5‘1. {1| IE!“ I 7‘1‘1'115‘345‘ “111'“2351 ‘ '1" 1 II .. ‘{,1-1 1111,11: .‘111'1I I '1'“ 15' "r '1. M” 111'.“ 1115.5‘11: , v 1 . ' ”I '4”. 1.21.“? "‘61:" 1'“%";' 3311' 311-3111.": 13”" 153543181'5111‘53'3‘ ' "' '3' “gas, “M ,3 1 31133.1 1.: .11- 111,11 “ I555”- 51515151 11113:,"15155 I'fi'k'hfft"ax,:" 1'. #515 mfifira L . :1: v ' I'll", 'Eil ‘ ‘ (“H-15451511511331; .3311? 11111311111 113.111, 1’1." , .5. ‘. I 1 1' '1". "‘ "'71" '1-1 $3.01. “"35: 31$“g51 .1131"1 11131" ‘31. ‘1.13:i-E"“*"(5 “I." - ..":1':.&. ‘1'1":" 1» .II‘E KM '11-’15'111111Iilu2‘IL!“ “ 1&1 1 115' 1511-5551 ‘,_45 9111515; 1.1. 1 XI11115151555 I “‘m‘L'“;u1 ' 1"" ‘ ‘1 ‘31.?31111‘1-111 117 '1. ,1 11.1 I ‘1' M “:2th 51551.15 1.151. 11.55 ‘51, 71 7. ‘12n""""'""' I ‘5 5'15. 5 1. t' . 1131 5x50155691} £1,555,155 " 11:3111'III1111 51175531115 ‘21:“. 55515111: .I A iigivisz-ti‘“ . ‘cifi‘ -. 1 .‘fi'wc'a 5 ‘5, 1. 5 5X1 ‘y‘fégtfiéfira‘n w, l I '\‘k ,\ . 1 5.51, I II . . .E' I‘ I'I.I‘5515,EL" 44"]; fig.) \figfié gge‘sz‘fifi- . , . _ 1 115'. I 11" .' x‘ I". . % iég“ “E 5 ER“;&1§7‘53‘33“3“ 1 " I ‘1 ' 1.3111; 5,5 ... z “" «W1 53%;; uékificnfisfi 5.5:. 1. . ;. 1 1 ‘ '1, _ 'V..L\'.‘ .1. _‘_, L 5 .1111. “1155:1155 “151: “f3 "5‘1 awffi" 5 '1.” 33:3" """3:?m$£i{g- I M51 '13.. 3.1.13.5.55, 313? - é-Q‘fifrffi ’ ‘ L 5.55':""'“"'"'""' “31$": 5.5555: .4511 1 eri‘f-I ff E5“- .rur :n w... {5; fig; -. '3 :3? » 3-3:" 57;“ vi 'Lm‘fi '?~"211 R," 513.11% 31: z 3; 0-1:: . 1.: ‘1" ~~. 1:11:53: 1'], h' 51.1 1 115511 ~ 1:11 15.31:. w37‘1;1.d7.l515.;11;15 " 9‘1 . 1 7 1 ._ “*9 '713::‘2'3‘1‘113311133" 1 11.13:! “i" ‘331. 1.: 1 1 ' "1171' ‘E'RE'i'Zjflkfiié'fi 11$ 1.113.:5I‘1'LI‘. 51.14 £25.?“ ' - ‘E I '1 3.5191155 Ith— +5I5‘l';‘4~15;55..g$1{b.311 15:16:: 1, '1. . 311.511.557.51“ 171521.15 1"" 1.12123 £1; 3-‘1 ‘ 'uIEI :i‘E‘Z'EE‘ . 1' $53.; 111;: {A 4" LCfi-N 1331‘ $24”: 3 — \V '8". .1, . I '- if .‘1 ' tnld'k’k‘1'k‘ku'é1'kll "kin.“ " “WE: 43,1532; .52..“ff’; 1‘41, $""17"k.1‘? a... _ #331 .- ,4." “ ' . ' “1% ‘33: .1’73'13’11 g1??? 153115 "9%: "311‘ $155 ~l‘1'2'155é'é'fi'a" 1.5 """W"1‘:;7'95_.1¢5I11I5I Jan 3“ i 1L1" I}: 1.." 13335: “$511. “Bin-'1 .0 1 A591. 152%.? 13. 1 1.1 $3"? L17:1 5.5 .5. 5.155555 55555.55 " " L'E'E'“ "515‘?" """"'{“i "1'6" ,4th 1155. “'73,!“ M \§ “L'm ’flufl-w‘r LL‘!‘ 1'] ' "I" ' .1 1.9113311311111555 1:1. ‘1. “3‘ @713": 3391:. {ya $a‘1k13 'i$31 ‘1; I '1 .1 >' 11.111‘15M “5'15'3'133 1 555 I. .1 III 0555 1 I535. M“ ”II . 1“? $1311.51?“ 11’ 5 “a ' 55591515: 1» “3.31.1.2 ”1‘55: ”:14; a. ..‘,"5 '. 1‘,' , "S‘E'fiw~ 5555 ‘55} 'L'L WIRE 515535 {231% I 1| ' 5L 5" I “3" ‘ M I" "Fifi-11L 5'1 7.11! GI: ":56 733% 1:“ 1 ' "'1' .I' .1 S‘IIISIE‘IIIII . 5.1151151 75 5 535%“ :1111' 1 , 1175I 1- _. 5, . ,I . 1_ 1111.155 II 1115 5355 [5; 1i; . 5I -‘ 1 . 1 5' _ %$WH11$5~ 11‘ 1 1'1‘ 1555- 51155555315 . E‘ 1 1 J 155151555555”: £35154 “‘4 1 In {11' "1h. I'EE' ' .{EIE'QE‘ '11" WW 1'5 " ‘7‘1'1 HEN 3-.'11"‘.1 “‘. 2395:: "'E' 1' :“11'131! . 1 5 L‘"E"" 151E111" ($153111: as" 555111111W1‘111‘! 111 511. 115115 11.1.1‘ {11 51111.5 1" “1115:1211'1111'1'91“ $.11.:13:§1.:~1"‘1' “131““? “fl!“ «'11 1 "'I'S'E £23141}. 3'. ‘15! '."L1 M11”: 311' ’L’ ’ 1: 'E>i f3. . ’ 1 27‘" ~—" ' ~-L.. m Augmmv‘. . _ 2,: . 2—1...- 47.. I6"! '31" 55:5 ' 5".‘..1: 111(11'91 .113' . "1-7.1 . . 55' 1 5:5 111' 15I1I'EEA‘EE -. . _ IILESIE'IIg'I 15.115“. 5 11115511 55511555 1.I11-111151I515I515.I5III155III. ‘ ' 11' ' ‘. l ‘. . t1. ‘ M“ ".E " fl ' % 21‘1“.» 1' EJEW$EW$%" I éfl. WM 'i‘ 11:13:11: 5555'EE1I5I55 ‘11'11111.515 iii." I111 111111715715 11:1555551 I :F“ 'E: I“ h%'$1 'Y ' 5 5' I ': ;_ n—i-k...‘ «.—.—- . an“ ’ ESR SITY LIBRAR llllllllHlllllllllllllllll ill l l H 3 1293 015914 lfllllll I Hi: THESIS 5‘.~i'-_ “WW" ’ J LI first This is to certify that the thesis entitled The Synthesis and Reactions of Sterically Hindered Secondary Amines and Their Corresponding Lithium Amides presented by Ihor E. Kopka has been accepted towards fulfillment of the requirements for Ph 0 D . degree in ChemiStry Major professor Date Sept. 17, 1981 0-7639 MSU LIBRARIES m RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. THE SYNTHESIS AND REACTIONS OF A SERIES OF STERICALLY HINDERED SECONDARY AMINES AND THEIR CORRESPONDING LITHIUM AMIDES By Ihor Elias Kopka A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1981 ABSTRACT THE SYNTHESIS AND REACTIONS OF A SERIES OF STERICALLY HINDERED SECONDARY AMINES AND THEIR CORRESPONDING LITHIUM AMIDES By Ihor Elias KOpka A series of highly branched secondary amines, including bis- (l,l~diethyl-2—pr0pyl)amine was prepared by coupling propargylamines with propargyl chlorides. Hydrogenation of the resultant dipropargyl- amines was accomplished with W2 Raney nickel in the presence of potas- sium hydroxide in ethanol. The resultant amines are among the most hin- dered secondary amines reported to date. The pKa values of the conjugate acids of the series of secondary amines exhibit a regular decrease with increasing size of the alkyl groups. The more hindered members of the series are inert to methyl iodide. Bis(l,l-diethyl—Z-propyl)amine reacts with boron trifluoride etherate to give a primary amine adduct and 3- ethyl—Z—pentene. N-Metallation of the amines with phenylsodium, n—butyl or phenyllithium was unsuccessful. Successful N—metallation was achieved with n—butyllithium in the presence of N',N',N,N-tetramethylethylenedi- amine. The rate of metallation was determined in the presence of 50 and 100 mole percent of tetramethylethylenediamine. The stability of the series of lithium amides was determined in tetrahydrofuran and diethyl ether solution at 240C. The stability of the amides first decreased, then increased with increasing amide bulk in both solvents. The series of lithium amides was reacted with 2-bromobutane, 2—iodobutane and 2—bromo—6-heptene in tetrahydrofuran to give a mixture Ihor Elias K0pka of olefins. There is an increase, then a decrease, in the l-alkene/- 2-alkene ratio with increasing amide bulk. Concomitantly, the trans/ .Eig-Z-alkene ratio decreased, then increased, with increasing amide bulk. Threo and erythro-defZ-bromobutane were synthesized and reacted with the series of lithium amides. Product butene analysis indicated that all the amide dehydrohalogenation reactions occur by an antifelimination mechanism. To Molly, Puff and Rudy ii ACKNOWLEDGMENTS I would like to thank Mike Rathke for his patience in dealing with my sometimes headstrong ways as well as for giving me the freedom to pursue interesting avenues in organic chemistry. I would like to also thank him for his friendship and critical insight throughout the course of my graduate career. I would like to thank Keki, Manfred and Andy for their expert technical help in constructing many glass articles which were indiSpen- sable to my research efforts. I want to thank Zak Fataftah for his friendship and assistance for a portion of this research project. I would also like to thank past and and present members of Rathke's Rascals who more often than not provided an entertaining atmosphere in the lab. Appreciation is extended to the Dow Chemical Company, the National Science Foundation and Michigan State University for financial support during my graduate studies. Finally, I want to thank my mother, father and sister for their unconditional love during good times and bad times. iii TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . . . . ix LIST OF FIGURES . . . . . . . . . . . . xi LIST OF ABBREVIATIONS . . . . . . . . . . xii CHAPTER I THE SYNTHESIS OF A SERIES OF STERICALLY HINDERED SECONDARY AMINES INTRODUCTION . . . . . . . . . . . . . 2 RESULTS . . . . . . . . . . . . . . 6 Hydrogenation of Propargylamines . . . . . . . 10 Acidity Constants of Amine Hydrochlorides . . . . . 12 Reacting Hindered Amines with Methyl Iodide . . . . 14 Reaction of Amines with Boron Trifluoride . . . . . 15 Reaction of Secondary Amines with Chlorine . . . . 15 DISCUSSION . . . . . . . . . . . . . 19 EXPERIMENTAL . . . . . . . . . . . . . 27 PrOpargyl Alcohols . . . . . . . . . . 28 Propargyl Chlorides . . . . . . . . . . 28 Propargylamines . . . . . . . . . . . 28 Preparation of 3—Ethyl-1-pentyn-3—ol . . . . . . 28 Preparation of A-Methyl-B—isopropyl-1-pentyn-3-ol . . . 29 Preparation of 3-Chloro-3-ethyl-l-pentyne . . . . . 29 iv TABLE OF CONTENTS—-Continued Page Preparation of 3-Chloro-3-methyl-1-pentyne . . . . . 30 Preparation of 3-Chloro-3-methy1-l—butyne . . . . . 30 Preparation of 3-Chloro-5-methyl-3-isopropyl-1-pentyne . . 31 Preparation of 3-Amino-3-ethyl-l-pentyne . . . . . . 31 Attempted Coupling of 3-Amino—3-ethyl-1-pentyne with 3-Chloro-3- ethyl-l-pentyne in Aqueous KOH Solution . . . . . . 33 Attempted Coupling of 3-Amino—3-ethyl-l-pentyne with 3-Chloro-3- ethyl-l-pentyne with Equimolar Quantities of either KB or KOC(CH ) O O O I O O O I O O O O O 33 3 3 Preparation of Bis(l,l-diethyl-Z-propynyl)amine (lg) . . . 34 Preparation of Bis(1,l-dimethyl-Z-propynyl)amine (12) . . 35 Preparation of (l'-ethyl-1Lmethyl——2-propynyl)(1,1——dimethyl-2— propynyl)amine (l3) . . . . . . . . . . 35 Preparation of Bis(l-ethyl-1-methyl-2-propynyl)amine {13) . 35 Preparation of (l-Ethyl-1-methyl-2-pr0pynyl)(1',l'—diethyl-2- propynyl)amine {12) . . . . . . . . . . . 36 Preparation of Bis(cyclohexylethynyl)amine . . . . . 36 Preparation of Bis(l,l—diethylallyl-l,l-diethyl-Z-propynyl)amine 36 Preparation of (l,l-Diisopropyl-Z-propynyl)(1,1-diethy1-2-propy- nyl)amine . . . . . . . . . . . . . 37 Hydrogenation of Bis(l,l-diethyl-Z-propynyl)amine (16) in Abso- lute Ethanol with 10% Palladium on Charcoal . . . . . 37 Hydrogenation of Bis(l, l- -diethy1- -2- -propynyl)amine (16) in Abso- lute Ethanol with Platinum Oxide . . . . . . . 38 Hydrogenation of Bis(l,l——diethy1-2-pr0pynyl)amine (16) to Bis- (l, 1- -diethyla11y1)amine (23) in Ligroin . . . . . 38 Hydrogenation of Bis(l,l-diethyl-Z-propynyl)amine in Absolute Ethanol with W2 Raney Nickel Catalyst . . . . . . 39 Product Analysis of Bis(l,l-diethyl-Z-prOpyl)amine (21) . . 40 Product Analysis of Bis(l,l-dimethyl-Z-propyl)amine (as) . . 40 TABLE OF CONTENTS-—Continued Page Product Analysis of (1'-Ethyl-1'-methyl—2-propyl)(l,l-dimethyl- 2-propyl)amine {25) . . . . . . . . . . . 41 Product Analysis of Bis(l-ethyl—l-methyl-Z—propyl)amine (26) . 41 Product Analysis of (l-—Ethyl- 1-methyl—2-propy1)(l',l'-diethyl- 2~propyl)amine (27) . . . . . . . . . 41 Product Analysis of Bis(l-ethylcyclohexyl)amine 628) . . . 41 Reaction of Bis(l,l-diethyl-Z-propyl)amine (21) with Methyl Iodide . . . . . . . . . . . . . . . 42 Reaction of BF3-OEt2 with Bis(l-ethylcyclohexyl)amine {28) . . 42 Reaction of BF °OEt and BF gas with Bis(l,l—diethyl-2-propyl)- . 3 2 3 am1ne (21) . . . . . . . . . . . . . . 42 Titration of Amine Hydrochlorides in 90% Ethanol with 0.1075 N KOH under Argon . . . . . . . . . . . . 43 Melting Points of Saturated Amine Hydrochlorides . . . . 44 Preparation of Bis(l,l-dimethyl-Z—propyl)chloramine . . . 44 Preparation of Bis(l,l-dimethyl—Z—propyl)chloramine with Chlorine and Sodium Hydroxide . . . . . . . . . . . 44 Preparation of Bis(1-ethyl—1-methyl-2-propyl)chloramine Using Chlorine and Sodium Hydroxide . . . . . . . . . 45 Preparation of Bis(l.l—diethyl-Z-propyl)chloramine using Chlorine and Sodium Hydroxide. . . . . . . . . . . . 45 Attempted Chlorination of Pentane in 90% Trifluoroacetic Acid: 10% Sulfuric Acid with Bis(1-ethyl-1-methyl——2-propyl)chloramine at 00 C . . . . . . . . . . . 45 CHAPTER II THE SYNTHESIS AND STABILITY OF HINDERED SECONDARY LITHIUM AMIDES INTRODUCTION . . . . . . . . . . . . . . 47 RESULTS . . . . . . . . . . . . . . . 53 vi TABLE OF CONTENTS--Continued Stability of Hindered Lithium Amides in THF and Diethyl Ether DISCUSSION . . . . . . . . . . . . . . Lithium Amide Stability in THF and Diethyl Ether . EXPERIMENTAL . . . . . . . . . . . . Reaction of Methyllithium with Bis(l,l—dimethyl-Z—propyl)amine (£3) 0 C D C O O O O O O O C O 0 Quantitative Determination of Secondary Lithium Amide by Quench- ing with tert-Butyl Acetate and Cyclohexanone Attempted Metallation of Q24) with Phenyllithium Preparation of Phenylsodium . . . . . . . . . Attempted Metallation of (24) Using Phenylsodium Metallation of Secondary Amines with n-Butyllithium and TMEDA CHAPTER III BIMOLECULAR ELIMINATION REACTIONS OF ALKYL HALIDES WITH LITHIUM AMIDES INTRODUCTION RESULTS . . . . . . . . . . . . DISCUSSION . . . EXPERIMENTAL . . . . . . . . . . . . . Dehydrohalogenation of 2-Bromobutane with Lithium DiiSOpropyl- amide in THF 0 O O O O O O O O O I O MEtallation of THF and Attempted Reaction of Metallated THF With Z-BromObutane o o o o o o o o o 0 Reaction of 2-Iodobutane with Hindered Lithium Amides in THF Preparation of Zfde-Bromobutane from 2-Butanone . Reaction of Zjde—Bromobutane with (21)-Li in THF at 0°C . vii Page 58 61 66 68 69 69 71 71 72 73 76 88 106 121 122 123 123 124 125 TABLE OF CONTENTS--Continued Page Preparation of 2-Butyl p-Toluenesulfonate . . . . . 126 Preparation of 2-Butyl Trifluoroacetate . . . . . 126 Reaction of 2-Butyl p-Toluenesulfonate and 2—Butyl Trifluoro- acetate with Secondary Lithium Amides in THF . . . . 127 Preparation of l-Pentene-S—ol from Allyl Bromide and Ethylene Oxide . . . . . . . . . . 127 Preparation of 5-Chloro-1-pentene from l—Pentene—S-ol . . 128 Synthesis of l-Heptene-6—ol . . . . . . . . 129 Synthesis of 2-Bromo—6-heptene from l—Heptene-6-ol . . . 129 Reaction of 2—Bromo-6-heptene (46) with Hindered Lithium Amides . . . . . . . . . . . . . . 130 Synthesis of Deuterium Bromide . . . . . . . . 131 Threo and Erythro-3fdf2-Bromomobutane Synthesis . . . 131 Determination of the Stereoisomeric Purity of Threo and Ery— thrO-3-g-2’Br0m0butane o o o o o o o o o o 133 Determination of the Kinetic Isotope Effect for Dehydrohalo- genation of Threo and Erythro-3jd72-Bromobutane with Hindered Lithium Amides in THF . . . . . . . . . . 138 Reaction of Lithium Amides with Threo and Erythro-3-d-2—Bro— mobutane in THF at O C . . . . . . . . . . 139 Reaction of Alkyl Lithium Amidesowith 2—Bromobutane in THF — 12—Crown-4 Ether Solution at —78 C . . . . . . . 139 BIBLIOGRAPHY . . . . . . . . . . . . . 141 viii Table II III IV VI VII VIII IX XI XII XIII XIV XV LIST OF TABLES Page Coupling of Propargylamines with Propargyl Chlorides . 9 Hydrogenation of Bis-Propargylamines in Ethanol with KOH and Raney Nickel . . . . . . . . . 11 Acidity Constants of RZNHZCl in 90% Ethanol . 13 Amide Formation in Hexane (T50 and T90) with n-Butyl- lithium O O O O O O O O C O 57 E2 Reaction of 2-Hexenyl Halides with NaOH in Methanol 82 Olefin Products from Reactions of 2-Substituted Butanes with 0.05 M tert-BuOH and tert-BuOH at 50°C . . . 84 Olefinic Products from Reaction of 2— Iodobutane with 0.25 M Potassium tert-Alkoxides in Dimethyl Sulfoxide at 500 C O O O 0 O O O O O O O O 86 Olefin Products from Reactions of 2-Halobutanes with Anionic Bases . . . . . . . . . . . 87 Product Ratios for the Dehydrohalogenation of 2- Bromo- butane with Secondary Lithium Amides in THF at 00 C . 90 Reaction Times Required to Obtain 50% and 90% of Total Butenes Evolved in the E2 Reactions of 2-Bromobutane . 91 Product Ratios for the Dehydrohalogenation of 2- Iodo— butane with Secondary lithium Amides in THF at 00 C . 93 Product Tatios for Dehydrohalogenation of 2— Bromo- 6- heptene with Secondary Lithium Amides in THF at 00 C . 96 Product Ratios for the Dehydrohalogenation of Threo-3- d-2-bromobutane with Lithium Amides in THF at 0°C . 100 Product Ratios for the Dehydrohalogenation of Erythro Bede-bromobutane with Lithium Amides in THF at 0°C . 