proton may be calculated to be 3. 43 A above the plane of the ring. The chemical shift in ppm at the proton is given by: 3e2a2 6 =mczr3 ' (33) where e is the charge on the electron, a is the radius of the benzene ring, In the mass of the electron, c the velocity of light, and r the distance of the proton from the center of the ring. Hatton and Schneider (96) have compared the behavior of polar molecules in toluene and methyl cyclohexane. It was suggested that the difference in solute chemical shift between solutions in these solvents made it possible to neglect the chemical shift due to van der Waals interactions and due to the reaction field since these would be the same. The observed chemical shift of the polar compound may then be explained by complex formation. The authors have proposed a complex in which the more positive end of the molecule is oriented above the plane of the ring, as shown in Figure 11. :N Figure 11. Orientation of a polar molecule with respect to the benzene ring. Schneider (97) has investigated the effect of the shape of the solute molecule in benzene solutions by comparing chemical shifts for rod-like 70 and planar solute molecules. The results eliminate the possibility of a shape effect and support the proposed benzene-polar solute complex. Stephen (98) has used statistical mechanics to study the effect of molecular interaction by treating molecules as point dipoles. For solvents whose molecules show a diamagnetic anisotropy two effects appear to be important. Magnetic moments induced in a solvent mole- _ cule produce a magnetic field at neighboring solute molecules. This effect is particularly important in aromatic compounds, and the contribution to the shielding constant from the magnetic anisotropy effect has been calculated to be 0. 70x10"6 in aromatic solutions. The second effect is a change in the electronic distribution of a molecule by an induced electric field in a nearby molecule. This effect is most important in polar molecules, and its contribution to the shielding constant is larger and of opposite sign than the first effect. For water, this contribution is -4. 5x10'6. Chemical Shift due to van der Waals Forces. Bothner-By (99) has calculated the shielding correction for a randomly oriented field due to London disperson forces as: A.:-wwa (M) where E is the electric field. For nonpolar liquids A0- is -0. 1 ppm. Bothner-By has determined the-'liquid association shift' for several organic solvents. The term 'liquid association shift' was first intro— duced by Schneider, Bernstein, and Pople (100) in a study of gaseous and liquid hydride molecules. If the shift from a gas to a liquid is given accurately by the bulk susceptibility correction, then 211’ ’6 6 (caIC) 6g 3 7CV x O ' (35) 71 The liquid association shift, 8:, is defined as the difference between the observed shift and the calculated shift, as follows: fii : 62 - 63(ca1c)’ (36) For neopentane, the liquid association shift has been calculated as -0. 16 ppm. All 82 values are negative, so the "liquid association" term leads to a doWn-field shift as expected from an effect which is primarily due to van der Waals interactions. Bothner-By has further suggested that the magnitudes of the anomalous shift may be calculated empirically using the equation .1 'B. = X Yd . (37) where x1 and y‘j are characteristic numbers as signed to the solute and solvent proton, respectively. The solvent parameter, ,yj, is a measure ofthe electric field in the neighborhood of the solvent mole- cule which arises from instantaneous and from permanent moments in the molecule. The solute parameter, xi, is the product of two factors. On the one hand is involved the accessibility of the solute proton to solvent molecules because of orientation of anisotropic solvent mole- cules. The second factor involves the sensitivity of the shielding of the solute protons to the surrounding electric field. The latter will be affected by the nature of the H-X bond and the ease of polarizability of the bond in the solute molecule. Buckingham' e_t a_1. (101), have chosen a nonpolar solute, methane, for study in nonassociated solvents relative to gaseous methane. Corrections have been applied for bulk diamagnetic susceptibility, and the shift should then be proportional to van der Waals forces and to the magnetic anisotropy of the solvent molecule. 72 Van der Waals forces involve interactions between permanent dipoles, between permanent and induced dipoles, and between neutral atoms or molecules. Qualitatively van der Waals forces for the solvent molecule involve two effects: (a) interaction between solute and solvent molecules in equilibrium configuration, resulting in electrons of the solute molecule being attracted by nuclei of surround- ing solvent molecules to produce a low-field shift of solute protons, and (b) destruction of axial symmetry of the solute by departures from equilibrium solvent configurations. The contribution to the shielding constant due to van der Waals forces is negative, and thus the chemical shifts are low-field shifts. Since the energy due to van der Waals forces cannot be evaluated directly, Buckingham gt a_._l. (90), have used an indirect measurement of molecular interaction and have compared the chemical shift of solute protons with the heat of vaporization of the solvent at the boiling point. A linear relationship was found for solvents in which only van der Waals effects are operating. For these solvents the expected low-field shift is observed with increasing interaction with the solvent as estimated from thermal data. Various solvents give anomalous shifts, and these may be explained on the basis of magnetic anisotropy. Chemical Shift due to Polar Effects. The magnetic screening of the nucleus by electrons is reduced in all directions by a uniform electric field (102). According to Marshall and Pople, this reduction is only partly due to the paramagnetic term of Equation (21), since the diamagnetic Lamb-type term is also reduced. The paramagnetic term represents the effect of mixing the ground and excited states by the magnetic field. The effect of the electric field is expressed in the following equations for the parallel and perpendicular shielding con- stants: 73 o— _ e2 ‘439 a4Ez] ) 3mcza 40 e2 ) (38) e2 1 193 a‘Ez — 3mcza 15 e2 ’ where O'Eis the electronic charge, m the nuclear mas, c the speed q l of light, a the atomic radius, and E the electric field. . The polar effect a of Equation (22) can be computed from E the reaction field introduced by Buckingham (90). When a polar mole- cule dissolves, it polarizes the surrounding medium, and this polari- zation leads to an electric field, the reaction field, at the solute. If the molecule is sufficiently symmetrical, it may be shown that the mean reaction field is parallel, and proportional to, its dipole moment, 11. The reaction field is given by —-> 1 '9 _ _§_-____ .8. R -[Ze + 2.5-] a ’ (39) where a is the polarizability of the solute molecule, 6 the dielectric constant of the medium, and p. the dipole moment. There are also local solvent effects on the solute proton. Fields from induced charges on the solvent molecules near polar groups probably produce electric fields at the solute even though the solvent itself may be nonpolar. In this case the reaction field is given by the approximation, 1 ----)N6. R- 6+1 . (40) Therefore, the chemical shift will be linearly dependent upon the solvent parameter in Equation (40), if other effects such as anisotropy, van der Waals forces, difference in shape, or strong specific interactions are absent. The effect of the electric field on the screening constant is 74 given by Buckingham (89), as follows: 0': 2x10’5 --»2x10'1?‘9Ez - 10'181'3z - . .. , (41) where E2 is the component of the total field, E, along the X-H bond. For E = 105 esu (the field 7 A from a proton), the term in Equation41 proportional to E is about 0. 2 ppm and is about twenty times that of the term proportional to E2. Buckingham (_e_t 341. , studied acetonitrile as a polar solute in various solvents. In order to consider the polar effect alone, mag- netically anisotropic aromatic, rod-shaped and halogenated molecules were omitted. To allow for the shift due to van der Waals interactions in an approximate way, a five percent solution of acetonitrile in rig-hexane was used as a reference. A plot of shift versus (e-1)/(Ze +2.5) showed a linear relationship for the solvents used. Some of the scatter may be due to inaccuracies of the bulk susceptibilities, but some is due to the assumption that van der Waals forces are the same for all solvents and equal to that of nV-hexane. Buckingham e_t :11. , have calcu- lated o—E for acetonitrile in Liv-hexane as -0. 11 ppm and in acetone as -0. 50 ppm. (90). Abraham (103) in reference to Buckingham's study has suggested a direct comparison of chemical shifts of methane and acetonitrile. By referring shifts to ing-hexane, corrected shifts are obtained which are a function of the reaction field of the solvent introduced by Buckingham. A plot of the corrected shift versus (e-l)/(e +0.9) for the solvent gives good correlation except for benzene as a solvent. Buckingham, Schaefer, and Schneider (101) measured the chemi- cal shift of the acetonitrile protons in several solvents with neopentane as an internal standard. The results agree with those of Buckingham g} a_1. (90), and Abraham (103). 75 With the exception of the magnetic anisotropy term O‘a for disc- shaped molecules and the polar contribution CE in some molecules, all contributions to the screening constant are negative, causing a shift to lower applied magnetic field. For solvents with large diamagnetic anisotropies and in the absence of hydrogen bonding, 07a is larger than van der Waals and polar contributions. Gas phase measurements of chemical shifts have been made by Rayne (_e_t a_1.(104), and their experimental results are in good agreement with calculations of shielding due to van der Waals forces and polar effects, when corrected for bulk diamagnetic susceptibility. Calculated contributions to the shielding constant show that the reaction field effect is from 3-4 times that calculated for van der Waals interactions for some gases, and ten times that calculated for van der Waals inter- actions for gases made up of polar molecules. The bulk susceptibility corrections represent about 80-90 percent of the medium effect within the pressure range used. Diehl and Freeman (105) have derived an expression for the reaction field of a polarizable dipole in a non- spherical cavity. Measurements of acetonitrile and paraldehyde give support to the shape theory as more accurate than the earlier assumption of a spherical cavity made by Buckingham (96). Hydrogen Bonding The presence of hydrogen bonding is evidenced in nuclear mag- netic resonance by a low-field shift of the proton involved in the hydrogen bond. If the lifetime in the associated and non-associated states is sufficiently long, then a separate signal is observed for each environ- ment. For very short lifetimes, a signal which represents the average in these states is seen; this is the case with hydrogen-bonded and non hydrogen-bonded protons. Both solvent dilution and temperature changes 76 may affect the population of protons in the bonded and non-bonded states and thus will change the position of resonance, but a second resonance is never seen. The resonance position of a proton which becomes involved in the hydrogen bond, XH‘ ° ° Y, is obviously affected by the induced electric currents in the hydrogen-bonded species. The magnetic field is affected by (1) induced currents in Y, and (2) by changes in currents in the XH bond (88). Regarding the first effect, anisotropy in Y involves interatomic bond distances longer than ordinary intramolecular bond lengths. From Equation (32) it is seen that the shielding constant is inversely related to the cube of the bond length. Effects from the anisotropy of the Y atom will ordinarily result in a positive contribution to the shield- ing constant and a high-field shift of the bonded proton (106). Free diamagnetic precession of electrons is possible only if their electrical environment is axially symmetrical about the direction of the magnetic field. The second effect increases the asymmetry of the molecule, and the screening is reduced. Calculations by Schneider Show that a characteristic shift of 4 ppm due to a hydrogen bond may be produced by an electric field of 1. 5x106 esu. Such a field would arise from an electron at 1. 7 A from the hydrogen-bonded proton and is in the correct order of magnitude. The effect on the bond itself, caused by the presence of an electron-attracting Y atom, is complex. Pimentel and McClelland (107) state that, although the chemical shifts of the protons of water and hydrogen sulfide in the liquid state are approximately proportional to the electronegativities of the central atom, the proton chemical shifts of a number of gases are not qualitatively in the order of the electronegativity of the central atom. The association shifts ( 0:6- O'g) between the gas and liquid near the melting point are all negative and show greater shifts for substances forming strong hydrogen bonds. 