$OME CARBONIUM ION REARRANGEMENTS Thesis for the Degree of Db, D. W MICHIGAN STATE UNIVERSITY Donald Oscar Rickter 1964 LISRARY Ivfii_li_:;';.lfl Siate University _A “a... M1CH|GAN iSi‘At'E UNNLng l‘ DEPARTMENT OF: CHEMISTRY EAST LANSING, MICHIGAN ABSTRACT SOME CARBONIUM ION REARRANGEMENTS by Donald Oscar Rickter PART A: THE REACTION OF‘t—AMYL CHLORIDE WITH ALUMINUM CHLORIDE IN A SOLVENT The aluminum chloride-catalyzed rearrangement of labeled t-amyl chloride goes largely by equation (1). The existence of other mechan— isms was shown by the relative ratio of equilibrations C-2 <::> C-3 and C-1 é::> C-b. Path (1) alone gives a ratio of 2.00, while 1.55 is the actual ratio observed.1 One explanation1 is that 13% of the reaction proceeds via equation (2). + + + + C-C-C-C ——> C-C-C-C -> C-C-C-C —-> C-C-C-C (l) I <_ t <_ t <_ I C C C C C + I + + C-C-C-C -—> C-C-C ——> C-C-C-C (2) v <_ I <_ v C C C However, it has been shown2 that bimolecular reactions occur: c *c c y t v I . N . :AI c-c=c-c + c-c-c-c —> c—g-c-c H°~> £5341> 555—» + <—— , <———— <——————— <+—————— c-c-c-c %& C + y C c l K . l * l c-c-c-c H'~> ———> c-c=c-c + c-c—c-c (3) , < <——— + . c-c-c-c 1 c Reaction conditions were found under which 1-13C- and 2-13C-t-amyl chlorides were rearranged without appreciable bimolecular reaction. Donald Oscar Richter Ethylene chloride was used as the solvent for the labeled t—amyl chlorides and aluminum chloride at O0 for one to two minutes. The t-amyl chlorides were recovered from the reaction mixture, purified by preparative gas chromatography, and analyzed by mass spectrometry. Scrambling of the labels, C-l 2::> C-b and C-2é:j> C-3 had occurred. The experimental errors in the mass spectrometric data were too large to permit calcula- tion of the rate ratio k2 3/k1 2. ) ) PART B: HIGHER-ORDER WAGNER—MEERWEIN SHIFTS Examples of 1,2-hydride shifts are well-known, but the existence of a 1,3-hydride shift (Sf. a 1,2; 1,2 mechanism) was only recently3 established. Numerous cyclic systems have transannular 1,5- and 1,6- hydride shifts4, and several organic ions of special geometry undergo 1,5-hydride shifts5:5, but the question remained: Do higher-order (l,h; 1,5; 1,6; 332') hydride shifts occur in simple aliphatic systems? Butylamine-l,l-d2 and pentylamine-l,l-d2 were prepared and converted to perchlorate salts. Deamination in water at 25° gave numerous products, which were analyzed by gas chromatography and infrared spectroscopy. The pentyl system gave the following percentage yields: pentenes, l9; l-pentanol, 22; 2-pentanol, lb; 3-pentanol, h; l—pentyl nitrite, 2; 2- and 3-pentyl nitrites, b; l-nitropentane, l; 2- and 3-nitropentanes, O.l; l-pentyl nitrate, l; and 2— and 3—pentyl nitrates, 0.6: 2a total of 68%. The butyl system gave butenes, l6; l-butanol, 22; 2-butanol, 20; l-butyl nitrite, 6; 2-butyl nitrite, 5; l-nitrobutane, l; 2-nitrobutane, l; l-butyl nitrate, b; and 2-buty1 nitrate, 1: :3 total of 76%. Preparative gas chromatography was used to obtain milliliter samples .1. ./u «(v ‘0- Donald Oscar Rickter of deuterated 1-pentanol and l-butanol from the deaminations. The n.m.r. Spectra of the neat alcohols showed no a-protons in the labeled l-pent- anol and a slight amount of a—protons in the 1-butanol. The mass spec- tral analyses of the trimethylsilyl ethers of the alcohols showed that there were actually a few percent of o-protons in both alcohols. Thus some rearrangement of the deuterium did occur. There was no loss of deuterium by exchange with the solvent. The finding of 3—pentanol was unexpected, since Streitwieser6 stated that it was not a product of the deamination of pentylamine. The distribution of deuterium in the secondary pentanols indicated that there had been successive 1,2—hydride shifts down the chain; 1.3., there were 75% 2-pentyl-l,l-d2, 21% 3- pentyl-l,1-d2, and h% 2-pentyl—S,S-d2. From the mass Spectral data it was not possible to tell if a small amount of l,h- or 1,5-hydride shift loccurred in the two systems. PART C: ALIPHATIC 1,3-METHYL SHIFTS Several rearrangements could be explained as either 1,3- or 1,2; l,2-methy1 shifts. One case7, said to require a 1,3—methyl shift, C / \ ) C C\\+ <+——? //C- C -C ‘P O~CP~O ‘9 (ll-I- O 0-. 0- O-O 0- 0 actually could be explained as a series of 1,2—shifts: C C C C C C C C C C C C C C C C C C c--—- ——>-——— —>--—- ——> I I C-C—C-C-C -——> C-C-C-C-C . +I <_ I <_ I C C C I I I C C C No other examples have required postulation of a 1,3-methyl shift. A system which would favor such a mechanism is a primary carbonium ion which could shift a methyl to form a tertiany ion. The simplest such case is neopentylcarbinyl cation. Neopentyl cyanide was prepared by nucleophilic displacement on the chloride in dimethylsulfoxide at 100-1300 for 2h hours. There was a hb% yield of colorless solid. Reduction with lithium aluminum hydride gave neopentylcarbinylamine. The amine was converted to its perchlorate salt and deaminated in water at 25°. The products were numerous. In a typical run b3% was neopentylcarbinol, 21% t-butyl alcohol, 1h% di- methylisopropylcarbinol, 1.3% methyl-tfbutylcarbinol, 2% t-butylethylene, and six other compounds. Gas chromatographic analysis showed that there was less than 0.2% of dimethylpropylcarbinol, the alcohol which would form in a 1,3-methy1 shift. REFERENCES (l) Roberts, J. D., R. E. McMahon, and J. S. Hine, J. Am. Chem. Soc., 33, 1237 (1950). (2) Karabatsos, G. J., F. M. Vane, and S. Meyerson, ibid., §§, 733 (1963). (3) Karabatsos, G. J. and c. E. Orzech, ibid., Q, 2838 (1962). (h) Prelog, V. and J. G. Traynham, Molecular Rearrangements, ed. P. de Mayo, Chapter 9, Interscience, New York, 1963. (5) Cohen, T., R. M. Moran, and G. Sowinski, J. Org. Chem., 26, 1 (1961). (6) Streitwieser, A., Jr., ibid., 22, 861 (1957). (7) Mosher, W. A. and J. C. Cox, Jr., J. Am. Chem. Soc., Z2, 3701 (1950).. SOME CARBONIUM ION REARRANGEMENTS By Donald Oscar Richter A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 196h ACKNOWLEDGMENTS The author is grateful for the opportunity to do research under the direction of Dr. Gerasimos J. Karabatsos. It was a stimulating and rewarding eXperience to work for a professor who encourages inde- pendent thinking and development of responsibility. He and his stu- dents, especially Chester Orzech, were very helpful in discussions when progress was stymied. A research assistantship of the U. S. Atomic Energy Commission (June 1962 to March 1963) and a fellowship of the Petroleum Research Pund--American Chemical Society (March 1963 to September 196k) are thankfully acknowledged. Seymour Meyerson and John Balha of the American Oil Company are to be thanked for the mass spectral analyses and calculations. Finally, I am grateful to my wife, PHYLLIS CARLSON RICKTER, whose Patience, Help, and Dedication made my doctoral studies possible. ii TABLE OF CONTENTS' PART A THE REACTION OF E-AMYL CHLORIDE WITH ALUMINUM CHLORIDE IN A SOLVENT Page INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . 2 RESULTS . . . . . . . . . g . . . . . . . . . . . . . . . . . 7 msaBSKN. .. ... ... .. ... .. ... .. ... .. 12 EXPERIMENTAL . . . . . . . . . . . . . . . . . . . . . . . . . 1A Studies of Dichloromethane . . . . . . . . . . . . . 1h Studies of 1,2-Dichloroethane . . . . . . . . . . . 15 Synthesis of 1-13C——t—Amyl Chloride . . . . . . . . . 16 Synthesis of 2- 13C- -t-Amy1 Chloride . . . 17 Runs with t-Amyl Chloride and Aluminum Chloride in 1,2-Dichloroethane at 0° . . . 19 First Run with 1-13C-t-Amy1 Chloride and 2-13C-t- Amyl Chloride . . . . 20 Second and Third Runs with 13C— —t-Amyl Chlorides . . 21 APPENDIX — Solvents for Carbonium Ion Rearrangements . . . . . 22 PART B HIGHER-ORDER WAGNER-MEERWEIN SHIFTS INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . 26 RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . L3 EXPERIMENTAL . . . . . . . . . . . . . . . . . . . . . . . . . 51 Synthesis of Pentylamine . . . . . . . . . . 51 First Synthesis of Pentylamine-l,1-d2 . . . . . . . 52 Second Synthesis of Pentylamine-—1,1-d2 . . . . . . . 52 Deamination of Unlabeled Pentylamine . . . . . . . . 52 First Deamination of Labeled Pentylamine . . . . . . 5h Second Deamination of Labeled Pentylamine . . . . . 56 Preparation of N- —Pentylacetamide . . . . . . . . 60 Preparation of Labeled N- -Penty1diacetamide . . . . . 62 Preparation of Butylamine . . . . . . . . . . . . . 63 Deamination of Butylamine . . . . . . . . . . . . . 63 iii TABLE OF CONTENTS (Cont.) Page Preparation of Butylamine—1,1-d2 . . . . . . . . . . 6b Deamination of Butylamine-1,1-d2 . . . . . . . . . . 65 Preparation of Silyl Ethers . . . . . . . . . . . . 66 Acetylation of Butylamine-1,1-d2 . . . . . . . . . . 66 Preparation of Butyl Nitrites . . . . . . . . . . . 68 PART C ALIPHATIC 1,3—METHYL SHIFTS INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . 70 RESULTS AND DISCUSSION . . . . . . . . . . . . . . . . . . . . 75 EXPERIMENTAL . . . . . . . . . . . . . . . . . . . . . . . . 80 Synthesis of Neopentyl Cyanide . . . . . . . . . . . 80 Reduction of Neopentyl Cyanide . . . . . . . 81 First Deamination of Neopentylcarbinylamine . . . . 82 Second Deamination of Neopentylcarbinylamine . . . . 83 Preparation of Neopentylcarbinylammonium-l,1-d2 Perchlorate . . . . . . 83 Third Deamination (Using Deuterated Amine). . . 83 Fourth Deamination (Using Deuterated Amine) . . 8h Preparation of Known Compounds for Gas Chromatography 85 REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . 87 iv TABLE Ia. IIa. IIIa. Iva. va. VIIa. VIIIa. Ib. IIb. IIIb. IVb. VIb. LIST OF TABLES PART A Runs with unlabeled t- ~amyl chloride in 1,2-dichloro- ethaneatOo...—................ Gas chromatographic analyses of unlabeled solutions Runs with labeled t- -amy1 chloride in L 2— dichloro- ethane at O0 . . . . . . . . . . . . Gas chromatographic analyses of products . . . . . . . Mass Spectral data for labeled E-amyl chlorides . . . Positions of labeling (atom percent13C) in E-amyl chlorides . . . . . . . Retention times Comparison of possible solvents PART B Percentage yields of deamination products from pentyl— amine and pentylamine-l,1-dz . . . . . Mass spectra of N-—penty1diacetamide and N-pentyl- l, 1- -d2- diacetamide . . . . . . . . . . . . . Mass spectra of trimethylsilyl ethers of labeled l-pentanol from deamination of pentylamine-l,l-d2 and of unlabeled l-pentanol Mass spectra of trimethylsilyl ethers of labeled 2-pentanol and 3-pentanol from deamination of pentyl— amine-1,l-d2, unlabeled 2-pentanol, and unlabeled 3-pentanol . . . . . . . . . . . . . . . . . Percentage yields of deamination products from butyl- amine and butylamine-l,l-d2 . . . . . . . . . . Mass Spectrum of N-butyl-l,1-dz-diacetamide . . Page 10 11 15 2h 30 31 32 33 39 110 TABLE VIIb. VIIIb. IXb. XIb. XIIb. Ic. IIC. LIST OF TABLES (Cont.) Mass Spectra of trimethylsilyl ethers of labeled l-butanol from deamination of butylamine-1,1-d2 and of unlabeled l-butanol . Mass Spectra of trimethylsilyl ethers of labeled 2-butanol from deamination of butylamine-l,1—d2 and of unlabeled 2-butanol . Products of deaminations of pentylamines . Products of second deamination of pentylamine-l,1-d2 . Minor products of deamination of pentylamine-1,1-d2 Products of deaminations of butylamines. PART C Deamination products of neopentylcarbinylamine and neopentylcarbinylamine-l,l—dz Compounds shown to be absent in deamination products of neopentylcarbinylamine vi Page 111 L2 55 58 61 67 76 77 u. Figure lb 2b 3b 5b 6b 7b 8b N.m.r. N.m.r. LIST OF FIGURES Page Spectra of pentylammonium perchlorates (80% solutions in deuterium oxide): (A) 10% unlabeled salt and 90% 1,1-dideuterated salt, (B) 1,1-di- deuterated salt . . . . . . . . . . . . . . 3b . spectra of.L12ntanols (neat samples): (A) un- labeled, (B) product of deamination of pentyl- amine-l,1-d2 . . . . . . . . . . . . . . . . . . . 35 . spectra of secondary pentanols (neat samples): (A) unlabeled 2-pentanol, (B) unlabeled 3-pentanol, (C) products of deamination of pentylamine-l,1-d2 . . 36 . spectra of butylammonium perchlorates (82% solu- tions in deuterium oxide). (A) A. 9% unlabeled salt and 95.1% L l- dideuterated salt, (B) 1,1- dideuterated salt . . . . . . . . . . . . . . 37 . Spectra of l—butanols (neat samples): (A) un- labeled, (B) product of deamination of butyl- amine-l,l-d2 . . . . . . . . . . . . . . . . . . . 38 spectrum of 2-butanol (neat sample) from deamina- tion of butylamine-1,l-d2 . . . . . . . . . . . . 38 Outline of the deamination of pentylamine—1,1-d2 . . . . A6 Outline of the deamination of butylamine-1,1-d2 .. . . . A8 vii PART A THE REACTION OF‘E—AMYL CHLORIDE WITH ALUMINUM CHLORIDE IN A SOLVENT INTRODUCTION The rearrangement of labeled E-amyl chloride with a catalytic amount of aluminum chloride is complicated. The mechanism is not as simple as equation (1). l 2H:N + l 2Me:~ + l 2H:N + C—C—C-C <__“—> C-C-C—C <._—"""‘> C-C-C-C <"———_—> C-C-C-C (1) I I Y I C C c 7 c If this were the only mechanism of equilibration of C-1 4:) C-Ii and C-2 é::> C-3, the relative rates of equilibration would be 1:2. Ex- perimentally1 the ratio is h2,3/k1,4 = 1.55 rather than 2. The explana- tion given1 is that a second mechanism is operating: 13% of the equilibration is via the neopentyl cation. % + % ' + + % ’ C-C-C-C < > C-C—C > C-C-C-C (2) 1 l 1 C C C Equation (2) does explain the relative rates. It equilibrates C—l with C-b but not C—2 with C-3. However, it is not necessary to postulate the intervention of high-energy primary carbonium ions. Bimolecular reactions also account for the experimental results and have been shown to occur.2 The simplest of several bimolecular pathways is that Shown below -- dimerization followed by a series of 1,3-Shifts of hydrides and methyls and then cleavage back to C5 molecules: c C c % ' % ' % ' + c-C-C-C > CLg-C-c H g N) c_;c—c-c + <—_ "c-c-c—C <—' C—C—C-C * 4' 1 1 C-C—C-C C c E: I * C C c + C o o N 3‘- ' 0 7L I 'X“ -£E=13—> *c—C-C-C _Mg;___> c-C-c-c‘ H"” > C—C-C-C <--—-—— , <+—————— , < + , +c-c-c C-C-C-C C-C-C-C I I I C c c C % ' % ' €7——e> c-c=C-c + c-c-c-c (3) When 2-methyl-2-chlorobutane-1-13C was stirred with aluminum chloride, several products (besides polymer) were formed.2 Analysis of the volatile products by gas chromatography Showed, in a typical experiment, 63% E-amyl chloride, 22% E-butyl Chloride, 7% 2-methyl-2- chloropentane, 2% 3-methyl-3-chloropentane, L% 2—methyl-3-chlorobutane, 0.9% iSOpentane, and 0.6% methylpentanes. The E-amyl chloride was separated by preparative gas chromatography and studied with n.m.r. and mass Spectrometry. The data and calculations led to the conclusions that the 13C label was completely scrambled between positions 1 and b and that about 9% of the molecules were dilabeled: C c c C 13c I I I I I 13c-C-C-C -——> 13c-c-c-c + C-c-C-c13 + 13C-C-c-C13 + 13C-c-c-C I I I I I C1 c1 c1 c1 c1 (A) \ 2/3 1/3 / \ 77% 23% / 91% 9% The reaction of 2-13C-E-amyl chloride gave similar results. The products were the same, and the labeled E-amyl chlorides were as follows: C C C C I I I I C—13C-C43 —> C-13C-c-C + C-C-13c—C + C-13C—13C—C I I I I c1 C1 C1 C1 (5) \ 1/2 1/2 . 