101 Deuterium Content in C15 and Trans-Z-Butene from the Dehydrohalogenation of Threo and Erythro-Bjde-bromo- butane O O O O O O O I O O O O 102 ix I fa“: LIST OF TABLES—-Continued Table XVI XVII XVIII XIX XIX Product Ratios for the Dehydrohalogenation of 2-Bro- mobutage with Lithium Amides Using 12—Crown-4 in THF at -78 C o o o o o o o o 0 Summary of Product Distribution Using Moderately Hin— dered Secondary Lithium Amides . . . . . Summary of Product Distribution Using Very Hindered Secondary Lithium Amides . . . . . Apparent Primary Deuterium Isotope Effect for 2-Bro- mobutane . . . . . . . . . . Mass Spectra of 2-Butenes for 3jd72—Bromobutanes . Butenes Produced by Dehydrohalogenation of Erythro and Threo-3jd72-bromobutane . . . . Page 104 107 109 117 134 135 LIST OF FIGURES Figure Page 1 Lithium Amide Stability in THF at 24°C. Recovered _4_4_ After Quenching by tert-Butyl Acetate and Cyclohexanone at -78 C. All Amides Made with TMEDA (1 equivalent) . . 59 2 Lithium Amide Stability in Ether at 24°C. Recovered 44 After Quenching by tert—Butyl Acetate and Cyclohexanone at -78OC. All Amides Made with TMEDA (1 equivalent) . . . 60 3 Variable Transition State Spectrum . . . . . . . . 76 4 IsotOpic Purity of Threo and Erythro-3-df2-Bromobutane . 99 5 Syg_and Anti Elimination Product Scheme for 41 and 48_ . 111 xi LIST OF ABBREVIATIONS DMF N, N—dimethylformamide 12—Crown-4 1, 4, 7, 10-tetraoxacyclododecane LDA lithium diisopropylamide LiTMP lithium 2, 2, 6, 6-tetramethylpiperidine NCS N—chlorosuccinamide NBS N—bromosuccinamide THF tetrahydrofuran TMP 2, 2, 6, 6-tetramethylpiperidine TMEDA N, N, N', N'—tetramethylethylenediamine xii CHAPTER I THE SYNTHESIS OF A SERIES OF STERICALLY HINDERED SECONDARY AMINES INTRODUCTION Secondary amines with branched alkyl groups are useful chemical reagents in organic chemistry. Applications of hindered amines depend on their increased substrate selectivity resulting from steric factors. 2,2,6,6—Tetramethylpiperidine (TMP) is probably the most hindered com— mercially available secondary amine routinely used in organic chemistry. This amine is used commercially as a photostablizer in polyethylene and polyacrylamide polymers. The amine inhibits degradation of polymers by quenching singlet oxygen which is produced by ultraviolet irradiation of atmospheric oxygen.1 Alkali metal amides derived from hindered secon- dary amines are proton selective and largely non-nucleophilic bases. Although bases of moderate steric requirements such as lithium diethyl— or diisopropylamide do not give a metallated derivative with alkylboranes 1_or vinylboranes 2, lithium tetramethylpiperidine2 (LiTMP) produces bo— ron stablized carbanions in both cases (eqs 1 and 2). R BCH + LiTMP benzene > R BCH Li + TMP (1) 2 3 o 2 2 25 c , , THF . . RCH CH=CHBR + LiTMP e RCHL1CH=CHBR + TMP (2) 2 2 250C 2 The failure of the smaller lithium amides to give proton abstraction is probably due to their coordination to the boron atom (eq 3). A variety R BCH + benzene \ 7 2 3 2 24°C R2B(CH3)NRé (3) 3 of related applications of LiTMP have been reported. Olofson3 has de- scribed the use of LiTMP for a practical synthesis of arylcyclopropanes 3_from benzylhalides (eq 4). Lithium diisopropylamide was less effec- tive (39% yield), presumably because substitution reactions of the star— ting benzyl halide are more likely with this less hindered amide. .. CH§==CH ArCHZCl + LiTMP -—————> (ArCH) 2 > Ar (4) 3_ 56% yield Other useful applications of hindered secondary amine derivatives have been demonstrated. The N-chloro derivatives of tetramethylpiper- idine and tert-butyl neopentylamine4 in sulfuric—trifluoroacetic acid solution were shown by Deno5 and others6 to produce a radical cation which initiates the regioselective chlorination of hydrocarbons at the sterically less hindered penultimate carbon atom (eqs 5 and 6). 30% H SO 2 4 (CH )C 70% CF C00H (CH ) C , + 3 3 \‘NCl + CH3CH(CH3 )CH 2CH3 3 > 3 3 \ijH -+ RC1 (5) (CH3 )3C 0C1!2 hv (CH3)3CCH2 4 .2 Ratio of 1° Cl 2° Cl 3° Cl From 4_ 1.72 5.98 1 From 6' 0.25 0.70 1 ((CH ) CH) NCl + CH 33CH(CH )CH CH 70% CF 3°°°H5((CH3 ) 22CH) NH + RC1 (6) 3 2 2 2 3 30/ H so 6 2 4 7 4 The reaction is catalyzed by either photochemical or chemical (Fe+3 salts) means. Radical cation.§, when compared to the less hindered radical cation 1, exhibits increased selectivity for primary and secondary hy- drogen abstraction. This is presumably due to the greater steric hin- drance of radical §_over radical 2, It would be useful to know if secondary amines with more hindered alkyl groups would show even greater substrate selectivity. An a,a’- enolizable non-symmetrical ketone may be deprotonated to form two re- gioisomeric enolates §_and.9 (eq 7). The lack of regiocontrol in kin— etic ketone deprotonation to give 8 and_9 is a significant problem which limits the use of such ketone enolates in organic synthesis. <-> O(—) 0 0 l I RCHZCLCHRIRZ Base 4, RCH =(LCHR1R2 + RCHZC = CRlR2 (7) 9 loo It is possible that hindered amide bases might favor proton abstrac- tion from the less hindered side of the ketone to a greater extent. A second reason for preparing highly hindered secondary amines is that their lithium amides might be stronger bases than the less hindered secondary lithium amides. C. A. Brown7 investigated the effect of increased alkyl group size upon alkoxide basicity. He found that the base strength of alkoxides increase with alkyl group size. Potassium tricyclohexylmethoxide is a stronger base, by about 1.2 pKa units,than potassium tert-butoxide. Brown attributed this to a decrease in solvation or ion-pair formation in the more hindered base. It seems likely that such an effect might also occur with me- tal amide bases in ethereal solvents. 5 A major goal of this study was the development of a simple, in— expensive route to a series of secondary amines of increasing steric bulk which are more hindered than commercially available secondary amines. We then planned to investigate synthetic applications of this series of amines in reactions where steric factors might lead to greater regioselectivity in product formation. The best procedure for the preparation of hindered secondary amines is probably that reported by Hennion.8 Hindered primary amine 19 reacts with 3-chloro-3-methyl-1-butyne ll_to give the coupled sec- ondary bis(propargylamine) 12 (eq 8). 40% KOH, H 0 Copper Bronze, Cu2C12 \/ HCEECC(CH3)2C1 + HCEECC(CH3)2NH 0 ll. 19. 8 days, 30 C (8) (HCEECC(CH3)2)2NH 47% yield 12 Compound 12_was semihydrogenated to 13_using 12% palladium on char- coal as a catalyst, then hydrogenated to the saturated amine 14, using W2 Raney nickel in absolute ethanol (eq 10). palladium on charcoal Raney nickel \ = 1 S _12 ,v (CH2 CHC(CH3)2)2NH 7 petroleum ether _13 ethanol (9) (CH3CH2C(CH3)2)2NH _14 41% yield Recently, we modified Hennion's procedure and prepared the highly hind- ered secondary amine bis(triethylcarbinyl)amine 21, We determined that 6 this was the most hindered secondary amine that could be synthesized in reasonable yield using Hennion's modified procedure. The sequence used to prepare this series of secondary amines is summarized in eqs 11 through 15. We found that the coupling procedure shown in eq 9 does not give significant yields of secondary propargyl amines when more hindered starting materials are used.9 NH3(liq) R2 CO + NaCE—ICH > RC ECRZCOH (1 1) HCEECRZOH + HCl > HCEECRZCCl (12) _ NH3(1iq) HC=:CR2CCl + NH2 / HC==CR2CNH2 (13) __ __ DMF __ HC=:CR2CC1 + HC=:CR2CNH2-———-—9 (HC::CR2C)2NH (14) Ni/H2 = \ (HC__CR2C)2NH Eton , (CHBCHZRZC)2NH (15) The best yield (52%) of 21_was obtained with an extra equivalent of the propargyl amine as an HCl acceptor in place of KOH in eq 14 with dimethylformamide (DMF) as the reaction solvent. We also found that W2 Raney nickel in ethanolic KOH solution was the most effective hydrogenation catalyst (eq 15). RESULTS The sequence used to prepare the series of secondary amines is summarized in eqs 11-15. Most propargyl alcohols like 3-methyl-1-bu- tyne-3-ol, 3—methyl-l-pentyne-3-ol and 1-ethynyl-l-cyclohexanol were com- mercially available. Both 3-ethyl-l—pentyne-3-ol and 4-methyl-3-iso- propyl -1—pentyne—3-ol were synthesized from the corresponding diethyl 7 and diisopropyl ketones by previously reported methods.10’11 The key steps in the synthesis (eqs 13 and 14) are based on the observations by Hennion that tertiary propargyl chlorides react with nucleophiles (eq 16) in the presence of strong base (hydroxide or amide anion) to give clean substitution at the tertiary carbon by a vinylidene carbene mechanism, as illustrated in eq l6. _______5 / HCEECCRZCI -————————9 (—)CEECCR201 <—————-— :C=C=C\ —————4> (16) ; HCEECCRZNu The coupling procedure shown in eq 9 did not give any coupled second- ary amine product when applied to more hindered reactants. The coupling of 3-amino-3—ethyl—1-pentyne with the corresponding propargyl chloride was attempted with a variety of bases (eq 17). None of the coupled product 16 was obtained with 40% KOH in H 0 or with NaH, KH or potassium 2 tert-butoxide in tetrahydrofuran (THF).12 However, using an extra equivalent of l§_gave,after 3 days reaction at 40C in DMF, a 60% GLC yield of 16,9 The coupling reaction fails when the copper bronze, cup— rous chloride catalyst is omitted or when a saturated secondary amine is used. The coupling reaction works, but in low yield,when the Olefinic 12 amine is used (eq 18). 0 Cu , CuzCl2 __ __ Base \ HC:=CC(CH3CH2)2NH2 + HC==CC(CH3CH2)2C1 —7 1§_ (l7) (HCEECC(CH3CH2)2)2NH 1§_ 8 R(CH3CH2)ZCNH2 + HC:=CC(CH3CH2)C1 -————{> (18) R Yield -—-------9 R(CH3CHZ)ZCNHC(CH3CHZ)2C==CH ‘CH2CH3 0% —CH=CH2 17% Results obtained for the coupling reactions of various propar- gyl chlorides with propargyl amines are presented in Table I. As ex- pected, the lowest yields were obtained with the more highly substituted secondary amines. Results obtained with even more hindered reagents defined the ultimate steric limitations of the reaction 9 (eq 19). HC::C(CH(CH3)2)2CNH2 + HC:=C(CH3CH2)2CC1 \V cu2°12, CH(CH CH CH 3)2 I 2 3 Copper Bronze >; HCEEO—C———NH-'C—CEECH (19) DMF, 2 weeks CH CH CH(CH3)2 2 3 5% yield Table I. Coupling of Propargylamines With Propargyl Chlorides -— :: \ == 2: HC:=CCR1R2NH + HC_.CCR3R401 17 HC__CCR1R2NHR3RAC__CH R1 R2 R3 R4 Yield Compound CH3 CH3 CH3 CH3 702 .1g CH3 CH3CH2 CH3 CH3 60% .1; CH3 CHBCHZ CH3 CH3CH2 552 _13 CH3CH2 CH3CH2 CH3 CH3CH2 552 _15 CHBCHZ CH3CH2 CH3CH2 CH3CH2 48% .19 "CH2(°H2)3°H2‘ ”CH2(°H2)3°H2‘ 66% 17 All yields are isolated yields. 10 Hydrogenation of Propargylamines Hydrogenation of the series of propargylamines in Table I posed a number of problems. Hydrogenation of pr0pargylamines or the respective amine hydrochlorides in Table I with a platinum catalyst in a variety of solvents gave almost exclusively the corresponding primary amines (eq 20). PtOZ/EtOH :2 ‘:'_: \ HC__CCR1R2NHR3R4CC__CH . , CHBCHZCRIRZNH2 + CH3CH2CR3R4NH2 H2, 60 p31 (20) Pd on carbon gave similar results in ethereal solvents; however, hydro- genation of 16_in absolute ethanol with palladium on charcoal gave the heterocycles 19 and 29_(eq 21).13 Hydrogenation of the other pro— pargylamines in Table I with Pd/C was not attempted. CH CH CH H Pd/C EtOH 3 3 3 2 'Lé ) E E H2, 50psi t t t (21) Et N Et Et N t H H .12. 29 A successful preparation of bis(triethylcarbinyl)amine 21_was achieved in 20% yield by low pressure hydrogenation with W2 Raney nickel. The major side reaction of the hydrogenation was hydrogeno- lysis of l§_to 22_(eq 22). Raney Ni/ H2 16 ) ((CH CH ) C) NH + (CH CH ) CNH ——- 3 2 3 2 3 2 3 2 22 anhydrous EtOH A ( ) 21 22 11 More reactive grades of Raney nickel (W4 and W6) gave mainly incom- pletely hydrogenated secondary amines along with hydrogenolysis pro- duct 22. Finally, a 72% yield of 21 was obtained by adding a 1 mol ex- cess of potassium hydroxide to the W2 catalyst to suppress hydrogen- olysis. Results obtained with a series of propargylamines by this last procedure are given in Table II. Table II. Hydrogenation of Bis—Propargylamines in Ethanol with KOH and Raney Nickel. Hz/Raney Ni (W2) — = \ HCZZCCRIRZNHR3R4CC__CH , CH3CH2CR1R2NHR3RACCH2CH3 EtOH, KOH Yield R1 R2 R3 R4 Compound (%)a CH3 CH3 CH3 CH3 '24 80 CH3 CH3CH2 CH3 CH3 _25 78 CH3 CH3CH2 CH3 CH3CH2 26_ 75 '7 CH3CH2 CH3CH2 CH3 CH3CH2 2; 75 CH3CH2 CH3CH2 CH3CH CHBCH2 21_ 72 -CH (CH ) CH - —CH (CH ) CH - 28 80b 2 2 3 2 2 2 3 2 ——- aIsolated yield. bNo KOH used in the hydrogenation. 12 Acidity Constants of Amine Hydrochlorides The acidity constants of the amine hydrochlorides obtained from the secondary amines of Table II were determined by potentiometric ti- tration with standard alcoholic potassium hydroxide (carbonate free). All titrations and solution transfers were done under argon to avoid carbonate formation with with atmospheric CO Titrations were at- 2. tempted in dioxane and dioxane-water solutions of varying concentra~ tions.14 The hindered amines and their hydrochlorides were not soluble even at very low concentrations (0.005 M) in 95:5 (v/v) of dioxane: water. Ethanol (90%) was chosen as the solvent since the amines and amine hydrochlorides are both completely soluble in this solvent at the concentrations employed. The procedure for these titrations was taken from a previous report. A carbonate free solution of KOH was prepared by dissolving po— tassium metal in absolute ethanol under argon and diluting with degas— sed distilled water to the desired concentration (0.1075 M), determined by titration with standard HCl solution. The pKa of the amine hydro- chloride was calculated as the pH of the solution at half the equiva— lence point volume (eq 23). The titration was monitored using a combi- nation calomel—glass electrode at 250C attached to a digital readout pKa= pH1 equivalence volume (23) ’2 pH meter. Results of the titration study are reported in Table III, with the hydrochlorides of diiSOpropylamine and 2,2,6,6-tetramethyl- piperidine being included for reference. A regular increase in acidity (apparent decrease in pKa) is observed with increasing steric bulk 13 Table III. Acidity Constants of R NH C1 in 90% Ethanol 2 2 RZNHZCl pK a [(CH3)2CH]2NH2C1 9.8a NH Cl 10.1b 2 [CH3CH2C(CH3)2]2NH2C1 9.9 [CH3CH2C(CH3)2NH2C(CH3) (CHZCH3)2]C1 9.4 [(CH3CH2)2C(CH3)]2NH2C1 8.7 [(CH3CH2)3CNH2C(CH3)(CH3CH2)2]Cl 8.0 [(CHBCH2)3C]2NH2C1 7.1 HZCH3 l NHZCl 9.0 a Lit.16a pKa = 11.07 (water solvent). bLit.16b pKa = 11.24 (water solvent). 14 of the alkyl groups. The most highly substituted amine in the series,21, appears to be the weakest base (pKa of conjugate acid=7.1) of any sat— urated aliphatic amine yet reported. Treatment of amine 21 with an ex- cess of either aqueous hydrochloric acid or hydrogen chloride in ether gave a dihydrochloride. This was confirmed by titrating the acid salt of 21 with standard KOH and observing two distinct inflection points in the titration curve at titrant volumes corresponding to the calcu- lated values for the di and monohydrochloride of 21, The second mole- cule of hydrogen chloride was not removed even under prolonged heating under vacuum (2 weeks,1000C). The monohydrochloride of 21 was readily obtained by treating the amine in hexane solution with slightly less than one equivalent of aqueous HCl, shaking vigorously for 5 minutes and decanting the hexane layer. The aqueous layer was dried under vacuum and the amine monohydrochloride was used without further puri- fication. Reacting Hindered Amines with Methyl Iodide Amines 24 and 21, the least and most hindered members of the series in Table II, respectively, were reacted with methyl iodide in methylene chloride in sealed vials for 1 week at 7000. No precipitates, indicating formation of the quaternary iodonium salts, were formed. NMR analysis of the reaction mixtures showed only unreacted methyl iodide and free amine, When 2,2,6,6—tetramethylpiperidine (TMP) was mixed with an equivalent amount of methyl iodide in chloroform under similar conditions, a white precipitate corresponding to the methyl io— donium salt of TMP was formed within a few minutes. 