77 In addition the diamagnetic circulation of the proton may be interferred with by the Y atom. This would cause deshielding,~ as seen by the Lamb formula for shielding due to diamagnetic currents: eZ p ‘ —2—f— d7”. (42) local diam - 3mc r where p is the electron density of the proton, and the integral is over all space. Buckingham (89) has calculated the change in shielding caused by an electric field applied to a covalently bonded hydrogen atom. His calculations predict a deshielding of the proton on a hydrogen bond. Stephen (98) states that a change in the intramolecular shielding constant will occur by distortion of the electronic distribution from nearby electric fields. For solutions in an isotropic solvent, measure- ment of the shielding constant leads to a correct intramolecular shield- ing contribution after correcting for bulk susceptibility and extrapolat- ing to infinite dilution. An excellent discussion of hydrogen bonding is found in an article by Cannon (108) . Relaxation Mechanisms Longitudinal Relaxation. The spin-lattice relaxation time, T1, is the half time for establishing thermal equilibrium along the mag- netic field axis among nuclei of different magnetic quantum numbers (87). In liquids this time is 10’2 to 107‘ sec. Thermal equilibrium of spins occurs through interaction with local magnetic fields. Energy is transferred from nuclei to the lattice by this mechanism. The extent of interaction depends on the magnitude of the local fields and on the rate of change in these fields. Paramagnetic molecules reduce the spin-lattice relaxation time and cause line broadening. 78 Transverse Relaxation. The spin- spin relaxation time, T2, results as precessing nuclei lose their phase coherence (87). One cause of transverse relaxation is the fact that different nuclei may experience different local fields, and there is a spread in the values of the resonance frequency with a consequent line broadening. Another method of producing transverse relaxation is spin- spin relaxation, which is an exchange of spin energy between two nuclei precessing at the same frequency. Inhomogeneity of the applied magnetic field contributes to phase incoherence and to a decrease in transverse relaxation time. Proton Excharge In a hydrogen bonded system the proton spends time in both XH. . - Y and X- . oHY states. If the lifetime is sufficiently long, two separate signals are seen, but if the lifetime is short or the exchange rapid, then an averaged signal may be seen. The spin-lattice relaxation time may be ignored if widths of signals for slow exchange are small compared with their difference in frequency. As developed by P0ple, Schneider, and Bernstein for the case of slow exchange (88), there is a broadened signal at 00A with width given by: _1_ _ 1 + 1 g - _ — , T2.41 TZA TA (43) where T is the transverse relaxation of nucleus A in the absence of 2A exchange, and TA is the mean lifetime at site A. For intermediate exchange assuming equal populations and life- times and large transverse relaxation times, the shape function depends only on {llv - v )E’C’Av . (44) 79 If C” is the mean lifetime in one site, then the signals coalesce when «I? t) -<- 217 Av ’ (45) where Av is the separation of the two resonance signals in cps. In the case of rapid exchange the mean frequency is ‘”mean = pAwA + pBmB ' (46) and the linewidth is given by -1— — 3‘5— + pB , (47) T'z T21; T25 where PA and pB are fractional populations at sites A and B, respectively, and Tap, and T23 are transverse relaxation times of nucleus A and B in the absence of exchange. Nuclear Spin- Spin Couplirg Nuclear spin coupling involves the interaction between nuclear spins through coupling with the electron spins and the electron orbital motion (88). The so-called Fermi-contact coupling contributes the largest quantity to the coupling constant between protons not directly bonded. This contribution is given as (3) 1611' If 1 J +2 'YN'YN,-Z-E-x[0)226 (rkN)6(er,)Sk-Sjl 0“, _ 2 ( ' _ _ _ NN 3h k J. _ (48) and requires for its evaluation a knowledge of the ground state wave function. 80 McConnell (109) has used a molecular orbital approximation to calculate the spin-spin interactions. Contributions from interactions between electrons in non-s orbitals and nuclear moments have been considered to be negligible for protons. Orbital contributions for protons not directly bonded to one another are small. Electron-coupled spin- spin interaction of protons, then, arises primarily through the Fermi interaction expressed in Equation (48). Valence bond calculations for nuclear spin interactions have been made by Karplus e_t a_1. (110). Such calculations give results in rather good agreement with experimental values. Contact electron spin coupling has been further investigated by Karplus (111). The assumption of perfect-pairing gives results which show some agree- ment with experiment and indicate that the contact term (Equation 48) is the principal factor in observed proton couplings. An angular dependence of electron-coupled proton interactions has been pointed out by Gutowsky (it 11. (112). The coupling decreases with increasing HCH angle for protons bonded to the same carbon atom. For nuclei other than the proton, there is a larger contribution of the dipole interaction between electronic and nuclear spins and of the interaction between the electron orbital motion and the nuclear moment. In particular for fluorine nuclei, the orbital contribution is still only a small fraction of the observed coupling (88). EXPERIMENTAL Instrumental Nuclear magnetic resonance spectra were obtained on the Varian A-60 spectrometer (113). Chemical shift measurements were obtained with both the room temperature V-6303 probe and the variable tempera- ture V-603l probe (114). These measurements are reported to 1:1 cps. Chemical shifts are reported in cps from tetramethylsilane. For dilution shifts, differences are given as positive for upfield and negative for downfield shifts. Resolution, or the linewidth at half-maximum ampli- tude, of 0.6 cps is possible with either probe. Integration of peak areas is used for the determination of equilibrium constants and per- centage enol values are accurate to about _+_2 percent. Variable temperature work was performed with the Varian V-6057 variable temperature system with the V-6040 controller, with regulation to 12°C at the sample. Low temperature calibration was checked by the use of the methanol chemical shift and at high temperature by the ethylene glycol shift. Measurements were made within about 2-3 minutes following introduction of the sample in the magnet. Temperature regulation at the sensor is 11°C, and the temperature calibration accuracy is 13°C. A sweep width of 500 cps was used generally for recording spectra, although this was changed to 50 cps for coupling constant determinations. Coupling constant means were repeated at least six times to reduce error to—i0.03. For recorded spectra, the radiofrequency field and intensity were varied as necessary. An increase in either of these factors within a spectrum has been indicated by the initials, ii, on the particular r esonanc e peak. 81 82 Compound Preparation a-Chloroacetylacetone was synthesized according to the method of D'Amico (115). Fractional distillation at 14 mm and 41.0—44. 5°C gave the product. a-Bromoacetylacetone was synthesized according to the method of Schwarzenbach and Felder (116) after preparing the copper complex of acetylacetone according to the method of Ciocca (117). The product was fractionally distilled as a yellow liquid at 13 mm and 60°C. Both n:buty1 a-chloroacetoacetate andAtV-butyl e-chloroacetoacetate were prepared by modifications of the procedure of D'Amico (115). Sulfuryl chloride (34 g.) was added dropwise to g—butyl acetoacetate (40 g.) with stirring at 0°C over a two hour period. The mixture was neutralized with ten percent aqueous sodium bicarbonate (150 ml.) and extracted with ether (150 ml) in, three portions. The extract was dried, the ether removed by distillation, and the product vacuum distilled at 5 mm and 84-86OC as a pale yellow liquid. 1;;Butyl acetoacetate was treated in the same manner as ’nV-butyl acetoacetate. - The ether extract was washed withvwater to neutralize it and then dried. The ether-was removed by distillation, and the product vacuum distilled at 3 mm and 59-610C as a pale yellow liquid. 8-Bromoethyl'acetoacetate was synthesized according to Donaruma (118) from 8-bromoethanol and ethyl acetoacetate by ester exchange, using PbO as a catalyst. Fractional distillation at 8 mm and 114—119°C gave the product. Ethyl y-bromoacetoacetate was prepared by the method of Burger and Ullyot (119). The product was distilled at 84-59C and 5 mm as a pink liquid. Isohima's preparation, in which ketene is reacted with ethyl a-cyanoacetate, was used to make ethyl a-cyanoacetoacetate (120). The ethyl e-cyanoacetate, along with an equimolar amount of pyridine, was heated to 800 before intro— duction of ketene, and ketene was fed in for a period of three and 83 one—half hours. The reaction mixture was shaken occasionally. The solution was neutralized, and the pyridine removed by distillation. The compound was fractionally distilled at 88-890C at 6 mm as a color- less liquid. Ethyl d-bromoacetoacretate was prepared according to the procedure of Kharasch (it a_1. (121), by the action of bromine on the parent ester. The product was distilled at 94-94. 5°C and 11 mm as a colorless liquid. Other compounds were commercially available, and their source is indicated in Tables XXVI and XXVII for 8-diketones and 8-ketoesters, respectively. Commund Purification Compounds were purified by the usual recrystallization, fractional distillation, and vapor phase chromatographic techniques. . Purity was checked by melting point, boiling point or refractive index measure- ment and by gas chromatography. Refractive indices were obtained on the Bausch and Lomb, Type 33-45-56 refractometer. Vapor phase chromatography was performed on the Perkin-Elmer Model 154A (20 percent silica column) and Aerograph Model A-700 (30 percent silica column). Preparation of Solutions Solutions were weighed on an analytical balance to 10. 1 mg, and mole fractions were determined to within i0. 001. Tetramethyl- silane was added as an internal reference to each solution, and the quantity added was the same in each case and did not affect the calcu- lated mole fraction to within 0. 001. Spectra from unsealed and evacu- ated sealed sample tubes were compared, and no difference in linewidth or shape was found in most cases. For acetic acid solutions and. those 84 using acetylacetone, tubes were evacuated and sealed but unsealed sample tubes were used in the remaining cases. Physic a1 Pr ope rtie 5 Physical constants for the 8-diketones and 8-ketoesters used in this work are given in Tables XXVI and XXVII, respectively. Included among these constants are boiling points, melting points, and refr active indic e s . Solvent Purification Benzene, chloroform, acetic acid, and dioxane were purified according to Vogel (122). Other solvents were purified by fractional distillation, following drying . Table XXVI. 85 Physical Constants of 8-Diketones B. p. M.p. 0 Compound 0C /mm 0C nDC Source Acetylac etone 38/10 - 1. 400319' 1 Eastmana a-Bromoac etylac etone 60/13 - - Synthesized3 ’ 4 a-Chloroac etylac etone 200V - 1. 4749 20' 3 Synthesized2 Dibenzoylmethane — 70-5, 77, 78:k - Eastmana 5, 5-Dimethy1-l, 3-cyclo- 148-9 - K and K‘3 hexanedione Hexafluoroacetylacetone 95V - 1. 332720° 4 Peninsular? Trifluoroac etylac etone 130V - 1. 388319° 8 ' Columbia“5 1, 3-Indanedione - 55/<1;nmS - 'Aldriche a-Methylacetylac etone 193v - 1. 437819° 8 Chem. Proc . O— l -Phenyl- 1, 3 -butanedione - 408 - Ea stmana 2-Phenyl- 1, 3 -indanedione - 145 - Aldriche Thenoyltrifluoroacetone - 30 / <1mmS - Columbia‘S m< Sublimation 4(- Vapor phase chromatography aEnol, enol, keto, respectively “Eastman Organic Chemicals, Distillation Products Industries BK and K Laboratories, Inc. yPeninsular Chem. Research, Inc. 6 €Columbia Organic Chemicals Co. , Inc. :Aldrich Chem. Co. , Inc. Chemicals Procurement Labs. , Inc. Table XXVII. 86 Physical Constants of 8-Ketoesters Compound C inm nD Source 8-Bromoethy1ac etoacetate 112-4/7 1. 475020' 3 Synthesized5 Butyl acetoac etate - l. 428020° 0 Eastmanc1 1:: Butyl acetoacetate 62/<1 1. 419820° 2 A Aldrich Chem. Co. , Inc. Butyl n-chloroac etoac etate 84-5/3 1. 446320° 3 Adapted from D'Amico; t: Butyl e-chloroac etoac etate 59-61/3 1. 445020° 3 Adapted from D'AmicoZ Ethyl acetoacetate 47/2 1.417521'1 Ethyl u-allylacetoacetate 67/<1 1. 438017° 6 Chem. Prov. Y Ethyl a-i’s’gamylacetoacetate 69/<1 1.434920° 0 K and K Labs. , Inc. Ethyl benzoylacetate 132/1 1. 520820° 6 Eastman“ Ethyl c-bromoac etoacetate 94-5/11 1. 462220" 3 Synthesized8 Ethyl 'y-bromoac etoac etate 84-5/ 5 - Synthesized6 Ethyl a-iflggbutylacetoacetate 251v 1.428820° 0 K and K Labs. ,‘ Inc. Ethyl (brig-butylacetoacetate 200v 1.433019° 9 K and K Labs. , Inc. Ethyl a-chloroac etoacetate 68/<1 1. 