92% 8% A The dilabeled molecules prove that bimolecular reactions occur. various bimolecular mechanisms can be written. Equation (3) is not an adequate representation. It does account for dilabeled products from C-l labeled E-amyl Chloride but not the dilabeled products from the C-2 labeled compound. Equation (6) results in dilabeled products from the C-2 as well as the C-1 labeled compound. C c c c-‘C—C-c Et"" > c-*C-C-c-C ilEEifii—> c—C-c-c 3' %' 'Q-___—_—- 91' c-C-C-c C- 9'0 c-Vc-c C c- c-c C c C M9"“> C-C-C—C H‘“’ > C-C-C-C > c-C-c-c + *C*-C < *C-c < + C (6) C—*C-c Cfié-C c-*C=*c—c The effects on isotopic equilibration are the same in equations (2) and (3) (1.3., C-l <::> C-h but C-2 > C-3) and the same in equa— tions (1) and (6) (1.3., C-l <::> C-b and C-2 <;:? C-3). The formation of E-butyl chloride was explained2 by disproportiona- tion of the intermediate 0101121+ to the tertiary butyl cation and a branched hexene: + c C C . C-c-c I o o I I '_ o I c-g-C-C H""> M€°~> C-C-C-C 4-3—11! c-c-C-C—c \I < < I <_ I I c-C-C-C C-g-C-C C C I c c ' C Me :~> H:N> C-C-C > , ‘I' ' .- .. .. = - - <——-—- <——-— c—C-c-c-c <——— C E c + c c C C C (7) 1 C )‘(3 .-.! .r' ¢-‘ u_c "ob “ I K) In 0., .\V. :“~‘ " I\| ' I ’ 1 c c c C c I o I o I I I C-q-C-C “8'23 C-C-C+ M‘a'"> C—C-C-C ——-> c-c + C-C=C-C—C (8) ' I < I < I I C-C-c-c c l C T c I c-c-c—c c-c-c-c C , + c Equation (7) gives 2-methyl—2—chloropentane, while (8) gives 3-methyl-3- chloropentane. The presence of 3-chlorO-2-methylbutane is probably due to this path: A1c1 - Cl C C C C < > C C C C < > ' (9) C C C The 2-methylbutane arises from hydride abstraction by either of the, carbonium ions in equation (9). While it has been clearly shownZthat bimolecular reactions occur, the intervention of neopentyl cation has not been excluded. The amount of rearrangement 213 neopentyl cation had not been established. The mechanistic picture would be Simplified by finding suitable conditions, such as running the reaction in a solvent, under which.£—amyl chloride rearranges but bimolecular_reactions do not occur. (The criterion for occurrenCe of bimolecular reactions is the formation of C4 and C6 fragments; 113., E—butyl chloride and methylchloropentanes.) Under such conditions the ratio kz’Z/kl,4 would be higher than the 1.55 found1 earlier, since bimolecular reactions lower the ratio. The amount of rearrangement by way of neopentyl cation would be measured by’the de- crease in the ratio below the value 2.00. A suitable solvent for the study would be polar enough to dissolve a catalytic amount of aluminum chloride and a high concentration of 6 E-amyl chloride. It would not react with the catalyst, the substrate, or any of the products. Finally, it would be easily separated from the rearranged E—amyl chloride and would not mask the presence of E—butyl chloride or methylchloropentanes. (Solvents are discussed further in the Appendix.) ‘ I .-..-- ..I-, 0 - ’9‘ V. .. " a. F \ \- v. ’v‘ p ¢I._ .. Q I " I "‘ ... v... "v RESULTS Tables Ia and Ila Show conditions under which E—amyl chloride and aluminum chloride were in solution with little or no bimolecular rear- rangement occurring. Similar conditions were used with labeled E-amyl chloride (Tables Illa-Iva). In run 6 there was no detectable reaction. The isotOpiC distribution was shown by mass Spectrometry (Tables va and VIa) to be the same before and after the treatment with aluminum Chloride. In runs 7 and 8 the amount of catalyst and the duration of treatment were both increased. The result was scrambling of label--to a small extent, about equal to the experimental uncertainty, in run 7, and to a very large extent in run 8. Mass spectrometry data (Tables va and VIa) indicated that equilibration in run 8 went beyond statis- tical distribution of label! Apparently the data contain experimental error almost as large as the values obtained. An attempt to calculate1 the rate ratio was unsuccessful; the ratio came out k1 4t - % log(-0.2A6) ’ = 1 = 0.562 - 1.8A6 log 1. h2,3t - 5 log(0.l90) No polymerization was apparent in any of the runs. (NO polymer precipitated out and no water-insoluble residue was left after distil- lation of the volatile products.) 8 Table Ia. Runs with unlabeled E-amyl chloride in 1,2-dichlor0ethane at 0°. Run Conc. Of Wt. Mole Ratio Reaction Products N ?— t-AmCl A1013 of A1C13 Time 0 Formed TWt.%) (mg.) to EfAmCI (sec.) 1 6.6 70 0.032 300 _<_ 0.1% 2 6.6 31 0.0111 20 < 0.01% 3 6.6 17 0.0077 60 < 0.01% b 15.0 6 0.00h6 9O 5 0.05% 5 h.b 0.0ll 15 < 0.0l% Table IIa. Gas chromatographic analyses of unlabeled solutions (using 30% silicone DC-550 column in Perkin-Elmer leL at 8 p.s.i.) Time - Run 1 Run 2 Compound (mln'g Product Product Solvent at 115 CHZClCHZCI 55 93.9 % 93.0 % 99.7 % EeAmCl NO 5.8 6.8 0.00 t-BuCl l2 . ‘Unknown W 12 :T 0'3 0'2 0'2 Unknown X 5 0.03 0.007 0.00 Unknown Y 7 0.00 0.007 0.00 Unknown Z 9 0.00 0.07 0.1 Time Run 3 Run 3 Run A Run 5 Compound (m1n.) ‘ at 80° Reactant Product ‘ Product Product CHZClCHZCl in 93.0 % 92.6 % 86.5 % 96. % jg-AmCI 12 7.0 7.11 13. 1;. E'BUCI 1‘ } 0.02 0.02 0.00 0.2 Unknown W h Unknown.X 5 0.0h 0.0h 0.01 0 Unknown Y 6 0.1 0.09 0.15 0 Unknown 2 10 0.02 0.02 0.0b O Table IIIa. Runs with labeled E-amyl chloride in 1,2-dichloroethane at 0°. Conc. of Mole Ratio Reaction $3“ t-AmCl of A1013 Time 131,332? ' TWt.%) to EfAmCl (sec.) 6 lb 0.0056 15 < 0.1% 7 15 0.0087 60 < 0.1% 8 15 0.012 120 1% 2-methyl- 2—butene Table Iva. Gas chromatographic analyses of products (20% SE—3O prep. column at 52° and 8 p.s.i. in Perkin4Elmer 15hL) Compound ($12?) Run 6 Run 7 Run 8 2-Methyl-2-butene 6 < 0.1% < 0.1% 1% _t-AIIUl chloride 17 15 15 11. CHZClGiZCl 13 85 85 85 10” Table V3. Mass spectral data for labeled E-amyl chlorides 'Run 6 Run 7 Run 8 React- Prod- React— Prod- React- Prod- ant uct ant uct ant uct Sample Number 131-1 132-3 135-1 136-1 135-1 139-1 + + + C5H9 1C5H10 IC5H11 69 8.0 7.0 6.9 7.0 9.3 70 67.0 58.5 55.8 56.6 65.5 71 189.2 181.2 180 187.9 193.8 72 181.5 181.6 185.6 153.6 188.5 73 - .9 - 9 - .8 - .8 2.9 % labeled 53.6 53.5 58.0 51.3 % doubly labeled 1.0 C3H4Cl+,C3H5Cl+,C3H6Cl+ 75 1.9 1.9 2.0 2.0 2.2 76 69.8 70.9 72.1 78.8 109.8 77 195.7 198.0 203 215.8 230 78 137.1 136.1' 188.8 188.6 83.7 79 -l.O -1.3 -1.3 -1.3 .2 % labeled 58.2 53.1 55.8 31.8 % doubly labeled 0.1 C4H7Cl+, 0411801+ 90 2.8 2.8 2.7 2.8 2.8 91 26.5 26.9 27.8 28.8 27.1 92 17.3 17.3 18.8 19.2 19.9 93 - .2 - .2 - .8 — .3 0.2 98 .1 .1 - .2 .0 0.1 % labeled 81.8 80.9 82.0 88.3 % doubly labeled -' 0.8 11 Table VIa. Positions of labeling (atom percent 13C) in E-amyl chlorides C-1 C-2 _ C-3 C-8 Run 6 (Reactant) 26.0 27.9 0 0 (Product) 25.8 27.5 O O Run 7 (Reactant) 26.8 28.3 0 0 (Product) 28.5 27.8 1.9? 0.5? Run 8 (Reactant) 26.8 28.3 0 0 (Product) 15.2 16.6 11.3 10.8 .7 DISCUSSION Conditions were found for the rearrangement of E-amyl chloride in the absence of bimolecular paths. Fifteen percent solutions of E-amyl chloride in 1,2-dichlor0ethane were stirred at 0° for 60 or 120 seconds. The amount of catalyst was about one molecule of A12C16 per two hundred molecules of alkyl halide. The experiments with labeled E-amyl chloride were not completely satisfactory. It was difficult to isolate the substrate in pure form from the reaction mixture. There was appreciable dehydrochlorination before or during the mass spectrometric determinations. This problem complicated the calculations. The determination of the amount of 13C label at positions 3 and 8 were made by subtracting values with rela- tively large experimental uncertainties. The isotopic enrichments at C—1 and C-2 were only about 25%, so each percent of error constitutes an uncertainty of 1/25 or 8%. However, after extensive scrambling of the label the uncertainties are worse——of the order of twice as large. It would be possible to improve the situation considerably if compounds with higher 13C-enrichment were available. Another possibil- ity is to use dilabeled molecules. For example, methylation of prop- iohyl Chloride-l-13C with methylcadmium,3 condensation of the methyl ketone withumethylmagnesium iodide—13C, and distillation from hydro- chloric acid, should give a good yield of E—amyl chloride labeled at positions 1 and 2. (There would actually be four kinds of molecules. If the labeled barium carbonate and methyl iodide each had Sixty atom- percent 13C, the E-amyl chloride molecules would be 36% dilabeled, 28% labeled at C-1, 28% at C-2, aha 16% unlabeled. This is better than the 12 a: 13 30% C—1, 30% C—2, and 80% unlabeled molecules obtained by synthesizing, the 1-13C- and 2-13C-compounds separately and then mixing them.) Actually bimolecular reactions are not absolutely excluded by this investigation. There could be a subliminal amount. The threshold of detection by gas chromatography is high enough to allow 0.1% E-butyl chloride in a 15% solution. This represents about one molecule out of 150 molecules of E-amyl chloride. The mass spectrometric data for run 8 indicated the presence of dilabeled material (approximately 1.0% of the 05+ fragments, 0.8% of the C4+, and 0.1% of the C3+). The values were of the same magnitude as the uncertainties to which they were subject. EXPERIMENTAL Studies of Dichloromethane Methylene chloride from Matheson, Coleman, and Bell was purified17 by Shaking it with 5% aqueous sodium carbonate and with water, drying over calcium chloride, and distilling at 39-800. Five ml. of the distillate was tested with 1.3 g. of anhydrous aluminum chloride (Mal- linchrodt AR grade). The liquid was colorless at first, yellow after one or two minutes, and bright yellow-orange within fifteen minutes. Another purification method was tried. The solvent was shaken with concentrated sulfuric acid and then twice with water. The re- sulting emulsion was dried over calcium chloride for three hours and distilled. The distillate reacted with aluminum chloride as before. Dhying the distillate over phosphorus pentoxide did not change the situation. The more elaborate procedure of Harmon18 (successive treatments with sulfuric acid, water, sodium bicarbonate solution, and calcium chloride, distillation from phosphorus pentoxide, and storage over Linde 8A molecular sieve) was not tried. Prins27 purified methylene chloride by repeatedly boiling it with, and distilling it from, aluminum chloride. The pure solvent remained colorless when treated with aluminum chloride. Gas Chromatographic studies of solutions were made. None of the available columns would completely resolve the peaks due to E—butyl Chloride and the solvent. The best separation was obtained with a 30% Silicone DC—550 column at low temperature and low pressure. There was inversion in the order of appearance of the first two compounds. 18 15 Table VIIa. Retention times (minutes) on Silicone DC-550 analytical columns in Perkin-Elmer Model 158L. 20% 20% 30% 30% 30% Compound B.P. 80° 59° 78° 85° 32° 8 p.s.i. 8 p.s.i. 8 p.s.i. 8 p.s.i. 5 p.s.i. 0112012 800 3.8 5.1 ' 5.5 11.6 30 3-13.101 510 3.6 5.8 5.0 12.6 27 (CH2C1)2 83° 8 18 17 88 115 E—AmCl 860 9 16 18 35 85 Studies of 1,2—Dichloroethane Analytical gas chromatographic conditions were found that separated completely the peaks due to E—amyl chloride and this solvent. (See Table VIIa.) The peaks came off in opposite sequence on 30% and 20% silicone DC 550 columns. Commercial 1,2-dichloroethane (Matheson, Coleman and Bell) was purified by shaking with concentrated sulfuric acid and with water, drying over calcium chloride, and distilling.19 The distillate turned to a pale yellow after five or ten minutes over aluminum chloride. (Thomasz° reported a pale straw color and a 3° rise in temperature for this mixture.) The solvent for the runs described was distilled from phosphorus pentoxide before use. Despite repeated distillation, however, the solvent always had detectable impurities (See Table IIa.) III (\I I. In ‘§ ~v .V' v. Ii: )“I 7‘. (_I t 16 Synthesis of 1-13c-t—Amyl Chloride The method was that used by Roberts1 and by Vane2b. Preparation of methyl iodide—13C from labeled barium carbonate was considered. (Nystromzz reduced carbon dioxide with lithium aluminum hydride in diethylcarbitol to methanol in 81% yield. Tolbert23 prepared methyl iodide-14C from labeled methanol, red phosphorus, iodine, and water on a vacuum line in 95% yield. He obtained yields of 80-90% when he hydrogenated carbon dioxide—14C over a copper-alumina catalyst at 860 atmospheres.) However, commercial methyl iodide-13C (Bio-Rad Laboratories) was finally obtained and used. Ordinary methyl iodide (1.30 9.; 0.00915 mole) was used to start a Grignard reaction with 1.105 g. of magnesium (0.0855 g. at.; Domal) in dry ether under dry nitrogen. Then 5.20 9. (0.0368 mole) of labeled methyl iodide in 12 ml. of ether was added with stirring. The result- ing methyl Grignard reagent had 62.8 x gfgfigg = 50.0 atom percent13C. After 1.5 hours of stirring the solution, 2.96 g. (0.0810 mole) of methyl ethyl ketone in 15 m1. of ether was added Slowly. The reaction mixture stood overnight. The complex was decomposed with a solution of 5.18 g. of ammonium Chloride (0.096 mole) in 31.5 ml. of water. The colorless layers were separated. The aqueous layer was extracted four times with ether. The ethereal solutions were dried and distilled at 38-35°. The residue was vacuum distilled: 3.3 g. at h7?510 (80 mm.). I The distillate was 75% E-amyl alcohol (thus 2.5 g.: 69% yield from the ketone) and 25% ether by gas chromatography (30% silicone DC-550 at 79° and 8 p.s.i.). 