15 Reaction of Amines with Boron Trifluoride Addition of boron trifluoride etherate to either TMP or 2§_in hexane solution gave within a few minutes a white precipitate of the corresponding adduct of boron trifluoride (eq 24 and 25). z (24) 312 w 111 21': BF : OEt ~ 3 2 > ~ /H NH N (25) 2 2 \BF 3 Under the same conditions amine 21 formed only the adduct of a prim— ary amine 22 and the olefin 3-ethy1—2-pentene (eq 26). Compound 22_was 5 min. _21 + BF3:OEt2 ~—:;;;-—9'CH3CH=C(CH2CH3)2 + (CH3CH2)3CNH2BF3 (26) 29 also formed when pure BF3 gas was added to a neat solution of 21, Reaction of Secondary Amines with Chlorine Fataftahl8 showed that 2§_will react with bromine in carbon tet- rachloride to give the bromonium salt 29 (eq 27). Treatment of 29_with one equivalent of standard sodium hydroxide gave a 66% yield of the bro— mamine 31 within 30 minutes (eq 28). Repeating this reaction by l6 Br ~\‘;] 2 \ - + 3 H (27) NH CCl4 / 11 Br— 2 2 Br .£§ 30 NaOH, H20 \ (28) CC14 7T7 H NB . _ r Br+ Br 2 2 30 31 dropwise addition of bromine to a two phase reaction mixture of stan— dard aqueous sodium hydroxide and a hexane solution of the amine gave quantitative yield of 21, This procedure is much simpler than other me- thods for forming haloamines. For example, N-bromosuccinamide in ether requires 1 week to convert 28 to 21. Neither N-chlorosuccinamide nor sodium hypochlorite converts 2§_to the correSponding chloramine. There- fore Fataftah's procedure was adapted to forming chloramines. The only modification in technique was addition of chlorine gas from a syringe to a slowly stirring solution of hydroxide and hexane-amine solution at 000. Care must be taken to insure that the chlorine does not react with the aqueous hydroxide solution to form hypochlorite ion. Stirring slowly insures that chlorine reacts first with the hexane-amine solution top layer, forming the chloronium salt. The chloronium salt subsequently reacts with hydroxide, forming the chloramine. This reaction was applied 17 to four different hindered amines and the results are shown below in eqs 29 and 30. r' 'TCl Cl , hexane NaOH 2 2 47 a (9) 0 C N H O N R Cl $2 $3 C12, hexane $2 + 13 NaOH CH CH C—NH-CCH CH -$> CH CH C-NH—CCH CH 2 3 2, l 2 3 00C 3 21 Cl H 2 3 H20 R1 R4 1 4 (30) 1‘2 i3 \ - — = =H 9‘; ‘ , CH3CHZC NCl C|CHZCH3 R1 R2 C 3 95/ y1eld R R 1 2 R=CH 1 3 93% " R2=CH2CH3 R1=CH2CH3 94% H R2=CH2CH3 All of these hindered chloramines were isolated by rough drying of the hexane layer with anhydrous sodium sulfate and removing the hexane under reduced pressure. The NMR spectrum of each chloramine could be clearly distinguished from starting amine. As the reaction proceeds, the N—H proton signal disappears and the methyl (methylene) proton signal shifts downfield for the chloramine. Benzene was used as an internal standard for determining the yields. The chloramines were stable for at least one day at 00 except for the chloramine derived from 21, which began to decompose within 301mh1 18 at ambient temperature ( as determined by NMR). We attempted to chlo- rinate pentane using these hindered chloramines according to several published procedures. The first was a standard photochemical initiated procedure for forming radical cations from chloramines.19’20 The se- quence of steps involved in the chlorination of a hydrocarbon substrate is outlined in eqs 31—35, 6 + H2804 and/or CF3COOH H RZNCI + H ; R2N\+ (31) Cl +/H hv or FeIII + R2N\. a RZNH - + Cl' (32) C1 + -+ SH + - ‘ ° RZNH 7 RZNHZ + s (33) SH + Cl- .1; S- + HCl (34) + s- + RZNHCl (or 012)————) 301 + RZNHI (or Cl') (35) The ultimate goal of this effort was to use these sterically hindered chloramines to selectively chlorinate the terminal methyl group of a long chain hydrocarbon. Studies have demonstrated that sterically hindered amine cation radicals show greater selectivity in hydrogen ra— dical abstraction from primary and secondary carbons over that of ter- tiary carbon atoms.20 Unfortunately, we were unable to observe any chlorination of pentane using the chloramines listed in eq 30. Both photochemical and transition metal initiated halogenation conditions were used. Neither 19 pentane nor chloramine was recovered; only a thick tar was obtained. Subsequent GLC analysis showed no low boiling organic products. Con- trol reactions with chloramines of dimethyl as well as diisoPropylamine synthesized using Fataftah's procedure gave halocarbons in yields and proportions corresponding to literature values. DISCUSSION Hennion studied the reaction of trimethylamine and tertiary pro- pargyl chlorides (eq 36).21 Reaction products are prOpargyl ammonium chlorides when R and R' are methyl. If R and R' are larger than methyl, the products were allenyl ammonium chlorides. . + C1" ?, _1R or R small >RR'C(NMe3)CEECH HCECCCI + Me3N (36) R R or R' large - :5 RR' C=C=C (N+Me3) C1 The reaction fails if Cu2C12 and/or copper bronze are not included in the coupling reaction mixture. Thus, copper is a necessary catalyst, which may help to stabilize the vinylidine carbene and hold it close to the unsaturated amine so that the coupling reaction can occur.9 This explanation seems to be supported by the observation that coupled pro- duct yields fall as the primary amine varies from propargyl to the sat- urated alkyl amine 22, The saturated amines in Table II were obtained by hydrogenation of bis(propargyl)amines with W2 Raney nickel in ethanol. More reactive Raney nickel (W4 or W6 grade) gave increased amounts of cleaved primary and unsaturated secondary amines. Hydrogenolysis occurs at the diallylamine stage, since bis(diethylallylcarbinyl)amine 22_was isolated by hydrogenation 12 in ligroin. Proton sources like water or 20 ethanol when added to the palladium on carbon catalyst mixture in ligroin rapidly cleaved 22_to the saturated primary amine 22_(eq 37). CH CH CH CH palladium/carbon 3 ‘2 I'2 3 l6 )fiCH =CH-C—NH—C-CH=CH -—-———9 (CHACH ) CNH (37) ——- petroleum ether, 2CH CH SH CH 2 J 2 3 2 H2 3 2 2 3 23 22 The highest yields of saturated secondary amines were obtained when po- tassium hydroxide was added to the Raney nickel—ethanol solution. The base apparently suppressed hydrogenolysis of the amine in the protic solvent. Some incompletely hydrogenated Olefinic secondary amine re- mained in all the hydrogenation reactions. These impurities ranged from 7% for the most hindered amine to about 3% for the least hindered amine _24 in the series. Distillation using a spinning band column gave the saturated secondary amines in 98% or better purity. The acidity constants of the saturated secondary amines were de- termined and are reported in Table III. Ethanol (90%) was chosen as the solvent for the titration reactions because it dissolves all the re- actants, and because pKa's have been reported for amine hydrochlorides 15’16 The effect of in- in different ethanol—water solution mixtures. creasing ethanol concentration on apparent pKa values of typical alipha- tic amines is to lower the observed pKa by approximately 1—2 units (re— lative to water) and to compress the range of pKa values for a given series of amines.16 Thus the differences in pKa values in Table III 21 are expected to be even greater in water solution. Presumably, solvation 22,23 of the highly hindered ammonium salt accounts for this effect. Si- milar observations, of a smaller magnitude, were reported by Brown for highly branched bases.22 Hall24 has also argued that solvation occurs by hydrogen bonding between Nf—H groups in the ammonium ions and water molecules. He found that primary and secondary amines are correlated by single lines in the Taft equation (eq 38) only in cases where the amine groups have low steric hindrance. RR = o 0 (38) Primary and secondary amines with more sterically hindered groups do not fall on the slope 0* of the Taft equation. The deviation for the lar- ger groups must be a steric effect because electronic effects are ac- counted for in the 0* values. As the degree of steric hindrance in— creases, the base strengths become lower relative to their predicted values. This steric weakening of base strength cannot be ascribed to B-strain because the tertiary amines should show this same decrease in basicity but to a greater extent. 0n the contrary, tertiary amines are correlated by a single line in the Taft equation, irrespective of the steric hindrance. Furthermore, primary amines simply are incap- able by definition of exhibiting B—strain. Steric hindrance of sol- vation can, however, explain the result. Brown22 also reports that 2,6—di—tert-butylpyridine does not form a simple hydrochloride, giving instead a dihydrochloride. He at- tributes this effect to the inability of chloride ion to form a hydro- gen bond to the nitrogen bound hydrogen. A second molecule of hydrogen 22 chloride is then required to form a hydrogen bond to the chloride ion. We observed that treatment of 21 with excess hydrochloric acid or hydro- gen chloride in ether gave a dihydrochloride. The second molecule of hydrogen chloride was not removed even after prolonged heating under vacuum. The monohydrochloride was easily obtained by adding excess amine to a standard aqueous solution of hydrochloric acid. The unre— acted free amine was removed by ether extraction of the aqueous phase. Corrections normally applied in determining the acidity con- stant of an amine hydrochloride,such as ion activity and auto-hydro— lysis of solvents, were not employed since we were unable to extrapo- late the apparent pKa's in 90% ethanol to 100% water. The hindered a- mines employed in the study were insoluble in concentrations as low as 0.001 M in 80% ethanol solution. Additionally, if the factors men— tioned above were applied in determining the actual pKa's of the amines, the apparent pKa's would fall within 0.1 or 0.2 pH units of the actual pKa values.16 This is close to the limits of accuracy for experimental measurments. Little is known about the reactions of sterically hindered sec— condary amines. Work by Klages25 in 1963 provides an example of how steric effects reduce the nucleophilicity of di-tert-butylamine 22. Klages reacted 22_with methyl iodide for 5.5 months and obtained 23 and 22, presumably by initially forming 2; (eq 39). I- 2[(CH ) C] NH )CNC[[(CH>C] NH+CH] 3° 2 > 33 33 33223 H 2 CH3I 2.2. 2—3— (CH 23 I I-/CH > 2(CH ) c-N-C(CH ) + (CH ) C—N 3 3 3 A 3 3 3 3 1‘1\ (39) H H+ CH 3 34 35 Niether amine 21_nor 2§_reacted with methyl iodide within a two week period. This observation is not unexpected, in view of the results ob- served with di-tert-butylamine. Boron trifluoride etherate and boron trifluoride gas react with diethylamine to give the addition product_§§ which disproportionates when heated above 2500C giving_31 and_3§ (eq 40).26 BFB'OEt2 +/’H +/,H _ (CH3CH3)2NH > (CH3CH2)2N\‘BF_ ‘——;? (CH3CH2)2N\\H BF4 (40) 36 3 37 + (CH3CH2)2N-BF2 33 Klages25 found that di—tert—butylamine reacts with boron trifluoride to give fiQ_and.gl, probably by the initial formation of the simple adduct_32 (eq 41). BF H + _ 3 \ /’ RZNHZ , R2N\ ————9R2NHZBF3 + 1221:1131:2 (41) R= t-butyl 24 Fataftah18 demonstrated that 2§_will react with boron trifluor- ide etherate to give a crystalline white solid with a sharp melting point (154.5-1550C). Elemental analysis agreed with values calculated for structure 42, The formation of 42 and not disproportionation pro— ducts analogous to 49_and_41 might simply be due to a difference in re— action conditions (eq 42). BF3 OEt2 (42) Under the same conditions as Fataftah applied in eq 42, we reacted_21 with either boron trifluoride etherate or the corresponding gas at 00C. We obtained the primary amine BF adduct and the elimination product 3 3-ethy1—2-pentene. We were unable to isolate the boron trifluoride adduct of the hindered secondary amine under any conditions. Fataftahl8 demonstrated that the bromamine of 2g could be formed within 1 week using N-bromosuccinamide in ether. He developed a new procedure in which bromine was added to the hindered amine and then the bromonium salt was deprotonated to form the bromamine. This reac— tion was completed quickly and in high yield in either hexane or car— bon tetrachloride using standard aqueous base as the deprotonating agent. This reaction was discovered when bromine was added to TMP. An orange compound, presumably the bromonium salt, was formed, but the yield was only 66%. The mother liquor was examined by NMR and showed the pre- sence of TMP‘HBr. The hydrobromide may have been formed by the reaction 25 of TMP with the corresponding bromonium salt as shown in eq 41. ngm + or Br H Br H HNh Br- it ,Based on reports in the literature,6 more hindered haloamines might exhibit increased selectivity for the less substituted hydrogen atoms on a long hydrocarbon chain. Attempts at both chemically and pho— tochemically induced chlorination of pentane with gl—Cl, 24-C1 and géfCl failed. Control reactions using chloramines listed in eq 42, prepared by Fataftah's procedure, gave chlorpentanes in proportions described in the literature.6’20 H2804 / hv, pentane R R NCl ) l-Cl + 2-Cl + 3—Cl pentane 1 2 o O C R1= R2= CHZCH3 2 76 22 (42) R1: R2= CHC2H6 4 76 20 R1: R2=-(CH3)2C(CH2)3C(CH3 2—- 10 70 20 An explaination for the lack of chloropentane is that the reaction is not going through a radical cation process, but is instead forming nitre- nium ions.26 Gassman has shown that silver catalysed solvolysis of a number of secondary bicyclic chloramines proceeds through a nitrenium cation intermediate. Gassman concluded that the corresponding azabicy- clic compounds would be ideal systems for establishing the existence of alkyl migration to divalent electron—deficient nitrogen. N—Chloriso- quinuclidine's solvolytic behavior was studied (eq 43) and it was 26 'concluded that £3_goes through a nitrenium cation intermediate. \ N fl ;b 01., N + OCH3 __ MeOH '* —————— 4- N’ '---> N N It may be that in acidic solution, our hindered chloramines are able to (43) eject a negative chloride ion to relieve the steric strain of the sp hybridized nitrogen and proceed to a less sterically crowded sp ni- trogen cation (eq 44). Whether this actually occurs is speculation, but 9 N-Cl‘ . 'H+' . . oOSO3H ——9 OCNE'. (44) + Cl'°'H""O-303H this does provide an explaination for why we do not observe chlorination of pentane nor any other identifiable hydrocarbon products in the workup of the reaction solution. The fate of the nitrenium cation is uncertain, but it is reasonable to assume that the cation might undergo an intra 27 molecular nitrenium ion rearrangement reaction like the one observed in eq 43. Ultimately, the amine cation would probably react in a random fashion with any hydrocarbon or amine in the highly acidic solution to form a polymeric tar. This explanation is reasonable in view of the fact that neither amine nor pentane was recovered. EX?ERIMENTAL Melting points were determined on a Thomas-Hoover melting point apparatus and are uncorrected. Elemental analyses were performed by Spang Microanalytical Labs, Inc. IR spectra were obtained as neat films on NaCl plates with a Perkin—Elmer 237—B spectrophotometer, and 1H NMR spectra were taken on a Varian T-60 spectrometer with MeASi in— ternal standard. 