442520° 8 Eastman“ Ethyl e-cyanoac etoacetate 88-9/6 1. 471020° 3 Synthesized7 Ethyl e-ethylacetoacetate 57/<1 1. 422019° 0 Bios Labs. , Inc. Ethyl e-fluoroac etoacetate 210V 1. 408520° 4 Chem. Proc. 7 Ethyl trifluoroac etoacetate 160V 1. 376220° O Peninsular Ethyl a-methylacetoacetate 73/11 1. 425320° 0 Aldrich Chem. Co. , Inc. Ethyl a-iggpropylac etoac etate - 1. 422321° 6 Ea stmana Ethyl e-n—propylacetoacetate 70/2 1.425120° 1 K and K Labs. , Inc. Methyl acetoacetate 62/<1 1. 418420' 5 Eastman“ v “Vapor phase chromatography. Eastman Org. Chem. , Distillation Products Ind. Butyl acetoacetate and ethyl e-i’sgpropylacetoacetate were synthesized for us by Distillation Products Industries. YChemicals Procurement Labs. , Inc. Peninsular Chem. Research, Inc. RESULTS AND DISCUSSION Nuclear Magnetic Resonance Spectra Chemical Shifts. A list of molecular formulas for the fi-diketones studied is given in Table XXVIII, along with the numbers of the figures (Figure 12-Figure 20a) in which the corresponding proton resonance spectra are shown. Also numbers have been assigned to each compound for the purpose of identifying points on graphs in the thesis. Formulas for the [S-ketoesters studied are shown in Table XXIX, with similar figure and numerical assignments. Figures 20b through 35b show proton magnetic resonance spectra for the B-ketoesters. Of the compounds studied nuclear magnetic resonance spectra have been recorded previously for acetylacetone (23, 71) a-methylacetylacetone (71), ethyl acetoacetate (58), 1, 3-indanedione (123), and thenoyl trifluoroacetone (123). Assignment of resonance peaks to the various proton groups in a compound is indicated on the figure showing the NMR spectrum of that compound. These assignments have been verified by integration of peak areas and also, in some cases, by changing the position of equi- librium by addition of solvent. A case in point is the verification of the enolic resonance peaks in ethyl a-methylac etoacetate by the addition of hexane. For trifluoroacetylacetone and hexafluoroacetylacetone, the position ‘of keto ‘resonance peaks was confirmed by high temperature studies. Identification of enol versus keto resonance peaks has been aided occasionally by fractional distillation of the fi-dicarbonyl. The first fraction is primarily enol tautomer, since the intramolecularly hydrogen-bonded enol tautomer is less polar than the keto tautomer. A spectrum taken shortly after vacuum distillation is seen in Figure 36 87 Table XXVIII. 88 Molecular Formulas of fi—Diketones No. Compound Formula Figure 1 Ac etylac etone CH3COCH2C OCH3 12 2 a- Bromoacetylacetone CH3COCH(Br)COCH3 l3 3 a-Chloroacetylacetone CH3COCH(C1)COCH3 l4 4 Cyclic iSOpropylidene /O-CO\ malonate(in CHC13) (Me)z-C\ /CHZ 15a O-CO 5 Dibenzoylmethane C6H5COCHZCOC6H5 16 6 Hexafluoroac etylac etone CF3COCH2C OCF3 15b 7 Trifluoroac etylac etone CF3COCH2COCH3 17 c éo 8 1, 3-Indanedione (in CHC13) \CHZ 18a C< \o 9 a-Methylacetylacetone CH3COCH(CH3)COCH3 18b 10 1 -Pheny1- l , 3-butanedione C6H5COCH2COCH3 19 (in CC14) 4 O 11 2- (I) -1, 3-Indanedione >CH¢ 20a (in Acetone) C \ \ 12 Thenoyltrifluoroac etone I l S 89 02 9: Sm wmm T m 3.x. ix 1.1 (l < in 1...; \\ (2: L. M. a .5 x Jmmo 2 ouo nTB M \ mOWoonmo .oGOHvomH>uoom mo Eflhoomm mzz acuonnm .NH 0.3.th 9O Figure 13. Proton NMR spectrum of a-bromoacetylacetone. e /OI~ CI-_l3 C: k CH3 (3:0 \> CHk 0He & __,“/\~_ // // f 951 301 H -—> 143 138 91 Figure 14. Proton NMR spectrum of a-chloroacetylacetone. cgfc’SH / CH3kC= o CHk 31, r him . A I T '7 922 H cps -—-> 292 139 133 92 All I wwm owm omw :3 m \ Nmmo m.mo & N QED .ocou—OUdTFoUMOHOfiHHMxofi mo gauuomm fizz nobonnw .an 0.2.th Al: I mod QNN 1‘: 4 .1. M‘ (’ \I In! «mo -Anmov .AnHUEU G3 3.25de 233390.?“on 0383 Ho 5.9.30on mzz Gogonnm ..mmd 0.2.th 93 Figure 16. Proton NMR spectrum of dibenzoylmethane (in CC14) 6 Ring protons l _l i -CHe 4 Ring protons K f—_H OR3 A I 993 452 423 H —-> 3.82 94 hww $8 I n OIUM EU x 030 a 030 "I \u u 0 EU 30\ 0 .oQOuoomTfoomououaGmnu Ho 5530on M22 couch“ .bd 0.”:th N2 £2 SN «.1283 SM dd +- 4b- 95 All mmo m o: mNA on” z n:0... n u 0 EU a N I «I ..\Uu EU 30 \4 O «"0an0 .oGOuoomH>uuodH>£uoEud mo guuovmm g2 Gououm 5.2 oudmfm T mmum - 2: OS 6 ‘ \\ Kill mucuoum 93% «\V N "mu .AnH USO G3 snowboampdaum J mo 5:30on “22 atuoum :an oudmfm 96 ONH NNH Hmm A.| m hmm unmq mew owo iwumo in fl 1 xii mew j J (21 EU 33.0.2" mcwm m _ r t m 30mm 93% N .AJUU GS ocofipodmufinum Jugaosmud mo 5.30on MEZASHOHQ. .mL 0.2.th 97 o 430$ Sm ANS mama: own 3m N: E 3 4 fij (xi \\ II/mll oflU 030 .. ml mo. 0 mu / «:0 1.3M Hmama I Efim u N M ED ouonmo & .ouquoodouood H>nuooaoanua mo guacamm MEZ nouounm .n—om 0.“:th .2... m3 N3 A... mmom 41:; \\ . Aocoao om camonm mfim m as 6:039:90:qu .Huacogmum Ho 5930on m Z GBOHAH +Gfluo§m wcfim w .mom “:5th 98 1 Nwm 1.. U .11).). . Willi! EU mm¢ mamas r L 1< Goaonm mcflm H P '— mGOHOHN mcmm N wow <11 \vao .AmmU as 02530093335 Kocgfism «0 5.9.3033 MEZ couounm .HN 0.3th 99 Table XXIX. Molecular Formulas of fi-Ketoesters No... Compound Formula Figure 13 B-Bromo ethyl acetoac etate CH3COCHZCOOCH2CHZBr 20b 14 Butyl acetoacetate CH3COCHZCOO(CHZ)3CH3 22a 15 ItsButyl acetoacetate CH3COCHZCOOC(CH3)3 22b 16 3' Butyl a-chloroacetoacetate CH3COCH(C1)COO(CH2)3CH3 23a 17 tsButyl u-chloroacetoacetate CH3COCH(C1)COOC(CH3)3 23b 18 Ethyl acetoacetate CH3COCHZCOOCHZCH3 24 19 20 21 22 23 24 25 26 27 28 29 3O 31 32 33 34 Ethyl u-allylacetoacetate Ethyl a-i’svoarnylacetoacetate Ethyl benzoylacetate Ethyl a-bromoacetoacetate Ethyl y-bromoac etoacetate Ethyl a-i/sgbutylacetoacetate Ethyl a-nfl-butylacetoac etate Ethyl a-chloroacetoacetate Ethyl a-cyanoacetoacetate Ethyl a- ethylacetoac etate Ethyl a-fluoroac etoacetate Ethyl trifluoroac etoacetate Ethyl a-methylacetoacetate Ethyl a-i’sgpropylac etoac etate Ethyl a-nfl-propylacetoacetate Methyl acetoacetate CH3COCH(CHZCH=CHZ)COOCHZCH3 25 CH3COCH[(CHZ)3- (CH3)Z]COOCH3CH3 26 C6H5COCHZCOOCH2CH3 27 CH3C OCH [CHZ- CH (CH3)z]C OOCHzCH3 29 a CH3COCH[(CHZ)3CH3]COOCHZCH3 29b CH3COCH(C1)COOCHZCH3 30 CH3C OCH(CN)COOCH2CH3 31 CH3COCH(CHZCH3)COOCHZCH3 32a CH3COCH(F)COOCHZCH3 33 CF3COCH2C OOCHZCH3 32b CH3C OCH(CH3)COOCHZCH3 34a CH3COCH[CH(CH3)Z]COOCHZCH3 34b CH3C OCHZCOOCH3 35b 100 o ow NC o? Tmmom a2 2% Mme Illéw . HS w a a 2 4f a a 2 030. 030 N xmo 3Q q -.Anmovx noofio J 0.6 «$0 . 30.5 0030 00 333 3 nu mo E53009... MEZ Gobounm .nNN 0udwfih JmQU o mm s: 0241 m mom 13. so... . o? W14 1‘ a \\ <2 £0 /|)\U . A. a u m an mo- :22 -~ - g . moo omo . no "no .r\)\’\ \ 0 mo «mo 7 s 2 Sn sumo \ o n oxnmo .330003000 Arno. mo gnuo0mm mzz :oaoum ..mmN 0.“:th 101 memo wfi 92 T ado SN St. 2.. ‘ I n IxU0 EU £30 30 n n .unmovwh o 00 mo 5002 x .000000000000ou030Id audnhmmo E3302: .mSHZ :ououm I .nmm 050mm” mm 3; oi 130m 0m... mam 3. l \\ J W 2 030 a N J $00 #20 I n I E 0 0x 0 .000000030000H330I0 333nm. mo Ennaoomm “32.2 doabnnm .0MN 0Hsmwh 102 Np SH M? T 30m SN 93 New W! L \\ l/j emu N a > u "\oonmu mu H a u no on» . j \n A. J omo N x mo nmo N afim onoxnmo .oumgomgvum 150 mo guuounm MEZ Gououm :vm 0.“:th 103 m» f: m2 mm: Tmmom SN mwm (L in :5 1 #1 . D"... I \ Nmonbofio xmo I ml ...\Uo EU \ mo ~30 drum 2% . . . é «moumo {I N:Uumo o: \ ..va «mo flmo \ 15m ouunmu x .mufioomgoomadmnd 3:3 mo 5.9.3033 fizz couounn .mN ousmmh r 104 ornlmmom mm mp m3 SN mwm New H." £0 :0 x 15H n no In 1|.(/\ Amwa‘ ifim $ ouufiu .Oagvomouuomlgmmm‘wnd 1230 Ho Gun—300mm fizz cauoum _ .om 0.3de 105 n. on» mac m mm‘m. owm omN ”Q Nwm va Sufi. H 15m as o «mo atm 1% 4__ «mo R Exam umgoom 305:5 350 m0 59.302? .522 Gououm o: N F J 30 Eon—cum mcwm .5. 953m 106 _c AI 3 r f. own m3 mmm mum 2;. i! 1...? 4 i a \\ ./2\m J omo emu \INTHU > N N j a 5M ":0 x30 1.3m \ -ounmoum .opmuoomguud 050.33% 14.30 Ma 8330on fizz Gen—on“ . .nmN uuummh d Tdmgm E. S. «S «3 :3 mom 3» -.T L , a a \\ 4? x. q \ o 30 n .. a «wflfi "Ummo £0 a mo cram xmu n we N xlfim Ono 5:0 x .3 «oomauoomofiounuu 150 we 55.3023 #342 acaoum . .mw~ 0.3m?“ /l\\j.\|/|\ \ I.“ Mme -EnEUV 0 Ox 0 .KNHUUNOHOUMHAQSQOmfl ~>£u® HO EJHHUUQW mzz :Ouohnfi .mw&m‘ mvHJervN 107 o mm mm wfi. mom A: Ar: mmum H Jul.“ fl ‘ 030 Nnmo > a x30 H g M PI|<|L I. . \I\ M. n ..n n u mu A 38 «mo 15m N Ono. nEU . w > S > . vH fiaduQUdouuomH Baum H 50 H0 #330on HHSHZ acuoum non 0.2;th Lmfiwlj mm mp 34 mph Al. I a 030 30 mo Efim J M nmo . H TEE (2 M \ -Nmmov ouox mo .mamuoomgoumgusnmtuw 350 Ho Efiuuuomm MHZZ douounm .mom uuanh 108 E 2. 02. 3: Al 30m , om” com 5 ”£0 15m \ xnmo cram .. «I .. mo mop. n n OUVHZU «mo ifim .oumuoomOHUUMOHOHnoud H330 Ho 53.32am MSHZ :3ch xx . (2 IO .om oufimHh. 109 b 1 NwH All MQUHIH MOW £0 15M no ”mo mo\ 0 1.6. .vumuoomouoomo:m>oud H530 mo Edyuoomm ”H22 Gououm ~30 Efim Sum 30 0 .3. 98th 110 All mmwum N own CNN mNN om H u VHHHU Pumo 020 :33 J mHHU N:0 \4 > H Sum .oampoomouoUMOHodeH: TEE Ho E33009“ H322 Gouounm .nHNm oudenH _o 3 E. o2 SN omm ml mmom $5 $1 ~ 5% H \\ % . . uul n H > 0 mo 3 H 5m «wufoé ”No afim \ can ":0 x .oumuoomgoomewHuoud H550 Ho 5.9.302? H322 GOHOHAH .mNm 0.2..th 111 o Almaom mp mp me NHN ova me hmm n N EU #35. N 0 30 \N afim _ I .. 2 W fi N EU 20 mo xafim x l n 0-0x most .oumuoomopoomonosaud T33 Ho 5:302? MEZ cououm .mm ousmHh ; 112 OT ado: Vm MN. «VNH @HVH XVH ®¢N $5.5 . 0 .— . \\ i j w ... > «a . $0 \ afim -NAnmuv N ouuxnmo .mumuooflogodamoumofluu 350 Ho £330on “22 Gouounm .33.” 0.19th ad Hg mu F: oan Al I NHN gym I __: \. =1 N .. mo :8 n n 30 x > o 3%» H 5H @205. £0 Ono 5W 15m _ x .oumuoumofiumafiugnu 350 m0 guuuomn fizz Gavan“ .9: 0.“th 113 o Allmmu TH nNH LT 4 40113 N m H 0 0.0 n \H l VH THU .00m000moH00m Trims» Ho E5502? H322 GOHOHnH 5mm onderA 4.1303 em 2 3H jowm we» ..lil! J) V. . i N )3 \\ c: A H fiomo Vamu HEHM r\)\/\ TH H 0 \ K flukmmov- M0 N H 5m n ouuvH m0 .mumufiumoamomfwmoumhfl 350 Ho guuoomm H222 Gauonna .mmm unsth 114 I! EB» $3 2: ado QMN con 0H H ammo 15m : 1%]. £0 HEHMH N u -.1 mob» 30: ouukm 0 I, :0 5333308 HmcoHu0muu maHBOHHOH 3300d300m0u030ud HER—0 «0 55.3023 #342 acuoum . .om ouanh for ethyl a-chloroacetoacetate and may be compared with that of the pure compound at equilibrium in Figure 30. In general it may be noted that the tautomeric equilibrium lies on the side of the enol tautomer for [B-diketones and far to the side of the keto tautomer for most B-ketoesters. Molecular models show that there is greater van der Waals non-bonded interaction in Bediketones between ac etyl protons in the most favorable diketo configuration thus shifting the equilibrium toward the enol tautomer in these compounds. Separate resonance signals are detected for the methyl protons of the acetyl groups of the keto and enol forms, as well as for the keto and enol a-protons and the enolic OH proton. Giessner-Prettre (58) has reported that the OH resonance line of ethyl acetoacetate is not visible in binary solutions without the addition of hydrochloric acid. The present study finds the OH resonance peak not only for ethyl aceto- acetate, but also for all the o.- substituted acetoacetates, both in pure form and in various solvents. Chemical shift measurements for pure fi-diketones and B-ketoesters (in cps from tetramethylsilane) are given in Tables XXX and XXXI, respectively. For the fi-diketones and fi-ketoesters neither the enol nor keto acetyl-methyl resonance peak varies appreciably in chemical shift. However, the deshielding effect of electron-withdrawing groups both in the a-position and on the acetyl group on the keto and enol a-proton is considerable. Resonances for the enolic OH proton occur at very low fields and show considerable variation among the compounds studied. For the most part the enolic OH resonance occurs at much lower field for B-diketones than for B-ketoesters. Methyl protons of the ethoxy group show little variation among the fi-ketoesters, whereas methylene protons are deshielded by electron- withdrawing groups in the a-position. Those compounds for which the ethoxy protons give different chemical shifts for enol and diketo .msHH E00 mnHo :H .....HHHHHm 33820 H I Hos0 30% NH .540 Hnmo E 2 Homé n .. mm m Nwm .. u u u 0so90000H05HHHHuH>oc0~HB - - - - OS .. - .. oaoHsmcmeqH-m .H Av-~ 300 E Ha_ HS. ..8 - - 23 Ha - HmN SH 02 088333; .H-H>:2HnH-H ooHH - VHS 23 - mmm - omH mNH 053816815025 I n u u n HmH u n H0GmxoflUHH$ 0c0wp0smchum.H ,b H u .. Sum oom u wmm omH NmH 0sou000H>u00moHoHHHHHHH u I own owm u 2% u .. 0:30 0330 omonosdmosm Pm mm: H. m3 300 a: Pm me: 0 3H. - 83 gm - - - - 323810235 HwHUU HHS 030G305 NHnZDV moH u .. .. u hoH .. .. 0G0H0HH>HHOHAHOmH 0HH0>U n u Hmo .. Hom .. mi: me 0soH0umH>u0umoEonso . . mum u mom u hmH mmH 0soH—000H30000H0HHH010 .. - 3. o 3. m - S N om H SH 283 92.30 04 H0 mooH .0 n n Gdomzho £0 0 £8 emu xHHo meo 0... mo 0 m0 HU 0 H>VHH< souounHud #3004 mocoHBHHoé 3:0 E 3030 HmoHEoHHo :8on .xxx 033.. 117 OUmfiUnpm me can mum .. oNN .. .. mmm 0N. 0umu0umow0omoaohauh fitnm 00» u 03. .. mom n “E; “EA wmm E. 0330030000800310 afifim x mmm mp O Q I o: Nwm me .. u Movw xho 0pmu0uma>owc0n Trim {$8 2: N .., m3 3 - A $8 om Ni ,. Mom - wfi E 2%. 3 Bfioomfioomafimoflé 150 Nmoumu Sm Nmaumo mom NmUHmoumo 1.: o: - SN - 02 p: gm 2 Bfluomopoomazmé 350 .. 