17 A mixture of 2.2 g. of the labeledzfleohol (0.025 mole) and 6.9 ml. of concentrated hydrochloric acid (0.082 mole) was heated for 60 min. Collected was 2.17 g. colorless liquid at 75~79°. Gas chromatographic analysis showed it was 95% E-amyl chloride, 0.5% 2—methyl-2-butene, and 8% ether. The yield was thus 2.06 g. (78%). Synthesis of 2-130-E-Amyl Chloride Propionic acid. The reaction system consisted of Pyrex glassware connected by Tygon tubing, with all joints and stoppers tightly wired. A flask containing barium carbonate-13C was fitted with a dropping funnel and an outlet tube going to a sintered glass tube in concentrated sulfuric acid in a gas-washing bottle. This bottle was followed by a second one which held only glass wool. The exit led to a 10 mm. tube which was to be just below the surface of the stirred Grignard solution. The tube was wide open at its tip. (Use of a sintered glass tip in the first seven runs resulted in yields below 50%.) The reaction flask had a Hershberg stirrer, a dropping funnel, and condenser leading to a calcium sulfate drying tube and then to a trap filled with saturated barium hydroxide solution. The system was flushed with dry nitrogen while being flamed and cooled. Then 6.86 g. of magnesium (0.266 g. at.; Domal) was added to the reaction flask. Thirty ml. of redistilled ethyl bromide in 100 ml. of dry ethyl ether was added with stirring in 0.5 hr. Another 0.5 hr. of stirring was needed for the last of the metal to react. Another 850 ml. of dry ether was added to decrease the concentration below 0.5 M. Dry nitrogen was bubbled through the solution while dry ice was added to an acetone bath around it to lower the bath temperature to -30°. l8 Seventy-five ml. of 80% perchloric acid was dripped onto 26.2 g. (0.133 mole) of barium carbonate in the first flask. The acid was added in 12 min. For 10 min. nitrogen was used to flush the carbon dioxide through the Grignard solution. Eighty ml. of 10% aqueous sulfuric acid was added to the cold reaction mixture in 8 min. The layers were separated. The aqueous layer was extracted three times with 50-m1. portions of ether. The combined ether solutions were dried over calcium sulfate, filtered and distilled. Obtained was 8.82 g. of prOpionic acid at 80° (80 mm.). Gas chromatographic analysis (10% silicone column in Beckman 00—2 at 70° and 30 p.s.i.) showed that it was 99% pure (1% ether). Thus the yield was 85%. Some of the carbon dioxide was recovered as a precipitate of bar- ium carbonate. It weighed 0.91 g. when dry. (This was 1.7% of the original amount.) 1—13C-Propionic Acid The above procedure was followed with 26.3 g. of barium carbonate— 13C (57.7 atom % 130; 0.133 mole). The yield was 7.18 g. at 79—800 (77 mm.) or 72%. Another 3.5% of the carbon dioxide-13C was recovered as 0.91 g. of dry barium carbonate-1°C.. Gas chromatographic analysis (20% silicone column in Beckman GC—2 at 100°) Showed that the distil- late was 98.7% prOpionic acid and 1.3% ether. 2-13CfE—Amyl Alchol Diazomethane was prepared from 10.3 g. nitrosomethylurea,24 dried over potassium hydroxide, and used to esterify the labeled prOpionic acid in ether. 19 A methyl Grignard reagent was made from 8.60 g. of magnesium (0.19 g.at.; Domal), 20 ml. of methyl iodide, and 100 ml. of dry ether. The ester solution was added slowly to the Grignard solution with stirring, which was continued for another two hours. The mixture stood overnight, was stirred another hour, and was then treated with 35 m1. of saturated ammonium chloride solution. The ether layer was decanted. The white aqueous paste was extracted twice with 25 ml. portions of ether. The ether solutions were dried, filtered, and distilled through a packed column. After removal of most of the ether, the liquid was distilled under vacuum. The product was 3.38 g. at 53-58° (90 mm.). Gas chromatographic analysis (15% LAC 886 in an Aerograph at 70° and 27 ml./min.) Showed that no methyl propionate was present and the alco- hol was 95% pure (5% ethyl ether). Thus the yield was 78.6%. Infrared analysis confirmed the absence of carbonyl compounds. 2—13C-EfAmyl Chloride A mixture of 3.05 g. of the labeled alcohol (0.038 mole) and 9.0 m1. of concentrated hydrochloric acid (0.108 mole) was heated for 85 min. Obtained was 2.35 g. colorless liquid at 78-780. Gas chromatographic analysis showed it was more than 99.5% E-amyl chloride with a trace of ether. The yield was 68%. Runs with2E-Amyl Chloride and Aluminum Chloride in 1,2-Dichloroethane at 0°C The procedure was based on that of Vane.2b A solution of 1.75 g. of E-amyl chloride (16.8 m. m01e) in 25 g. of-purified 1,2-dichloro- ethane in a 50 ml. flash in an ice bath was stirred magnetically while 70.0 mg. (0.525 m. mole) of aluminum chloride (Mallinckrodt AR grade) 20 was added through a side—arm. (The catalyst was weighed in a 7 mm. Pyrex tube that was sealed at one end and tightly closed with a Tygon cap. The cap was made by welding Tygon tubing with hot crucible tongs.) After 5:00 min. of stirring, the reaction was quenched by the addition of 0.10 ml. (0.79 m. mole) of dimethylaniline. The ice bath was re- moved, and the liquid was vacuum distilled into a receiver cooled with a slush of dry ice—trichloroethylene. The distillate was 25.37 g. of colorless liquid (95% recovery). It was analyzed by gas chromatography. The results of this first run are in Tables Ia and Ila. There was no evidence of polymer-formation. Remaining in the reaction flask after the distillation was 0.23 g. of pale yellow solid on the walls; it was water-soluble and smelled like the amine. Found in the first of two liquid nitrogen traps following the receiver was 0.26 g. of colorless liquid with composition similar to that of the distillate. In later runs there were different concentrations of t—amyl chlor- ide, amounts of catalyst, and duration of run. First Run with 1-13C-thmy1 Chloride and 2-13C—thmyl Chloride A mixture was made by adding 107%; x 0.80 ml. = 0.83 ml. of 1-13C- t-amyl chloride with 50.0 atom percent 13C to 1%??? x 0.80 ml. = 0.37 ml. of 2-13C—t-amyl chloride with 57.7 atom percent 13C. Measurement was with a 1.00 ml. syringe graduated in hundredths of 3 ml. The 0.80 ml. was injected into a 20% silicone SE-30 preparative column (1 in. x 100 in.) at 520 and 8.5 p.s.i. in a Perkin-Elmer model 158L. Collected from 15.0 to 21.0 min. was 881.3 mg. of t—amyl chloride. A solution of 356 mg. (3.33 m. mole) of the labeled t—amyl chloride in 2.02 g. of 1,2—dichloroethane was treated with 2.5 mg. (0.0188 m. mole) 21 of aluminum chloride for 15 sec. and then quenched with 9.0 ul. (0.071 m. mole) of dimethylaniline. Ninety-five percent of the organic halides was recovered for purification by preparative gas chromatography. The 2.88 g. was injected in three installments into a 20% SE—30 column at 500 and 8 p.s.i. Unfortunately the large solvent peak preceded the solute peak and tailed into it. The material collected at 17 to 28 minutes was only 80% pure. About 0.15 ml. of it was put through the column a second time, giving 35.6 mg. for mass spectral analysis. The final sample was shown to be free of solvent by analysis with a 30% Silicone DC-550 column. Second and Third Runs with 13C—_t—Amy1 Chlorides A mixture of 0.99 g. of 1-13C- and E??? x 0.99 g. = 0.86 g. of 2-13C- Efamyl chloride was purified by preparative gas chromatography as before. A solution of 350 mg. (3.26 m. molefi of labeled t—amyl chloride in 2.02 g. of 1,2-dichloroethane (freshly distilled from phosphorus pentox- ide) was stirred with 3.8 mg. (0.0285 m. mole) of aluminum chloride for 60 sec. Six ul. of dimethylaniline (0.087 m. mole) was added to quench any reaction. The volatile liquids were pumped out and separated by two passes through the preparative gas chromatograph. A solution of 836 mg. (8.07 m. moles) of labeled t-amyl chloride in 2.86 g. of 1,2—dichloroethane (freshly distilled from phosphorus pentoxide) was stirred with 6.6 mg. (0.050 m. mole) of aluminum chlor- ide for 120 sec.. Quenching was by 8.3 ul. (0.065 m. mole) of dimethyl- aniline. Ninety percent of the yolatile liquids was recovered by vacuum distillation. The distillate was purified with preparative gas chroma— tography. APPENDIX Solvents for Carbonium Ion Rearrangements Methylene chloride is a non-nucleophilic solvent that has been used for Friedel-Crafts acylations,5 for cationic polymerizations of isobutylene (with aluminum chloride),4 and for studies of carbonium ions.18 It was the solvent for a Prins reaction of hexachloropropene and 1,2-dichloroethy1ene14 at 5°. Methylene chloride adds to 1,2—di- chloroethylene at 80—600.15 This solvent has dielectric constant 9.18 at 200 and 10.02 at 0°,16 nearly the same as that of t—amyl chloride (9.3 at 16°).6 It would have been used in these studies if it had been possible to separate it from t—butyl chloride on the gas chromatographic columns available. Ethylene chloride has dielectric constant 10.65 at 2006 and 11.66 at 1°.7 Waterman et al.26 found that it reacted with 10% aluminum chloride (thus 0.0825 mole A1C13 per mole of CHZCICHZCl) at 85-550 to give resin in 80% yield. A more remarkable reaction was.reported by Sisido and Yosikawa.8 The same concentration of aluminum chloride was used at 260 for 75 min. The products included bibenzyl and m—bis (fl-phenylethyl) benzene! Apparently the chlorocarbon condensed to benzene, which underwent Friedel-Crafts alkylations. The total yield of condensation products was about 12%. In spite of these reports, ethylene chloride has been used as a solvent for Friedel-Crafts acylae tions5 and Fries rearrangements.76 It did not react under the condi- tions used in this study. There was the disadvantage of difficulty of separation. The best preparative gas chromatography column gave a large solvent peak followed by a'small solute peak in its tail. 22 23 Other solvents were not tried. Possibly one of the Friedel-Crafts solvents with high gas chromatographic retention time would be more satisfactory. Several possibilities are tabulated below with three major components of the reaction mixture. The gas chromatography data1° are for two columns: A, 20% fi,fl'-oxypropionitri1e on Chromosorb and B, 20% tri—m—tolyl phosphate on Chromosorb. Both were at about 50 m1./ min. in a Burrell Kromo—Tog K-2. Carbon disulfide has very recently29 been used in a study of isomerization of butyl bromides. When 0.31 mole of aluminum bromide and 7.6 moles of carbon disulfide per mole of n-butyl bromide were stirred at 0°, eighty percent of the butyl bromide was isomerized in about ten minutes. Small amounts of t-amyl bromide and neopentyl bromide said to be present were rationalized with bimolecular mechanisms incorrectly based on those of Karabatsos it al.2d Chloroform2°>3li32 and nitro compounds form soluble solid complexes with aluminum chloride. The complexes with nitromethane3° and nitro- benzene2° are excellent catalysts for alkylating aromatic compounds. Aluminum chloride rapidly exchanges chlorine with carbon tetra- chloride and chloroform, even at their melting points.28 Olah's comprehensive new source book on electrophilic organic reactions97 includes numerous discussions of solvents for aluminum chloride. F~mf ya. .u h A a A. v ..- . o . -Av- - Hi hi (\3 ran UA-v-‘Z M RR Q «v-4 ..‘“ 28 Table VIIIa. Comparison of possible solvents. Dielectric Constanta SolubilityERetention Times Compound B.P. 00 20° Oigij8§ A5égin.)1;,3o 2—Me—2-butene 38.8 ---. -—- -—~ 1.8 1.2 t-BuCl 51.2 10.95 --- --- 2.3 --- t-AmCl 86.0 —-- 9.3 --- 8.0 8.0 _ (16°) cs2 86.3 2.69 2.68 0.2 2.0 - 2.2 CHC13 61.2 5.16 8.81 1.00 at 00 10.6. 5.5 cm4 76.8 2.28 2.28 0.78b 8.0 8.3 @12012 80.1 10.0216 9.08 --- 6.8 2.8 (CH2C1)2 83.5 11.66(1°)7 10.65 < 8% 22.0 7.0 CH3N02 101.0 -—- 37.5 600 63.1 7.9 c120c012 121.09 -—— 2.86 —-- --— 12.5 C6H5N02 210.9 39.3811 35.7 v.3. -—— --— aReference 6 unless otherwise noted. bVa1ue given for 8°. Others state28 that solubility is less than 0.0033m (0.88 g./1.) at 00. PART B HIGHER-ORDER WAGNER-MEERWEIN SHIFTS 2S INTRODUCTION Examples of 1,2-hydride shifts have been well—known for years; the question of whether there is a 1,3-hydride shift in an open chain system has been settled only recently. Karabatsos and Orzech33 showed that rearrangement of l-propyl cation is almost all by a 1,3-hydride shift rather than by two successive 1,2-shifts. Their experimental method was to deaminate deuterium-labeled propylamine, collect the 1-propanol, and study its nnmzspectrum. The results clearly showed that equation (1) rather than (2) describes the rearrangement. 1,3—H:nu + 4- H20 1 2—H:~ 1 2-H:~ H2° + + + CH3CD2CD2 —L——> CH3CDCD3 —2————> CHZCHDCD3 > HOCHZCHDCD3 (2) After noting the importance of 1,2- and 1,3—hydride shifts, it is reasonable to look for 1,8- and 1,5-hydride shifts. Do they occur? If so, to what extent and under what conditions? There are numerous examples of 1,5- and 1,6—hydride shifts in medium-sized rings -- dating back twelve years to independent discover— ies of transannular hydride shifts by Cope34 and by Prelog.35 The electron—deficient carbons have been generated by several means: hydroxylation of olefins with performic acid, solvolyses of cycloalkyl , tosylates, and nitrous acid deaminations of cycloalkylamines. This field has been reviewed recently.36 Few higher-order (1,8; 1,5; 1,6; ESE“) hydride shifts have been reported for aliphatic noncyclic systems. Deamination of some steroids 26 27 With especially favorable geometry goes 213 a 1,5—hydride shiftz59 R' R' .R" ’ ".va " NaNO 2 aq. HOAc H0(CH2)ZO (3) > H0(CH2)20 o —5-a'H S bCHzNHZ Letsinger and coworkers4935° have found 1,5—hydride shifts in per'- substituted naphthalenes. For example, 0H ¢' on H H— c’¢ . Hdhh H '9 OI‘ + + H HOAgr+ 12 < > <:> H2504 OH + H0 + + H-t’OH €22 \ PH C82 H 20 8022 (Quantitatively) Cohen et 31. have referred to four other nontransannular 1,5—hydride shifts63 and have demonstrated a new example:54 28 1 1 .