13C NMR spectra were taken with a Varian OFT—20 spectrometer and are calibrated in parts per million downfield from Me4Si; the J values are given in Hertz. Mass spectra were taken with a Hitachi RMU-6 mass spectrometer. Gas—liquid chromatograph (GLC) was performed with a Varian 920 gas chromatograph. All reagents and sol- vents were dried and purified before use. Absolute ethanol (Gold Shield) was used for all hydrogenations. Dimethylformamide (DMF) was dried over calcium hydride and distilled before use under vacuum. An- hydrous cuprous chloride (Cu2Clz), used for converting the propargyl alchohols to the propargyl chlorides, was obtained as a 95% pure powder from Ventron. Cuprous chloride was prepared immediately before use in an amine coupling reaction by a published procedure.27 Copper bronze powder was obtained from the Illinois Bronze Powder Co. Raney nickel alloy was obtained from Ventron, Inc. 28 Propargyl Alcohols. 3-Methyl-1-butyn-3-ol, 3-methyl-1-pentyne- 3-01, and l-ethynyl-l—cyclohexanol were commercially available (Aldrich). All other pr0pargyl alcohols were synthesized from the corresponding 10,11 ketones by previously reported methods. Propargyl Chlorides. All propargyl chlorides were prepared from the corresponding propargyl alcohols and used without further puri- fication.29 All were dried and stored over anhydrous potassium carbonate. The purity of the propargyl chlorides was determined by GLC (10% Car- bowax 20M on Chromasorb W, AW-DMCS). Propargylamines. l—Ethynylcyclohexylamine and 3-amino-3-ethyl- 1-pentyne were commercially available (Aldrich). All other propargyl- amines were made by a previously reported method.28 Preparation of 3-Ethyl-1-pentyn—3-ol. This compound was pre- pared according to the procedure of Vaughn and co-workers,10 and its preparation is representative of the preparation of tertiary propargyl alcohols. A 5—L, three—necked, round bottomed flask was fitted with an efficient mechanical stirrer mounted through a glass bushing along with two gas inlet tubes for acetylene and ammonia, both of which were immersed in liquid ammonia. The third neck of the flask was fitted with a dry ice condenser which was connected to a KOH drying tower by rubber tubing. The flask was flame-dried, purged with NH gas, and 3 insulated with a S—L heating mantle. The flask was charged with 3500 mL of liquid ammonia, the stirrer was started and a rapid stream of acetylene was passed into the solution for about 30 min to saturate the solution. Additional ammonia gas was condensed in from time to time to keep the solution volume at about 4 L. Sodium (115 g, 5 mol) was cut into strips and added at such a rate that the entire solution 29 did not turn blue. The addition required about 1.5 h, depending on the rate of passage of acetylene. Stirring and addition of acetylene was continued for l h. 3-Pentanone (430.6g, 4.95 mol, 98%) was added dropwise over 1 h to the ammonia solution. The solution was stirred overnight. Then the heating mantle was removed, and the reaction mixture was allowed to stand until all the ammonia had evaporated. The solid residue was decomposed by adding about 1500 mL of ice and water. The mixture was carefully acidified with 50% H2804 (300-500 mL). The organic layer was dissolved in 400 mL of ether and washed with 200 mL of brine. The original aqueous phase and the brine wash were then extracted with two 200~mL portions of ether. The combined ethereal solutions were dried over anhydrous MgSO and filtered, and the ether was removed under re— 4 duced pressure. The product was distilled under reduced pressure through a 15—cm Vigreux column. The yield of 3-ethy1-l-pentyn-3-ol was 520 g (93% yield): bp 135-1360C; vmax 3413, 3300, 2970,1470, 1375, 910 cm—1; NMR (00013) 0 1.0 (6 H, t, J=7), 1.65 (4 H, q, J=7), 2.3 ((l H, s), 2.38 (l H, s); nZSD 1.4207. Preparation of 4-Methyl-3—i30propy1-l-pentyn-3—0l. Application of the above procedure to 2,4-dimethyl—3—pentanone gave 133 g of 4-me— thyl—3-isopropyl-l-pentyn—3-ol: 89% yield; bp 690C (16 mm); vmax 3490, l 3315, 2970, 1475, 985 cm- ; NMR (00013) 6 0.98 (12 H, d, J=7), 1.92 ( 3 H, m, J=7), 2.35 ( 1 H, s); n25D 1.4435. Preparation of 3-Chloro-3—ethyl-1—pentyne. The following pro- cedure for the conversion of 3—ethy1-l-pentyne-3-ol to 3-chloro-3- 2 ethyl-l-pentyne is representative for preparing the chlorides. 9 A 1-L, three-necked flask provided with a magnetic stirrer, thermometer, 30 and dropping funnel was charged with 56 g (0.5 mol) of calcium chlor- ide, 40 g (0.4 mol) of Cu2Cl2 (95% brown powder), 400 mg of copper bronze powder, and 430 mL (5 mol) of cold concentrated hydrochloric acid. The mixture was flushed with argon and cooled (ice bath) with stirring. One mole of 3-ethy1—l—pentyn-3—ol (113.3 h) was added drop- wise over 30 min. Stirring was continued for 1 h (O-SOC). The upper organic layer was separated and washed immediately with three 100-mL portions of cold concentrated hydrochloric acid, with two 100-mL por- tions of water, and once with lOO-mL portion of saturated aqueous sodium carbonate. The colorless product was thoroughly dried with two portions of anhydrous K2C03. Analysis of the sample by GLC (10% Carbowax 20M on Chromosorb W) showed the sample to be 96% pure. The chloride was used without further purification. The total isolated yield of the chloride was 73%: vmax 3290, 2970, 1950, 1460, 1380, 1 1315, 835 cm- ; NMR (00013) 6 1.47 (6 H, t, J= 7), 1.92 (4 H, q, J= 7), 2.58 (1 H, s), n25D 1.4387. Preparation of 3-Chloro-3-methy1-l-pentyne. Application of the above procedure to 3-methyl-l-pentyn-3-ol (196 g, 2 mol) gave 176 g of 94% pure product (76% yield). This solution was used without further purification: Vmax 3290, 2975, 1950, 1460, 1380, 1210, 815, 750 cm-1; NMR (CDC13) 0 1.05 (3 H, t, H= 7), 1.73 (3 H, s), 1.93 (2 H, q, J= 7), 2.53 (l H, s), nZSD 1.3749. Preparation of 3—Chloro—3-methyl-l-butyne. Application of the above procedure to 2—methy1-3-butyn-2-ol (168 g, 2 mol) followed by distillation at atmospheric pressure through a 20-cm Vigreux column gave 62 g of 97% pure product: 30% yield; bp 74—7600; vmax 3290, 21106m‘1; 25 NMR (CDC13) 0 1.82 (6 H, s), 2.57 (1H, s); n D 1. 4156. 31 Preparation of 3-Chloro-5-methyl-3—isopropy1-1-pentyne. Appli- cation of the above procedure to 4-methyl-3-isopropyl-l—pentyn—B-ol (137 g) gave 115 g of the corresponding propargyl chloride (70%), 94% pure by GLC. The chloride was used without further purification: bp 1 S7-6OOC (15 mm); Vmax 3290, 2960, 1475, 1390, 810 cm- ; NMR (00013) 6 1.10 (12 H, d, J=6), 2.13 (2 H, m, J=6), 2.55 (1 H, s); n25D 1.4559. Preparation of 3-Amino-3-ethyl-l—pentyne. The following pro- cedure for converting 3-chloro-3—ethyl-1-pentyne into 3—amino-3-ethy1— l-pentyne is adapted from a published procedure 28 and is represent— tative for preparing all other primary propargylamines from the corre— sponding propargyl chlorides. Sodium (24 g, 1.04 mol) was converted to the amide (catalyzed by 0.3 g of FeCl3) in 1 L of liquid ammonia (anhydrous) in a 3-L, three-necked, round—bottomed flask provided with a mechanical stirrer, dry ice condenser, and a long stem gas inlet tube for introducing ammonia. Then 130.6 g of 96% pure 3-chloro-3- ethyl-1-pentyne (0.96 mol) diluted with 4 volumes of anhydrous ether was added dropwise over a 1.5 h period with continuous stirring. The flask was insulated with a 3—L heating mantle and allowed to stir overnight. The ammonia was allowed to evaporate, and chopped ice (500 g) and ether (150 mL) were then added. The ether layer was sep- arated and the aqueous layer extracted once with 100 mL of ether. The combined ethereal extract was washed with cold water and filtered. The extract was acidified with concentrated HCl (60 mL). The ether layer was discarded and the mixture was extracted once with 50 mL of ether. The aqueous solution was then treated with 29 g of NaOH in 30 ml of water to release the amine which was recovered by extraction with ether. Distillation gave 81.2 g (73% yield) of pure 3-amino-3-ethyl 32 1-pentyne: bp 36—38OC (2 mm); Vmax 3360, 3290, 3280, 20806m‘1; NMR 25 (CDC13) 0 1.0 (6 H, t, J= 7), 1.53 (6 H, m), 2.27 (1 H, s); n D 1.4392. Prepararation of 3-Amino-3-methyl—l-butyne. About 105 g of 97% pure 3-chloro—3-methyl-l-butyne (1.0 mol) was added to sodamide (1.1 mol) in liquid ammonia. After workup, 35 g (42% yield) of 3-amino- 3-methyl-l-butyne was obtained: bp 79-800C (760 mm); vmax 3370, 3290, 3210, 1620 cm-1 (1 H, s); n25D 1.4180. ; NMR (CDC13) 0 1.4 (6 H, s), 1.67 (2 H, br s), 2.25 Preparation of 3-Amino—3-methyl-l-pentyne. 3-Chloro-3-methyl- 1-pentyne (120 g, 97%, 1.0 mol) was added to sodamide (1.1 mol) in li— quid ammonia. After workup and distillation, 59 g (61% yield) of 3- amino—3-methy1-l—pentyne was obtained: bp 53-54OC (100 mm); vmax 3360, 3300, 3290, 1640 cm-1; NMR (00013) 6 1.0 (3 H, t, J= 7), 1.53 (4 H, m), 2.25 (1 H, s); n25D 1.4302. Preparation of 3-Amino—4-methyl-3-isopropyl-l-pentyne. 3-Chlo— ro—4-methyl-l-pentyne (51 g, 0.32 mol) was added to sodamide(0.35 mol) in liquid ammonia. After workup and distillation, 14.5 g (41% yield) of 3-amino—4-methy1-3-isopropyl—1-pentyne was obtained: bp 56—57OC 1 (3 mm); Vmax 3370, 3290, 3250 cm— ; NMR (00013) 6 0. 98 (12 H, d, J=6), 25 D Preparation of 3-Amino-3—ethyl-1epentene from 3-Amino-3—ethyl- 1.27 (2 H, br s), 1.85 (2 H, m, J= 6), 2.2 (l H, s); n 1.4501. 1-pentyne. This experiment was adapted from a previously published procedure.8 Sodium metal (2.3 g, 0.1 mol) was added in small pieces with stirring to a solution of 22.2 g (0.2 mol) of 3-amino—3-ethyl—1- pentyne in a 500 mL, three-necked, round-bottoned flask containing 200 mL of anhydrous liquid ammonia. Ammonium chloride (0.1 mol, 5.4 g) was then added slowly. Alternate additions of sodium and ammonium 33 chloride were repeated until a total of 11.3 g (0.47 mol) of sodium and 27 g (0.5 mol) of ammonium chloride had been added. A constant total volume was maintained by periodic addition of ammonia. Ether (50 mL) was added, and the liquid ammonia was allowed to evaporate overnight. The mixture was filtered, and the solid was washed with two 50 mL portions of ether. The combined ether solutions were dried over anhydrous K2C03. Distillation gave 9.44 g (42% yield) of 3—amino- 3-ethyl-l-pentene; bp 128-1290C (760 mm); vmax 3350, 3290, 3075, 1685 cm-1; NMR (CDC13) 5 0.88 ( 8 H, q, J=7), 1.42 (4 H, q, J= 7), 4.8-5.9 (3 H, m). Attempted Coupling of 3-Amino-3-ethy1-l—pentyne with 3—Chlo— ro-3-ethy1—15pentyne in Aqueous KOH Solution. This experiment was adapted from a previously published procedure.8 3-Amino-3-ethyl—1— pentyne (4.5 g, 40 mmol),2 mL of 40% KOH solution, 10 mg of copper bronze powder, and 8.6 g of 3-chloro-3-ethyl—1-pentyne were mixed to- gether with 10 mg of CuzCl2 and maintained at 25-300C. After 24 h, an additional 30 mL portion of KOH solution and 10 mg of copper bronze powder were added. Five additional 30 mL portions of KOH were added, one after each 24 h period. .After the eighth day, an aliquot of the sample's organic layer was analyzed by GLC (10% Carbowax 20M Chromo- sorb W, AW-DMCS treated, 6 ft column). Analysis of the separated compo— nents indicated that the propargyl chloride had hydrolyzed to the pro— pargyl alcohol, and the primary amine was recovered quantitatively. No high-boiling coupled products were seen. Attempted Coupling of 3-Amino—3-ethyl-l-pentyne with 3-Chloro- 3).. Potassium hydride suspension (2 mmol, 0.5 mL of 5.4 M KH) was injected 3—ethyl-lepentyne with Equimolar Quantities of either KH or KOC(CH into a flame—dried, lO—mL, round-bottomed flask fitted with septum 34 inlet, a flow control valve and a Teflon-coated stirring bar. The flask was flushed with argon. The mineral oil was removed via syringe by wa- shing with three 2 mL aliquots of dry pentane. Copper bronze powder (15 mg) was suSpended in 2 mL of dry THF and injected into the flask. Then 1 mmol of 3-amino—3-ethyl-1-pentyne was injected. The flask was thermostated at 220C, and a gas manometer was attached. Then 0.16 g (1 mmol) of 3—chloro-3-ethyl-1-pentyne was injected dropwise. About 0.8 mmol of H was evolved over a 5 h period. After the reaction was 2 quenched with H O (0.5 mL), no coupled secondary amine was detected by 2 GLC. Identical results were obtained with NaH and KOC(CH3)3. In each case, there was partial H evolution, but no coupled amine was formed. 2 Preparation of Bis(l,l—diethyl-Zepropynyl)amine.116). The fol— lowing procedure for the coupling of 3-amino-3-ethyl-l-pentyne with 3-chloro—3-ethyl-l-pentyne is representative for the formation of dipro— pargylamines. A 500 mL, round—bottomed flask equipped with a magnetic stirring bar, septum inlet, and gas inlet valve was flame-dried under argon. Copper bronze powder (220 mg) and freshly prepared CuzCl2 (220 mg) were added followed by 109 mL of DMF (dried over CaH2 and distil- led) and 3-amino-3—ethyl—l—pentyne (29.8 g, 260 mmol). The flask was flushed with argon for 10 min and cooled to 40C in a cold room. Then 18.3 g (133 mmol) of 95% 3-chloro—3-ethyl-l-pentyne was injected. After 72 h, the solution was quenched with 30 mL of 20% aqueous NaOH (150 m- mol). Water (100 mL) was added and the solution was steam distilled. The organic layer was separated from the distillate, and the aqueous layer was extracted with three 50 mL aliquots of ether. The combined ether extracts were dried over MgSOA. The ether was removed under reduced pressure and the residue distilled through a short Vigreux 35 column. The primary propargylamine was recovered(12.7 g, 111 mmole, 87% of extra equivalent), and the coupled amine was distilled under vacuum; bp 61-64OC (0.5 mm). There was obtained 12.9 g (48% yield) of bis(1,1— diethyl-2-propyny1)amine: vmax 3290, 2965, 2925, 2870, 1460, 1375, 1170 cm"1; NMR (00013) 6 0.93 (13 H, t, J= 7), 1.73 (8 H, q, J= 7), 2.25 (2 H, 3); mass spectrum, m/e (relative intensity) 206 (M+ + 1), 176 ( 17), 82 (100), 67 (16), 55 (15), 41 (15). Preparation of Bis(l,l—dimethyl-Z-propynyl)amine (12). 3-Chloro- 3—methyl-1-butyne (0.5 mol, 51 g) was reacted with 3-amino-3—methyl- l-butyne (1.1 mol, 92.5 g) for 24 h and was worked up as previously de- scribed. After distillation, 52.5 g (70% yield) of bis(l,1—dimethyl— 2-propynyl)amine was isolated: bp 60-650C (20 mm); vmax 3290, 2300, 1465, 1375, 1360, 1210, 1065 cm-1; NMR (CDC13) 6 1.28 (12H, 5), 2.23 (2 H, 3); mass spectrum, m/e (relative intensity) 149 (M+), 134 (46), 118 (3), 91 (3), 68 (100), 67 (16), 41 (30). Preparation of (1'-Ethyl-1'-methyl—2jpropynyl)( 1,1-dimethyl- 2-propynyl)amine (13). 3—Chloro-3—methyl—l-pentyne (0.78 mol, 113 g) was reacted with 3-amino-3-methyl-l-butyne (2 mol, 168 g) for 72 h, and the mixture was worked up as previously described. After distil— lation, 79 g (62% yield) of the product was obtained; bp 60-6200 (5 mm); v 3290, 2290, 1380, 1065 cm"1 max ; NMR (00013) 6 1.0 (3 H, t, J= 7), 1.35 (l H, s), 1.55 (9 H, s), 1.58 (2 H, m), 2.3 (2 H, 8); mass spectrum m/e (relative intensity) 163(M+), 68 (100). Preparation of Bis(l-ethyl—l-methyl—Z-propynyl)amine (14). 