02. mom - mom m2 SH 3% E Bfioomoaoom TEE £3an No $338 3 E - EN - «02 ME - - 83038838208 130.“ £0 1:3 mm NEUO T353 vmm own a wow n 0*; and .. .. 0umu0omou0omoHoEond audmu: x dram”: em MS. MS. - 2; 02 N: - - 3803803.. 130nm mmUNANmB 3:5 no "mo :33 mm , £00 :53 Sum om» mom - mom 0.2 w: - - 383333 130 Hm~mo TEE SN N: Sm .. 2N ofi o: 03 - 39088on :33 0820a N m n N n .860 030 030 xmo x mo x mo a mu mo mu 353800 maouohnmud Hhu0o< szumn 000000000M1u 0.30% Ga mumEm H00fi50£0 ceaomnm .UCCA 0308 um 118 .928 000.00 0&0 00 0000000 000050000 u .0“ 0000008 n 8 .0000 n 0 .000M n x M000 m2 flofimuv om £02 £00 2: “@0050 Sm 001000010000 3; x oxmmovmo 2. “0:00:00 0m £05 m0 ~00 0300-3 2: £0 150 .3 mm mmufimmov m2: £02008 fl. “00 0:3 mm M00000 2: cm NAnmUEU om won on» wmh own 01:. how 0mm 0N5 Mhh owm new 000 NAN NNM wow. oom mom OHM own mNN mNH wNH 0N0 omH mmH omH 00; mm; wNH NMH o: N“: 02 3%. «0qu wvm vme xmmN 0NHN 0mm moN 0mm own mvm mu 2. «l. 00w Mon 02‘ m: x 3‘ mm on 2. Mb MN. 00000 00000 00 3000002 00000 00000 000030.00 hmsd 130M 00000 00000 00100 nmmwm ..0 0030mm 0000000000000>£00Eu0 1000M 0000000000 000.00.300.00 130M 000000000000000500u0 0>£0M 00000 00000 0010000 .0 0030M 00000 00000 00000» 0 ..0 130M 00000 00000 000 0020-0 330M 00000 00000 001005 Ismnd Karma 00000 00000 001030m.m\0:0 0500“ nnnnnnnnnnnn 119 tautomers include the a-bromo, a-fluoro, a-chloro, and the trifluoro- ethyl acetoacetates, along with ethyl benzoyl acetate. Substitution of fluorine atoms for protons on the acetyl methyl group of fi-diketones results in a deshielding of the a-protons for the keto tautomer of 20 and 30 cps in tri- and hexa-fluoroacetylacetone, respectively. Both chlorine and bromine in the a-position result in deshielding of the keto a-protons of about 80 cps from the resonance position in acetylacetone itself. In fi—ketoesters the keto a—proton is once again deshielded by fluorine, chlorine, and bromine atoms in the u-position by 118, 91, and 94 cps, respectively, from its resonance in unsubstituted ethyl acetoacetate. Bromine substituted in the fy-position (on the acetyl group) deshields the B—proton only 15 cps, and in the B-position on the ethyl group has little effect. In B-bromo ethyl acetoacetate the a-proton is deshielded more than in ethyl 'y-bromoacetoacetate. Among the a-alkyl substituted compounds little difference in the resonance position of the a—proton is observed. A trifluoro group in place of a methyl group only deshields the keto a-proton 16 cps for the [S-ketoesters. The enolic a-proton is deshielded in the trifluoro- and hexafluoro- acetylacetone compounds by about 25 and 50 cps, respectively, from its position in acetylac etone. In thenoyltrifluoroacetone it is deshielded by nearly 50 cps. For the trifluoro-B-ketoesters the a-proton occurs 38 cps below its resonance position in ethyl acetoacetate. Bromine in the y-position deshields this proton by about 20 cps, and a phenyl group in place of the acetyl methyl group deshields the u-proton by 40 cps. For B-ketoesters the position of the OH resonance line varies over about 100 cps. For the series of compounds substituted in the a-position, greater deshielding occurs in the order chlorine >.bromine 5 cyano. For alkyl substitution in the a-po sition the hydroxyl resonance peaks have chemical shifts in the range 720—773 cps with no apparent 120 general trend with increasing bulk. It appears surprising that an a-methyl group results in deshielding of the OH proton by 28 cps. The trifluoro compound shows some shielding of the OH proton from ethyl acetoacetate. In ethyl u-fluoroacetoacetate, there is evidence that some enol tautomer is present, but the enol OH does not appear at the expected low fields. There is a peak of the correct intensity at 240 cps which might be a non-intramolecularly-bonded OH resonance. However molecular models do not indicate any interference with the internally hydrogen bonded species. Variable temperature studies have not been helpful in assigning the enol OH resonance position. Within the B-diketone compounds similar trends are observed. The position of OH resonance is at much lower applied fields than in the esters, and there is more variation among the compounds studied. A trifluoro and also hexafluoro group in place of the acetyl methyl group results in shielding of the OH proton. Deshielding of the OH proton occurs in the order a-chloro > a-bromo > a-methyl. That the substitution of a methyl group for an a-proton causes decreased shield— ing to the extent of 75 cps is surely puzzling. Table XXXII indicates the enol OH chemical shift in cps from tetramethylsilane for pure compounds in the order of decreased shield- ing and presumably increasing interaction between OH and the carbonyl oxygen atom. Forsen and Nilsson (74) have plotted the carbonyl stretching frequency against chemical shift of the enol OH and found that a lower carbonyl stretching frequency corresponds to a lower chemical shift of the enol OH. They point out that a lower carbonyl stretching frequency is observed for cases of association of the type AH- - ° C=C.-. In Figure 37 the chemical shift of enol OH for B-dicarbonyls has been plotted against the stretching frequency for the enol chelate carbonyl. Chemical shift measurements for line A of the figure are 121 Table XXXII. Chemical Shifts of Enol OH for B—Dicarbonyls >ikn Order of Increasing Interaction Between OH and C=O _ _q Compound Chemical Shift B-Bromo ethyl acetoacetate 712 Ethyl y-bromoacetoacetate 716 Ethyl trifluoroacetoacetate 720 Butyl acetoacetate 730 Ethyl acetoacetate 730 [ECButyl acetoacetate 733 Ethyl a-chloroac etoacetate 737 nsButyl a-chloroacetoacetate 740 ELButyl a-chloroacetoacetate 747 Ethyl a-iggamylacetoacetate 754 Ethyl a-Arfinutylacetoacetate 757 Ethyl a-methylacetoacetate 758 Ethyl a-iggamylacetoacetate 762 Ethyl a- ethylacetoacetate 764 Ethyl a-bromoacetoacetate 764 Ethyl u-IIL-propylacetoacetate 768 Ethyl a-allylacetoacetate 770 Ethyl benzoylacetate 770 Ethyl u-mbutylacetoac etate 771 Ethyl a-i/aobutylacetoac etate 773 Ethyl u-i/ggpropylacetoacetate 779 Hexafluoroacetylacetone 780 Ethyl a-cyanoacetoacetate 807 Trifluoroacetylacetone 847 Thenoyl trifluoroacetone (in C82) 898 continued 122 Table XXXII - Continued Compound ' Chemical Shift o.-_Chloroacety1ac etone 922 a- Bromoacetylacetone 951 Acetylacetone, : 934 1-Phenyl-l, 3-butanedione (in CCl4) 980 a-Methylac etylacetone 990 Dibenzoylmethane (in CC14) 993 * Chemical shifts are in cps from TMS. 123 00000000370 .000 00 0w0a 00 00:0 000000000070 00.0 00 000000 :0 000000 00.00 :0 000000 00 0000000000 00590 0000 :0 0.00083: 000.0. .0 000 A 0000 . 00.00 0w00 00.00 0N00 0000 0 _ _ . 4 < 0 . 0 0 0 d 0 ow¢fl 400m0 1-0Nm0 000.003 000000 00000000000010 :4 £00.00 2.83 2.3 308305-00 I O 001. 00000070 0000 000.00.0000000000008-0 I D 40000 iomeA. 0 0 0. -. 8.: . m 2 O o .0000 00000 J.0000 onAV _ 0 r .1 _ 0 _ n 1 4. F n _ 0000 0000 000 0N0 000 0000 000 00> 0N0. 00.000000030000070 .000 00000000000 90000000000 000000.000 mdmu0> 0.00 0000 00 000000 00005000 .hm 0st0h :83 OuU 124 taken from the present work, and an average value for the stretching frequencies has been taken from Tables VI and VII in the literature. Values of stretching frequencies for ethyl trifluoroacetoacetate, hexafluoroacetylacetone, and trifluoroacetylac etone obtained in the present work are 1675 cm‘l, 1629 cm'l, and 1600 cm'l, respectively. Line B of Figure 37 represents the results of Forsen and Nilsson (74). As may be seen from the figure, there is a general trend of increasing carbonyl frequency with a more shielded enol OH proton, in agreement with Forsen and Nilsson's results. However, the fi-dicarbonyls in the present study do not give a strict linear relationship. Certain deviations from linearity may be noted, in particular ethyl trifluoroacetoacetate, trifluoroacetylacetone, and thenoyl trifluoroac etone. Structure. Among the unsymmetrical B-diketones there is the possibility of two different enol tautomers, as shown in Figure 38 for trifluoroacetylacetone. Park e_t a_t_l. (3), prefer structure (a) with no justification given. Belford St a_1. (38), assume that structure (a) is H\ H o/ \o o’/ \ a 2; t 2. / \ \ / \ / CF3 |/ CH, CF, (I:/ \CH3 H H (a) (b) Figure 38. Possible enol tautomers for trifluoroacetylac etone. more acidic than (b), and that structure (b) is probably in greatest concentration. Bellamy and Beecher (32) state that the trifluoro group induces enolization of the adjacent keto group and that the trifluoro 125 group is then far removed from the carbonyl undergoing chelation. Thus they prefer structure (a). Nakamoto (86) has found that the trifluoro group strengthens the adjacent C=O bond for chelates, which would tend to favor structure (b). However, the infrared evidence is inconclusive. Nuclear magnetic resonance studies to date do not give firm evidence to distinguish between structures (a) and (b). Chemical shift measurements for the series acetylacetone, trifluoro- and hexafluoro- acetylacetone indicate that the resonance position of the OH proton for the trifluoro compound occurs midway between its position in the other two compounds. It appears possible that the proton is exchanging sites and that the resonance in trifluoroacetylacetone actually represents an average between structures (a) and (b). A For other unsymmetrical fi-diketones observed in the present work, only one enol form appears to be present, although the possibility of rapid exchange is not precluded by the NMR evidence. Tautomeric Equilibrium. The percentage enol tautomers in the B-diketones and fl-ketoesters is given in Tables XXXIII and XXXIV, respectively, along with equilibrium constants at 33°C. For the fl-diketones substitution of an electron-withdrawing group such as chlorine in the a-position results in increased enolization; however, bromine results in a marked decrease in enolization. - Molecular models show that a chlorine atom is about the size of a methyl group. Steric considerations in the a-chloro- and a-methylacetylacetone should, therefore, be approximately equal. Bromine, however, is considerably larger, and van der Waals non-bonded interactions become far greater than with chlorine. Also bromine is less electronegative than chlorine and would consequently result in a less acidic a-hydrogen atom. The high enol content of both trifluoro- and hexafluoroacetylacetone indicates 126 Table XXXIII. Percentages of Enol Tautomers and Equilibrium Constants for fi-Diketones as Determined by Nuclear Magnetic Resonance Percent * Protons Compound Enol Ke Integrated e k Acetylacetone 81 4 . 3 CH /CH7_ a-Bromoacetylacetone 46 0 . 85 CHE/CH;< a-Chloroacetylac etone 94 16 CHgg/CH;c Cyclic isopropylidene 0 0 - malonate (in CC14) Dibenzoylmethane (in CCh) 100 - - Hexafluoroac etylac etone 100 - - k Trifluoroac etylacetone 97 32 CHE/CH3 1, 3-Indanedione (in CHC13) O O - a-Methylac etylacetone 3O 0 . 43 o.-CH§/ci-CH;< l-Phenyl- 1, 3 -butanedione 100 - - (in CC14) 2-Phenyl- l , 3-indanedione 0 O - (in dioxane) Thenoyltrifluoroac etone 100 - - (in 082) * enol Ke - [keto] , and all measurements are at 33 j; 2°C. 127 Table XXXIV. Percentage Enol Tautomers and Equilibrium Constants for fi-Ketoesters as Determined by Nuclear Magnetic Resonance , - Percent * Protons Compound Enol Ke Integrated B-Bromo ethyl ac etoacetate 6 O. 06 CHE/CH;< Butyl acetoacetate 15 O. 18 CHk/ethyl CH3 ,2,- Butyl ac etoacetate 17 0. 21 (CH3);?/(CH3)§< Butyl a-chloroacetoacetate 20 O. 25 CHk/OHe t: Butyl a-chloroacetoac etate 46 0. 85 CHE/CH1}: Ethyl ac etoac etate 8 0. 09 CHf/CH;< Ethyl a-allylacetoac etate ~3 A0. 03 peak heights Ethyl benzoylacetate 22 o. 28 CHe/(cnfgk Ethyl a-i/s‘gamylacetoac etate N3 NO. 03 peak heights Ethyl a-bromoac etoac etate 5 0. 05 peak heights Ethyl a-i’svgbutylacetoac etate N2 ~0. 02 peak heights Ethyl a-EJbutylacetoac etate “’2 ”0 . 02 peak heights Ethyl a-chloroac etoac etate 15 0. 18 CHR/ethyl CH2 Ethyl a-cyanoac etoac etate 9 3 13 OHe/CHSH-k Ethyl a-ethylacetoac etate N1 ~ 0. 01 peak heights Ethyl a-fluoroacetoacetate 15 0. 18 CHk/ethyl CH2 Ethyl trifluoroac etoacetate 89 8. 1 CHe/CH;< Ethyl u-methylacetoac etate 5 0. 05 CHk/ethyl CH2 Ethyl a- i/sgpropylac etoacetate m l ~0. 