+ + N2 H 3+}, 0 + CH -N . H 0 I 2 _._>2 3+! 2 > 8811801129! (5) ‘N\CH2-¢ N\CH20 + o 0 otho In most of these examples a 1,5-hydride shift is facilitated by a molecular framework that favors the formation of a six-membered cyclic transition state. The simplest systems to check for 1,8- and 1,5-hydride shifts are 1-butylamine and l-pentylamine, respectively. If each amine is.1abeled with deuterium at carbon no. 1 and deaminated, the reaction paths of interest are + CH -CD ? CH —CD H HONO 2 2 2 2 (CH2)n—CH2 (CHZ)n_‘CHZ H 0 H 0 V 2 V 2 (o) (n = l or 2) HOCDZ(CH2)n+lCH3 H0CH2(CH2)n+1CD2H Rearrangement is indicated by the presence of protons on the carbon bearing the hydroxyl (easy to spot by n.m.r.). RESULTS The deamination of aqueous solutions of pentylamine gave l-pentanol, 2-pentanol, 3-pentanol, pentenes, pentyl nitrites, pentyl nitrates, and nitropentanes. Similarly, butylamine gave l-butanol, 2—butanol, butenes, butyl nitrites, butyl nitrates, and nitrobutanes. Tables Ib, Vb, and IXb-X£R>summarize the product analyses of some of the deaminations. Identi- fication of products was by gas chromatography supplemented by infrared analysis. Known samples of alcohols, olefins, and nitrite esters were used for comparison. Preparative gas chromatography was used to obtain milliliter samples of l-pentanol and l-butanol from deamination of pentylamine-1,1—d2 and butyl- amine-1,1—d2. The n.m.r. spectra (Fig. 2b and 5b) of the neat alcohols showed no a-protons in the l-pentanol and only a slight amount in the 1-butanol (‘1’= 6.83 p.p.m.). The deuterium contents of the starting material and products were determined by mass Spectrometry. The labeled amines were run as the diacetylated derivatives (N—alkyldiacetamides), and the alcohols as tri- methylsilyl ethers. Tables IIb—IVb and VIb-VIIIb outline the analyses obtained. 29 30 Table Ib. Percentage yields of deamination products from pentylamine and pentylamine-1,1-d2 (second deamination) Pentylamine-1,1-d2 Pentylamine (Second Deamination) B.P. Pentenes 7 l9 20-38° 2-Am0NOe'c and ‘3 1, 96° B-AmONO 98° 25 2-Am0H% and 3-AmOH 18 119° 8 116° 21 l-AmOH 22 138° l-AmNOz 1 172° 1-Am0N02 1 157° 2-AmN02~X- and :f 0.1 150° 3-AmN02 152° 0.6 188° 180° 2-Am0N02% and 3-Am0N02 l-AmONO } .. ,. 2 108° 1 1 1 K'ch Totals 59 68 L- *The analyses did not distinguish between 2—penty1 and 3vpenty1 compounds. Mass spectrometry and gas chromatography (10 ft. Carbowax 20M at 68°) showed that the secondary pentanol was 22% 3- and 78% 2-pentanol. 31 Table IIb. Mass spectra of N-pentyldiacetamide and of N-pentyl-l,l—d2— diacetamide. Mass Ion Unlabeled. Labeled Mass Ion Unlabeled Labeled 58 27.7 21.1 118 C5H5N02+ 79.1 10.2 59 8.62 18.6 115 5.33 8.21 60 02116110+ 118. 129. 116 -0.07 81.1 61 0.59 12.9 117 —- 5.89 62 0.18 0.50 118 —- 0.08 70 18.0 8.23 125 5.68 -- 71 11.1 13.3 126 0.80 72 C3H6N0+ 519. 20.7 127 0.59 6.18 73 201. 37.0 128 78.1 1.00 78 2.25 897. 129 C7H15NO+ 96.3 9.66 75 0.18 209. 130 6.12 57.1 76 0.07 1.82 131 0.03 86.0 77 0.69 0.73 132 -- 7.77 --- 133 -— 0.35 96 1.38 —— —-— 97 0.93 0.70 153 8.50 -- 98 1.63 1.08 158 0.31 0.62 '99 0.97 1.51 155 0.0 3.86 100 C4H5N02+ 59.0 8.83 156 0811141102+ 100.0 0.86 101 12.0 12.3 157 -0.10 5.80 102 92.6 183. 158 -0.07 100.0 103 -0.07 7.73 159 -- 0.12 108 -- 8.88 160 -- 0.08 105 -- 1.35 --- 106 -— 0.27 171 0911171102+ 17.8p 0.89 107 -- 0.19 172 1.89 0.62 110 0.69 “ 173 0.03 1§;§p 111 7.09 —- 178 -- 2.98 112 9.13 1.66 175 -- 0.19 113 1.11 5.56 32 Table IIIb. Mass spectra of trimethylsilyl ethers of labeled l-pentanol from deamination of pentylamine-1,1-d2 and of unlabeled 1-pentanol. Labeled Unlabeled Mass Raw Peaks Monoisotopic Raw Peaks Monoisotopic 99 13.0 13.0 88.7 88.8 100 19.2 17.7 6.8 0.1 101 16.8 18.8 182.1 177.5 819.8 102 28.0 20.6 757.2 ' 27.0 7.1 103 165.0 162.8 600.0 589.2 = 71.9% 108 83.6 27.1 58.2 0.2 105 511. 502. = 66.3% 27.1 1.8 106 89.5 —0.2 1.3 -0.2 107 20.0 0.8 0.2 0 115 3.0 2.8 22.5 116 5.0 8.6 5.3 117 15.3 18.8 6.7 118 5.0 3.3 0.8 119 6.1 5.6 3.2 120 1.1 0.8 -0.1 129 0.9 0.8 26.0 130 2.0 1.8 0.1 131 19.0 17.8 1.8 132 2.5 0.1 0.1 133 2.2 1.1 0.7 183 -— -- 17.0 17.0 188 1.7 1.7 2.7 0.5} 2326.6 185 3.3 3.1} 2310.0 2309.1 = 99.3% 186 77.0 76.6 2283.7 297.0 -2.8 187 2196 2168 = 96.5% 828.0 -0.05 188 288 58.8 0.6 189 93.9 28.0 2.3 150 7.0 2.9 0.1 151 0.9 0.8 0.2 159 0.9 (3.9 10.3 9.6 160 2.8 2.3 19.2 17.6p 161 1.3 0.9 3.0 0.1 162 15.5 15.3p 1.0 0.2 163 2.7 0.5 0.2 0.1 168 0.8 0.1 -- -- aMonoisotopic peak intensities are corrected for presence of hexamethyl- disiloxane (0.3 volume percent of the labeled sample and 5.0 volume percent of the unlabeled sample). 33 Mass spectra of trimethylsilyl ethers of labeled 2-pentanol and 3-pentanol from deamination of pentylaming-l,l-d2, un- labeled 2-pentanol, and unlabeled 3-pentanol. Table IVb . Labeled Mixturg: 4+7Un1abe1ed Mass Raw Peaks Monoisotopic Monoisotopic Monoisotopic . 2-Pentanol 3—Pentanol 99 8.3 7.5 28.3 100 3.1 1.1 -0.5 101 21.9 71.3 65.7 102 19.8 80.8 0.5 103 62.8 36.6 5.2 108 38.1 -0.2 0.5 105 8.2 0.8 0.3 115 10.0 9.8 11.8 50.1 116 6.2 5.0 + 8.9 13.9 117 135.0 133.6 6.3% C5H130Si + 2233.2 10.9 118 96.0 81.3 3.8% C5H12D0Si + 0.1 119 1918 1900 89.8% C5H11D20Si 0.6 120 207 -0.7 121 76.0 -0.1 122 8.0 0.6 123 0.8 0.3 129 6.0 6.0 18.0 8.1 130 2.7 2.0 + 1.3 7.6 131 279.6 277 88.8% C6H15OSi + 0.2 2226.6 132 88.5 11.8 2.0% 06H14DOSi + 0.6 —1.0 133 297 283.5 89.6% CsHlsDZOSi 0.5 1.5 138 35.0 0.2 135 12.0 0.2 136 1.0 0.3 183 -- -- 8.3 3.5 188 -- -— + 1.0 1.3 185 129.0 129.0 20.3% C7H170Si + 598.6 286.6 186 35.0 18.8 3.0% C7H15DOSi + -0.7 -0.3 187 587 886.3 76.7% C7H15D20Si 0 0 188 72.0 -0.5 189 25.0 0.1 150 2.0 0.0 159 1.1 1.1 22.3 19.5 160 1.9 1.7 0.1p 0.1p 161 25.8 25.1 0.9 0.03 162 3.9 ' 0.2p 0 0 163 1.9 0.7 0 0 aMonoisotopic peak intensities are corrected for presence of hexamethyl- disiloxane (0.8 volume percent in the labeled mixture, 0.8 volume per- cent in the unlabeled 2-penty1 ether, and 9.8 volume percent in the unlabeled 3-penby1 ether). 38 W/ Figure 1b. N.m.r. spectra of pentylammonium perchlorates (80% solutions in deuterium oxide): (A) 10% unlabeled salt and 90% 1,1— dideuterated salt; (B) 1,1-dideuterated salt. B 1:7, 5 6M7 8' 9. 10 Figure 2b. N.m.r. spectra of l-pentanols (neat samples): (A) unlabeled; (B) product of deamination of pentylamine-1,1-d2. 36 C WM%~.~ Wm, W?“ T = 8 5 6 7 8 9 10 Figure 3b. N.m.r. spectra of secondary pentanols (neat samples): (A) unlabeled 2-pentanol; (B) unlabeled 3—pentanol; (C) products of deamination of pentylamine-1,1-d2. “NWMW{ ‘88 8 M 1888‘“) VWA “VAHWANwNVfi \ Wx‘whflw ”'UVK‘L' .pamwkfirk “NM L,\v_.2 Figure 8b. N.m.r. Spectra of butylammonium perchlorates (82% solutions in deuterium oxide): (A) 8.9% unlabeled salt and 95.1% 1,1- dideuterated salt; (B) 1,1-dideuterated salt. 38 1-1. 6""? a 9' 10 5b-B T: 8 5 6 ‘7 8 9 10 U8" 8* 6b- .___~3,3__j T= '8 5 6 ' 7 8 9' 10 'Figure 5b. N.m.r. Spectra of 1-butanols (neat samples): (A) unlabeled; (B) product of deamination of butylamine-1,1-d2. Figure 6b. N.m.r. Spectrum of labeled 2-butanol (neat sample) from de- amination of butylamine—1,1-d2. 39 Table Vb. Percentage yields of deamination products from butylamine and butylamine-1,1-d2. Butylamine Butylamine-1,1-d2 B.P. Butenes . 6 16 —6° to 8° 2-Buty1 nitrite 23: 8 5 68° 2-Butan01 c_a_. 15} 23 20 100° l-Butyl nitrite 10 6 76° l-Butanol 18 22 1180 2-Buty1 nitrate 0.7 1 128° 1—Buty1 nitrate 2 130° 2-Nitrobutane 0.2 :} S 1800 l-Nitrobutane 0.8 1 153° Butylammonium perchlorate 0.2 0.8 -- Totals 56 77 80 Table VIb. Mass spectrum of N-butyl-l,l-dZ-diacetamide W Mass Raw Peaks Monoisotopic 113 2.0 2.0 118 1.9 1.8 115 13.0 12.9 116 155.8 158.6 117 158.8 189.1 118 38.6 28.8 119 3.6 0.9 120 1.6 1.2 180 1.0 1.0 181 18.3 18.2 182 6.0 8.8 1.2%do 183 12.0 11.5 3.0%d1 3.0%dl 188 368.0 367.0 95.8%d2 97.0%d2 185 30.0 -0.3 186 9 0.3 187 0.3 -0.2 157 0.8 0.8 1.3%do 158 8.3 8.2 7.1%d1 159 58.6 58.2p 91.6%d2 160 18.7 9.6 161 1.7 0.8 162 1.3 1.1 81 Table VIIb. Mass spectra of trimethylsilyl ethers of labeled l-butanol from deamination of butylamine-1,1-d2 and of unlabeled l-butanol. Labeled Unlabeled Mass Raw Peaks Monoisotopica Raw Peaks Monoisotopicb 99 5.0 5.0 3.9 3.8 100 5.0 8.5 0.9 0 101 7.3 6.7 19.0 17.8 156.8 102 18.8 18.0 685‘7 28.0 25.3 103 ‘22.5 20.9 118.3 109.9 =70.1% 108 62.8 60.2 12.2 0.8 105 581.0 578.8 = 83.8% 6.3 0.8 106 57.6 0.08 0.3 0 107 23.0 0.7 -- -- 108 1.0 -0.05 -- -- 129 -- -- 1.8 1.8 130 7.0 7.0 0.3 0.1 131 5.0 2.7 2808.6 578.0 572.1 = 99.7% 132 79.2 78.8 68.5 -0.2 133 2730.0 2720.5 = 97.1% 26.7 0.9 138 328.0 -1.6 2.8 0.6 135 111.6 -0.2 1 0 136 7.0 0.5 -— -- 137 0.6 0.6 -— -- 185 0.6 0.6 3.3 3.3 186 19.0 18.9 3.3 2.8p 187 26.2 23.7 128.8 f 0 188 18.1 10.2p 19.6 0 189 3.7 1.3 9.8 0 150 0.8 0.1 1.0 0 aA very small impurity of hexamethyldisiloxane was not corrected for. bCorrection was made for the 1.9 volume percent of hexamethyldisiloXane present. 82 Table VIIIb. Mass spectra of trimethylsilyl ethers of labeled 2-butanol from deamination of butylamine—1,1—d2 and of unlabeled 2—butanol. Labeled Unlabeled Mass Raw Peaks Monoisotopica Raw Peaks Monoisotopicb 118 0.9 0.8 -- -- 115 9.3 .7 6.0 5.1 116 8.0 3.8 2052 2.0 1.1 LZI'b 117 539.0 535.9 817.0 815.2 = 98.5% 118 112.2 58.1 88.8 0 119 1870.0 1851.1 = 70.7% 17.8 0.8 120 159.0 -0.2 1.0 0.9 121 58.0 -0.1 -- -- 122 3.8 0.8 -- -- 129 0.7 0.6 1.8 1.8 130 1.9 1.7 800.7 1.0 0.8 183 0 131 136.8 127.2 186.0 180.8 a 98.6% 132 38.3 21.8 22.0 -0.2 133 661.0 689.8 = 98.6% 10.6 1.0 138 78.6 -0.05 0.8 0.5 135 26.9 -0 08 0.1 0 136 1.8 0.2 -- -- 188 -- -- 0.1 0.1 185 0.8 0.8 6.9 6.9 186 2.6 2.5 1.0 O.9p 187 203.8 5.0 116.7 0 188 31.1 Op 18.3 '0 189 15.6 0 9.2 0 150 1.7 -0.2 0.9 0 aCorrection was made for the 3.1 volume percent of hexamethyldisiloxane present. bCorrection was made for the 1.8 volume percent of hexamethyldisiloxane present. DISCUSSION Pentyl.System The mass Spectrum of the diacetylated starting material (Table IIb) shows that the labeled molecules are mostly dilabeled. The major frag- mentation path is loss of methyl (from an acetyl group), giving peaks at 156-158. Calculation from their intensities gives 98.5% d2, 5.1% d1, and 0.8% do. The 0.8% is probably not real. Recalculation without it brings the figures to 98.9% d2 and 5.1% d1. The parent peak (at mass 173) is less reliable; its low intensity makes for larger experi- mental error. It indicates that the labeling is 91.2% d2, 3.6% d1, and 5.2% do. Dropping the d0 and renormalizing give the values 96.2% d2 and 3.8% d1. The l-pentanol from deamination showed no a-protons in the n.m.r. (Figure IIb), but the more sensitive mass Spectral analysis (Table IIIb) showed there was actually some monodeuterated compound. The fragmenta— tion of trimethylsiloxypentane can be as follows: CH3 1 I ' CH3CH2CHZCHZ -——L-- CH2081 -——+—*- CH3 ' ' CH3 : Q“ “~‘ Loss of methyl (Si-C cleavage) leaves an ion of mass 185, while the loss of butyl (C—C cleavage) leaves mass 103. The peaks in the region of 185 indicate 97.1% d2 and 2.9% d1. The interpretation of the 103 region re- quired the mass spectrum of the unlabeled l-pentyl ether. It had a peak intensity at 103 which was 71.9% of the total intensity of the 99-105 region. The corresponding parent-less-butyl peak from the labeled ether (at 105) had 66.35% of the total intensity of its region. Thus 83 88 66.3/71.9 = 92.2% of the a-methylenes were dideuterated. Further calcu- lations gave 5.0% d1 and 2.8% do at this position. Comparison of these figures with those from the 185 region Showed that the distribution of label was as follows: 92.2% C4H9CDZOH 2.8% C4H7DZCHZOH 2.1% C4H8DCHDOH 2.9% C4H9CHD0H. Recalculation without the contribution of the monodeuterated species gave: 95.0% C4H9CD20H 2.9% C4H7D2CHZOH 2.1% C4H8DCHD0H. The loss of deuteriums from the a-carbon is probably due to reversible hydride shifts within the pentyl system. Less likely are alkyl shifts such as 1,2-propyl or 1,3-ethy1. There was no loss of deuteriums by exchange with the solvent. The data of Table IVb are consistent with the amount of monodeu- _teration previously found. The peaks at 131 and 133 are clearly due to the loss of C2H5 or CzHSDz from the 3-pentyl ether with about equal probability. The data indicated very little deuterium at position 3, which means little reversible 1,2; 1,2- or 1,3-hydride Shift. The amount of 3-pentyl ether can be calculated on the basis that the 2—pentyl ether readily loses propyl but not ethyl, while the 3-pentyl ether loses ethyl and cannot lose propyl. Using sensitivities derived from the Spectra of unlabeled compounds, the calculation of the composition of the labeled mixture was: 8S _ intensities at 131 to 133 _ 571.9 = % Mm - - 28.6 ' m 23'? intensities at 117 to 119 = 2118.5 % 2'Am = 27.3 27.3 = 77.5 Normalizing and omitting the 0.8% of hexamethyldisiloxane bring the figures to 23.0% 3-pentyl and 77.0% 2-pentyl. The finding of 3-pentyl compounds was not expected, Since Streitwieser44 has stated that no 3-pentan01 is formed in the deamina- tion of l—pentylamine. Gas chromatographic analysis of the intractable pair of secondary alcohols showed that 21.7% was 3-pentanol and 78.3% 2-pentanol. The peak at 117 shows that there iS some 2-pentyl-5,5-d2 ether present with the 2-penty1-l,l-d2 ether. This suggests that there were successive 1,2-hydride shifts down the pentyl chain -- past the mid- point -- to give all three secondary carbonium ions: 75% 2-pentyl-l,l-d2, 21% 3-pentyl-l,l—d2, and 8% 2-pentyl-5,5-d2. The over-all pattern for the deamination of pentylamine-1,1-d2 is summarized in the following outline. Butyl System The n.m.r. spectrum of the primary alcohol from the deamination had a slight amount of o-protons. (See Fig. Vb.) The mass spectrum of the diacetylated starting material (Table VIb) Shows that 97.0% is dideuterated and 3.0% monodeuterated. (These figures are from the parent—less-methyl region of masses 182-188.) It was not possible to calculate the results from the mass spectral analysis of the labeled 1—butyl Silyl ether. The spectrum of the un- labeled ether was not comparable with that of the labeled ether in the 86 Figure 7b. Outline of the deamination of pentylamine-1,1—d2. 