3-Chloro-3-methy1-1-pentyne (0.91 mol, 123 g) was reacted with 3-amino- 3-methyl-1-pentyne (1.95 mol, 187 g) for 72 h and was was worked up as previously described. After distillation, 89g (55% yield) of product was isolated: bp 50—520C (5mm); vmax 3290, 2960, 2920, 2860, 1510, 36 1375, 1180 cm-1; NMR (CDC13) 6 1.0 (4 H, m), 1.41 (7 H, s), 1.60 (4 H, m),2.25 (2 H, 8); mass spectrum, m/e (relative intensity) 178 (Mf4 1), 148 (42), 82 (48), 68 (100), 53 (39), 41 (35). Preparation of (1—Ethyl-1—methy1-2-propynyl)(1',l'-diethyl- Zfipropynyl)amine (15). 3-Chloro-3-ethyl-1-pentyne (0.75 mol, 98g) was reacted with 3-amino-3-methyl—l—pentyne (1.5 mol, 149 g) for 72 h and was worked up as previously described. After distillation, 79 g (55% yield) of product was isolated: bp 67-680C (5 mm); vmax 3310, 2970, 2880, 1470, 1380, 1170 cm-1 ; NMR (00013) 6 0.95 (10 H, t, J=6), 1.5 (3 H, s), 1.65 (6 H, q, J= 6), 2.2 (2 H, 5); mass spectrum, m/e (relative intensity) 192 (M++ 1), 162 (36), 82 (100), 68 (67), 53 (25), 41 (29). Preparation of Bis(cyclohexylethynyl)amine. 1—Chloro-l-ethynyl- cyclohexane (1.08 mol, 154 g) was reacted with (1- ethynylcyclohexyl)- amine (2.16 mol, 265 g) for 48 h, and the mixture was worked up as previously described. After distillation, 155 g (65% yield) of product was obtained: bp 105-1060C (2 mm); vmax 3290, 2300, 1070 cmfll; mp 71- 7200; NMR (00013) 6 1.55 (13 H, m), 2.0 (8 H, m), 2.35 (2 H, s); mass spectrum, m/e (relative intensity) 230 (M++ l), (M+), 229 (17), 200 (21), 186 (52), 172 (73), 118 (50), 80 (100), 67 (49), 41 (58). Preparation of Bis(l,l-diethylallyl-l,l—diethyl-Z—propynyl)amine $2222. The procedure for preparing this compound was identical with the procedure for preparing the other secondary dipropargylic amines, ex— cept that a 2:1 molar ratio of 3-amino-3-ethyl-1-pentene to 3-chloro— 3-ethy1—l—pentyne was used. Characterization and yield determination was accomplished by GLC: vmax 3290, 3045 cm—1; mass spectrum, m/e (rela— tive intensity) 208 (M++ 1), 188 (5), 178 (50), 84 (100), 82 (85), 67 (13), 55 (44): NMR (CDC13) 0 0.88 (13 H, q, J= 7), 1.4 (8 H, q, J= 7) 37 2.18 (1 H, S), 4.73-6.18 (3 H, m); yield (GLC) 17%. Prgparation of (l,l-Diisopropyl—Z—prgpynyl)(1,1—diethyl—21pro— pynyl)amine. 3-Chloro-3-ethyl—l-pentyne (0.12 mol, 14.6 g) was added to 80 mL of DMF containing 0.16 g of Cu Cl 2 2, 0.16 g of copper bronze, and 3-amino-4-methy1-3—isopropyl-1—pentyne (0.24 mol, 30.2 g) at 00C. The reaction mixture was allowed to react for 1 week at 400 and then warmed to 230C for 24 h. After workup, the mixture was distilled un- der vacuum to collect the high-boiling organic components. GLC analy- sis (10% Carbowax 20M) and collection of the highest boiling peak showed it to be the title compound: vmax 3290, 2950, 2920, 2860, 1455, 1375, 1150, 1060 cm_l; NMR 6 1.0 (19 H, m), 1.8 (6 H, br m), 2.05 (1 H, s), 2.15 (l H,s); mass spectrum, m/e (relative intensity) 223 (M+), 190 (60), 96 (100); yield (GLC) 5%. Hydrogentation of Bis(l,1—diethyl-2-propyny1)amine (16) in Ab- solute Ethanol with 10% Palladium on Charcoal. A 50 mL, round-bot- tomed flask equipped with a magnetic stirring bar, septum inlet, and gas inlet valve was attached with rubber tubing to a mineral oil gas buret. A 10 mg sample of 10% palladium on charcoal (Engelhard Ind- ustries Inc.) and 4 mL of absolute ethanol (Gold Shield U.S.P.) were added. Hydrogen gas (Matheson 99.9%) was flushed through the system and the buret was charged with the same. The solution was cooled to 00C with an ice bath. Them 1 mmol (0.23 mL) of 16 was added to the rapidly stirring solution. Hydrogen uptake was monitored with a gas buret and product formation via GLC (10% Carbowax 20M on Chromosorb W) at 1600C. Hydrogen uptake (74 mL, 2.9 mmol) ceased within 1 h. The GLC trace showed two distinct high-boiling products and a low-boiling product eluting with the solvent. Preparative GLC and subsequent 38 spectral analysis identified the high-boiling components as 3,4—di- methyl-2,2,5,5—tetraethy1-3—pyrroline.19 and 3-methylene-4-methyl-2,2- 5,5-tetraethy1—3—pyrrolidine 29, Repeating the experiment with tridec- ane as an internal standard established the yields of 12_and 29_as 48% and 15%, repectively. Spectral data for 12; vmax 2970, 2925, 2875, 13 2340, 1465, 1420, 1385, 990 cm-1; c NMR (00013, 1 29.79, 8.69, 7.15; H NMR (CDC13) 0 0.80 (13 H, t, J= 6), 1.43 (14 H, SiMe4) 0 134.2, 70.69, s, superimposed on m); mass spectrum, m/e (relative intensity) 209 (M+), 180 (100), 152 (22), 136 (27). Spectral data for 295 vmax 3055, 2069, 2035, 2860, 1650, 1460, 8856m‘1; NMR (00013) 6 0.83 (13 H, t, J= 6), 2.3 (8 H, m), 2.4 (l H, m), 4.63 (2 H, t, J= 3); mass spectrum, m/e (re— lative intensity) 209 (M+), 180 (100). Hydrogenation of Bis(l,l-diethyl-Z-prOpynyl)amine (16) in Ab- solute Ethanol with Platinum Oxide. The same experimental conditions were used as in the palladium—catalyzed procedure, except that 5 mg of PtO2 was substituted for 5 mg of 10% palladium on charcoal. A total of 93 mL (3.8 mmol) of hydrogen was take up. By GLC, only trace amounts of of products (<2%) having the same retention times as 19 and 29_were seen; the rest of the starting material was hydrogenated to 1,1—di- ethyl—l-aminopropane. Hydrogentation of Bis(l,l-diethyl-Z-propynyl)amine (16) to Bis- (l,l-diethylallyl)amine (23) in Ligroin. A 10 mmol sample of l§_(2.05 g) was dissolved in 30 mL of ligroin in a 250 mL centrifuge bottle. Then 20 mg of 10% palladium on charcoal was added. The bottle was placed on a Parr hydrogenation apparatus and hydrogenated for 10 h; the initial H pressure was 50 psi and the pressure dropped 37 psi. The 2 GLC analysis revealed three peaks, one of which was 23} Its spectral 39 properties were identical with the semihydrogenation product obtained by using W4 or W6 Raney nickel in ethanol. Addition of 10 mL of abso- lute ethanol to the ligroin solution and continuation of the hydrogen- ation completely hydrogenolyzed the bis(l,1—diethylallyl)amine. Spec- tral data for the diallyl amine are as follows: vmax 3390, 3045, 1630 cm‘1 ; NMR (CDC13) 5 0.75 (13 H, t, J= 7), 1.43 (8 H, q, J= 7), 4.67—6.0 (6 H, m); mass spectrum, m/e (relative intensity) 209 (M+), 180 (100). Hydrogenation of Bis(l,l—diethyl-Zepropynyl)amine in Absolute Ethanol with W2 Raney Nickel Catalyst. Raney nickel alloy was activa- ted by a literature procedure.3O W2 Raney nickel (2 g) was added to a solution of 40 mL of absoute ethanol and 10 mmol (2.05 g) of 16 in a 500 mL centrifuge bottle. The bottle was placed in a Parr hydrogen- ation apparatus and purged with hydrogen five or six times. The bot- tle was pressurized to 60 psi and the shaker turned on. The pressure dropped 5 psi in 18 h. The solution was filtered to remove the cat- alyst and the ethanol evaporated under reduced pressure. Bulb to bulb distillation [62-64OC (0.2 mm)] gave 0.54 g (20%) of the saturated a- mine_21. GLC analysis (5% 0V-101 Chromosorb W, AW-DMCS treated) showed the sample to be 95% pure; the remaining 5% was bis(l,1-diethy1allyl)- (1,l,1—triethy1carbiny1)amine. The same experiment was performed with W4 and W6 Raney nickel under identical conditions. Tridecane was added as an internal stan- dard in each case. Analysis of the products of each experiment by GLC showed that as the reactivity of the catalyst increased, the degree of hydrogenation of the dipropargyl secondary amine decreased. Thus W2 Raney nickel was the most satisfactory catalyst for the hydrogen 40 ation of l§_to the saturated secondary amine 21, The same experiment was performed under identical conditions, except that 20 mmol (1.12 g) of potassium hydroxide was dissolved in the ethanolic solution 0f.l§ before the W2 Raney nickel catalyst was added. The catalyst was fil- tered and the ethanol was remove under reduced pressure. Water (20 mL) was added to the viscous residue. The solution was transferred to a separatory funnel and the aqueous layer extracted with two 20 mL por- tions of ether. The ether layer was pooled, dried over anhydrous po— tassium carbonate and then evaporated under reduced pressure. About 1.52 g (71% yield) of the saturated amine was obtained (93% pure). The remaining unsaturated secondary amine was separated by spinning- band distillation. All subsequent hydrogenations with the remaining dipropargylamines were carried out under identical conditions with the same prOportions of amine, solvent, catalyst and base and were scaled up to a 50 mmol scale. In the case of bis(cyclohexylethyny1)amine, hydrogenation without KOH gave a higher yield and increased purity of the saturated amine than when KOH was present. The amount of semihydro— genated amine decreased from a maximum of 9% for the most hindered amine 21 to 3% for the least hindered amine 24. Product Analysis of Bis(1,l-diethyl-Zepropyl)amine (21): NMR (CDC13) 0 0.78 (19 H, t, H= 7), 1.4 (12 H, q, J= 6); mass spectrum, m/e (relative intensity) 214 (M++ 1), 184 (8), 86 (100), 57 (32), 56 (10), 43 (14), 41 (23); yield 7.6 g (71%). Anal. Calcd for C14H31N: C, 78.79; H, 14.64; N, 6.56. Found: C, 78.68; H, 14.57; N, 6.60. Product Analysis of Bis(1,l-dimethyl-Z-propyl)amine (24): NMR (CDC13) 6 0.83 (7 H, t, J= 6), 1.06 (12 H, s), 1.13 (4 H, q, J= 6); mass spectrum, m/e (relative intensity) 156 (MHi l), 142 (7), 128 (7), 58 41 (100), 43 (30); yield 6.4 g (80%). Anal. Calcd for C10H23N: C, 76.35; H, 14.74; N, 8.90. Found; C, 76.16; H, 14.57; N, 8.83. Product Analysis of (l'—Ethyl-l'—methy1-2—propyl)(1,1—dime- thyl-Z—propyl)amine Q5):NMR (CDC13) 6 0.80 (10 H, t, J= 6), 1.1 (6 H, s), 1.4 (6 H, q, H= 6); mass spectrum, m/e (relative intensity) 172 (M++ 1), 142 (17), 86 (11), 72 (100), 58 (64), 43 (33); yield 6.7 g (78%). Anal. Calcd for C11H25N: C, 77.12; H, 14.71; N, 18.18. Found: C, 76.84; H, 14.57; N, 8.09. Product Analysis of Bis(1-ethyl-l-methyl-ijropyl)amine (26): NMR (CDC13) 6 0.52 (l H), 0.8 (12 H, t, J= 6), 1.05 (6 H, S), 1.3 (8 H, q,J= 6); mass spectrum m/e (relative intensity) 186 (M++ l), 156 (15), 86 (16), 72 (100), 55 (17), 43 (46); yield 7.0 g (75%). Anal. Calcd for C12H27N: C, 77.76; H, 14.68; N, 7.56. Found: C, 77.70; H, 14.58; N, 7.46. Product Analysis of (1-Ethyl-l-methyl-2-pr0py1)(l',l'-diethylr Z—propyl) amine (27): NMR (CDC13) 6 0.49 (l H, s), 0.78 (15 H, t, J= 6), 1.05 (1H, s), 1.35 (10 H, q, J= 6); mass spectrum, m/e (relative in- tensity) 200 (HI+ 1), 170 (14), 112 (2), 86 (100), 72 (95), 57 (26), 43 (39); yield 7.5 g (75%). Anal. Calcd for C13H29N: C, 78.31; H, 14.66; N, 7.03. Found: C, 78.43; H, 14.54; N, 7.08. Product Analysis of Bis(l—ethylcyclohexyl)amine (28): NMR (CDC13) 6 0.80 (7 H, t, H= 6), 1.40 (24 H, m); mass spectrum m/e (relative in- tensity) 237 (M+), 184 (4), 128 (4), 98 (11), 86 (100), 72 (15), 57 (20) yield 9.5 g (80%). Anal. Calcd for C H N: C, 81.01; H, 13.08; 16 31 N, 5.91. Found: C, 81.11; H, 12.94; N, 5.79. 42 Reaction of Bis(l,l-diethyl-Z-propy1)amine (21) with Methyl Iodide. Methyl iodide (0. 75 mL, 10 mmol) was added to a 50 mL, round- bottomed flask containing a stirring bar, side arm septum, gas inlet valve and 25 mL of THF. Then 10 mmol (2.52 mL) of_21'was injected. The flask was sealed, and the solution was stirred for 2 weeks. The sol- vent was evaporated under reduced pressure and the product analyzed by GLC and NMR spectrosc0py. Analysis showed that no ammonium salt was formed; 21_was recovered quantitatively. Reaction of BFQ°0EtO with Bis(1-ethylcyclohexy1)amine (28). BF3°OEt2 (1 mmol; distilled and stored under argon, 0.125 mL) was added to hexane (1 mL) under argon in a round-bottomed flask equipped with a gas inlet valve, septum side arm, and stir bar. A white air—stable solid was quickly formed. Drying under vacuum overnight gave a white crystalline powder free of ether (by NMR): NMR (CDC13) 6 1.05 (6 H, t, J= 7), 1.6 (20 H, br m), 1.95 (4 H, br q, J= 7), 5.6 (1 H, br 5); mass spectrum, m/e (relative intensity) 237 (M+), remainder identical with spectrum of free amine. Anal. Calcd for C16H31NBF3: C, 62.96; H, 10.23; N, 5.69; B, 3.54: F, 18.67. Found: C, 61.11; H, 10.34; N, 4.39; B, 4.83; F, 19.22. Reaction of BFgoOEtgand BFgGas with Bis(l,l—diethyl-Z—pro— py1)amine (21). BF3°0Et2 (1 mmol, 1.125 mL) and 21 (1 mmol, 0.25 mL) were injected into a pyrolysis tube, and the glass tube was flame sealed. The neat solution reacted for 1 h at 250C. The tube was opened and the liquid removed. The remaining air—stable solid was an- alyzed by NMR and mass spectroscopy. It was identified as the amin— olysis product (CH3CH2)3CNH2BF3: NMR (CD013) 6 0.93 (6 H, q, J= 6), + 5.76 (2 H, s, D 0 exchangeable); mass Spectrum, m/e 184 (M ). 2 43 Repeating the same reaction but with BF gas (24 mL) gave a liquid 3 fraction as well as the BF3 amine complex. The liquid was identified by NMR and mass spectrometry as 3-ethyl-2—pentene. Titration of Amine Hydrochlorides in 90% Ethanol with 0.1075 N KOH under Argon. The procedure for the titration reactions were taken from a previously reported procedure.15 Potassium hydroxide was pre- pared by dissolving potassium metal in absolute ethanol under argon and diluting with degassed H O to the required concentration. All sol- 2 utions were stored in polypropylene containers under argon. All tran- sferring of solutions was done via cannula, and the titrations were done under argon. The amines were converted to the hydrochlorides by stoichiometric titration with standardized aqueous HCl (0.0954 N). Wa- ter was removed from the amine hydrochlorides by drying for 1 week un- der high vacuum in a desiccator over phosphorus pentoxide. The titra- tion was followed by using an Orion l601-A Digital Ionalizer with a Markson combination pH reference electrode at 20°C. All of the titra- tions were carried out so that the solution was 0.01 N at the equiva- lence point. The pKa was calculated as the pH of the solution at half the equivalence point volume. All titrations were carried out three times, and the theoretical equivalence point was within 0.25% of the experimental value. In each case, only a single inflection point was observed. For a discussion on the effect of ethanol on the ap- parent strength of organic amine bases and for a method frequently used for the extrapolation of the pKa from solutions which are progres- 16a sively less alcoholic, see the paper by Hall et al. 44 Melting Points of Saturated Amine Hydrochlorides. The melting points are as follows: diisopropylamine'HCl= 214-2160C, tetramethyl- piperidine‘HCl= >270°c, bis(1,l—dimethyl-Z-propyl)amine:HC1= 196-197°c, (1'-ethyl-1'-methyl-2-propyl)(l,l-dimethyl-Z-propyl)amine-HC1= 158-1600C, bis( 1-ethy1-l-methyl-2—propyl)amine°HC1= 142-1440C, (l-ethyl-l—me- thyl—2-propyl)(l',1'-diethyl-2-propyl)amine~HC1= 144-14500, bis(l-ethyl- cyclohexyl)amine°HCl= 190-19200 and bis(1,l-diethyl—Z-propyl)amine'HCl = >2600C. Preparation of Bis(1,l-dimethyl—Z—propyl)chloramine. The fol— lowing procedure for the synthesis of the chloramine adduct of 24 is representative of the n—chlorosuccinamide procedure. To a 25 mL round—bottomed flask were added 6 mmol of N-chlorosuccinamide (0.67 g) and 10 mL of ether. The heterogeneous mixture was stirred for 10 min at 220C and 5 mmol of 24 (1.01 mL) was added to the mixure. Monitoring by NMR showed that the reaction was completed in 4.5 h. The succin- amide was filtered from the solution and the ether layer washed 3 times with 3 mL aliquots of water. The ether was dried with anhydrous sodium sulfate and removed under reduced pressure. Preparation of Bis(1,l—dimethyl-Z—propyl)chloramine with Chlor— ine and Sodium Hydroxide. This procedure for forming the chloramine of _24 is representative for forming chloramines of 21_and 26, A 25 mL Bantamware round—bottomed