01 peak heights Ethyl a-anropylacetoac etate ”1 NO, 01 peak heights Methyl acetoacetate 0 0 - * enol K3- [keto] ' and all measurements are at 33 1; 2°C. 128 that the electronegative fluorine atoms are successful in withdrawing electrons from the vicinity of the a-protons. Park St 3:1. (3), explain the very high enol content in compounds such as trifluoroac etylacetone by additional stabilization of the £i_s_ form by possible CH- - oF bonding and in the -tr_ar_isenol by OH. . .F bonding. Filler and Naqvi (18) point out, however, that the discontinuity in the observed percentage of tautomers precludes this explanation. They suggest the alternative explanation that the enols become increas- ingly acidic as the number of fluorine atoms increases which would lead to stronger hydrogen bonding. In dibenzoylmethane, 1-phenyl- 1, 3-butanedione, and thenoyltri- fluoroacetone, the presence of the aromatic ring results in increased enolization. In each case, however, the aromatic ring is unable to assume a position parallel to the intramolecular six-membered ring of the enolic tautomer because of steric interaction with the enol u-proton. The conjugated system has evidently not been extended. Stabilization of the enol tautomer must then result from an electron- withdrawing effect of the rings. Neither 1, 3-indanedione nor 2-pheny1- l, 3-indanedione gives evidence of enolization in solvents such as chloroform, benzene, dioxane, and dichloroethane. This is not sur- prising in view of the fact that molecular models indicate the formation of an intramolecularly—bonded system is precluded by steric consider- ations. Likewise, cyclic isopropylidene malonate cannot form an intramolecular hydrogen bond and is completely ketonic in carbon di- sulfide, carbon tetrachloride, benzene, and chloroform. Among the B-ketoesters a trend similar to that in the fi-diketones is observed. ‘ Alkyl groups substituted in the a-position decrease the degree of enolization, whereas both fluorine and chlorine in this position increase the percentage of enol tautomer. However, bromine in the a-position results in less enol tautomer, and this decrease may be 129 explained as in the case of a-bromoacetylacetone. A larger alkyl group on the alkoxy end of the fi-ketoester results in an increase in enolization, which may be explained by steric interaction between the alkoxy protons and those of the ac etyl methyl group for the diketone tautomer. The substitution of a cyano group in the u-position results in very great increase of the enol form. A molecular model indicates that the possibility of N. . .H bonding in the enol tautomer is not a likely one. Further, no steric hindrance is apparent in the keto tautomer. It would appear that the equilibrium is shifted toward the enol form because of electron withdrawal from the vicinity of the a-proton. Solvent Effects Carbon Tetrachloride. Infinite dilution chemical shifts of protons of fi-diketones and B-ketoesters in carbon tetrachloride solution are given in Tables XXXV and XXXVI, respectively. Chemical shift values are given in cps and are positive for a shift to high field on dilution. It may be mentioned that the addition of carbon tetrachloride to fi-dicarbonyls results in an increase in the percentage of the enol tautomer. For the acetyl, alkoxy, and a-alkyl-group protons of both B-diketones and fi-ketoesters there is little effect on chemical shift on dilution. The a-proton for the keto tautomer usually shows an upfield shift of about 10-15 cps, but in ethyl a-fluoroacetoacetate the upfield shift is 29 cps. Enol a-proton resonance peaks shift upfield on dilution by about the same amount as the keto a-protons. . Exceptions are hexa- fluoro- and a-chloroacetylacetone in which no shift is observed. The enol OH proton resonance line moved upfield on dilution in carbon tetra- chloride by from 6 to 21 cps in the fi-dicarbonyls. Only u-chloro-, trifluoro-, and hexafluoro-acetylacetone do not show this upfield shift of the OH proton. 130 Table XXXV. Infinite Dilution Proton Chemical Shifts in fi-Diketones in Carbon Tetrachloride)" = Acetyl Compound CH: CH3},< a—Hk o-H OHe Other Ac etylac etone 2 l 10 12 21 - a-Chloroac etylacetone -2 0 l8 - 3 - Hexafluoroac etylacetone ~ 6 - O l - Trifluoroac etylac etone 2 — - 9 -2 - a-Methylacetylacetone 0 3 18 - 16 O- a-CH?’ e k = keto e = enol >‘.< = Chemical shifts are in cps relative to the pure compound. ‘t .chomcsoo 0.3m 05. on o>HHmHoH mnHo 5. ohm mHHHHHm HmoHEozo u .0. .HoovamE n E .Hoao u o 6qu u H HmHIHUOVIO I I I o o I I. I - vumwuumouoom T3302 ... H .H .NI .N .0 HH I mH N o H o oumuoomoHoomHifiHoummmMud HEHHM I N h I NH N Ho 0 HI onHoomoHo “.mH>HHHoEId HiffimH om oH I I NH N. o I I Ho H I onHo omoHo omouooHHHuH Trim .. VH I I I I mm H I m N I ougoomogoomouogfid HHEHH I I NH I mH m E H HI oumuo omoHoomH>£HoIo H>HHM I OH I. . 3 I I o I m N H N uHo oumwoomoHoomoaoHHuId H>~HHM I E .m .mI NH I «H H E H MI 330 omoHoomH>Hdn IzmId HEHHH I N .o . H I NH I mH H E o m I oumuoomouoomH>HsnmImmId H>HHH I I I I I in > on o S aim «3% 8332 3:3 H 5m I o .NI .H I \I I oH H E o NI onHoomoHoomH>EmmmmId HEHHM I N .o .H oH I MH H H o m I ouwuoomoHoomH>HHmIo 38.an I I o w mH H H H m I oumuoumopoom HEHHM VHMHMHHUV HI I m N o o o I I oumuoomoHoom 13de NHIHO 25 HI I o N. 9 HI HI I I onHoomoHo om Hiram HoHHO mason—Una THO mud HIHIo oHovH HocmH NHHU «.30 s o o x a H s a H vHHda THO H 304 H Hum Honso EoO oHuHHoHaHomuuoH conumU CH muvumoouovHIa a: .33ch HmUHEonHO Gouounm coHudHHQ ouHcHHGH .H>XXX oHan. .w 132 Giessner-Prettre (58) explained the shielding of the keto o-CHZ protons on the basis of breaking up of associated groups of ketone molecules which were assumed to be held together by means of inter- molecular hydrogen bonds. However, the present study shows that the enol protons also become more shielded in nearly all B-dicarbonyls. Increased shielding of both the enol OH and enol u-CH protons may be explained as due to breaking up of the intramolecular hydrogen bonds on dilution. The keto protons, if involved in intermolecular hydrogen bonds, become dissociated on dilution to about the same extent as the intramolecular hydrogen bonds. In either case, one assumes that the respective peaks represent an average of the chelate and non-chelate enol form or of the bonded and non-bonded ketone form. Figure 39 shows a graph of chemical shift in cps versus mole fraction for solutions of ethyl acetoacetate in carbon tetrachloride. A spectrum showing the effect of dilution on the proton resonance peaks of acetylacetoneis shown in Figure 40. In Figure 41 are plotted chemical shifts in cps for the OH resonance peaks of a series of compounds at infinite dilution in carbon tetrachloride, relative to the OH resonance peak in the pure compound, versus mole fraction. In each case the shift is taken as positive if the proton is shielded relative to the OH peak for the pure compound. If the shielding does represent principally a break- ing up of the intermolecular hydrogen bonds, then carbon tetrachloride affects this dissociation in the increasing order of hexafluoroacetyl- acetone (least), a-chloroacetylacetone, ethyl acetoacetate, ethyl u-allyl- acetoacetate, ethyl a-ethylacetoacetate, ethyl egg-butylacetoacetate, u-methylacetylacetone, and acetylacetone (greatest). l-Phenyl-l, 3-butanedione on dilution in carbon tetrachloride shows a slight upfield shift for both enol and keto acetyl methyl protons, a negligible shift for keto and enol a—protons, and a downfield shift for the enol OH proton resonance line, which is a very broad resonance peak. 133 Figure 39. Chemical shift of protons of ethyl acetoacetate in benzene and carbon tetrachloride. 4O O-——Benzene Q Q: g __ 0— cc14 G “(2)-9‘6) ‘ ‘ _‘fi ‘ I_‘J 1;] LJ U _ 80.. \e \G\ \ ‘9 ‘ ‘ x \ ~ 1200 a B swab: Q 5‘ a “- C_I:_I3C(OH)- BB 5 B a ‘E®“*‘C§300- 160._ cps N Rs. ‘9 2:00 _ \®\ 240. 9~H%_9@~ H-‘ 13 a B B E} Emfi-‘h—‘EthyICl-Iz 280L HIT] —e'-'J——*e)— ” — - jafil'fififi '5“ '“ — - we — 1. -C(OH)-CI—_I- 1_._.__-__________...-_.._____.__..___H 1r— :1; 720“ a S {3 ‘1' u a H. ...—— EnolOH ...-are” - a/ flag/’6 He’ 7601 9’ f 0 o.'1 o.'2 0.3 0:4 035 0:6 0.T7 0.33 039 1.0 Mole Fraction Solute OS o2 Eh I. 3 mm... mom 11111 J JAM < s 1 N mo <1 Wnofu x Jmo Mao x m o 134 I. \ onmo .H .Ho> om\va «HUD CH ocouoomH>poom Ho 83.30on MHZZ couounm .Qow. 9:3me o2 0.2 SN lads: «mm w? \W . IIJNI‘V u n u H.H o oxmo are a o omo \ onmo .ocoHoomH>Hoom Ho 9330on ”:22 Gouonnw ..moHu oHSmHh 1.35 Figure 41. Chemical shift relative to pure solute of the OH proton in [B-dicarbonyls in carbon tetrachloride versus mole fraction solute. O - B- Diketone s A - fl—Ketoester s Chemic a1 Shift (CPS) -2 I l__ 0 0210.2 0.3 0T4 015 0.6 o.'7 0:8 0:9 1.0 Mole Fraction Solute 136 Benzene. Chemical shift measurements of fi-diketones and [3-keto- esters at infinite dilution in benzene are summarized in Tables XXXVII and XXXVIII, respectively. The shielding of the protons caused by. the usual ring current effect is illustrated by these tables, in which positive shifts refer to resonances which occur at higher applied magnetic fields. Alkoxy, acetyl, and a-keto protons are shifted upfield at infinite dilution by as much as 33, 57, and 50 cps, respectively. The enol a-proton among the esters shows only modest upfield shifts, but among the B-diketones it is shifted up to 46 cps at infinite dilution. Protons of the a-alkyl substituent show little ring current effect. Figure 42 shows a spectrum of ethyl acetoacetate in benzene to illustrate these points. . l-Phenyl-l, 3-butanedione on dilution in benzene gives the usual upfield shifts of enol and keto acetyl methyl proton resonance lines. Both enol and keto u-protons are shielded on dilution. The striking exception to increased shielding is seen in the enolic OH resonance position which becomes de shielded on dilution in benzene. For a series of B-dicarbonyls the infinite dilution shift of the OH proton is from -10 to -45 cps. In Figure 43 are shown infinite dilution chemical shifts relative to the pure OH resonance position for a series of fi-dicarbonyls in benzene versus mole fraction solute. Exceptions to the downfield shift of enolic OH protons are observed in both trifluoro- and hexafluoroacetylacetone. This apparent anomaly will be discussed later. The large downfield shift of the enolic OH proton resonance line, coupled with upfield shifts of other protons suggests that the solute mole- cule is so oriented that the OH proton lies approximately in the plane of the benzene ring. The upfield shift of other protons in [B-dicarbonyls is explained by an orientation of the solute molecules which places these protons above or below the plane of the ring. The formation of complexes with benzene is not without precedent. Reeves and Schneider (95) have proposed that a complex of benzene and 137 Table XXXVII. Infinite Dilution Proton Chemical Shifts in fi—Diketones in Benzene * Acetyl Compound , CH;a CH3 a-H a-H 0H3 Other Acetylacetone 25 33 50 39 ~45 - o-Chloroacetylacetone 25 29 39 - -29 - Hexafluoroacetylacetone - - - 40 +37 - Trifluoroacetylacetone 57 54 - 46 +27 - a-Methylacetylacetone 24 29 47 - -34 20 a-CH},< 32 III-CH;a \ k = keto e = enol * Chemical shifts are in cps relative to the pure compound. 138 .Hoqsomgou oHSnH 0ch 0H o>HHmHoH mmo E 6.3 mHHHaHm HmoHEoHo v”. .HoovamE III a .Hono H o .oHovH u x nHIHDO I MN I I I Hum mm I I I oumuoumouo Um. H5302 03 .0: 3% I : I .... .3 Q... I 2 mm 2 m: m: 838830230803-.. 15m I w omI I wN «N I .wH MN mumpoomOHoomH>5oEId HEHHM 6mm va I I oH I «N S. I I VHH m UHom 3.30 omouo omo~odHHH3 HKEHM I I I I HVHV MN I 0N wN onHoomoHoomonoDHHId HEHHMF I o .0 omI I 0N 0H 8 mH HN manuoomopoomHEHHoId 15M 0N \IN I I mNI I 2“ Nm HN Mom Mom oamuoomouo omonoHrHoI.e HEHHH I 5 HM I I mH NIH 5 0H ON 330 omoHoomHifidQuwad HEIHHH I m .N. .NI mm I I 0 pH as pH ON oumuoomoHoomHfisnmImIHId HEHHH o: I I NNI m HN I I vHNH o .uHHVH onuoomH>oucofl 15M I N. ...... .m mm- I 2 2 E E m: Bflmomsoumifimmflé 15m I m .e .H I «m. I S om 2 2 NM 33333312qu 15m m: 6H N I mm- m m... 2 mm xfi .22 888388 35m 2.2.3on 3 anmoTw I S. N mm mm MN I I 333338 13m Is ...mo 27w I om- m mm S MN I I 383833 13m .H 05.0 QDOHw I. d oHIHO oHIHO Id vHHIH Id 03% Hocm ~30 «HIHU USSOQEOU - 13:... «mo Elmo... 15m xx ocouoom 5 mpoumooHovHIu 5. 323m HmoHEoHU coHoHnH coHHdHHQ vHHaHHGH HHS/XXX oHnHmIHI 139 o. 130m em 8 m2 8.: «mm «0.... a? [4. IIIJ ‘SXAIIH‘{<((I \\ gm). HIHU THO o o \/ Nmu g I > H "\UIMHIHO H H M IO ...mo \ N 1» n H 5H \ VHTHU ouoInmo vH H.Ho> H_m\mv oqoncon GH oumuoomgoom H.230 Ho Estevan MHZZ aoHounH .nHNv vudeh 0 Al mama «.5. OMN :H mmH oON 15m o¢N Nom \\ 1H3. i 13H. . HIHO Exam a £0 .opmuoomopoom 15.6 Ho 93.30on fizz Goponnm .MNHV ousth 140 Figure 43. Chemical shift relative to pure solute of OH proton in B-dicarbonyls in benzene versus mole fraction solute. h 36 "‘ O— B-Diketones A— fi-Ketoesters 32.. T P 286‘ Z4 .. 