9.2 2% n-CQHQCDZOH 95 .0% 2. 8% n-C4H7DZCHZOH 2.9% 2.1% n—C4HBDCHDOH 2.1% 22%{297 9%n-C4H9CHD0H 8.9 n—C 37170110110 2 I CHgCHZCHZCH g7CDZNH2 OH 8% 2% (98.W %n—C3H5D2CHCH3 ' * 18% 0” CH3CH2CH2CH2CHDNH2 %n-C3H7CHCH2D 3.8% (5.1%)19 5H 8% 96.0% CZH5CHCH2CHD2 1 OH 8.0% C2H5CHCH2CH2D 1 OH : 19% Pentenes 9% Other products (pentyl nitrites, pentyl nitrates, and nitropentanes) *Parenthetical figures are from parent—less—methyl peaks of pentyl— diacetamide. 87 103 region. Comparison with the published spectrum93 of l-butoxytri- methylsilane indicated that the unlabeled sample had impurities with masses 101 and 102. The 2—buty1 Silyl ether was simpler than the 2-penty1 (since there is only one secondary butyl cation). Calculations from the intensities in the 117 region (loss of ethyl) Show that 70.7/98.5 = 71.8% of the ions are dideuterated, 2.7% are monodeuterated, and 25.5% are undeuter— ated. Thus 25.5% of the labeled 2-butyl ether molecules had two deuter- ium atoms in the ethyl group. The 131 region (loss of methyl) gave the distribution 82.3% d2, 2.8% d1, and 15.0% do. This means that 15.0% of the methyl groups lost were CDZH. It does not tell what fraction of the no. 1 carbons of butyl were dideuterated, since methyls were lost from the trimethylsilyl as well as from butyl. If the isotopic purity is assumed to be the same in the 2-butyl ester as in the l-butyl ester (97.0% (12 and 3.0% d1), then 97.0—25.5 = 71.5% of the 2-buty1 groups con- tained CDZH. This implies that 15.0/71.5 = 21.0% of the methyl groups lost came from butyl and 79.0% from trimethylsilyl. The information obtained on the deamination of butylamine is summarized in Figure 8b. Streitwieser and Schaeffer94 reported that deamination of butyl- amine-1,1—d2 in anhydrous acetic acid gave no ethyl rearrangement. They converted the esters formed into alcohols and analyzed the l—butanol for deuterium by mass spectrometry. (No results were given for 2-butanol.) The l-butanol from the reaction gave intensities 12.3 at 31 (CHZOH)+, 26.1 at 32 (CHDOH)+, and 63.6 at 33 (c0208)+. The l-butanol-l,l-d2 from which the butylamine was made gave intensities 13.3, 25.8, and 66.8,reSpectively. Thus there seemed to be no rearrangement (within experimental error). However, mass spectrometric studies of alcohols 88 Figure 8b. Outline of the deamination of butylamine-1,1-d2. 97.1% 1—C4H7D20H 2.8% l—C4H8DOH 0.1% 1-C4H90H 22% 71.6% C2H5CHCD2H CH3CH2CH2CD2NH2 5H 97.0% 2%..) 25.5% 02830201083 agGQGQGmMQ OH 3.0% 2.8% CZH5CHCHZD 1 OH 3 16% Butenes 18% Other products (butyl.nitrites, butyl nitrates, and nitrobutanes) 89 are known to be fraught with difficulties. This thesis reports more products of deamination than previous studies in the literature. In the definitive study,85 3.5 moles of butylamine in 1 1. of water was deaminated with 3.5 moles of hydro- chloric acid and 10.5 moles of sodium nitrite. No reaction occurred "in the cold". Heating to boiling gave these percentage yields: l-butanol 25.0 2-butanol 13.2 l—butyl chloride 5.2 2-buty1 chloride 2.8 butenes 36.5 high—boiling material 7.6 butyl nitrites (traces) Total 90.3 (Reportsaz,83 of isobutyl alcohol's being a product have been disproved.54:55) Streitwieser94 obtained a 80% yield of butyl esters (mostly acetates but including 8 to 8% nitrates), which were hydrolyzed to 65% 1—butanol and 35% 2—butanol. There was also an unspecified amount of butenes: 71% l-butene, 9% cig—2—butene, and 20% trans-2-butene. Deamination of butylamine in aqueous acetic acid at 70° gave butenes (60% 1—butene, 15% £5, and 25% iii—r18) in 26% yield.95 (No other products were discussed.) Pentylamine has been deaminated by converting its hydrochloride to the nitrite and then heating. The yields of products were:67 alcohols ("perhaps one-third" secondary) 5 . olefins 3 nitroso-secondary amine primary amine primary ammonium chloride :— OHHOO :— \lfl\0|—‘O Total 8 The formation of nitrite and nitrate esters is readily explained by the reaction of the alcohols with nitrous acid and its decomposition 50 product, nitric acid. Some of the nitrite ester may be due to nucleo- philic displacement by nitrite ion, which according to ambident anion theory,68 should form more nitrite ester than nitroalkane when it at- tacks either a carbonium ion or a diazonium ion. The low material balances found in this study are probably due to loss of olefins. Deaminations are known to give a larger proportion of elimination compared to nucleophilic displacement than these results indicate. EXPERIMENTAL Synthesis of Pentylamine37 Valeronitrile (Eastman white label) seemed pure by gas chromato- graphic analysis: one peak (at 20.0 min. on 30% silicone DC-550 column at 105° and 8 p.s.i.). One-tenth of a mole of it (8.31 g.) in 30 ml. of dry ether was added dropwise (in 35 min.) with stirring to ice-cold dry ether (250 ml.) and 3.8 g. of lithium aluminum hydride (lumps; 0.10 mole). The white suspension stood overnight. Four ml. of water, 3 m1. of 20% aqueous sodium hydroxide, and 18 m1. of water were added slowly with ice—cooling and stirring. The clear and colorless ether layer was decanted from the white solid. The solid was stirred three times with ether. The ether solutions were combined, dried over calcium sulfate, and filtered. The filtrate was refluxed one hour with barium oxide and then distilled at 35° through a 5-inch packed column. The residual 13 ml. was filtered and distilled, giving ether and three fractions: 152-2 1.03 g. 92—102° 152-3 2.85 g. 102.0-102.5° 152-8 1.81 g. 23, 80° (reduced pressure) Total 5.29 9. Each of the fractions was about 95% pure pentylamine by gas chroma- tography (30% Silicone DC—550 at 111° and 8 p.s.i.). The yield of amine was 58%. Another run, using twice as much material and similar procedures, gave 11.83 g. of distillate at 98-103° which was 97% pentylamine and 3% ether by gas chromatography. The yield was therefore 68%. 51 52 First Synthesis of Pentylamine-1,1-d2 Valeronitrile (9.91 g.; 0.119 mole; Eastman white label, not dis- tilled) was reduced with 5.01 g. (0.119 mole) of lithium aluminum deuteride (from Metal Hydrides). The solvent was anhydrous ethyl ether that was freshly distilled from lithium aluminum hydride. Obtained was 6.78 g. of amine, 90-102°. It was 95% pentylamine and 5% ether by gas chromatography. (Yield 60%). Second Synthesis of Pentylamine-1,1-d2 The valeronitrile was distilled after it was found to have one impurity. Gas chromatographic analysis with a 20% SE—3O preparative column Showed two peaks: 99% valeronitrile at 25 min. and 1% unknown at 62 min. The column was at 72° and 8 p.s.i.) The fraction at 89— 90° (102 mm.) was collected. Bad frothing during distillation was controlled with one drop of General Electric Antifoam 66. Quantities of reactants in this run were just twice those of the previous run. The yield of labeled amine was 18.50 g., which was pure by gas chromatography (68.5% yield). Deamination of Unlabeled Pentylamine One fortieth of a mole (2.18 g.) of pentylamine was added to an equivalent amount of perchloric acid (2.10 ml. of 71.6% acid). The liquid was removed with a Rinco evaporator, leaving white crystals, which were dried in a vacuum desiccator. The yield was 8.82 g. (98%) with m.p. 225-2310. One fiftieth of a mole (3.75 g.) of pentylammonium perchlorate was deaminated,4° giving 0.22 g. of alcohols (23' 12% yield). Gas 53 chromatographic analysis of them (30% Silicone DC-550 at 110° and 8 p.s.i.) gave: time (min.) ca. 89% 1—pentanol 12.6 ca. 86% 2-pentanol 7.9 ca. 8% less volatile unknown 23.5 The deamination product distilled from the reaction mixture was mostly water under a frothy emulsion of alcohols. Several methods were tried to dry the emulsion layer. A successful way was to add the emulsion to molecular Sieve 8A (Linde), using 10 g. of Sieve for each gram of water, let it stand for 85 min., and pump off the alcohol at room temperature and < 1 mm. Hg. (This gave 85% recovery of alcohol as a single phase from a mixture of 2.0 m1. of alcohol and 0.5 m1. of water.) One twentieth of a mole (8.36 g.) of pentylamine was reacted with an equivalent amount of perchloric acid (8.20 ml. of 71.6% acid), giv- ing a colorless liquid and a chunk of white solid. To this mixture was added 3.75 ml. of 71.6% perchloric acid (0.085 mole), 25 ml. of water, and then, dropwise, a solution of 7.50 g. of sodium nitrite (0.109 mole) in 10.0 ml. of water. The resulting reaction mixture was blue. It was distilled to give an organic layer and an aqueous layer. Satura- tion with potassium fluoride and separation of layers gave 2.5 ml. (2.02 g.) of organic liquid. (This would be a 86% yield if only pentyl alcohols were present.) During the reaction and distillation the gases and more volatile liquids were collected in dry ice traps: blue liquid. After evaporation of the blue N203 (b.p. 3.5°C) there was 0.8 m1. of yellow liquid. (This was a 7% yield of pentenes if it was hydro- carbons only.) S8 The distillate was fractionated by preparative gas chromatography. (The analysis is included in Table IXb.) Washing with 5% hydrochloric acid did not change the gas chromatographic analysis; little amine was present and little hydrolysis of nitrite esters occurred. First Deamination of Labeled Pentylamine The amine was converted to its perchlorate salt in two batches. In the first, 2.9 ml. of 71.6% perchloric acid was added dropwise to 3.19 g. of pentylamine-1,1—d2 (0.035 mole). Vacuum pumping on the product left 6.88 g. of colorless crystals with m.p. 218-219° (98% yield). More salt was made from 1.87 ml. of 71.6% perchloric acid and 1.62 g. of pentylamine-1,1-dz (0.018 mole). This salt was added to 6.20 g. from the first batch, giving an estimated 9.88 g. (0.050 mole) of pentylammonium-1,1-d2 perchlorate, which was dissolved in 3.75 ml. of 71.6% perchloric acid and 20 m1. of water. To this pale brown solu- tion was added a solution of 7.50 g. sodium nitrite (0.108 mole) in 15 ml. of water over a half-hour period. The reaction mixture was blue-green. Distillation from a 105-125° oil bath gave 2.5 ml. yellow organic liquid and 6.0 ml. colorless aqueous layer at 76-970. The organic layer was dried over magnesium sulfate. Two liquid nitrogen traps were used to collect volatile liquids from the reaction and distillation. In the first was collected a deep blue liquid. Evapora- tion of nitrogen oxides left 0.65 g. of liquid which was analyzed by gas chromatography (See Table IXb.) The organic layer was separated into five fractions by preparative gas chromatography: pentenes, 2- pentanol, l—pentanol, and two unknowns. The 2-pentanol fraction con- tained a large amount of impurity (at least twice as much as in the 55 Table IXb. Products of deaminations of pentylamines (gas chromatographic analyses and separations, .using a 100 in. x l in. 20% SE-3O on Chromosorb W column). 72° and 8 p.s.i. 95° and 3 p.s.i. C d Time Pentylamine Pentylamine-l 1—d2 Pentylamine-l l-dz ompoun 5 (min.) (distillate) (First Run) (Second Run) (Distil- (Distil- . late) (Trap) late) (:12?) Pentenes ‘}. 3.1 1% 8% 15% é 8.8 Unknown 7.2 0 0 7 O -- 2°-.'.117101\10"'r 11.8 86 V * 85 39 39 19 2°~AmOH 12.0 0 l-AmONO 15 13' 26 31 8 22 l-AmOH 21 35 22 0 38 32 2°-Am0N02* , * 81 2 2 O 2 S8 +20-AmNO2 l-AmONOz 88 .2 1 7 o 8 {8, +1-AmN02 *The analyses did not distinguish between 2-penpyl and 3—pentyl compounds. 56 previous deamination.) The identity of the impurity was uncertain. One possibility with the right retention time was pentylamine. The 2-pentanol fraction was washed with 5% aqueous hydrochloric acid, dried, and analyzed again by gas chromatography: There was no change. One curious result was the gas chromatogram of a known sample containing 0.2 m1. of pentyl- amine, 0.2 ml. of l-pentanol, and 0.2 ml. of 2-pentanol. These compounds had retention times 16 min., 20 min., and 12 min., respectively, but the mixture gave only the two alcohol peaks. An infrared spectrum of the impure 2-pentanol had very strong absorption at 6.05 u, indicating nitrite ester. The two likely nitrites, l—pentyl and 2-pentyl were synthesized by a method based on that for l— butyl nitrite.38 Yields of 76% and 88%, respectively, were obtained on a scale of 0.05 mole (cf. 81—86% for 5 moles38). Although nitrite esters are said39 to be extremely easy to saponify, several experiments showed that the nitrites of l—pentanol, 2-pentanol, and 3-methyl—l-butanol resisted hydrolysis in refluxing 20% aqueous sodium hydroxide for several hours. Refluxing for about 18 hours was required to saponify 95% of l-pentyl nitrite. Second Deamination of Labeled Pentylamine A sample of 18.22 g. of pentylamine-1,1—d2 (0.160 mole) was added slowly (in 30 min.) to 13.0 ml. of 71.6% perchloric acid (0.160 mole), giving a pale pink slush, which was pumped on (< 1 mm.) until it solidi- fied to a mass of pale pink crystals, m.p. 209—229°. After storage in an evacuated desiccator for two days, the crystals were white, weighed 28.98 g. (95% yield) and had m.p. 223—2280. Recrystallization from 66 ml. of heptanol (on steam bath) by addition of 1500 m1. of purified 57 petroleum ether gave 26.80 g. (93% recovery) of white crystals, m.p. 225- 229°. (The "30-600 pet. ether" from the stockroom was shaken with con- centrated sulfuric acid (twice in this case) until the acid layer was colorless and then distilled slowly. The 80—55° cut was used.) Another 3.08 g. of pentylammonium—1,1-d2 perchlorate, m.p. 223- 225°, was obtained by careful addition of perchloric acid to the ether fractions from the distillation of the labeled pentylamine and recrystal— lization as described above. N.m.r. Spectra (Fig. 1b) were run on satur- ated solutions (80 wt.