flask was fitted with a Teflon stirring bar, septum side arm, and a gas inlet valve. The flask was charged with 4 mL of hexane, 2 mL of 1.18 N NaOH and 2 mmol of 24 (0.40 mL). The flask was cooled to 00C and 55 mL of chlorine gas (2.2 mmol) was added by sy- ringe. The reaction was complete within 10 min. The solvent was re— moved under reduced pressure and the chloramine was added to an NMR 45 tube with an equal molar amount of benzene. The yield of chloramine 24 was 95% by NMR. There were no signals at 6 1.07 or 1.3 (relative to TMS), corresponding to the methyl and methylene protons of the free amine 24, Spectral data; vmax no N-H bend at 1500cm-1 ; NMR (CC14) 6 0.89 (6 H, t, J= 6), 1.27 (12 H, s), 1.57 (4 H, q, J= 6). Preparation of Bis(l-ethyl-l-methyl-Zepropyl)chloramine Using Chlorine and Sodium Hydroxide. Hexane (10 mL), 5 mL of 1.18 N NaOH and 1.12 mL of 26_ (5 mmol) were added to a round-bottomed flask as pre- viously described. The yield by NMR was 93%. Spectral data; vmax no N-H bend at 15106m'1; NMR (0014) 6 0.93 (12 H, c, J=6), 1.32 (6 H, s), 1.63 (8 H, q, J= 6). Preparation of Bis(l,l-diethyl-Z-propyl)chloramine using Chlorine and Sodium Hydroxide. Carbon tetrachloride (2 mL), standard NaOH (2 mL, 1.18 N) and 2 mmol of 21 (0.51 mL) were added to a round bottomed flask as previously described. Then 55 mL of chlorine gas was added by syringe to the stirring solution at 0°C. The reaction was complete within 30 min. The NMR spectrum could not be taken in chloroform be- cause a chemical reaction occured in which the NMR tube became hot and the reaction mixture turned red. The NMR spectrum showed that the chloramine decomposed within 1 h at room temperature. Spectral data; NMR (CC14) 6 0.9 (18 H, t, J= 6), 1.77 (12 H, q, J= 6); vmax no N-H bend at 1510cm-1. Attempted Chlorination of Pentane in 90% Trifluoroacetic Acid: 10% Sulfuric Acid with Bis(l-ethyl-l-methyl-Zepropyl)chloramine at 932;. Chloramine 26 (1.1 g, 3.5 mmol) was added to a quartz flask un- der subdued light containing 25 mL of 90:10 (v/v) mixture of trifluo- roacetic acid-sulfuric acid under argon. The flask's contents immedi- ately turned red. Then a five fold molar excess of pentane (3.4 mL) was 46 added to the mixture by syringe. The flask was irradiated by a 300W sunlamp placed 18" from the cooling bath ( 50C) immersed reaction mix- ture. The flask was irradiated for 211until a test for positive C1 became negative (formation of I with aqueous KI). The contents of the 2 flask darkened to a soot brown during the irradiation. The products were isolated by pouring the reaction flask's contents into 100 mL of chopped ice. The organic layer ( 0.5 mL) was washed with water and analyzed by GLC using a 50' x 1/8” 20% SE-30 Chromosorb W column. No products corresponding to either 1, 2 or 3-chloropentane were observed. The organic layer consisted of pentane and high boiling organic compo- nents (> 270°C) Identical results were obtained for chloramines corresponding to 223 233 and 22, No chloropentanes were obtained when FeCl3 was sub- stituted as the free radical initiator in each of the previously de— scribed reactions. CHAPTER II THE SYNTHESIS AND STABILITY OF HINDERED SECONDARY LITHIUM AMIDES 47 48 INTRODUCTION Lithium dialkylamide bases have achieved widespread appli— cations in organic synthesis during the last decade. Since the early investigations by Levine 31 in which lithium and sodium diiSOprOpyl- amide (LDA and NaDA) were used to metallate picoline, hindered lith- ium amides have enjoyed widespread application as powerful nonucleo- philic proton selective bases. Their most common application has been as bases for proton removal from weakly acidic compounds. Ketones and esters33 are completely and rapidly deprotonated by lithium amide bases in ethereal solvents to give the corresponding enolate anions (eqs 44 and 45). 0 on 0s12 Li ,, 0 (>6 . 6 6 I I THF, -78°c (93%) -Li ' I I 3L1 CH3ECH3 01130-012 ; 0112:: -OR ____, szc-OR (45) THF, -78°C R'X R Hindered lithium amides are very useful for the regioselective generation of ketone enolates. By adjusting the conditions under which a ketone enolate mixture is formed, it is possible to establish 49 either kinetic or thermodynamic control. Kinetic control will be ob- served when the enolates, once formed, are interconverted only slowly. This situation is seen when very strong bases such as hindered lithium amides are used in an aprotic solvent in the absence of excess ketone. The small lithium cation is tightly coordinated to the oxygen atom of the enolate anion and this tends to decrease the rate of proton ex- change reactions (eq 46). o<-> 3") 11 B<-> gm R2= -CH2R % K RZCHC-CHZR ,, RZCH-C=CHR (46) a Kb 2. i The conditions for kinetic control usually favor the less substituted enolate_§, probably because removal of the less hindered hydrogen is more rapid, for steric reasons, than removal of the more hindered pro- ton. On the other hand, at equilibrium, it is the more substituted enolate 5 that is usually the dominant specie. This is because the stability of the carbon-carbon double bond increases with increased substitution. For weakly acidic compounds such as ketones, lithium amide bases are well suited to kinetic enolate formation because of two principle features. They are very strong bases (pKa W 35-38) which completely and rapidly enolize relatively weakly acidic compounds. A consistant relationship is found in that kinetic control in enolate formation usu- ally favors the less substituted regioisomer. For relatively acidic compounds such as diketones or B-ketoesters, secondary or tertiary alkoxide bases are sufficiently basic (pKa W 16-19) for quantitative enolate formation. The resulting carbanion is more stable than the 50 isomeric anion in which only one of the carbonyl substituent can de- localize the negative charge (eq 47). O O — O O O H (CH3)3CO( : H I II C 47 CH3CCH2C-OR , CH3CCH- -OR. ‘——;(CH2CCH2 -OR ( ) (CH3)3COH (-) thermodynamic very enolate minor Besides showing increased regioselectivity in ketone enolate formation, hindered lithium amides apparently are thermodynamically stronger bases than less hindered amide bases. Presently, 2,2,6,6— tetramethylpiperidine is the most hindered secondary amine commercially available. The corresponding lithium amide has found applications in organic synthesis which other less powerful amide bases have been un- able to fulfill. Rathke and Kow34 report that lithium tetramethylpip- eridine (LiTMP) or lithium tert-butylneopentylamine can remove the a- proton from an organoborane to generate carbanions (eq 48). _ DCH -B CH3 LiCH - 2 L'TMP 1 ‘ —L—> (48) N O O N Other applications of LiTMP, where less basic secondary lithiums fail to react,include the synthesis of cyclopropyl ethers by reaction of 35 excess alkene with chloromethyl ethers and LiTMP. The base removes HCl from the ethers with formation of ROCH: (eq 49). 51 0C H LiTMP 2 5 + C H OCH Cl \ 7- 5 2 -230—3 2306 (49) 55% Even weakly acidic isocyanides that are not metallated by n—butylli- thium can be metallated in the a-—position by LiTMP.36 The reaction has been used as the first step in a synthesis of 2—oxazolines (eq 50). 1 1 1. R3R4C=O R\ LiTMP R (-) 2. ROH ¢\ lCHN=Cz ————9 C—N=C: > N (50) R2 R2/ R1 R4 R R3 Olofson has shown that as lithium amide bases become more hindered, they become more effective bases for the synthesis of arylcyclopro- panes 37(eq 51). LiNR2 + ArCHZCl + ———> .} a R = Et 19% R = i-propyl 39% R = (i-pro)(cy) 41% (51) R = (cy)2 45% R = [(CH3)2CCH2—]2CH2 54% 52 Finally, the hindered amide LiTMP is a powerful enough base to convert THF to the enolate anion of acetaldehyde as well as to convert bromobenzene to benzyne. These two components react ip_situ to form anthracene38 (eq 52a—c). O I ______) -—)O CH2=C—H + CH2=CH (52a) (0) [O)\Ll A,70 2 Br @ + LiTMP ——‘> .1 + LiBr (52b) O(-)Li .I + CH =6-H —> 3 f» 1 ———> 2 /, OLi 1t .1 CH2 -——-> o + o —> —> CH 1 O(—)Li — © © © 66% overall yield 53 Because hindered secondary amides such as LiTMP exhibit some unique metallating properties, we undertook to synthesize the sodium and lithium amides of the very hindered amines in Chapter I. We will describe attempts at metallating the amines with various alkyl and aryl metal reagents. We will describe the successful lithiation of all the hindered amines in Table I by chelate assisted metallation with n- butyllithium. Factors will be described which effect the rate of amine metallation. Finally, the stability of hindered lithium amides in THF and diethyl ether will be investigated, along with the study of the re- lative stability of each amide in Table I. An explanation for the ob- served results will be suggested. RESULTS Initial attempts to metallate the series of hindered amines de- scribed in Chapter I focused on n—alkyl and aryl lithium metallation of the least hindered amine in the series, compound_24. We were completely unsuccessful in metallating_24 under a variety of conditions. Two cri- teria were used to verify metallation of 24, They are quantitative bu- tane evolution as measured by manometry and quantitative B—hydroxyester formation by ester enolate condensation with cyclohexanone (eq 53,54). 7H3 7H3 24°C (1H3 7H3 CH3CHzC-NH-CCH2CH3 + n—BuLi 4%; CH3CHzCa——¥t——?CHZCH3 CH3 CH3 hexane/pentane CHBALl CH3 saturated with (53) 24 butane + butane 54 O O I 24—Li + CH é—O(CH ) THF 1 CH g—O(CH.) 1. Cyclohexanone > -—' 3 3 3 -—78 C I 2 3 3 + Li 2. H30 (54) HO CH C-O(CH 2 3)3 44 Methyllithium in ether (2.1 M) was diluted with anhydrous ether to a 1.0 M solution. One equivalent of 24_was added and gas evolution was monitored over 7 h. Initially, about 40 mL of gas evolved within the first hour with a total of 60 mL evolving over 7 h. At STP one mmol of gas occupies about 25 mL, therefore the calculated gas displacement for the 3 mmole reaction is 75 mL. Apparently, the reaction was 85% complete within 7 h. By NMR, no methyllithium remained in the ether so- lution (6 = -1.7 using TMS as the reference standard). After removing the ether solvent under reduced pressure and treating the reaction mix- ture with Eggpfbutyl acetate, cyclohexanone and aqueous acid, we ob— served only an 8% yield of 44. A control reaction using LiTMP made from TMP and methyllithium gave a 97% yield of 44 by GLC analysis. The same reaction sequence was repeated with n-butyllithium and phenyllithium substituted for methyllithium. Very slow gas evolution from the butane saturated reaction mixture (20 mL butane in 8 h) was observed with n-butyllithium. Quenching the n-butyllithium-amine 55 mixture with Efbutyl acetate and cyclohexanone gave a 10% yield of the B-hydroxy ester 44. The same reaction sequence with phenyllithium in benzene gave, after 10 h, 9% of 34, Since alkyl and aryllithium reagents apparently metallate 23 poorly, we decided to use the more powerful metallating agent phenyl— sodium. Phenylsodium was synthesized from chlorobenzene and sodium me- tal dispersion. The reaction of 2fi_with phenylsodium in hexane for 12 h gave, after quenching with tert-butyl acetate and cyclohexanone, a 4% yield of_44. Thus phenylsodium is apparently as poor a metallating agent for 2fi_as are the previously mentioned alkyllithium reagents. Phenylsodium is insoluble in hexane, and the addition of 24_did not bring it into solution. There was no change in the appearance of the heterogeneous reaction mixture during the course of the reaction. The next course of action was to return to the soluble alkyl li- thium reagents and repeat the reactions with_24 using the amine ligand chelating agent N, N, N', N'—tetramethylethylenediamine (TMEDA) to assist in the metallation of 24, Ligands like TMEDA coordinate to the lithium ion of alkyl lithium reagents and break up the organolithium aggregates that are found in solution. The resulting monomeric alkyllithiums are generally more reactive metallating agents than the uncomplexed aggre— gates. TMEDA was added to n-butyllithium (1:1 mol equivalent), and enough hexane was added to make a one molar solution, which was then sat- urated with butane. Upon injection of 3 mmol of 24 into the reaction mixture, (thermostated at 22.000) there was an immediate and quantitative evolution of butane (74 mL), as measured manometrically. The hexane solvent was removed under reduced pressure and replaced with THF. The 56 flask was immediately cooled to -780C and quenched with tert-butyl ace— tate and cyclohexanone, as previously described, to give a 97% yield of '44. This was taken as confirmation that nearly quantitative metallation of 24 had been achieved. A control reaction with 23, TMEDA and n-butyl- lithium was performed to establish that neither the lithium amide nor n—butyllithium was metallating TMEDA. The reaction mixture was quenched with deuterium oxide and examined with proton NMR. Product analysis showed that the TMEDA had not been metallated. The same reaction sequence was repeated with diisopropylamine, TMP, 22, 22, 26 and 22, The progress of the metallation reaction was followed by monitoring butane evolution with a mercury filled manometer. Metallation of the hindered amines was confirmed by the quantitative formation of 44_obtained by quenching each amide solution with tert- butyl acetate and cyclohexanone in either THF or diethyl ether. At least a 96% yield of 44_was obtained with each amide solution studied. Table IV lists the data for these metallation reactions. The reactions were run at 1 molar concentration in n-butyllithium and hindered amine, using 1.0 and 0.5 mol equivalents of TMEDA. The data is based on the volume of butane evolved with respect to time. The numbers T and T repre- 50 90 sents 50% and 90% evolution of butane, respectively. 57 Table IV. Amide Formation in Hexane (T and T90) with n—Butyllithium 50 R NH n—butyllithium, TMEDA >_ R NLi + butane 2 hexane, 22.00C 2 100% TMEDA T50 T90 LDA <1 min <1 min LiTMP <1 min <1 min gégLi <1 min <1 min 227Li <1 min <1 min IgéfLi 1.5 min 4 min 2ZfLi 10 min 25 min 227Li 3 h 20 h 50% TMEDA T50_ T90 LDA <1 min <1 min LiTMP <1 min <1 min .247Li <1 min <1 min 2§fLi <1 min 1 min 2§fLi 10 min 35 min 2ZfLi 65 min 265 min 1—Li 7 h 40 h 58 Stability of Hindered Lithium Amides in THF and Diethyl Ether. The stability of these hindered lithium amides in THF and diethyl ether solution was examined. The hindered lithium amides were syn- thesized using n-butyllithium and one equivalent of TMEDA. The hexane solvent was removed under vacuum and the remaining amide-TMEDA complex was dissolved in either THF or diethyl ether at 24°C. At least six re— actions for each amide were performed in THF and diethyl ether. The ethereal solutions were quenched with tert-butyl acetate and cyclohexa— none at time intervals rangeing form 15 min to 13 h. A steady, contin- uous decrease in the yield of 4&_with time was observed, reflecting a slow decomposition of the hindered amide. The reaction products formed by attack of the lithium amides on the ether solvents are apparently in- capable of enolizing the tert—butyl ester. The yield of 4&_obtained in the quenching experiments provides a method for monitoring how fast each amide attacks THF and diethyl ether. Figure 1 shows the rate of decomposition of the metallated hin- dered amides in THF. Figure 2 shows the decomposition rate of the same metallated amides in diethyl ether. Amine 22_was not included in the study because the amine peak appeared in the GLC trace at the same position as the B-hydroxy ester 33, All reactions were carried out us- ing a one molar concentration of amide in ethereal solution at 24°C. In THF, amides 24fLi and 2§fLi appear to be the least stable of the six studied, whereas in diethyl ether, LiTMP is by far the least stable amide. Apparently, LiTMP attacks diethyl ether just about as fast whether or not TMEDA is present. In THF, LiTMP, LDA and 27-Li are fairly stable (T50> 6 h). In diethyl ether solution, LDA and lithium 100 90 8o \1‘ \‘I‘ 1. 