20~- 16<~ 122. Chemical (CPS) -121- 0 0:1 0:2 0:3 0:4 0:5 0T6 0:7 0.8 0L9 1.0 Mole Fraction Solute 141 chloroform exists. Hatton and Richards (124) concluded for amides in benzene solution that the more positive end (the methyl protons) of the amide lay above or below the plane of the ring and the more negative end (the carbonyl group) tended to be as removed from the n-electron cloud as possible. Hatton and Schneider (96) confirmed the existence of molecular complexes with benzene by measuring the temperature coefficients of solute chemical shifts. Schneider (97) determined, in a study of solute-solvent interactions in benzene, that a roughcorrelation exists between the ratio of dipole moment to molecular volume and the chemical shift of the solute molecule. A specific association of %' (acetylacetone)ethylenediimine in benzene was reported by Dudek and Holm (81). An attempt was made to determine the nature of the orientation of B-dicarbonyls in complexes with benzene. Johnson and Bovey's tables for the calculation of chemical shifts of protons in the vicinity of the benzene ring (94) have been used to attempt to relate the solute mole- cule chemical shift measurements to distance from and angle with respect to the benzene ring. A comparison of the infinite dilution chemi- cal shifts for ethyl acetoacetate has been made with the so-called 'isoshielding' lines around benzene and, . with the tables of chemical shift versus Z and P distances (perpendicular and parallel tothe benzene ring, respectively), gives a qualitative picture of the interaction. There is apparently no way of orienting the benzene ring with respect to the solute molecule in order to simultaneously give the large upfield shifts of both acetyl methyl and ethoxy protons, and the large downfield shift of the enolic OH proton. However, an association involving two mole- cules of benzene with each solute molecule does give a qualitative and quantitative explanation of the observed chamical shifts. With the aid of molecular models, the interaction proposed in illustrated in Figure 44, which pictures the acetylacetone-benzene complex. Reference to the 142 Figure 44. Proposed complex of acetylacetone with benzene. / O/H“‘~O \A\ H HH H J i, H H \ /‘/ \iy \/ \\C/ \’\\ / (7 H I A \A \ H H 4 A \ \’\ 143 figure gives an explanation to the two exceptions noted above, namely the trifluoro- and hexafluoroacetylacetone molecules. The substitution of fluorine atoms for the methyl protons of the ac etyl group would result in the highly electronegative atoms lying in the area of greatest n-electron density, which is a most electrostatically unfavorable situation. It may be noted that ethyl trifluoroacetoacetate gives the least deshielding of the enol OH proton, which may indicate association of benzene with only the alkoxy end of the solute molecule. Dilution in benzene results in an increase in the enol tautomer. It may be that the benzene- solute complex occurs only with the enol tautomer and shifts the equilibrium in favor of the enol tautomer. Hexane. The chemical shifts of protons of B-dicarbonyls are given in Table XXXIX at infinite dilution in aIii-hexane. Figure 45 is a spectrum of butyl acetoacetate in dill-hexane, and the resonance positions may be compared with those of pure butyl acetoacetate in Figure 22a. Negligible chemical shifts are observed for alkoxy protons. A small shielding effect is seen for the acetyl protons. Both enol and keto a-protons are shielded in solution to a considerable extent relative to the pure liquids. Generally the enol OH resonance peak is slightly shifted to lower fields in solution. An association of the solute molecule with an inert solvent such as hexane appears rather unlikely. Upfield shifts of the a-protons in both enolic and keto tautomers might be explained by breaking up of intermolecular association of solute with solute molecule. However, such association does not appear to affect the intramolecular hydrogen bond. Very dilute solutions of fi-dicarbonyls in hexane approximate the gas phase in terms of percentage of enol tautomer. At least we may say that the absence of upfield chemical shifts of the enol OH proton resonance peak indicates that strong hydrogen bonds exist in hexane 144 .Hocdomgoo 0.3:» 05. o» 02.3.30.” 0&0 CH 0.2m mHHHHm HmoHE0£O HoovamE I E HoG0 I 0 33H I 0H 3.300 H8300: 305 3 X3. Ammooovm I I 3 - I m I o 8800838 1502 I 0 0H NIH I I N. w H 0 253001304 I MI 0H H H o N. m 0H 330003000 35H I m I H H E H. N CH 0HmH0000H000H>£p0§Id 35H I MI I MN m E m m mm 0HmH00m0u00m0hoH£0Io 35M $50 030 030 2&0 £0 £0 83H Sam :03 .o “a 3.5950 coHoHnHIo HIHHHHM «EU HNHM00IConHu00HQIQ CH mcouounw Ho mHHHHHm H00H50£U .XHXXX 030R. 0". 145 0.: mm: III 0 2.: 0.3.. «E . I. III; . II- I. \s... ii $0.5 .I Hi HE. . I~H1HUOOU I _fij _ Eéi 3: .. # __AI\omO £0 Ono .00 ~00 H.Ho> mN\oHv 000x0aHIfltcH 030300003000 H300. Ho 85.3009... MHZZ couch“ .mv 0uderA 146 T gum o 0 mm m2 ...2 SN If (JIAM i fl I a ".0 £0 a mo 0 2 £00 000 m.mo 06:00 x \ if”: .H .Ho> mH\~INV nHUHIHU 5 030300003000 H5593? 53.30090 MSHZ acuounm .Hov 0.3mHh o Almmum ow N: 02 2: mom m? I H ..< Ifi Ia! . 4. H. ubmmo J VM00 00 00030 000 000 \ Inqmov 030300003000 H3353? 55.3025 #322 cououna .03.. 0udeh 147 solutions and the position of the equilibrium, which favors the enol tautomer in this solvent, appears to support this interpretation. Chloroform. Chemical shifts of protons of B—dicarbonyls at infinite dilution in chloroform are given in Table XL. A typical NMR spectrum is shown in Figure 46 forAtv-butyl acetoacetate in chloroform. Both alkoxy and acetyl protons show some deshielding in this solvent. In general, there is little effect upon keto or enol a-protons on dilution in chloroform. However, the keto a—proton of ethyl a-chloro- acetoacetate is considerably shielded on dilution, and this may be evi- dence for the existence of an intermolecular hydrogen bond. However,- the enol OH proton resonance position does not appear affected by chloroform. Therefore, one might assume the intermolecular hydrogen bond of ethyl o-chloroacetoacetate to be between keto molecules, perhaps of the type shown in Figure 47. c /\ 1 \ 0 CH3 Figure 47. Association of keto molecules in ethyl a-chloro- acetoacetate. The effect of chloroform on the resonance position of the enol OH proton is in general negligible in fi-ketoesters. The upfield shift of the enol OH proton in ethyl benzoyl acetate is unique among the B-ketoesters. In the B-diketones studied chloroform appears to cause shielding of the enol OH proton. The position of the enol OH resonance in fi-ketoesters on dilution in chloroform appears to indicate that the intramolecular hydrogen 6559930 03$ 03 o» 0>30H0H 090 CH 0.30 dem H00HE0HU a. 30H I H 8 3:0 I 0 4 1 0007.008“ I E I H m N v N I N I I I mo 0G030 00130 000 HOSHHHH H. HmHIHOIdv m I. N I I N H. E oI bI I 03030000300035.0830 HEHHM I MI I mH MI NI mI 0H .0NI 0H 030300003000ouoH£0Id HEHHM UAW! QMI I NN HI N I I I Hm I vHoH I 0N 03030 00 300000. HKEHM I N m m m I o m I vI HI 03030 00 030 00 13H «130v m o HuI I m m I I I 030300003000H>35m3x I . ON N o mI HVI I I ww 020300033004 H0HHO HIHO HIHId HIHId anEU 0nHIHO «$0 «HIHU Hocmm . 0559880 vH 1300.4I HWHHH 300.3% "I *EHOHOHOHHU CH GOHHDHHQ 03HCHHGH 30 mHI>GonHIH00HQIu CH mcouohnm Ho mHHHaHm H00HE0HU :IHX 030R. 149 bond is stronger than in carbon tetrachloride. For the (S-diketones, however,“ this resonance peak is shielded to about the same extent as in carbon tetrachloride. Evidently chloroform does affect the intra- molecular association of the (3—diketones more than the esters. l-Phenyl-l, 3—butanedione shows little change in chemical shift except for the enol OH, which moves far upfield on dilution. Thenoyl tri- fluoroacetone also shows increasing shielding of the enol OH proton on dilution in chloroform. ~ Korinek and Schneider (125) have obtained the difference between the chemical shift of chloroform at infinite dilution in inert hydrocarbon solvents and in donor solvents. On the assumption that the donor mole- cules do not show self-association, the resultant difference in chemical shift is the association shift of chloroform and may be correlated with the hydrogen-bond strength and the relative donor strength of the solvent molecule. . For the present study, it has been assumed that the fi-dicar- bonyls do not show appreciable intermolecular association. Further, the shift of the chloroform proton peak in inert hydrocarbon solvents relative to pure chloroform is constant at +10 cps at infinite dilution. Therefore, a graph of the chemical shift of the chloroform proton resonance at infinite dilution in B-dicarbonyls versus mole fraction of chloroform will be meaningful in a relative manner. It may be seen on the basis of such a plot in Figure 48 that the best donor molecule is ethyl acetoacetate and the poorest is trifluoroacetylacetone. It would appear then that the most extensive hydrogen bonding of chloroform with the solvent molecule occurs with ethyl acetoac etate. In general it may be noted from Figure 48 that the [S-ketoesters associate with chloroform to a greater extent than do the [B-diketones. It will be recalled that the B-ketoesters exist for the most part as the keto tautomer, whereas the fi-diketones are mostly enolic. It is sug- gested that chloroform tends to associate with-the keto form of the 150 fi-dicarbonyl molecule rather than with the enol tautomer. The absence of the upfield shift of the keto and enol a-proton as was observed in carbon tetrachloride may be taken as evidence that the a-proton is not involved in intermolecular hydrogen-bond formation. However, the 8- dicarbonyl may be associating with the solvent through a hydrogen bond between the carbonyl group and the chloroform proton. . The ex- ception is ethyl a-chloroacetoacetate, in which the keto a—proton is shielded in solution relative to the pure compound. The position of this compound in Figure 48 indicates that it associates with chloroform to a greater extent than do other fl-ketoesters, and it may be that such association is involved with the enol rather than the keto form. Carbon Disulfide. Infinite dilution shifts for protons of acetyl- acetone and ethyl acetoacetate are given in Tables XLI and XLII, respectively. Once again little change in chemical shift occurs in the alkoxy or acetyl methyl protons, whereas the keto and enol a-protons and enol OH resonance peaks are shielded on dilution, as in carbon tetrachloride. This may indicate dissociation of both intra- and inter- molecularly hydrogen-bonded groups. However, carbon disulfide favors the enol tautomer. Thenoyl trifluoroacetone shows shielding of the enol OH proton on dilution in carbon di sulfide. E1333. Infinite dilution shifts for protons of acetylacetone and ethyl ac etoacetate are in Tables XLI and XLII, respectively. In this comparatively inert solvent, the enol and keto u-protons are shifted upfield on dilution. The enol OH resonance position moves to higher applied fields in the B-diketone, but not in the B-ketoester. Yet this solvent favors enolization over the pure compound in both cases. Acetic Acid. Chemical shifts for this solvent are given in Tables XLI and XLII and show little change on dilution, except for the enol OH 151 Figure 48. Chemical shifts-of chloroform proton in fi-dicarbonyls as donor solvents. A Ethyl acetoacetate O Ethyl a- methyl acetoacetate D ,tv-Butyl acetoacetate O Acetylacetone CD Ethyl a -chloroacetoacetate [D Trifluoroacetylacetoacetate Y 4700 4604- ‘I O c13013 _ (Cpsl 450“- 440.. inert cyclohexane 430-- l A fl 1 Y 0 0.1 0.2 0:3 0:4 0.5 0.16 0:7 0.8 0.9 1.0 Mole Fraction B-Dicarbonyl 152 Table XLI. Infinite Dilution Proton Chemical Shifts in Acetylacetone . . * in Various Solvents Ac etyl; ' 0.: Proton Solvent (31:? 0135— 014? ~ CH6 OH Acetic acid -1 -1 -2 O 32 Ac etonitrile -1 O 0 -4 - lO Benzene 25 33 50 39 -41 Carbon disulfide 3 4 14 13 16 Carbon tetrachloride 2 l '10 12 21 Chloroform -4 -5 O 2 20 Dimethylsulfoxide - 2 3 - 5 -8 68 Dioxane 1 2 m 2 38 Absolute ethanol -1 -l m -l 595 95 Percent Ethanol - 3 -4 m -4 605 Ether 3 4 m 7 21 Hexane 8 7 l7 l6 6 Methanol -2 -3 -4 -3 625 e = enol, k = keto, b = broad, m = masked 31k Chemical shifts are in cps relative to the pure compound. 153 .Hossomgoo 0.39 03 3 0>HH0H0H who 5 0H0 33am H00H50HHU ... 0001005 .I. 5 .300 n 0 .oH0vH 0 0H H H I N - HI H N. H I 35.5on m I oH NH HI w H o 0c0x0HIH N I o H H H. H. o o .850 N o m- N. H. N. H. Hoamfim 820m mo H N HI H. H H. N. Honmfim 83034. N N N m N H H- 2:385 o m I o I N o N H 0308323385 N N m m I o m I H. I 0283020 w o 0H N N o NI 0030300300. 