%) of pentylammonium perchlorates in deuterium oxide. To 27.31 g. of pentylammonium-1,1—d2 perchlorate (0.188 mole) in dilute perchloric acid (10.8 ml. of 71.6% perchloric acid and 60 ml. of water) in a 3-necked, 28/80, 300 m1. flask, was added, from a dropping funnel, a solution of 21.60 g. sodium nitrite in 85 m1. of water. The reaction proceeded at room temperature for 2.5 hrs. Distillation at 73—980 then gave two liquid phases. The distillate was saturated with potassium fluoride and separated, giving 7.5 m1. of yellow organic liquid. During the reaction and distillation all effluent gases were passed through two dry ice traps. The second trap was empty. The first held 8 m1. of blue liquid which formed brown gas in air (N203 —> N02). After 15 min. of standing at room temperature, the blue liquid became 3.5 ml. of yellow liquid. The two liquids were analyzed immediately by gas chromatography. (See Table Xb for the results.) N.m.r. spectra were run with the neat samples of labeled pentanols collected by preparative gas chromatography. (See Figures 2b and 3b.) The primary and secondary pentanols had n54 1.8078 and 1.8010, respectively. Each sample (1.11 ml. 58 Table Xb. Products of second deamination of pentyl amine—1,1-d2. Analyses with 10 ft. x 1/8 in. column of 20% Carbowax 20M on Chromosorb W at 118° in Aero— Separations with 100 in. x l in. column of 20% SE-30 on Chromosorb W at 95° and 3 graph p.s.i. in Perkin-Elmer 158L Compounds ($128) Distillate Trap Distillate T139) Fraction Pentenes 3 2 5% 70% L1; 23 3 173-0 Unknown 7.8 0.6 O O -—- 2°-Am0N0* 10 5 * 12 39 19 173-2 20-AmOH 11.5 38 l—AmONO 17.8 8 2 8 22 (Shoulder) 'l-AmOH 22.2 86 12 38 32 } 173-3 2°-Am0N02* 18 2 0 % 2 58 173-8 +2°—AmN02 --- --- --- l-AmONOZ --- -—- ——— i 88 8 +l-AmN02 --- --- -—- 89 173-5 (Nothing (Nothing from 22 from 22 to 97 to 75 min.) min.) *The analyses did not distinguish between 2-pentyl and 3-pentyl compounds. S9 of labeled l-pentanol and 1.10 ml. of labeled 2-pentanol*) was refluxed for 16 hrs. with 1.2 ml. of hexamethyldisilazane (0.0058 mole for 0.010 mole of alcohol) and one drop of trimethylchlorosilane.48 The tri- methylsilyl ethers were purified by preparative gas chromatography (15 to 20 min. and 20 to 28 min. peaks at 1020 and 9 p.s.i.). The mass spectrometric data are in Tables IIIb and IVb. Unlabeled l-pentanol and 2-pentanol were converted to their trimethylsilyl ethers in the same way, but it was found that the long refluxing period48 was not necessary; the composition of the reaction mixture was the same after two hours as after twenty hours of refluxing. The problem of determining 3—pentanol in samples of 2-pentanol was solved as described below. The available gas chromatographic col- umns for Perkin-Elmer, Wilkens, and Beckman instruments were_tried with a known mixture of the two alcohols. Only one gave the slightest separa- tion. A 10 ft. x 1/8 in. column packed with 20% Carbowax 20M on Chromo- sorb gave 22.3% resolution41 at 98°. The peaks were at 20.6 and 22.0 min. The Wilkens Company42 obtained 78.8% resolution with a 20 ft. x 3/8 in. 30% Zonyl E-7 column at 92°. (Zonyl E-7 is the ester of pyro- mellitic acid with four molecules of H(CF2)nCHZOH.)42b (Later it was learned that complete separation could be obtained with 20 ft. Carbowax 800 columns.)43 No 3-pentanol was found in the deamination products of l-pentyl- amine when they were analyzed with the 10 ft. Carbowax 20M column. This result agreed with that of Streitwieser.44 Mass spectrometric analysis WIt was later found that 3-pentanol was also present. 60 (Table IIIb) of the trimethylsilyl ether of supposedly pure 2-pentanol from deamination showed that the 3-pentyl ether was also present. This prompted the reexamination of the 2—pentanol sample. It was 21.7% 3- pentanol by gas chromatographic analysis with the same Carbowax column but with a lower temperature, 68°. The alcohols had retention times 82.7 and 87.7 min. The resolution was 71 8%. Preparation of N-Pengylacetamide Two methods were tried for the preparation of N-pentylacetamide. Each gave a mixture of monoacetylation and diacetylation. The first method was that of Smith and Adkins:44 acetic anhydride and pentylamine at 0°. To 1.87 9. (0.0169 mole) of pentylamine was added dropwise 1.71 9. (0.0169 mole) of acetic anhydride (freshly distilled at 135.0-137.00) at 0°. The mixture was set aside for several days. The work-up was washings with 10 ml. of water, 10 ml. of 5% sodium hydroxide, 10 ml. of 5% hydrochloric acid, and 10 ml. of water, drying over calcium sul- fate, and distillation. Gas chromatographic analysis (20% SE-30 prepar- ative column at 128° and 8 p.s.i.) showed two peaks: 65% at 88 min. and 35% at 78 min. Preparative gas chromatography resulted in separa— tion but extensive decomposition-—brown liquids were collected when colorless samples were injected. Infrared analysis of the distillate Showed strong N-H absorption at 3.03 p. The second method46 was to use acetyl chloride in toluene at 60°. A solution of 1.87 g. pentylamine (0.0169 mole), 1.80 ml. acetyl chloride (0.025 mole), and 3.51 ml. pyridine (0.088 mole), in 17 ml. toluene was stirred magnetically in a 70-800 oil bath for 25 min. The reaction mix- ture (massive white precipitate and yellow liquid) was washed four times 61 Table XIb. Minor products of deamination of pentylamine-1,1-d2 (fractions collected by preparative gas chromatography). 1. Gas chromatographic analyses. (20% SE-30 at 110°) (20% Carbowax 20M at 128°) 173-8 Compound (81:?) 173-5 173-5 (81:?) 0 Unknown 1.0 0 2% 5.5 0 20—Am0H* 1.5 0 1% 9.7 73% 20-Am0N02* 3.0 0 12% 13.7 12% l-AmOH 2.0 1% 0 16.5 0 l-AmONOZ 8.1 83% 38% 17.3 0 1—AmN02 3.7 83% 51% 38.5 15% 2°-AmN02* 6.2 1% 0 est.20 0 Unknown 8.3 11% *The analyses did not distinguish between 2—pentyl and 3—pentyl compounds. 2. Infrared analyseS(lO% solutions in carbon tetrachloride) (wavelengths in microns) 173-8 173-5 No RONO (6.20) No RONO (6.20) Strong RONOZ: Strong RONOZ: 6.10 6.05 7.80 7.66 11.88 11.55 Medium RN02: Strong RN02: 6.80 6.82 7.20 7.18 7.32 62 with 20 ml. each time: water, 5% sodium hydroxide, 5% hydrochloric acid, and water. The organic layer was dried over calcium sulfate and distilled. Bad frothing was eliminated with one small drop of General Electric antifoam 66. Toluene came over at 68—650 (122 mm.), leaving a pale yellow liquid residue, 0.70 g. with n54 1.8897 (gi. lit.47 n55 1.8812 for N—pentylacetamide). Gas chromatographic analysis of this liquid showed only a trace of toluene (33' 0.3%), 29% at 17.5 min., and 71% at 29.0 min. (20% SE-3O prep. column at 160° and 8 p.s.i.). Several attempts to collect pure fractions gave decomposition at higher temperatures and very high retention times at lower temperatures (e;g;, nothing came through in 3 hours at 90° and 8 p.s.i.). Preparation of Labeled N-Penhyldiacetamide The second method46 was chosen, since it gave a greater propor- tion of diacetamide, which was more convenient for mass spectrometry. Some of the sample of pentylaminerl,l-d2 that was used for the second deamination had been set aside. A solution was made from 1.58 g. of the amine (0.0169 mole), 3.51 ml. of pyridine, and 17 m1. of toluene. Acetyl chloride (1.80 m1.; 0.025 mole) was added dropwise at room tem- perature. The reaction mixture (white solid plus liquid) was stirred and heated in a 70-100° oil bath for 30 min. It was washed with 20 ml. volumes of water, 5% sodium hydroxide, 5% hydrochloric acid, and water and dried over calcium sulfate. Toluene was distilled at about 700 (150 mm.), but the liquid residue still had about 15% toluene (by gas chromatographic analysis). Another hour of pumping (£3. 1 mm.) and heating (oil bath up to 150°) brought the toluene content down to 1.5%. Pumping (at the capacity of the vacuum pump) and heating (oil bath up 63 to 170°) were continued until all of the liquid vaporized and_condensed on the walls of the vigreux column. This liquid, 0.21 9. (n54 1.8898), was allowed to drain into a clean flask. It was 90% pentyldiacetamide and 10% pentylacetamide by gas chromatographic analysis. The mass Spectral analysis is in Table IIb. Preparation of Bugylamine Butyronitrile (Eastman white label) was distilled at 118.5—115.0°. A tenth of a mole (6.91 g.) in 30 ml. of ether was added slowly to 3.8 g. of powdered lithium aluminum hydride in 250 ml. of ether with ice-bath cooling. (The ether had just been distilled, from lithium aluminum hydride.) The mixture was stirred and refluxed for four hours. It was then chilled in ice and treated with 8 m1. of water, 3 ml. of 20% sodium hydroxide, and 18 m1. of water. The ether layer and three ether extracts of the aqueous paste were combined; dried, and distilled. Only 0.8 m1. of amine was obtained. However, the ether that had been distilled off was treated with perchloric acid until the mixture was barely acidic. Vacuum pumping removed the ether and water, leaving white crystals. After drying in a vacuum desiccator overnight the crystals weighed 11.85 g. and had m.p. 193-198° (lit.77 195-1970). This was a yield of 68.8% (g£. lit.37 57% yield of amine). Deamination of Butylamine No reaction occurred in the first three attempts, when the quanti- ties of reactants were 0.050 mole of butylammonium perchlorate, 0.18 mole of perchloric acid, and 0.108 mole of sodium nitrite. Decreasing the amount of acid brought success. 68 A clear, colorless solution was prepared from 8.83 g. of the per- chlorate salt of butylamine (m.p. l98-196°; 0.0508 mole), 3.75 ml. of 71.6% perchloric acid (0.085 mole), and 20 ml. of water. Sodium nitrite (7.50 g.; 0.108 mole) in 10 m1. of water was added in 20 min. at 25° with magnetic stirring. After two hours the reaction mixture was dis— tilled at 62-98° (oil bath 115-180°). The distillate was saturated with potassium fluoride, and the yellow oil separated. Obtained was 1.8 ml. (1.57 g.) of oil 220-A. Its analysis is in Table XIIb. Further distillation gave only water and a solid residue. Gases evolved during the reaction and distillation had been passed through two dry ice traps. The second trap was empty. The first held blue liquid. It was possible to weigh 0.7 g. of it in a vial cooled in dry ice and analyze it by gas chromatography. Its composition was 22% butenes, 1% unknown compound, 39% 2—buty1 nitrite, and 38% l-butyl nitrite. This sample (220-C) was sealed in a glass tube. Two days later the liquid contained a few sharp needles, presumably polymerized olefins. The residue from distillation was dissolved in water and made alkaline to phenolphthalein with less than 1 ml. of 20% sodium hydroxide. Ether extractions gave an amine which reacted with 0.12 millimole of perchloric acid to form 0.02 g. of amine salt (0.2% of the starting material). The total yields of deamination products are summarized in Table Vb. Preparation of Butylamine-1,1-d2 Ten grams of lithium aluminum deuteride (0.238 mole; Metal Hydrides) was used to reduce 16.82 g. of butyronitrile (0.238 mole) in the manner 65 used for the unlabeled amine. Obtained were 1.53 g. of butylamine-1,1—d2 (0.020 mole) and 29.87 g. of dry, white, crystalline perchlorate salt (0.168 mole; m.p. 198-1980), a total yield of 79.0%. N.m.r. Spectra (Fig. 8b) were run on saturated solutions(82 wt.%) of the salt in deuterium oxide. Deamination of Butylamine—1,1-d2 A solution was prepared from 26.55 g. of butylammonium-1,1—d2 perchlorate (0.151 mole), 11.3 ml. of 71.6% perchloric acid (0.136 mole) and 20 ml. of water. A solution of 22.6 g. of sodium nitrite (0.330 mole; Mallinckrodt AR grade) in 32 m1. of water was added dropwise with stirring. The reaction mixture was stirred in a 28—26° water bath for two hours. Distillation at 78-97° and saturation of the distillate with potassium fluoride gave 8.81 g. of yellow oil, 226—A. Further distillation gave 1.75 g. of organic layer 226—B. Ether extraction of the aqeuous phases of the distillates gave organic liquid (.35 9). Gases evolved during the deamination and distillations were passed through dry ice traps. There was 3.0 g. of yellow-green condensate. The solid residue from distillation was dissolved in water, made alkaline to phenolphthalein (with 3.2 m1. of 1M sodium hydroxide), and extracted with ether. Distillation, neutralization, and drying gave 0.2 g. of colorless amine perchlorate (0.8% recovehy). A preparative gas chromatographic column packed with 20% silicone SE-3O on Chromosorb W was used at 70°. An experiment with a 50-50 mixture of l-butanol and 2-butanol showed that the recovery of alcohols was only 50% at 8 p.s.i. but increased with decreasing pressure. The optimum was 3 p.s.i., where there was 79% recovery of 0.75 ml. of in- jected alcohols. 66 The deamination product 226-A was fractionated under these condi- tions. There were three injections (0.6, 0.7, and 0.7 ml.). Obtained were six fractions, including labeled 2-butanol at 18-26 min. and labeled l-butanol at 30—39 min. The analyses are summarized in Table XIIb. The n.m.r. spectra of the two alcohols are shown in Figures 5b and 6b. The total yields of deamination products are outlined in Table Vb. Preparation of Silyl Ethers Unlabeled trimethylsilyl ethers were prepared from l-butanol and 2—butanol by refluxing 0.30 g. of each alcohol with 0.50 ml. of hexa— methyldisilazane and a small drop of trimethylchlorosilane for two hours. The products were purified by preparative gas chromatography (20% SE- 30 at 133° and 8 p.s.i.). The ethers were collected at 37-88 min. and at 29—35 min. The impurities were 2-butanol at 17 min. and hexamethyl- disiloxane at 23 min. The collected samples had only traces of impur- ities by gas chromatographic analysis (10 ft. 20% Carbowax 20M in Aerograph). The labeled alcohols obtained from deamination were converted to trimethylsilyl ethers in the same way. All four ethers were analyzed mass spectrometrically by Mr. S. Meyerson. The results are in Tables VIb-VIIIb. Acetylation of Butylamine-1,1—d245 The 1.53 g. of butylamine-1,1-d2 (0.020 mole) mentioned above was dissolved in 8 ml. of pyridine and 20 m1. of toluene and treated with 6.6 ml. of acetyl chloride (0.093 mole). The semi-solid mixture was stirred magnetically for 30 min. in a 70—90° oil bath. It was then 67 Table XIIb. Products of deaminations of butylamines. 