70 OJ 4.) U) [121 g 60 X 0 1... '2, a, 50 e'. 8 1. 40 0) > O U 32 30 o\° 20 10 Figure 1. Time (Hours) Lithium Amide Stability in THF at 240C. Recovered 44 After Quenching by tert-Butyl Acetate and Cyclohexanone at -780C. All Amides Made with TMEDA (1 equivalent). 100 90 80 70 60 50 4O 30 % Recovered B-Hydroxy Ester 44 20 10 Figure 60 \ '\. * \ £ 0 O\D \o d \ \ \ \D . = LDA a £5== LiTMP \ v = LiTMP A no TMEDA v O = _2_4_—Li a = .2541 ° 1% = 2244 V 27 Li 2 \ 4\ O + _.. v\ A - v \A \V 1 I f ‘ ' I ' ‘ ‘ir ' V " j 'fi f f ' j ' I ' V ‘ V 0 1 2 3 4 5 6 7 8 9 10 ll 12 13 Time (Hours) Lithium Amide Stability in Ether at 24°C. Recovered 33 After guenching by tert-Butyl Acetate and Cyclohexanone at -78 C. All Amides Made with TMEDA (1 equivalent). 61 amides of 24, 2§,_2§ and 22 are all quite stable (T50> 12 h). Comparing the stability of all the amides in both THF and ether, one observes that,except for LiTMP, all of the amides studied are more stable in ether than in THF. A discussion of the amide formation re- actions and the amide stabilities in ethereal solvents is presented in the next section. DISCUSSION Metallation of simple primary and secondary aliphatic amines is a well established experimental procedure. Lithium and sodium amide have both been made by lithium and sodium reduction of liquid ammonia.1 Primary amines have been metallated by sodium and lithium alkyl reagents as well as by potassium hydride.40 However, the most useful metal amides are the sterically hindered secondary lithium amides. They are easily made by reacting commercially available amines such as diiSOpropylamine and tetramethylpiperidine with hydrocarbon-soluble alkyl lithium reagents. These hindered amides achieved a prominent rele in organic synthesis because they are very basic (pKam 35-38) non-nucleophilic organic bases. The goal of this study was to devise a procedure for the metal- lation of hindered secondary amines and to determine whether the cor- responding amides exhibited distinctive chemical properties related to their steric bulk. Neither phenyl or n—butyllithium metallated amine 2§_within a per- iod of time sufficient for less hindered amines. The criteria used to judge whether an amine was metallated were the enolization of tert-butyl acetate by the amide to give, after condensation with cyclohexanone, the hydroxy ester 33. None of the two lithium reagents mentioned above 62 gave any 44 when directly reacted with tert-butyl acetate and cyclohex- anone. Alkyl lithium reagents are powerful nucleophiles which attack esters, ketones and other carbonyl compounds at the electron deficient carbon more rapidly than they abstract a proton from the relatively a- cidic a-carbon atom. Control reactions with LDA and LiTMP showed that an amide base was required to enolize the tert—butyl ester. The ester enolate subsequently reacted with cyclohexanone in an aldol type reac- tion to give, after quenching with aqueous acid, the B—hydroxy ester 44, Phenylsodium was equally ineffective in metallating_24, probably because the metallating reagent was completely insoluble in the hexane solvent. Methyllithium is apparently metallating_24, based on methane evolution and the loss of the methyllithium proton NMR signal. However, only low yields of 44_were obtained in quenching experiments. This dichotomy can be explained when one looks at the stability of 24§Li in ether. The T50 for 24fLi in ether is about 10 h. Apparently, the metallated amine is attacking the ether solvent. This would explain the low yields of 44 in the quenching reactions with cyclohexanone and tert-butyl acetate. Based on this observation, methyllithium in ether is not the metallating reagent of choice for forming very basic hindered alkyl lithium amides. The formulae of alkyl lithiums are often written as "RLi", but this representation is not accurate. Alkyl lithium reagents are highly associated in solid and solution phase.42 In common ether solvents, there is evidence that tetramers solvated by ether molecules are domin- ant. Hindered alkyl lithiums such at t—butyllithium, are often more re— active metallating agents than simple alkyl lithiums. This increased reactivity has been correlated with a less aggregated lithium species. 63 We believe that with very hindered secondary amines, alkyl li— thium aggregates are too large to deprotonate the nitrogen atom (eq 55). CH 1 3 CH3-2rCH2CH3 NH 7‘ No reaction (55) CH C CH CH 3 ‘ 2 3 CH3 (RLi)4_6 + With less hindered amines, such as diisopropylamine and tetramethylpip- ridine, two factors contribute to metallation by n-butyllithium. First, the less hindered amine nitrogen is not as crowded by the surrounding alkyl groups as in the case of more hindered amines. This permits a less obstructed approach to the amine proton by the n-butyllithium mo— lecule. More important, however, the less hindered nitrogen atom can act as a Lewis base which can coordinate with lithium atoms of the alkyl lithium aggregate. This Lewis base might therefore be expected to break down the aggregate and, as a result, increase the effective carbanion character of the alkyl lithium reagent. The increased car— banion character would manifest itself by increasing the rate of amine H ¢fi'K) (RLi)1n + <»—2NH ———————) R-LilR 2 __—> N ski, 4,... .. L...<<) 2 metallation (eq 56). (56) 64 Amine_24 as well as the other amines in Table I are probably too hindered to act as effective Lewis bases to de-aggregate n-butyllithium. An example of the poor Lewis base character of 24 was described in Chap- ter I, where (H112 failed to form a quaternary ammonium salt with 24, 18 3 while TMP readily reacted with CH3I. In order to de-aggregate n-butyllithium, the chelating ligand TMEDA was added to the hexane— n-butyllithium solution. There was an immediate warming of the reaction mixture along with a very rapid quan- titative evolution of butane gas when TMEDA was added (eq 57). CH CH \3/ 3 N _ +Ku\CH (R-Li) + (CH ) NCH CH N(CH ) -——————4> R Li . 2 6 3 2 2 2 3 2 N. H J 2 .31. 3 3 °H3 (57) CH3CHzC-CH3 + 24-———§ NLl + RH CH3CHzC-CH3 CH3 It appears that as the amide bases become more hindered, the rate of metallation decreases. A pronounced slowing in the amine metal- lation reaction occurs with amines 2Z_and 22_when 1.0 equivalent of TMEDA is used. The two most hindered amines 2Z'and.2l require 0.5 and 24 h for complete conversion to the corresponding amides, respectively. All of the other amines are metallated within 4 min at 240C with 1.0 equivalent of TMEDA. There is a substantial decrease in the metalla- tion rate for 24, 21, and 22 when 0.5 equivalents of TMEDA is used; 0.75, 5 and 40 h are needed, respectively, to completely metallate the amine. 65 Steric considerations in amine metallation do not seem to play an important role until the very hindered amines 2Z_and 22_are used. With 1.0 equivalent of TMEDA, 22 is metallated 500 times more slowly than any other amine except_22. The additional methyl group on_2l com— pared to 2Z_dramatically increased the amine's resistance to metalla- tion by n-butyllithium. This is probably due to the increased steric congestion presented by the extra methyl group. The rate of metallation of the hindered amines is reduced when 0.5 instead of 1.0 equivalent of TMEDA is used. This is expected based on the fact that 0.5 equivalent of TMEDA would not completely dissoci— ate the alkyl lithium aggregates in solution. Thus the effective size of the alkyl lilifilmlreagent is increased while the effective carbanion character of the anion is reduced. Another factor which might effect the metallation rate of the very hindered amines is the thermodynamic basicity of the hindered amides. If the thermodynamic base strength of hindered amides increa- ses with increasing steric bulk, the metallation of 21 and 22_might be equilibrium controlled (eq 58). k2 RZNH + n-BuLi ‘ """"""""" > RzN-Li + butane (58) k 1 + TMEDA Though this proposal may seem unlikely, there is some evidence to sup— port the proposition that very hindered amides are more basic than less hindered amides. Fataftah18 showed that the lithium amide of 2§.metal— lates toluene (pKa= 41). Under the same conditions, LDA and LiTMP did not metallate toluene ( eq 59). 66 CH3 CH2 (SiE) CH(SiE)2 NLi ESiCl + a © © + © (59) 2 14% 7 \r N The major reason for the decrease in the metallation rate of very hin- dered amides is probably the increased steric repulsion between the hindered amine and the alkyl lithium reagent. The next Section will discuss the stability of lithium amides in THF and diethyl ether. Lithium Amide Stability in THF and Diephyl Ether. The results of the amide stability study provided interesting and unexpected results. The lithium amides show considerably different sta- bilities in THF and diethyl ether. All of the lithium amides in Table I are more stable in diethyl ether than they are in THF, except for LiTMP (Figure II). Except for LiTMP and 247Li, all of the hindered a- mides that were studied have a T 0 greater than 18 h at 240C. Both 5 LiTMP and 247Li are relatively unstable with a T of 3.5 and 11 h, re- 50 spectively. Suprisingly, amide 2§fLi is much more stable than amide 247Li in diethyl ether while in THF, amides 24fLi and 2§fLi are nearly identical in stability. Amides 2§fLi, 2§fLi and 227Li are all much more stable in diethyl ether (T50> 18 h) than in THF. As the substitution near nitrogen increases and the amides become more hindered, a large 67 relative decrease in their'stability:h1'FHF is seen. LiTMP's half life is 8.5 h, while the amides 247Li and 2§fLi are half destroyed within 1.5 and 1.7 h, respectively. But as the steric hindrance of the amides be- comes even greater, their relative stability improves. The T50 of 2§fLi and 2ZfLi in THF increase to 2.8 and 6 h, respectively. The stability of the lithium amide of 22_was not studied due to analytical problems. The observation that the stability of progressively more hindered amides increase, then decrease was unexpected. If amide base strength increa— ses with the steric hindrance of the amide, the most hindered amides in the series would be expected to be most basic. One could infer from the data in Tables I and II that the decom- position of the alkyl lithium amides is probably caused by the depro- tonation of the ether solvents. Figure I and II also suggest that the rate of ether deprotonation is directly proportional to the kinetic ba- sicity of the lithium amides. For the less hindered amides, kinetic base strength apparently increases with with increasing amide bulk. This is reflected by the drop in amide stability in THF for LDA through .227Li. Amides 2ZfLi and_2l-Li are apparently so hindered that they can— not deprotonate THF as quickly as the moderately hindered amides, re— flecting a decrease in the kinetic base strength of these amides. The generally greater stability observed for hindered amides in diethyl ether is not unexpected. Organolithium compounds are gener- ally more stable in diethyl ether than in THF because THF is readily me— tallated to give ethylene and the stabilized acetaldehyde enolate anion (eq 60). OLi ::_j: I 68 Diethyl ether is less polar than THF. The more polar the aprotic sol- vent, the better it is at solvating the small lithium counterion. This increased solvation of lithium generally increases the kinetic basicity of the anion base. If THF solvates lithium amides better than diethyl ether, one would expect to see an apparent increase in the kinetic bas— icity of hindered lithium amides in THF. Only when the amides become too bulky for THF to solvate the lithium ion would a drop in the kinetic basicity be expected. This rationalization would explain the results observed with 2pri and 22fLi in THF. Amide ion stability in ether solvents might also be influenced by TMEDA. Some TMEDA remains with the amides after the amine is metallated and the hexane solvent is removed. A metallation reaction of diethyl ether by LiTMP was performed without TMEDA. The amide decomposed some- what faster without than with TMEDA. Studies with more hindered amides were impossible without TMEDA, since TMEDA was necessary to help metal- late the hindered amines. No conclusions on the effect of TMEDA on the amide decomposition rate can be drawn until a reliable measure for TMEDA content in amide solution is obtained. This might be accomplished using NMR by noting the relative peak ratios for the hindered amide and TMEDA. EXPERIMENTAL All solvents and reagents were reagent grade quality. All alkyl lithium reagents were purchased from Aldrich and standardized before use by titrating with sec-butyl alcohol using 1,10—phenanthroline as an indicator.39 Diethyl ether and THF were distilled from lithium aluminum hydride and stored under argon. All of the hindered amines synthesized in Chapter I were at least 98% pure and stored over molecular sieve. 69 Diethyl and diisopropylamine, tetramethylethylenediamine and tetra- methylpiperidine were distilled from calcium hydride and stored over molecular sieve. Cyclohexanone and tert-butyl acetate were purchased from Aldrich and used without purification. Analysis of hydroxy ester 44_ was performed by GLC at 1200 using a 6'x 1/4" column with 5% OV—101 on Chromosorb W, AW—DMCS treated support. Reaction of Methyllithium with Bis(l,l-dimethyl-ijropyl)amine (24). A 10 mL flame dried round-bottomed flask was fitted with a rubber septum sidearm and Teflon stirring bar. The flask was connected to a mercury manometer by a ground glass joint and the entire system flushed with argon. The flask was immersed in a thermostated water bath at 240C. A magnetic stirrer was placed beneath the bath and the manometer purged with methane. Then 1.8 mL (2 mmol) of 2.1 M methyllithium was injected into the reaction flask by syringe. Then 2 mmol of 24 was added by Syringe (0-40 mL), the manometer readings were noted per— iodically and the manometer reservoir was adjusted as needed. After 14 h, the readings were off scale, indicating evolution of more than the theoretical 50 mL of methane gas. Other gaseous products had been evolved. An aliquot of sample was removed by syringe and a proton NMR was taken. Benzene (d6) was added as a reference standard ( 6= 7.25). No methyllithium was seen by NMR ( + \ B + RZCHCXR2 —7 BH + R2CCXR2 62 <-) (-> ( ) RZCCXR2 7' R2C=CR2 + X Criteria which permit differentiation among E2 and the two ElCB me chanisms [ a) where the first step is rate limiting and the second is fast, b) the first step is a rapidly attained equilibrium and the sec- cond is the rate determining step] have been summarized.5’46 Many elim— ination reactions have been examined and the ElCB mechanism has been shown to operate in only a few cases. l,l,l—Trifluoro-Z,2-dihalogenoe— thanes undergo alkoxide-catalyzed hydrogen - deuterium exchange much faster than dehydrofluorination.In these cases, the electron-withdraw- ing effect of the halogens and the poor leaving ability of fluoride are important factors favoring carbanion formation over halide elimination.45 mechanism has been obtained recently 45 for the elimination reaction of phenyl ethers like MeSOCHZCHZOPh and Strong evidence for an ElCB MeZS+CH2CHZOPhI-. These compounds contain a poor leaving group (-OPh) 78 and a substituent capable of stabilizing a carbanion. Proof of the character of the elimination mechanism for the ElCB reaction is obtained by measurement of the elimination rates of substrates labelled in the B-position with deuterium. For an E2 elimination reaction which passes through a symmetrical transition-state, the theoretical maximum (about 7) for the kinetic deuterium isotope effect kH/kD is predicted. As the character of the transition state becomes increasingly carbanionic, the isotope effect should decrease to 1.0 in the limiting ElCB mechanism, since any isotope label would be exchanged rapidly with the medium. The isotope effects kH/kD in the dehydrofluorination of CF CHCl and 3 2 CF CHBrCl are 1.26 and 1.41, respectively.45 3 The E2 elimination mechanism is by far the most commonly observed reaction pathway. During the past decade, considerable effort hasbeen directed towards determination of the transition state structure for E2 reactions and the effects exerted upon this structure by the nature of the reactants and the reaction conditions. There is a close relation- ship between the transition state structure of an E2 reaction and the re— action products especially with regard to positional and stereochemical orientation. Generally, E2 reactions proceed most readily with the leaving groups in an antiperiplanar conformation. The_§p£if elimination is favored by the need to minimize the repulsion energy between the mi- grating electron pairs. Numerous examples are available illustrating the preference for gppifeliminations. Both acyclic and alicyclic compounds show an gpplfpreference in base promoted elimination reactions (eq 63- 66).52 79 - H Ph Et0( ) 4;; \\ =C/ (63) I Ph/C \M Ph Me 8 N(Me)3 H Ph - EtO( ) Ph\ l/Ph 7‘ ,c=c\ (64) Me H Ph e OTs i-Pr (-) EtO e (65> Me (100%) H i-Pr i-Pr -p _ IV . ...