09:00 m H. m H m H. 0 ..HH. H $0320 £5.30 0mH oHH mm. H. 3 NM 0N H2 VHS 0:3ch H o H I H E o 0 03333004 H. N H.I NI H m- N. Hum 0:34 00 000 “00 “mo “00 .00 .00 .5300 4138.4. H.300 GOHOHnHIu mHG0>Hom mdoHH0> CH 030300003004 H>HHHMH CH mHHHHm H00HE0HHU CouoHnm HHOHHDHHQ 0HHGHHGH .HHAX 03.08 H. 154 proton in acetylacetone which is shifted upfield and is very broad. Ethyl a-methylacetoacetate gives essentially the same results as ethyl acetoacetate. In dibenzoyl methane the OHpeak is again very broad and is separate from the carboxyl proton peak of the solvent. Although an increase in percentage of keto tautomer is observed, only a slight upfield shift of the enol OH is observed. Trifluoroacetyl- acetone shows little effect, except for the enol OH proton resonance which has coalesced with the OH resonance peak of the solvent and consequently shows a dilution shift of close to 200 cps. Thenoyl tri- fluoroacetone gave results similar to those of the other B-diketones with separate resonance peaks for the enol OH and carboxyl OH groups of the solvent. Again little dilution effect is seen, other than for the enol OH resonance position. Protons of 1, 3-indanedione show little change from their position in chloroform. Little chemical shift is noted for the protons of l-phenyl-l, 3-butanedione in acetic acid as compared with the same resonances in carbon tetrachloride solution. Further, no change is noted on dilution, except for the enol OH proton which moves slightly upfield. Acetonitrile. As seen in Tables XLI and XLII, there is little effect upon resonance positions of protons in either ethyl acetoacetate or acetylacetone by dilution in this solvent. Yet acetonitrile favors the keto tautomer. Only the enol OH proton resonance position of acetylacetone varies on dilution, and this is deshielded somewhat with increasing dilution. DimethLlsulfoxide. Chemical shifts in Tables XLI and XLII indi- cate little dilution effect for most proton resonances of either acetyl- acetone or ethyl acetoacetate. The solute enol OH proton does show considerable shielding upon dilution. One might expect competition for the enol OH proton by this solvent, especially since dimethylsulfoxide 155 favors the keto tautomer. The fact that this resonance peak becomes extremely broad does appear to indicate association with the solvent. Alcoholic Solvents. Chemical shifts in absolute ethanol, methanol, and 95 percent ethanol are given in Tables XLI and XLII for acetylacetone and ethyl acetoacetate. Only methanol affects the position of the tauto- meric equilibrium, and it causes a decrease in the percentage of enol tautomer. These solvents show little dilution effect on ethyl acetoacetate, but in acetylacetone the enol OH resonance signal is coalesced with-the solvent OH resonance throughout the concentration range between the pure solute and infinite dilution. A spectrum of acetylacetone in methanol is shown in Figure 49. The half width of the coalesced OH signal becomes somewhat narrow on dilution. Calculations made from the shape of the OH signal indicate an exchange time of the order of 10" see”, if one assumes equal relaxation times for OH protons of solute and solvent molecule. Triethylamine. The proton resonances of methyl acetoacetate show no chemical shift in this solvent. at; Butyl acetoacetate and ethyl u- allyl- acetoacetate show only slight chemical shifts of enol and keto protons, but the enol OH resonance peak is evidently too broad for detection although the amount of enol tautomer increases on dilution. The keto a-protons are shielded somewhat on dilution. In ethyl acetoacetate only keto peaks appear visible, but these are considerably broadened, and there is evidence of chemical reaction by discoloration of the solu- tion. Among the fi-diketones, acetylacetone has been studied by Reeves (2.3) and Reeves and Schneider (57) in this solvent. Results are com- parable to the present study. Only the enol tautomer is found, and the most significant change in chemical shift on dilution occurs with the enol OH resonance peak, which moves upfield. In l-phenyl-l, 3-butanedione 156 HNH -Ir- Hem on n20 ovH monmo CNN AI. mmoHIH mmm qr W< NED HIHO H... _HH: 3:. .H...HHH_H___._H._.H _H___H._..._._.H H..:.HHH .H..H..H._HH.. ....HHHH H__._H.H..._H......__HHH.H..._HH__.H__H.H_H_.H_H...H.__._.H_..__...._ ImHo .HuHonmo .CmHm .ov Hocmfiofi HHH ocoaoomHEoom Ho 5.9.50on .522 nouounm .oHN ondeh 157 chemical shift changes on dilution are negligible except for the enol OH which in a saturated solution is 200 cps upfield of its value in carbon tetrachloride. In more dilute solutions this resonance peak broadens considerably. Dibenzoylmethane in a saturated solution of triethylamine shows little chemical shift effect on dilution, except for the enol OH which moves from the value of 993 cps in carbon tetrachloride to 660 cps in a saturated solution and to 267 cps in a more dilute solution. For thenoyl trifluoroacetone, the results are much the same as for other fi-diketones, with only the chemical shift of the enol OH proton resonance peak being affected by the solvent. Diethylamine. In l-phenyl-l, 3-butanedione the enol form is present, but the enol OH is so broad as to preclude its detection. The enol a-proton is also broad and is increasingly shielded on dilution, but the keto a-proton is masked by this solvent. The amine proton on the nitrogen is deshielded on dilution. In dibenzoyl methane the enolic a-proton is not much affected by dilution in diethylamine either in line- width or in chemical shift. The NH proton resonance peak, in contrast to its behavior in l-phenyl-l, 3-butanedione, moves upfield on dilution and becomes sharper. The enol OH proton resonance is too broad for detection. Among the fi-ketoesters only methyl acetoacetate does not appear to be affected by diethylamine. InAtV-butyl acetoacetate and ethyl acetoacetate there is evidence of chemical reaction, since even the acetyl methyl proton resonance signal is broad. No enol tautomer is seen. Glickrnan and Cope (f26) have reported the following reaction between a B-dicarbonyl and a secondary aminef cmcocmcoom + (CH3)ZNH —— CH3C|3=CHCOOEt . (50) N(CH3)Z For t-butyl acetoacetate in'diethylamine the keto cc-proton resonance «0 signal has completely disappeared, but for ethyl acetoacetate it is very 158 much broadened and combined with the NH proton. On further dilution this coalesced signal moves upfield. In ethyl a-allyl- and a-ifls‘gamyl- acetoacetate, the equilibrium does not appear to be affected by this solvent, and the principal chemical shift effect is broadening of the keto a-proton. The NH proton resonance signal moves upfield on dilution in these solutes. Acetylacetone has been studied in this solvent by Reeves (23) and Reeves and Schneider (57). The enol tautomer is favored, and there is evidence of complex formation with this tautomer. Thenoyl trifluoro- acetone apparently undergoes chemical reaction with this solvent. Other Bases. In both pyrrole and pyridine the enol OH proton of thenoyl trifluoroacetone is considerably shielded on dilution. Solvent Effect on Keto-Enol Equilibrium In Table XLIII and XLIV are given equilibrium constants for acetylacetone and ethyl acetoacetate in various solvents at 0. 1 mole fraction solute. For these fi-dicarbonyls benzene, carbon tetrachloride,» Liv-hexane, carbon disulfide, and ether increase the percentage of enol tautomer on dilution. This effect is greatest ing-hexane. The increase in the enol tautomer in these solvents is observed for all the fi-dicarbonyls studied. For B-ketoesters with alkyl groups substituted in the a-position, this increase is less pronounced. In Table XXXIX the percentage enol is given for several fi-dicarbonyls at 0. 5 mole fraction in/rxhexane. For chloroform the percentage enol at infinite dilution is given in Table XL. Those solvents in which little effect is observed on the keto-enol equilibrium are dioxane, chloroform, absolute ethanol and 95 percent ethanol. Those solvents which result in a decrease in the enol content of fi-dicarbonyls are methanol, acetic acid, acetonitrile, and dimethyl- sulfoxide. Equilibrium constants are given in Tables XLIII and XLIV. . ‘T‘“"-‘"““" 159 Table XLIII. Equilibrium Constants for Tautomerization of Acetyl- acetone in Various Solvents at O. l Mole Fraction of Solute* Percent * Solvent Enol Ke Hexane 95 19 Carbon tetrachloride 96 2.4 Ether 95 19 Carbon disulfide 94 16 Benzene 89 8. 1 Chloroform 87 6. 7 Dioxane 82 4. 6 Ethanol 82 4. 6 Pure 81 4. 3 95 Percent Ethanol 77 3.4 Methanol 74 2. 8 Acetic acid 67 2. 0 ACetonitrile 62 l. 6 Dimethyl sulfoxide 62 1. 6 enol * 0 Where Ke = and measurements are at 33 j; ZIC. keto ’ 160 Table XLIV. Equilibrium Constants for Tautomerization of Ethyl Acetoacetate in Various Solvents at O. 1 Mole Fraction of Solute * Percent * Solvent Enol Ke Hexane 39 0. 64 Carbon tetrachloride 28 O. 39 Ether 22 0. 29 Carbon disulfide 20 O. 25 Benzene 16 0. 19 Dioxane 11 0. 12 Absolute ethanol 10 0. 11 I Chloroform 7. 5 O. 081 Pure 7 . 5 0 . 081 95 Percent Ethanol 7. 2 0. 078 Methanol 5. 8 O. 062 Ac etonitrile 4. 9 O. 052 Dimethyl sulfoxide 2. 2 0. 023 Acetic acid 1.9 0.019 * _ enol e keto o , and measurements are at 33 :1: 2 C. 161 Figures 50 and 51 give the percentage enol at various mole fractions for acetylacetone and ethyl acetoacetate, respectively, in several solvents. It may be seen that the percentage enol for ethyl acetoacetate extrapolated to infinite dilution in pu-hexane is about 50 percent, that of the gas phase. This result may be compared with the findings of Powling and Bernstein (36) for hexane solutions of fi-dicarbonyls in which the infinite dilution free energy of tautomeriza- tion approached that found in the gas phase. A strict comparison of equilibrium constants obtained from nuclear magnetic resonance spectroscopy and from other methods is not possible, since the concentration of solutions is not indicated in the bromine titration and infrared techniques. However, Table X shows qualitative agreement for the percentages of enol tautomer in acetylacetone and ethyl acetoac etate by bromine titration with the results of Tables XLIII and XLIV by NMR. Qualitative agreement may also be noted between Kabachnik's (8, 9) results in Table XI for ethyl acetoacetate by bromine titration and Table XLIV by NMR. Results from NMR do not show good agreement with those of Conant and Thompson (55) in Table XIV for a-alkyl substituted B-dicarbonyls. The present NMR study does not show appreciable in- crease in the percentage of enol tautomer in hexane for such compounds. Good qualitative agreement is found between the work of Murthy e_t a_._1. (35), from ultraviolet spectroscopy in Table XV for the percentages of enol tautomer in acetylacetone and ethyl acetoacetate and the NMR results shown in Tables XLIII and XLIV. Dewar (60) has stated that solvents with higher dielectric con~ stants favor the keto form, since the carbonyl group of the keto tautomer is more polar than C=C-OH in the enol tautomer. Therefore, there is an increase in solvation energy for the more polar tautomer. 162 Figure 50. Percentage enol tautomer in acetylacetone in various solvents. 1%Hexane 2 Carbon tetrachloride 3 Carbon disulfide 4 Ether 5 Benzene 6 Chloroform 7 Absolute ethanol 8 Dioxane _ 1004— 9 Acetic acid 10 95 Percent Ethanol 11 Methanol 12 Dimethylsulfoxide 13 Acetonitrile 90- Perc ent l Enol T autom e r I 80. 7o. 601 50‘ J L o.'1 ofz 0.13 of4 o.‘5 T6 057 o.'é o.'9 1. Mole Fraction Solute Figure 51. 163 Percentage enol tautomer in ethyl acetoacetate in various solvents. 52+ 48” 44* Percent Enol 40 v- Tautomer 36* 7 32« 28.. 24.» 20‘" 7 16+ 1 EHexane 2 Carbon tetrachloride 3 Carbon disulfide 4 Ether ,9 5 Benzene o 6 Chloroform 7 Absolute ethanol 8 Dioxane 9 Acetic acid 10 95 Percent Ethanol 11 Methanol 12 Dimethylsulfoxide 13 Acetonitrile 0 D E] 44, c. I ° 1 E1 9 l A A El 0 El 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Mole Fraction Solute 164 Wheland (62) has used the van't Hoff—Dimroth relation as an explanation of the increase in enolization in nonpolar solvents. , This relation defines the quantity, G, as follows: 1 = Lenoly(sole) _ so k G_ [keto]/(solk) - Ke( sole )’ (51) where Ke is the equilibrium constant for enolization, and sole and solk are solubilities of enol and keto tautomer, respectively. The quantity, G, is independent of the solvent if solubilities are not too large. Therefore, the solubility of enol relative to keto is greatest in the least polar solvents. Powling and Bernstein (36, 127) have assumed that solvent inter- action could be described on the basis of the reaction field of Onsager and have taken the cavity to be that occupied by the solvent molecule. The enthalpy of enolization in the gas phase is given by AH =AH +(_€__-_1__a, (pi-pi). (52) g soln 2€+l M solvent where AHS is the enthalpy of tautomerization in solution, 6 the oln dielectric constant of the solvent, p and M the density and molecular weight of the solvent, respectively, and p4 and u; the dipole moments of enol and keto tautomers. Powling and Bernstein have obtained a linear relationship for the enthalpy and free energy of tautomerization in a graphof these thermodynamic quantities versus the solvent property of Equation (52). They have taken literature values for the free energies and enthalpies of tautomerization of the B-dicarbonyls. Their results are shown in Figures 52 and 53 for acetylacetone and ethyl acetoacetate, respectively. Also given in these figures are some re- sults from the present study. The solvent property has been calculated according to Table XLV, using values in Lange's Handbook (128). .xuuomonm uno>H0m 05 068.33. 3 .HHHom G§Hoo HHH 0339an 33530 03H. ...... .HwNHV uHoonHquHIH 90954 Scum... 165 . . . . HocmHIHHoSH 0mm.H+ 000.0I 0 HH 00NN0> 0 omNHw Mm . . . ocmonIH 00.0 00¢.HI HHN H omvmo 0 00N0w H o O I vaflm o o .l 0 ¢ 000 H... 0mm 0 0n n omN00> 0 oomHum ocmzum O O C H 0HIH.H+ 000.HI H0 0 omowfi 0 mmm Hum . . o m0.H o HN.N ocmonQ 0mN.H+ 0m> 0I *0 w. oom omN . . . . EHoHouoHHHU 00¢.H+ 000 0I 0H... HN 00N0HN H oome HN NU . . . . oUHuoHfiomHHou cosh 000.0... 00H HI mm M oomom H OmNmm N . . . . oHuHHHSmHHu £00.th 0mN.H+ 0m» 0I mm H. 00N0N H 00N¢0 N . . . ocouaom 000.0... 000.HI 0m N 00N0>w 0 omNbN N 4 o o o o ”HMHHMGOHOU 000.N+ 00H 0I HNH 0 oommwh 0 oomm Fm . . . . HVHom oHuoo< 00>.H+ 00H. 0I No.0 oommo H oommH 0 I» o oumuoomoonIH‘ occuoumH Ho 4 #2 H +UN: Q m uno>Hom H H M HNIoH d; H I u ... ... oHoE\HmovH «... .md inomoum HGo>Hom 05 Ho mode> HEN mHno>Hom msoHHm> HHH :oHHomnh oHoH>H H .0 Ha mH>conHmoHQIa Ho coHHmuHuoEoHHHmH Ho mmHmIHoHHHw ovum .2.va 3nt 166 Free energies are those calculated from equilibrium constants in Tables XLIII and XLIV at 33°C. For acetylacetone in Figure 52, line A includes the so-called inert solvents. The free energy of tautomerization extrapolated to infinite dilution is. -2. 