1. Gas chromatographic analyses, using a 100 in. x l in. 20% SE-3O column at 73° and 8 p.s.i. C d Time Butylamine Butylamine-1,1-d2 ompou“ 5 (min.) (Distillate 220—A) (Distillate 226—A) Gases 0.7 -- -- Butenes 2.1 2% 8% 2.8 2-Bu0H 7.6 38 38 l-BuONO 9.2 17 0.2 1-BuNH2 9.8 0 0 1-Bu0H 12 29 39 2—Bu0N02 27 3 8 15 2-BuN02 86 1 l-BuNOZ 68 1 0.8 2. Infrared analyses of products of deamination of butylamine-1,1-d2 fractionated by gas chromatography (wavelengths of strongest ab- sorption in microns) References (unlabeled compounds) 2-BuOH: 3.03, 3.50, 6 87, 7.33, 7.79, 8.87 81(no. 75) 1-BuOH: 3.1 3.5, 8.60, 8.80, 6.87, 7.27, 8.92, 9.35, 10. 2 81(no. 5278) 2-BuONO and l-BuONO: (not collected; seen only as trace impurities in the alcohols, 33. 6.18) 91 2—BuON02: 6.16, 7.85, 11.57 69 l—BuONOZ} 6.15, 6.86, 7.26, 7.85, 10.8, 11.8 87;81(no. 13895) 2-BuN02 6.52, 7.26, 8.80 69 1-BuN02: 8.65, 8.83, 6.12, 6.87, 6.83, 7.25, 8.59, 8.86 69; 89 68 washed with 20 m1. volumes of water, 5% sodium hydroxide, 5% hydrochloric acid, and water. Drying, filtration, and distillation gave colorless liquids: 0.30 9. 223-2 70° (0.8 mm.) n54 1.8867 0.16 g. 223—3 65° (0.8 mm.) 1.8886 0.26 9. 223—8 hold—up 1.8897 The literature gives comparable boiling points for butylacetamide:7° 230° (760 mm.) and 180—181° (21 mm.) but n55 1.8391. The only data for butyldiacetamide are nitrogen analysis (9.11%; calc. 8.91%)79 and boiling point (88-62° at 0.2 mm.)°° The infrared spectrum of 223-3 had a barely detectable N-H absorption and differed from the spectrum of butylacetamide.°l The gas chromatogram of 223-3 (20% SE-30 preparative column at 150°) showed no toluene, 0.9% at 10 min., and 99.1% at.27 min. Preparation of Bptyl Nitrites The directions38 for preparing butyl nitrite were modified for use with only 3.70 g. (0.050 mole) of each butanol. The yields obtained were 63% with l—butanol and 58% with 2—butanol. PART C ALIPHATIC 1,3-METHYL SHIFTS 69 INTRODUCTION Several carbonium ion rearrangements have been reported as pos- sible 1,3-methy1 shifts. The earliest was discussed in the 1933 doc- toral thesis of Laughlin and published after repeated verification.51 The reaction sequence began with the acid-catalyzed copolymerization of p-butyl alcohol and p—butyl alcohol: C C C C 1 ’ 1 2 H°~ C-C++IC=C >CCCCC—2————~—CCCCC (1) I l C C C C C c I II The major product was 3,8,8—trimethy1-2—pentene, formed by simple loss of a proton from I or II. However, another octene, III, was found. It was explained by the rearrangement of II, either by a 1,3-methy1 shift or by two successive 1,2-methyl shifts: + II .12_ME__—> C-C-C-C—C > C-C=C-C-C (2) <——— ' g y < 7 7 7 C C c C C C III C 1 + _ . + 11<112_ME_2L> C-C-C-C-C .11318212;> C C- C C- C > III. (3) g 1 < l I v <*_—' C C C C C The second system was also described by Whitmore and his students.52 0H 1 H+ + C-C-C—C > C-C-C-C + C-C=C-C + C=C—C-C (8) I I I I C C C ' C C C C 8 C C + c-C C C > C C C C E C C > C C C C c-C C (5) l 1 <*_' 1 1 <__’ 1 1 C C C C C C IV 70 71 C C C 1 C- C- C— C + C- C- C- C <-——> C-C-C-C-C-C <——+> C-C-C-C-C-C <-—-> C-C=C-C-C—C (6) 1 1 1 C C C C C C C C C C C V The only decenes found were IV and V. Careful fractional distillation separated them: h5% Yield of IV and 35% yield of V. The pure fractions were ozonized and identified as carbonyl derivatives. Whitmore's work52 has been criticized by Johnson,71 who treated t-amyl alcohol with sulfuric acid, hydrogenated the mixture of decenes, and found by analytical distillation that five decanes were present. Whitmore's failure to find more decenes was explained by resistance of some of them to ozonolysis, incomplete separation of decenes by distil- lation, and presence of smaller amounts of rearranged products because more dilute sulfuric acid was used. In 1950 Mosher and Cox53 reported a rearrangement which they claimed had to be explained as a 1,3-methyl shift. They said that successive 1,2-alel shifts would give a different product, IX, which was not found. OH c + C C + c-C-C-C-C< >CCCCC—>CCCCC (7) C-CC C-CC C-CC VI VII C + c 1 VI (ié— Mew > C-C-C-C—C <—> C-C=C-C—C (8) " C-CC C-CC VIII +C ' VI £1321) C-C---CCC im>cuccoc<_> C—C—C—C-C-C (9) ' "' " C-C C C-CCC CCC IX 72 The product obtained was a mixture of three parts of VII and four parts of VIII. Product analysis was by ozonolysis of the mixture and identi- fication of acetaldehyde, acetone, and ethyl isopropyl ketone as their derivatives. Ethyl t—butyl ketone was not mentioned. It would be worthwhile to repeat the rearrangement of VI and study the products with newer instrumental methods, especially gas chromatography. However, it is not necessary to presume that VI -——> VIII proves the existence of a 1,3-methyl shift. Other reasonable mechanisms can be written for VI -——> VIII. For example, c _ . + V _ . + _ . + VI (M) C—CE-C—C':_C M) C_(':_(':_(':_C <¥fl> C—(E—(E—(E—C c c c c c c c c 6: ('3 6: C + fflifl» c-E-{z-c-c (M c-c-é-c-c <£> c-c=c-c-c (10) c c c c c c c c c': ('3 ('2 VIII The literature is Sprinkled with rearrangements which are 1,2; 1,2— methyl shifts rather than 1,3-shifts. One example54 is that of 12a- methyl-9B,llB-oxidoprogesterone, X: Ac (ll) 73 Another case is the final steps in the biosynthesis of lanosterol, XV, from squalene, XIII:55 XIII XIV XV Elaborate work with 13C-labeled precursors led to the conclusion that the mechanism of XIV ——a> XV was 1,2; 1,2-methyl shifts, not 1,3. ~There seem to be no alkyl or aryl transannular shifts.56 For ex- ample, the search for a l,S-methyl shift in the acetoLysis of 5,5-di- methylcyclononanol tosylate was fruitless; infrared analysis.showed that there was less than five per cent (if any) 1,54methyl shift (equation 13). 57 w- ...as d- 5&6 Meinwa1d7° conclusively proved that a 1,5-methyl migration does (114) not occur in the acid-catalyzed rearrangement of cinenic acid. Letsinger49 found 1,5-phenyl and l,S—hydride but no l,S-methyl shift in Eeri-sub- stituted naphthalenes. 7b A model which is simpler than those mentioned above and which favors the possibility of a l,3—methyl shift in an acyclic system is the neOpentylcarbinyl cation: C OH 1 + ? + H20 I C-C-C-C > C-C-C-C-C > C-C-C-C—C (15) v v 1 C C C XVII XVIII XIX It is to be expected that l,2-hydride and 1,2-methyl shifts also occur: c I + + XVII > C—C-C-C ———+> c-c-c-c (16) < ' <—— ' ' c c c XX XXI Thus the products should include alcohols and olefins formed from ions XVIII, XX, and XXI. RESULTS AND DISCUSSION The deamination of neopentylcarbinylamine gave several products. Their analysis is outlined in Tables Ic and IIC. Identification of products was by means of gas chromatography. Samples of known alcohols were run at the same conditions as neat samples and as additives to the mixtures of reaction products. Generally the retention times varied slightly, but the peaks were easily recognized by their sizes, shapes, and order of appearance. In the first deamination, using conditions based on those of Roberts 33 al.,40 the major product was the unrearranged alcohol, neo- pentylcarbinol. Because the yield was low, the reaction conditions were changed. In the second deamination the acid was added to the aqueous mixture of ammonium salt and sodium nitrite.73 In this case the major product was t-butyl alcohol and the over-all yield was still low. Two deaminations were performed with a,a—dideuterated amine, using the original reaction conditions. After three hours of reaction the main product was neopentylcarbinol. Extended contact with aqueous acid increased the amount of t-butyl alcohol and dimethylisopropylcarbinol. Several attempts to identify the minor products were unsuccessful, because the quantities were too small for collection of fractions by gas chromatography. Some of the compounds shown to be absent are listed in Table IIc . Complete gas chromatographic separation of the two tertiary alcohols, dimethylisopropylcarbinol and dimethylpropylcarbinol was difficult to 75 76 Table Ic. Deamination products of neopentylcarbinylamine and neopentyl— carbinylamine-1,1-d2 (gas chromatographic analyses) Compound Time Firsg Second Third Rund Fourt (min.) Run RunC (% at (% at Run (%) (%) 3 hr.) 2hhr.) Gases 1.2 trace 1.6 -- 1.6 8 (Isobutylene)e l.h trace 0 O O E-Butylethylene 1.6 trace 0 2 0.2 7 (2,3—Dimethyl—l-butene) 3.0 o 1.5 0.3 o 2 gfgiifiit2¥iiitzlene :} u.5 1.6 0.13 1.2 0.3 6 7.5 2.3 l.b h.6 3.2 5 t—Butyl alcohol , ll O.h 56 21 51 '9 lb 0.3 6.h 0.8 13 h (t-Butyl nitrate) 17 2.2 1.5 O.h O 2 (2-Nitro—2omethylpropane) 18 1.6 1.1 b.6 1.7 2 22 0.2 O O 0.1 l Dimethylisopropylcarbinol 39 8.3 16 lb 23 2 Dimethylpropylcarbinol bl R (CH3)3CCH2CHZOH + (CH3)3CCH=CH2 V //////}” + + (CH3)3CCH2CHZ —> (CH3)3C + CH2=CH2 l (CH3) 3COH + (CH3) 2C=CH2 l,2-H:N :' OH 7 (CH3)3CCHCH3 > (CH3)3CCHCH3 + (CH3)3CCH=CH2 l,2-Me:~ OH + ! (CH3)2CCH(CH3)2 ————-fi> (CH3)2CCH(CH3)2 + (CH3)2C=C(CH3)2 + CH3 The purpose of the deuterium labeling was to determine whether there was a t-butyl shift. Before these experiments were completed, however, it was found that Saunders96 had done a similar study. He diazotized neopentylcarbinylamine-l-14C and separated the products by distillation. The neopentylcarbinol-14C had less than one percent rearrangement of its radiocarbon label. Only two products were found-- a 20% yield of neopentylcarbinol and a 16% yield of dimethylisopropyl- carbinol. EXPERIMENTAL Synthesis of Neopentyl Cyanide Dimethylsulfoxide (m.p. 18°; Matheson) was dried over calcium hydride overnight and vacuum distilled from a 60° water bath.5° Sodium cyanide (Baker reagent) was dried55 for 2h hours in a 110° oven. Neo— pentyl chloride (K and K Laboratories) gave a single peak on a silicone SE-30 gas chromatographic column, but this is not proof of purity; t—amyl chloride had the same retention time on this column. The neo— pentyl chloride was therefore tested by shaking it with 0.10 g silver nitrate. There was no visible reaction. (t-Amyl chloride reacts instantaneously.) Forty grams of dried sodium cyanide was stirred with 230 ml. of freshly distilled dimethylsulfoxide. The suspension was heated to 90° on a steam bath and stirred while 53.30 g. of neopentyl chloride was added without heating. There was no evidence of reaction. 0n heating with a mantle the reaction mixture (yellow liquid and white suspended solid) quickly darkened to yellow-brown after it reached reflux tempera- ture (85°). Stirring and heating (100 to 130°) were continued for 2h hours. When it cooled to room temperature it was a brown slush, too thick to pour, so 350 ml. of water was added to it. After separation of the organic layer the water layer was extracted with 100 ml. por- tions of ethyl ether. The organic layer was combined with the ether extracts, washed three times with 50 m1. portions of saturated aqueous sodium chloride, dried over calcium chloride, and distilled through a 6—inch vigreux column. After collection of fractions containing ether and neopentyl chloride, the water was turned off and the condenser 80 81 jacket drained, because the fraction at l3b-135° solidified as it cooled to room temperature. (Neopentyl cyanide boils61 at l37° and melts at 28-300.) Obtained was 18.79 g. The lower boiling fractions were re- distilled slowly and another 2.82 g. was collected at 133—135°. Thus the total yield of neopentyl cyanide was 21.61 g. (bb.b% yield) of colorless solid. Some unreacted neopentyl chloride (15%) was recovered. Reduction of Neopengyl Cyanide A colorless solution of 8.75 9. (0.0900 mole) of neopentyl cyanide in 25 m1. of dry ether was added in 25 min. to a stirred and ice-cooled mixture of 3.b g.(0.090 mole) of powdered lithium aluminum hydride in 250 ml. of dry ether. (The ether had just been distilled from lithium aluminum hydride.) There was no sign of reaction occurring. The mix- ture was stirred and refluxed (steam bath) for four hours. After cool- ing it with ice, it was successively treated withbml. of wateradded. drop- wise, 3 ml. of 20% aqueous sodium hydroxide, and lb ml. of water. The clear ether layer was decanted, and the white solid residue was extracted three times with 30 m1. volumes of ether. The ether solutions were dried over calcium sulfate, filtered, and distilled through a 5-inch packed column. After removal of most of the ether at 35°, there was 25 ml. of residue, which was distilled in a smaller apparatus. 0b- tained at 111-1120 (7&5 mm.) was only 3.96 g. of neopentylcarbinylamine with ma: l.bl30. (Theoretical yield: 9.11 9.; literature92 values D 112.8-.9° (7b5) and n25 D the lower-boiling fractions by gradual addition of 35% perchloric acid l.b122.) More of the amine was recovered from until the liquid was acidic to pHydrion indicator paper. The white, deliquescent crystals were first dried in an evacuated desiccator. When 82 heated on a Fisher hot stage, they lost liquid around 160°, and the re- maining dry, white crystals melted at 280-28b°. Recrystallization from 30 ml. of B-heptanol, using 750 ml. of 30-600 pet. ether, gave 8.50 g. with m.p. 290-29b°. The yield of the purified ammonium salt from the nitrile was b6.9%. First Deamination of Neopentylcarbinylamine (The proportions of reactants were the same as those.used by Roberts.4°) A mixture of 8.07 g. of neopentylcarbinylammonium perchlorate (0.0b0 mole) and b.80 g. of 71.6% perchloric acid (0.03b3 mole) was diluted with water to 70 m1. To this mixture was added 5.93 g. of sodium nitrite in 10 m1. of water, with magnetic stirring, during 20 min. The reaction flask was in a l 1. 