1 kcal/mole. Line B includes the alcoholic solvents and gives a free energy value at infinite dilution of -1. 8 kcal/mole. Line C includes all other sol- vents, and the free energy of tautomerization value at infinite dilution is again -1. 8 kcal/mole. For ethyl acetoacetate in Figure 53, the inert solvents (line A) may be extrapolated to give a free energy of tautomerization at infinite dilution of +0. 1 kcal/mole, approaching the gas phase value. Alcoholic solvents (line B) give an extrapolated free energy of tautomerization at infinite dilution of +0.6 kcal/mole, and other solvents give +0. 2 kcal/mole. In general, a strict compari- son'may not be made between results of Powling and Bernstein and those from nuclear magnetic resonance in the present study. Free energies used in Powling and Bernstein were taken from several sources, but principally bromine titration studies which are not too reliable. Long-Range Proton Spin-Spin Coupling Takahaski (129) has reported a long range proton spin- spin coupling constant, JAB’ for CH£COCHB(CH3)Z of 0.45:]:0.05 cps and for CH3CHfCOCH3B of 0.4810. 05 cps. Takahaski has observed that the long range coupling in these ketones may be due to n-electrons in the vicinity of the carbonyl group. Kowalewski and de Kowalewski (130) have measured long-range couplings in HCOOCHZCH3 as 0. 85 and 0.57 cps for JHCOO-CHZ and JHCOO-CH3’ respectively. In the present study of CH3COCH2COOCH2CH3, these couplings are 0. 5 and 0. 3 cps for chz—coocgz and chz-coocmcg,’ Long range spin- spin coupling constants for the fi-dicarbonyls respectively. are given in Table XLVI. The coupling constant, JAB’ between the 167 ... .13.... 303on SH H+wN $on .ImIH H Iv: NH HH 2 s w IN. w m. H. IN IN H o q 1 q / H H 1 H H 1 4 . H o .0 /B/ / I. N .0 / HNmH £on53 / pad mcHHBOnH Eoum HoonHsonUIIQ . . Hmmv :Houmcuom Hons wGHHgomIID HN 0 / , H83 £5.10 . 0.0 r w .0 H o.H . N .H . HuHoE\HmBHH oHHuuHcoHoo< mH T MGI H8232 HH H. H30 32. 033.4. o 1 onmonQ m T 0.H Hocmfio oHDHOmnH< N. 5.800.830 0 odoucom m . w.H opHHHdeHu donudU HN uofim m I opHnoHAomuHoH GonumU N 0 .N ocmxomufi H N.N .>HnonHounH aco>H0m msmuo> ocouoomH>Hoom H0w :oHHmuHuoEoufimH Ho >mnoco oourm .Nm ouSmHh 168 uG®>HOm 2 H + wN E: x IN. H IHIIIIW HH 2 HH .... H. I. H H. m H. H I... H O I. N .N B / 2 / I o .N / 33 5333me / Hug mcHHsvonH Scum HuoumHfioHngIQ . ... o / HNmH 532:5 H28 mzHHsonHIIU I w .H m/ H 601/ VHIHOB QEWO / . . Ql//I / if. 0 H / .,// . / 47. Hw H. N / a. .. N .. / 4s. / 63.333004 mH H8232 HH 3on 03004. 0 _ onmonQ 0 H0550 ouaHOmanI N. EMOHOHOHHHU 0 odoucom W 3.53:. E530 H.‘ Sfim m opHuoHHHomuHoH Gonumu N onNoHHIsmm H / Ln... Lo ”mmo oHoE\H3VH m4 ..Snomoum Hco>H0m msmuo> onHoomoHoom H>HHHo HoH noHHmNHuoEou—SHNH Ho 13.5%... oouh .mm 0.59me 169 Table XLVI. Long Range Spin-Spin Coupling in fi-Dicarbonyls 3:: Skin Coupling (lugs) Compound JAB JCE JDE JCD Other CH3COCH2COCH3 0. 44 - - - CH3COCHmo oHoE\H.mu 308>mu M mod AOoV omcmm pGSOQEoU md Ed MAN 0mm ... oudomuomgorw 0mm 0mm manoapmofifluu GM minfifisvm GofimNCoEOHSMH on”. no“ moflficmsfl Baggpoguosfi 42>..Cm 3nt 177 tautomer. The enthalpies of tautomerization are too high algebraically for acetylacetone and too low for ethyl a-chloroacetoacetate. EntrOpies of tautomerization indicate a more ordered system for the enol tautomer as compared with the keto tautomer. Both 0.- chloroacetylacetone and ethyl trifluoroacetoacetate have fairly large negative entropies of tautomerization. It is suggested that these com- pounds have very strong intramolecularly hydrogen-bonded species. SUMMARY The tautomeric equilibria of fi-dicarbonyls have been investi- gated by nuclear magnetic resonance spectroscopy. Chemical shift measurements have been made in a number of solvents, and equi- librium constants ([enol]/[keto]) have been determined. The equi- librium has been studied as a function of temperature, and thermo— dynamic quantities have been obtained from the experimental data in the case of several fi-dicarbonyls. 178 10. 11. 12. 13. 14. 15. 16. REFERENCES . K. H. Meyer, Chem. Ber., 4_5, 2843 (1912). . K. H. Meyer, Chem. Ber., 4:42 2718 (1911). . J. D. Park, H. A. Brown, and J. R. Lacher, J. Am. Chem. Soc., 15.: 4753 (1953). M. D. Ratnakar and K. K. Mahajan, Vikram, J. Vikram Univ., _2_, 37 (1958). . M. J. Malawski and W. Czerska, Roczniki Chem., 33, 491 (1960). . B. Eistert and W. Reiss, Chem. Ber., E, 92 (1954). . B. Eistert and W. Reiss, ibid., 8_7_, 108 (1954). . M. I, Kabachnik, S. T. Yoffe and K. V. Vatsuro, Tetrahedron, _1_, 317 (1957). M. I. Kabachnik, S. T. Yoffe, E. M. Popov, and K. V. Vatsuro, ibid., _1_2_, 76 (1961). M. 1. Kabachnik, S. T. Yoffe, E. M. Popov, and K. V. Vatsuro, Zh. Obshch. Khim., 31, 2122 (1961). M. I. Kabachnik, e_t a_1., ibid., 3_1, 2682 (1961). M. I. Kabachnik, gt a_1., Dokl. Akad. Nauk SSSR, 144, 802 (1962). B. Eistert and F. Geiss, Tetrahedron, l, 1 (1959). H. Henecka, Chem. Ber., 81, 189 (1948). M. S. Newman, ed. , Steric Effects in Organic Chemistry, John Wiley and Sons, Inc., New York, (1956). P. Rumpf and E. La Riviere, Compt. Rend., 244, 902 (1957). 179 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 180 G. 5. Hammond, W. G. Borduin, and G. A. Cuter, J. Am. Chem. Soc., 81, 4682 (1959). R. Filler and S. M. Naqvi, J. 0% Chem., _Z_6_, 2571 (1961). K. L. Wolf, Physik, z., 3_1__, 227 (1930). c. T. Zahn, Physik. z., 23, 570 (1933). M. Beyaert, Natuurw. Tijdschr., _1_§_, 197 (1936). R. J. W. Le Fevre and H. Welsh, J. Chem. Soc., 1909 (1949). L. W. Reeves, Can. J. Chem., _32, 1351 (1957). N. V. Sidgwick, J. Chem. Soc., 907 (1925). B. Eistert, F. Arndt, L. Loewe, and E. Ayca, Chem. Ber., 84, 156 (1951). S. Bratoz, D. Hadzi, and G. Rossmy, Trans. Faraday Soc., 22, 464 (1956). R. S. Rasmussen, D. D. Tunnicliff, and R. R. Brattain, J. Am. Chem. Soc., '_/_l_, 1068 (1949)- D. N. Shigorin, Izv. Akad. Nauk SSSR, Ser. Fiz., _1_4_, 395 (1950). D. N. Shigorin, Zh. Fiz. Khim., _Z_g, 584 (1954). R. Mecke and E. Funck, Z. Elektrochem., 62, 1124 (1956). C. L. Angell and R. L. Werner, Australian J. Chem., 6, 294 (1953). L. J. Bellamy and L. Beecher, J. Chem. Soc., 4487 (1954). H. F. Holtzclaw, Jr., and J. P. Collman, J. Am. Chem. Soc., _7_9_, 3318 (1957). Ch. Tamm and R. Albrecht, Helv. Chim. Acta, :12, 768 (1960). A. S. N. Murthy, A.Ba1asubramanian, and C. N. R. Rao, Can. J. Chem., _4_0_, 2267 (1962). 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 181 J. J. Powling and H. J. Bernstein, J. Am. Chem. Soc., _7_3_, 4353 (1951). K. Kuratani, Rep. Inst. Sci. Technol., Univ. Tokyo, 6, 217 (1952). R. Belford, A. E. Martell, and M. Calvin, J. Inorg. Nucl. Chem., _2_, 11 (1956). Sadtler Standard Spectra, Sadtler Research Laboratories, Philadelphia, (1962). R. S. Rasmussen and R. R. Brattain, J. Am. Chem. Soc., _7_l, 1073 (1949). I. M. Hunsberger, R. Ketcham, and H. S. Cutowsky, J. Am. Chem. Soc., E, 4839 (1952). K. Kuratani, Rep. Inst. Sci. Technol., Univ. Tokyo, _5_, 25 (1951). Z. Bankowska, Bull. Acad. Polon. Sci., Ser. Sci. Chim., E 401 (1962). L. J. Bellamy, Infrared Spfctra of Complex Molecules, John Wiley and Sons, Inc. , New York, 2nd ed. , (1958). R. E. Rundle and M. Parasol, J. Chem. Phys., _2_0_, 1487 (1952). D. N. Shigorin, Dokl. Akad. Nauk SSSR, _9_6, 561 (1954). P N. Shigorin, Zh. Fiz. Khim., 24, 932 (1950). N. Shigorin, ibid., 33, 954 (1950). T” P Grossmann, Z. Physik. Chem. (Leipzig), 109, 305 (1924). .9 Briegleb e_t a_1., Z. Naturforsch, _8_b, 219 (1953). o. Briegleb et al., ibid., 13, 184 (1952). G. Briegleb et al., ibid., Q, 1 (1951). M. I. Kabachnik e_t 9.1-: Dokl. Akad. Nauk SSSR, 99, 1169 (1954). W. J. Le Noble, J. Am. Chem. Soc., _8_2_, 5253 (1960). 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. '73. 182 J. B. Conant and A. F. Thompson, Jr., ibid., 54, 4039 (1932). P. B. Russell, ibid., 24, 2654 (1952). L. W. Reeves and W. G. Schneider, Can. J. Chem., _3__6_>_, 793 (1958). C. Giessner-Prettre, Compt. Rend., 250, 2547 (1960). E. M. Kosower, J. Am. Chem. Soc., 89, 3267 (1958). M. J. S. Dewar, Electronic TheorLof Organic Chemistry, Clarendon Press, Oxford, (1949). L. Pauling, Nature of the Chemical Bond, Cornell Univ. Press, Ithaca, 3rd ed., (1960). G. W. Wheland, Advanced Organic Chemistg, John Wiley and Sons, New York, 3rd ed. , (1960). J. C. Reid and M. Calvin, J. Am. Chem. Soc., 12:, 2948 (1950). Z. Bankowska, Roczniki Chem., _3_6_, 1159 (1962). R. C. Pearson and J. M. Mills, J. Am. Chem. Soc., Z_2_, 1692 (1950). R. G. Pearson and R. L. Dillon, ibid., _7_5_, 2439 (1953). E. S. Gould, Mechanisms and Structure in Organic Chemistry, Holt and Co. , New York, (1959). P. Rumpf and R. Reynaud, Compt. Rend., 250, 1501 (1960). W9. Walisch and H. A. Ruppersberg, Ann. Univ. Saraviensis, Sci., 33, 1 (1959). M. Calvin and K. W. Wilson, J. Am. Chem. Soc., £31, 2003 (1945). H. S. Jarrett, M. S. Sadler, and J. N. Shoolery, J. Chem. Phys., ;2_1., 2092 (1953). W. G. Schneider and L. W. Reeves, Ann. N. Y. Acad. Sci., 22, 858 (1958). F. J. Balta Calleja, Compt. Rend., 249, 1102 (1959). 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90 91. 92. 183 S. Forsen and M. Nilsson, Acta Chem. Scand., _1__3_, 1383 (1959). S. Forsen and M. Nilsson, ibid., _1_4_, 1333 (1960). S. Forsen and M. Nilsson, Arkiv Kemi, -1_7_, 523 (1961). S. Forsen, M. Nilsson, and C. A. Wachtmeister, Acta Chem. Scand., _16, 583 (1962). M. Nilsson, Svensk. Kem. Tidskr., _7_3, 447 (1961). S. Forsen and M. Nilsson, Arkiv Kemi, E, 569 (1962). S. Forsen, Svensk Kern. Tidskr., _7_4, 439 (1962). C. O. Dudek and R. H. Holm, J. Am. Chem. Soc., 8_3_, 2099 (1961). S. Forsen, figectrochim. Acta, 1_8_, 595 (1962). S. Forsen, Arkiv Kemi, 20, 1 (1963). S. Forsen, ibid., _22, 25 (1963). S. Forsen, ibid., 2.2, 41 (1963). K. Nakamoto, Y. Morimoto, and A. E. Martell, J. Phys. Chem., 66, 346 (1962). J. D. Roberts, Nuclear Magnetic Resonance, McGraw-Hill Book Co. , New York, (1959). J. A. Pople, W. G. Schneider, and H. J. Bernstein, High-resolution Nuclear Magnetic Resonance, McGraw-Hill Book Co. , New York, (1959). A. D. Buckingham, Can. J. Chem., _38, 300 (1960). .3 A. D. Buckingham, T. Schaefer, and W. G. Schneider, J. Chem. Phys., 33, 1227 (1960). A. A. Bothner-By and R. E. Click, J. Am. Chem. Soc., 7_8_, 1071 (1956). A. A. Bothner-By and R. E. Click, J. Chem. Phxs” 26, 1657 (1957). 184 93. A. A. Bothner-By and R. E. Click, ibid., 2_6_, 1651 (1957). 94. C. E. Johnson, Jr., and F. A. Bovey, ibid., 2_9, 1012 (1958). 95. L. W. Reeves and W. G. Schneider, Can. J. Chem. 3_5_, 251 (1957). 96. J. V. Hatton and W. G. Schneider, ibid., _4_0_, 1285 (1962). 97. W. G. Schneider, J. Phys. Chem., .63, 2653 (1962). 98. M. J. Stephen, M01. Phys., _1_, 223 (1958). 99. A. A. Bothner-By, J. M01. Spectryu _5_, 52 (1960). 100. W. G. Schneider, H. J. Bernstein, and J. A. Pople, J. Chem. Phys., 28, 601 (1958). 101. A. D. Buckingham, T. Schaefer, and W. G. Schneider, J. Chem. Phys., 34, 1064 (1961). 102. T. W. Marshall and J. A. Pople, Mol. Phys., _1_, 199 (1958). 103. R. J. Abraham, ibid., 34, 1062 (1961). 104. W. T. Raynes, A. D. Buckingham, and H. J. Bernstein, ibid., _3_6_, 3481 (1962). 105. P. Diehl and R. Freeman, M01. Phys., 4, 39 (1961). 106. W. G. Schneider, J. Chem. Phys., 28, 601 (1958). 107. G. C. Pimentel and A. L. McClellan, The Hydrogen Bond, Reinhold Publishing Corp. , New York, (1960). 108. C. G. Cannon, Spectrochim. Acta, 1_0_, 341 (1958). 109. H. M. McConnell, J. Chem. Phys., E, 460 (1956). 110. M. Karplus and D. H. Anderson, ibid., _32, 6 (1959). 111. M. Karplus, ibid., 32, 11 (1959). 112. H. S. Gutowsky, M. Karplus, and D. M. Grant, ibid., _3_1_, 1278 (1959). 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 185 . Instruction Manual, Model A-60 Analytical NMR Spectrometer System, Publ. No. 87-100—030, Instrument Division, Varian Assoc., Palo Alto. Preliminary Instruction Manual, V6057 Variable Temperature System for A-60 Analytical Spectrometers, Publ. No. 87-100-110, Instrument Division, Varian Assoc. , Palo Alto. J. J. D'Amico (to Monsanto Chemical Co.), U. 8. Patent 2, 704, 761, Adar. 22, 1955. G. Schwarzenbach and E. Felder, Helv. Chim. Acta, fl, 1044 (1944). B. Ciocca, Gazz. Chim. Ita1., _61, 346 (1937). L. G. Donaruma, J. Org. Chem., 2_6_, 4737 (1961). A. Burger and G. E. Ullyot, ibid., _1__2_, 342 (1947). T. Isoshima, Nippon Kagaku Zasshi, 1, 425 (1956). M.,S. Kharasch, E. Sternfeld, and F. R. Mayo, J. Am. Chem. Soc., _5_9_, 1655 (1937). A. I. Vogel, Practical Organic Chemistry, John Wiley and Sons, Inc. , New York, 3rd ed. , (1956). NIVIR Spectra Catalog, Va rian Associates, National Press, (1962). J. V. Hatton and R. E. Richards, M01. Phys., _3_, 253 (1960). G. J. Korinek and W. G. Schneider, Can. J. Chem., _3_5_, 1157 (1957). S. A. Glickman and A. C. Cope, J. Am. Chem. Soc., 61, 1017 (1945). J. Powling and H. J. Bernstein, J. Am. Chem. Soc., 1_3_, 1815 (1951). Lange's Handbook of Chemistry, Handbook Publishers, Inc. , Sandusky, 9th ed., (1956). I. Takahaski, Bull. Chem. Soc. Japan, _3_5_, 1046 (1962). V. J. Kowalewski and D. G. de Kowalewski, private communication. u11111111111111.1111)(11151611