25° water bath during the reaction. Only a faint blue color was seen in the reaction mixture. During the two hours of stirring after addition an oil formed above the aqueous layer. The mixture was saturated with sodium chloride, 20 ml. of ether was added, and the layers were separated. The aqueous layer was extracted three times with 20 ml. volumes of ether. The combined ether solutions were dried over calcium sulfate. As the ether dried, 1.95 g. of unreacted ammonium salt precipitated. 0n distillation the filtered ether solution gave 0.98 g. of colorless liquid with odor of camphor and 1.03 9. yellow solid residue. The distillate was analyzed by gas chromatography. (See Table Ic.) The deamination yield was only about 25% with recovery of 37% of the unreacted ammonium salt. 83 Second Deamination of Neopentylcarbinylamine Zollinger72 quotes the statement of Austin73 that the correct method of deamination is slow addition of mineral acid to a solution of the alkylammonium salt and sodium nitrite. This method was tried with about 1.6 g. of impure neopentylcarbinylammonium perchlorate (m.p. 2b5- 255°). This salt (0.008 mole) and 1.66 g. of sodium nitrite (0.02b mole) were dissolved in water, forming 33 ml. of solution. A capillary dropper was inserted through a cork in the sidearm of a 50 ml. reac- tion flask holding the solution. Dilute perchloric acid (0.96 g. of the 71.6% acid diluted to 5.6 m1.; 0.0069 mole) was added through the capillary, which extended below the surface of the stirred solution, in five minutes. The mixture was stirred for two hours at 25°. It was then satur- ated with sodium chloride. The yellow.oi1 layer was removed with a dropper and dried over calcium sulfate. It weighed 0.2 9. Analysis of this product was by gas chromatography. (See Table Ic.) Preparation of Negpentylcarbinylammonium-l,1—d2 Perchlorate An ether solution of 9.30 g. of neopentyl cyanide (0.0958 mole) was reduced with b.02 g. of lithium aluminum deuteride (0.0958 mole). The deuterated amine was converted to its perchlorate salt (m.p. 281- 28bO (dec.)), which was not recrystallized. The yield from nitrile to ammonium salt was 78.8%. Third Deamination (Using_Deuterated Amine) One twentieth of a mole (10.18 g.) of neopentylcarbinylammonium-l,l-dz perchlorate was stirred with 6.00 g. of 71.6% perchloric acid (0.0b29 8b mole) diluted to 200 m1. Almost all of the salt dissolved. A solution of 7.bl g. of sodium nitrite (0.1076 mole) in 10 m1. of water was added in twenty minutes. There was little evidence of reaction. After three hours a sample of the small oil layer was removed and analyzed by gas chromatography. (See Table Ic.) One hour later another 6.00 g. of 71.6% perchloric acid and then 7.bl g. of sodium nitrite in 10 ml. of water were added dropwise. Five hours later a third addition of 6.00 g. of acid was made. The mixture was stirred for thirteen hours and then treated with 7.bl g. of sodium nitrite. Indicator paper showed that the pH changed from 3 to 5 on addition of the nitrite. The reaction mixture was saturated with sodium chloride 2b hours after the reaction was started. The oil layer was separated and dried over cal- cium sulfate. There was 1.72 g. (See Table Ic.) Gases evolved during the reaction were passed through two dry ice traps. The second was empty. The first held blue liquid (largely nitrogen sesquioxide), which was washed by passing it through a 10% solution of sodium hydroxide. Obtained in a dry ice trap was about fifty microliters of yellow liquid. Gas chromatographic analysis showed that it had the same composition as the oil layer. Fourth Deamination (Usinngeuterated Amine) Five grams (0.02b6 mole) of neopentylcarbinylammonium-l,l-dz per- chlorate was deaminated under conditions similar to those of Roberts,4° avoiding the excess acid which rearranged the product in the third run. The 5.00 g. of salt and 2.9b g. of 71.6% perchloric acid was diluted to a volume of b0 ml. with water, the minimum amount which would permit the magnetic stirrer to turn in the sediment of undissolved salt. A 85 solution of 3.6b g. of sodium nitrite (0.053 mole) in 5 ml. of water was added in 20 min. The mixture became pale blue and evolved color- less and brown gases. The gases were passed through a vigreux column, a condenser, a gas-washing bottle containing 200 ml. of 10% sodium hydrox- ide solution, an empty bottle, and two dry ice traps. After two hours of stirring at 25°, the reaction mixture was saturated with sodium chloride. The 1.25 g. of yellow oil layer was removed with a capillary dropper and dried over calcium sulfate, giving 0.83 g. of oil. (See Table Ic for its analysis.) The dry ice traps were completely empty. The alkali washing solu— tion was saturated with sodium chloride and extracted three times with ether. Distillation of the ether extracts gave 0.30 ml. of liquid which was about half methyl-t-butylcarbinol and half dimethylisopropyl- carbinol. Ether extraction of the aqueous part of the reaction mixture gave 0.b0 g. of yellow slush which was washed with benzene, leaving 0.20 g. of colorless crystals (the unreacted ammonium perchlorate). The benzene solution contained approximately equal amounts of methyl-t-butylcarbinol and dimethylisopropylcarbinol. Preparation of Known Compounds for Gas Chromatography Pyrolysis of 2.2 g. of bi§(l,l—dimethylbutyl) oxalate in an oil bath at lb0-150° gave 0.15 g. of olefin in a receiver cooled in dry ice—trichloroethylene. There were two products: b0% 2-methyl-2-pentene at 2.8 min. and 60% 2—methylwl-pentene at 3.1 min. on the gas chromato~ graphic column used for product analysis. 86 Five millimoles of t-butyl alcohol was converted to its nitrite ester.38 A mixture of 0.07 g. of neopentylcarbinol, 0.12 g. of methyl-t- butylcarbinol, and 0.18 g. of dimethyliSOpropylcarbinol was treated by the same conditions.38 The products were tfbutyl nitrite and t-butyl alcohol from the primahy alcohol, a compound with retention time lb min. from the secondary alcohol, and apparently dimethyliSOpropylcarbinyl nitrite with retention time 33 min. 10. 11. 12. 13. lb. 15. l6. 17. REFERENCES Roberts, J. D., R. E. McMahon, and J. S. Hine, J. Am. Chem. Soc., lg, 1237 (1950)- (a) Karabatsos, G. J., F. M. Vane, and S. Meyerson, ibid., 83, b297 (1961). (b) Vane, F. M., Thesis, Michigan State Uni- versity (1963). (c) Karabatsos, G. J. and F. M. Vane, J. Am. Chem. Soc., 85, 729 (1963). (d) Karabatsos, G. J., F. M. vane, and s7 Meyerson, ibid., §§, 733 (1963). Kharasch, M. S. and 0. Reinmuth, Grignard Reactions of Nonmetallic Substances, 713, Prentice-Hall, New York, 195H. f Kennedy, J. P. and R. M. Thomas, J. Polymer Sci., b2, 189 (1961). Jensen, F. R., G. Marino, and H. C. Brown, J. Am. Chem. Soc., 81, 3303 (1959). Maryott, A. A. and E. R. Smith, Table of Dielectric Constants of Pure Liquids, Natl. Bur. Stds. Circ. 5lbg(l951). Heston, W. M., Jr., E. J. Hennelly, and C. P. Smyth, J. Am. Chem. Soc., 12, 1071 (1950). Sisido, K. and Y. Yosikawa, ibid., lg, 1608 (19b8). Dreisbach, R. R. and S. A. Shrader, Ind. Eng. Chem., b1, 2877 (l9b9). West, P. W., B. Sen, B. R. Sant, K. L. Mallik, and J. G. Sen Gupta, J. Chromatog., 6, 220 (1961). Lattey, R. T. and o. Gatty, Phil. Mag. [7], 1, 1929: Beilstein, y, 173 (1913). Musset, E., A. Nikuradse, and R. Ulbrich, Z. angew. Phys., 8, 8 (1956). Blau, M., W. T. Carnall, and J. E. Willard, J. Am. Chem. Soc., 1b, 5762 (1952)- Prins, H. J., Rec. trav. Chim., 56, 779 (1937). Farbenind., I. G., German Patent 715069 (1938); Beilstein, 1, AS (1958)- Morgan, S. O. and H. H. Lowry, J. Phys. Chem., 3b, 2b13 (1930). Vogel, A. I., Practical Organic Chemistqy, 176, Wiley, New York, 1956. 87 18. 19. 20. 21. 22. 23. 2b. 25. 26. 27. 28. 29. 30. 31. 32. 33. 3b. 35. 36. 37. 88 Harmon, K. M., F. E. Cummings, D. A. Davis, and D. J. Diestler, J. Am. Chem. Soc., 8b, 33b9 (1962). (a) Wiberg, K. 8., Laboratory Techniques in Organic Chemistgy, MCGraw-Hill, New York, 1960. (b7 Fieser, L. F}, Experiments in Organic Chemistgy, 288, Heath, Boston, 1955. Thomas, C. A., Anhydrous Aluminum Chloride in Organic Chemistry, 2b, Reinhold, New York, l9bl. Mouneyrat, M. A., Comptes rendus, 126, 1805 (1898). Nystrom, R. F., W. H. Yanko, and W. G. Brown, J. Am. Chem. Soc., Tolbert, B. M., ibid., 532, 1529 (19b7). Amstutz, E. D. and R. R. Myers, Organic Syntheses, Coll. Vol. II, b62, Wiley, New York, l9b3. Arndt, F, ibid., 165. Waterman, H. I., J. J. Leendertse, P. J. G. Colthoff, Chem. Weekblad, 32;, 550 (1935); c. A., 32, 111.7 (1936). Prins, H. J., Rec. trav. Chim., 56, 779 (1937). Wallace, C. H. and J. E. Willard, J. Am. Chem. Soc., 12, 5275 (1950). Sommer, J.—M., Bull. Soc. Chim., 126b, 516. Schmerling, L., Ind. Eng. Chem,, b9, 2072 (l9b8). Huntress, E. H., Organic Chlorine Compounds, Wiley, New York, 19b8. Boeseken, J., Rec. trav. Chim., 22, 306 (1903). Karabatsos, G. J. and C. E. Orzech, J. Am. Chem. Soc., 8b, 2838 (1962). ‘— Cope, A. C., S. W. Fenton, and C. F. Spencer, J. Am. Chem. Soc., 14. Saab (1952). Prelog, v. and K. Schenker, Helv. Chim. Acta, g, 2qu (1952). Prelog, V. and J. G. Traynham, "Transannular Hydride Shifts," Chapter 9 of Molecular Rearrangements, P. deMayo, ed., Inter- science, New York, 1963. Amundsen, L. H. and L. S. Nelson, J. Am. Chem. Soc., 73, 2b2 (1951). 38. 39. b0. bl. b2. b3. bb. b5. A6. A7. b8. b9. 50. 51. 52. 53. 5b. 55. 89 Noyes, W. A., Organic Syntheses, Coll. Vol. II, 108, Wiley, New York, 19b3. (a) Pexsters, M., Chem. Zentn, I., 1398 (1907). (b) Rodd, E. H., Chemistgy of Organic Compounds, Vol. IA, 336, Elsevier, Amsterdam, 1952. Roberts, J. D., and M. Halmann, J. Am. Chem. Soc., 15, 5759 (1953). Kaiser, R., Gas Phase Chromatography, Vol. II, 18, Butterworths, Washington, 1963. (a) Taft, E.,of Wilkens Instrument and Research, Inc., Private Communication, l96b. (b) Wilkens Instrument and Research, Inc., "Aerograph Research Notes," 1 (Spring l96b). Albert, Kg,of American Oil Company, Private Communication, l96b. Streitwieser, A., Jr., and D. van Sickle, unpublished results cited by A. Streitwieser, Jr., J. Org. Chem., 82, 861 (1957). Smith, M. E. and H. Adkins, J. Am. Chem. Soc., 88, 661 (1938). Smith, D. M. and w. M. D. Bryant, ibid., 5_7_, 61 (1935). Cope, A. c. and D. M. Gale, ibid., g, 3713 (1963). Langer, S. H., S. Connell, and I. Wender, J. Org. Chem., 85, SO (1958). Letsinger, R. L. and P. T. Lansbury, J. Am. Chem. Soc., 81, 935 (1959). Letsinger, R. L., W. J. Vullo, and A. S. Hussey, ibid., 81, 1965 (1959). '” " Whitmore, F. C., K. C. Laughlin, J. F. Matuszeski, and J. C. Surmatis, J. Am. Chem. Soc., 85, 756 (19b1). Whitmore, F. C. and W. A. Mosher, ibid., 85, 1120 (l9bl) and 88, 281 (l9b6). Mosher, W. A. and J. C. Cox, Jr., ibid., 12, 3701 (1950). Fried, J., work cited by N. L. Wendler in Molecular Rearrangements, ed. P. de Mayo, Part 2, 1025, Interscience, New York, 196b. Cornforth, J. W., R. H. Cornforth, A. Pelter, M. G. Horning, and G. Popjak, Tetrahedron, 5, 311 (1959) and R. K. Mandgal, T. T. Tchen, and K. Bloch, J. Am. Chem. Soc., 88, 2589 (1958). 56. 57. 58. 59. 60. 61. 62. 63. 6b. 65. ,66. 67. 68. 69. 70. 71. 72. 73. 7b. 75. 76. 90 Prelog, V. and J. G. Traynham in Molecular Rearrangements, ed. P. de Mayo, Part 1, 609, Interscience, New York, 1963. Blomquist, A. T. and Y. C. Meinwald, J. Am. Chem. Soc., 80, 630 (1958). Smiley, R. A. and C. Arnold, J. Org. Chem., 85, 257 (1960). Nagata, W., T. Sugasawa, Y. Hayase, and K. Sasakura, Proc. Chem. Soc., l96b, 2bl. Becker, H.—D., G. J. Mikol, and G. A. Russell, J. Am. Chem. Soc., §§, 3110 (1963). Birch, A. J., J. Chem. Soc., 19b9, 2721. Newman, M. S., T. Fukunaga, and T. Miwa, J. Am. Chem. Soc., 88, 873, (1960). Cohen, T., R. M. Moran, and G. Sowinski, J. Org. Chem., 88, l (1961). Cohen, T. and J. Lipowitz, J. Am. Chem. Soc., 89, 251b (196b). Sommer, L. H., H. D. Blankman, and P. C. Miller, J. Am. Chem. Soc., 15, S759 and 19, 803 (195A). Shriner, R. L., R. C. Fuson, and D. Y. Curtin, The Systematic Identification of Organic Compounds, Wiley, New York, 1956. Adamson, D. W. and J. Kenner, J. Chem. Soc., 193b, 838. Kornblum, N., R. A. Smiley, R. K. Blackwood, and D. C. Iffland, J. Am. Chem. Soc., 77, 6269 (1955). Kornblum, N., H. E. Ungnade, and R. A. Smiley, J. Org. Chem., 81 377 (1956). ) Meinwald,.L,.leLChem. Soc., 21, 1617 (1955). Johnson, G. C., ibid., 82, 1A6 (19b7). Zollinger, H., A20 and Diazo Chemistry, 93, Interscience, New York, 1961. Austin, A. T., Nature, 188, 1086 (1960). Karabatsos, G. J., C. E. Orzech, Jr., and S. Meyerson, J. Am. Chem. Soc. (in press). Streitwieser, A., Jr., J. Org. Chem., fig, 861 (1957). Ogata, Y. and H. Tabuchi, Tetrahedron, 88, 1661 (196b). 77. 78. 79- 80. 81. 82. 83. 8b. 85. 86. 87. 88. 89. 9o. 91. 92. 93. 9b. 95. 96. 97. 91 Doss, R. C., U. S. Patent 3031838 (1962); Chem. Abstr., 51, 7508 (1962). wadso, 1. Acta Chem. Scand., 11, 635 (1958). 9 Kornhauser, A. and D. Keglevié, Tetrahedron, 18, 7 (1962). Keglevi6, D., Private Communication (l96b). The Sadtler Standard Spectra, Midget Edition, Sadtler Research Laboratories, Philadelphia, 1962. Linnemann, E. and V. v. Zotta, Ann., 188, 3 (1872). Ray, P. C. and J. N. Rakshit, J. Chem. Soc., 191, 1b1 (1912). Meyer, V., J. Barbieri, and F. Forster, Ber., 18, 132 (1877). Whitmore, F. C. and D. P. Langlois, J. Am. Chem” Soc., 58, 3bb1 (1932). Haszeldine, R. N. and J. Jander, J. Chem. Soc., 1258, 691. Carrington, R. A. G., Spectrochim. Acta, 18, 1279 (1960). Brown, J. F., Jr., J. Am. Chem. Soc., 11, 63b1 (1955). Buczkowski, Z. and T. Urbanski, Spectrochim. Acta, 18, 1187 (1962). Washburn, w. H., Anal. Chem., gg, 155 (1958). Tarte, P., Bull. soc. chim. belges, ég, 2bo (1951). Howe, A. P. and H. B. Hass, Ind. Eng. Chem., 58, 251 (19b6). Sharkey, A. G., L. A. Friedel, and S. H. Langer, Anal. Chem., 29, 770 (1957). Streitwieser, A., Jr. and W. D. Schaeffer, J. Am. Chem. Soc., 79, 288 (1957). Smith, M. B. and w. H. Watson, Jr., ibid., 81, 317A (1962). Saunders, W. H., Jr., ibid., 18, 6127 (1956). Olah, G. A., Ed., Friedel-Crafts and Related Reactions Chapters b,5, and 25, Interscience, New York, 1963 and l96b.