<> ‘ ’77 + 66 Me H ( ) H H C]- Me (75%) Me (257°) Although anti stereochemistry for elimination reactions is pre- fered, there are quite a few examples where the eliminated groups are gyp_and coplanar. For some substrates such as trans-Z-phenylcyclopentyl tosylate, the elimination reaction with potassium tert-butoxide in tert 80 butanol is only 10 times slower than elimination of the corresponding cis isomer (eq 67). H H (CH3)30K ——————>./Ph< (CH33)OK -' Ph (CH3) 30H (CH3 ) 30H 2‘21: (67) OTs k1 cis —-—> anti k1= 0.1 k2 syn <——— trans In some cases, 5 n—elimination may be favored over gppifelimination. One example involves substrates in which trans leaving groups have a di- hedral angle of 1200 or 1500 due to restricted rotation in a rigid ring system, but_pi§ groups have a 00 angle. Thus gypf elimination from trans 2,3-dihalogenonorbornane is faster than gppifelimination from pig-endo- 2,3-dihalogenonorbornane, because coplanarity can be achieved in the . . . . . . . 54 tran51tion state for syn- elimination but not for the trans-elimination. cis, endo— k1<(R0'H+)n (70) 85 Bartsch proposed that the orientation observed in the presence of 42 results from free tert—butoxide base, while in the absence of 42, both free and associated base forms are the active species. A number of models have been preposed to rationalize the gen— eration of substanially higher proportions of l—alkene and lower trans- .Eii 2-butene ratios by associated base species. The most recent ex- planation for positional orientation control produced by associated bases involves the transition state structures in eq 71 for 2—bromobutane. (-)9tt-Bu (Hog-Bu (-)O:t—Bu (71) I \‘\ l\\\ . \\ : \sH I ‘\ : \\ 1 H ‘H 1* 11 | 1‘ RCH H CH3 H CH3 O-t:Bu ?—t-Bu O-t-Bu ‘ I I l H‘ , ”'K <—>= «”’K (-)I "‘1‘ X’ X’ Xv" Iu> lw In Zavada and Pankova57 demonstrated that the steric properties of asso- ciated and dissociated tert-BuOK are very similar for eliminations con— ducted in tert-BuOH, The base specie is assumed to be a homohydrogen— bonded tert-BuOH ion pair which provides substantial electrostatic interactions of the base counterion with the leaving group. For associ- ated bases (no crown present), the attractive base interactions are stronger in transition states forming l-alkene 4, and Sing-alkene.§, than in that leading to trans-2—alkene_§, because of the steric interac— tfirulwith the methyl group in_g. In going from dissociated to associated 86 base, the supression of trans-Z-alkene formation results in proportional increases in the production of 1-alkene and gisz-alkene. This leads to the observed enhancement of l-alkene proportion and a decrease in the trans[gi§ ratio with associated base. Hindered unassociated bases produce greater proportions of the thermodynamically less stable l-alkene in eliminations from 2-alkyl ha— lides and tosylates, than do unassociated bases of more moderate size. Since hindered unassociated alkoxide bases are more reactive than asso— ciated alkoxide bases, studies were performed to investigate the orien- tation control provided by the two types of bases. Reactions of 2-butyl iodide with a number of tertiary alkoxides in DMSO were examined.58 The results are shown in Table VII. Table VII. Olefinic Products from Reaction of 2-Iodobutane with00.25 M Potassium tert-Alkoxides in Dimethyl Sulfoxide at 50 C. Base % l-Butene tran§[gi§ 2—butene tert-butoxide 20.7 2.99 triethyl methoxide 20.9 3.13 di—tert-butyl-n-octyl 24.5 3.31 methoxide tricyclohexyl methoxide 27.2 3.04 tri-2-norbornyl methoxide 29.4 3.41 tert—butoxide (0.25 M in 29.9 2.09 tert butanol) 87 None of the sterically congested unassociated bases showed better orien- ation control than associated tert-BuOK in tert—BuOH. A decrease in orientational selectivity is observed with the hindered lithium amide base 2,6-di—tert-butylpiperidine_46. sults are shown in Table VIII. The re- Table VIII. Olefin Products from Reactions of 2-Halobutanes with Anionic Bases. trans Substrate Base Solvent Crown 1-butene cis ratio 2—iodobutane tert-BuOK diglyme none 17.7 3.43 " " " l8-crown-6 18.6 3.58 ' " MeZSO none 18.5 3.39 " 4§_ diglyme none 26.1 3.92 " " " 15-crown-5 20.6 4.13 " " " 12—crown—4 17.2 3.98 " tert-BuOK tert—BuOH none 34.4 2.17 2—bromobutane 46_ diglyme none 40.3 3.81 " tert-BuOK tert—BuOH none 50.0 1.51 88 Positional orientation in the reaction of 4§_with 2-iodobutane in di— glyme is similar to that previously observed in reactions of this sub- strate with tertiary alkoxides in DMSO (see Table VI). However, the importance of base ion pairing is shown by the decreased percentages of l-butene noted in the presence of suitable macrocyclic crown ethers. Free 4§_provides poorer directional orientational control than the previously examined dissociated but highly hindered tertiary alk- oxides. That this unexpected result is not due to some fundamental difference between dissociated nitrogen and oxygen bases is strongly suggested by recent investigations of orientation in eliminations from 2—iodobutane involving more ordinary nitrogen and oxygen bases. Bartsch demonstrated, using linear free energy relationships, that sensitivity of positional orientation to base strength variation is the same for amide ion and oxyanion bases. RESULTS Reagent grade 2—bromobutane, distilled and stored over copper wire in an amber bottle was added (1.0 eqivalent) to a 1.0 M solution of hindered secondary amide in THF at 00C. Elimination reaction progress was followed by GLC analysis of the product butenes using n-pentane as an internal standard. Aliquots of the reaction mixture was analyzed at 90 min intervals. Total butene yield increased with time, but the relative butene ratios remained unchanged during the course of the re- action. Butene isomer interconversion was not observed for any of the lithium amide elimination reactions studied. 1—Butene was subjected to LDA, gflfLi, ggfLi and glfLi in THF at 00C for 6 h. None of the l-butene 89 was converted to either gig or transz-butene. A similar lack of isomer interconversion was observed when transf2-butene was reacted under the same experimental conditions. Table IX lists the results of the E2 re- action of 2-bromobutane with various lithium amides. Almost complete mass balance was observed in these elimination reactions. The difference in the actual and the theoretical butane yield was accounted for (to within 3 or 4%) by unreacted 2-bromobutane. Each amide in Table IX was formed using 1 equivalent of TMEDA and n-butyllithium, except where no— ted. The rate of E2 reactions varied with the amide used. Table X lists the time required for the elimination reactions in Table IX to reach 50% and 90% completion. A control reaction was performed to insure that the decomposition product(s) of lithium amide attack on THF (see Figure I) did not itself effect dehydrohalogenation of 2—bromobutane. The lithium amide of 24 was reacted with THF for 2 days to insure that the amide had decomposed completely. One equivalent of 2-bromobutane was then added to the so— lution mixture along with pentane as an internal standard. After addi— tion of 2-bromobutane, no more than 2% of butene products were observed by GLC after 5 days reaction time. The product analysis of these elimination reactions was conducted by GLC. Results obtained by quenching the amide—halide solutions with 1 equivalent of H 0 were identical with results obtained from unquenched 2 solutions. Quenched amide-THF solutions were very viscous gels. This made sampling the solution mixture by syringe very difficult. All sub— sequent analyses for butenes were performed without quenching the amide solutions. 90 Table IX. Product Ratios for the Dehydrohalogenation of 2—Bromobutane with Secondary Lithium Amides in THF at 0°C. RlRZNLi + /\/ ___) W N / \ Br A 13. ‘ E GLC R Go B 00 00 ' 1 R2 A(/) (/) C(/) B/C Yie1d(%) R1= R2= Et 55 31 13 2.4 97 " no TMEDA 56 30 14 2.2 96 R1= R2= i-Propyl 67 22 11 2.0 80 ” no TMEDA 67 22 11 2.0 90 tetramethylpiperdine 86 7 7 1.0 81 " no TMEDA 85 8 7 1.2 80 R1= R2= EtMeZC- 90 6 4 1.5 82 £3 R1=EtMe2C—;R2=Et2MeC- 89 7 4 1.7 68 25. R1= R2= EtzMeC- 71 23 6 3.8 56 26. R1=Et3C-; R2=Et2MeC- 57 34 9 3.7 58 _2Z Rl= R2= Et3C- 55 36 9 4.0 74 All the amides were prepared using 1 equivalent of TMEDA in hexane with n—butyllithium. 91 Table X. Reaction Times Required to Obtain 50% and 90% of Total Butenes Evolved in the E2 Reactions of 2-Bromobutane. Amide T50 T90 EthylzNLi <1 min <5 min IsopropylZNLi W30 min N1.5 h ::;::::.::trmthfl- <5 .1 h ZéfLi N15 min ”2 h zéfLi N30 min 2.5 h géfLi 2 h 13 h ngLi 3 h 30 h 21-Li W11 h 90 h All amides were made using 1 equivalent of TMEDA in THF at 00C. 92 The results presented in Table IX reflect the actual butene composition obtained from amide base-induced B-elimination reactions. GLC analysis of the product butenes shows that the relative butene ra- tios remain unchanged from the inception to the completion of the dehydrohalogenation reaction for all of the amides studied. Several interesting trends appear in the elimination reactions. As the amide bases increase in steric bulk, the relative proportion of 1—butene increases (Hoffmann product), while the £522§4Ei§ 2-butene ratio decreases. This trend holds for moderately hindered amides. However, as the steric hindrance of the amides becomes even greater (amides ZéfLi, géfLi, glfLi and ZlTLi)’ the relative proportion of 1-butene begins to decrease steadily with increased steric hindrance. There is also a con- comitant drop in the ££§g§[gi§ 2-butene ratio with the very hindered amides (227L1,‘2§fLi, ZZTL1 and glfLi). The increase and subsequent decrease in the l-butene ratio with increasing amide bulk was unexpected. In order to determine whether this trend was unique to 2-bromobutane, several other substituted butanes were studied. The same experimental procedure and conditions were used to study the E2 reaction of 2-tosyl-butane and the trifluoroacetic ester of 2-butanol.61 Neither substrate gave measurable amounts of butenes with any of the amide bases in Table IX. However, 2-iodobutane readily reacted with the amides. Table XI presents the results obtained under conditions identical with those used in Table IX. Again, the same rise and fall in the 1-butene ratio with increas- ing amide bulk is observed. Also observed is a decrease followed by an increase in the ££gg§[gi§ 2-butene ratio. Moderately hindered amides SUCh as LiTMP and fl-Li offer better orientational selectivity in the 93 Table XI. Product Ratios for the Dehydrohalogenation of 2—Iodobutane with Secondary Lithium Amides in THF at 0°C. R R NLi + ’//‘\\r/"-——-—-9 4§§>\\//’ + x/A§§§/// + / \ 1 2 I A B c R1R2 A(%) B(%) C(%) B/C 'Yield R1=R2= Et 37 21 41 1.9 98 R1: R2= i-Propyl 40 22 38 1.7 100 tetramethylpiperidine 61 19 20 1.05 97 R1= R2= EtMeZC- 69 12 18 1.5 97 22: R1=EtMe2C—;R2=Et2MeC— 58 15 26 1.7 96 gg_ R1= R2: EtZMeC- 38 20 42 2.1 84 gg_ R1=Et3C—; R2=Et2MeC— 28 22 50 2.3 94 21_ Rl= R2= Et3C- 28 23 49 2.2 85 All of the amides were prepared using 1 equivalent of TMEDA in hexane with n-butyllithium. 94 dehydrohalogenation of 2-bromobutane, compared with 2—iodobutane. Rela- tively unhindered as well as very hindered amides have a similar lack of orientational selectivity in dehydrohalogenation of either 2-bromo or 2-iodobutane. .A number of other experiments were performed to determine the cause behind the puzzling decrease in orientational control with increa- sing steric bulk for the very hindered amide bases. The possibility of an initial a-elimination was considered. Isotopically labeled ZfiQfZ—bro- mobutane (eq 72) was synthesized and dehydrohalogenated with lithium di- ethylamide, LiTMP and glfLi. The product butene ratios for all three trials were, within experimental error, identical with the butene ratios obtained with undeuterated 2—bromobutane. ” 1. LiAlD4 OH PBr3 ?r - L - - CHBCHZC CH3 2 H 0 , 0130112? 0113 ———-> 01130112? CH3 (72) ‘ 2 D D Mass spectral and proton NMR analysis of the product butenes showed no deuterium loss or scrambling due to a-deuterium abstraction by the amide bases, as depicted in eq 73. Proton _) <—> H. sum ' CH3CH2(l3-CH3—-—) CH3$C-CH3 —33utenes 1r >< Br “ .EL—Li + ChBCHzC-CH3-——————dv (73) D 3 H D 95 Alternatively, the poor regioselectivity observed with hindered amide attack on 2-halobutanes might be due to an electron transfer initiated radical anion dehydrohalogenation reaction. A suitable alkyl bromide substrate, 2-bromo-6-heptene, was synthesized62 and reacted with a num- ber of lithium amides. This alkenyl halide is commonly used as a ra- dical trap in photochemical and chemical induced electron transfer re- actions. According to the results in Table XII, no more than 2% of the reaction products in any of the amide dehydrohalogenation reactions are a result of an electron transfer from an amide anion to the haloalkene substrate. The free radical trapped products are methylcyclohexane and 1,2-dimethylcyclopentane (eq 74). 0 Br (-) (_) o W —'—B"—) M r / CH M 3 CH3 ——» o + H3 519 (74) The same trends in l-butene/Z—butene ratios and Erans[gi§-2-butene ra- tios are observed for 46 as with 2-bromo and 2—iodobutane. Thus the l,6-/1,5-heptadiene ratio rises, then falls and the Erangflgig 1.5—hep- tadiene ratio falls, then rises, with increasing amide bulk. Reaction of lithium diisopropylamide with 2-chlorobutane in THF proceeded much more slowly than with any other butyl halide. Almost 24 hours was required for complete reaction and a total butene yield of 64% (eq 75) was obtained. 0 (75) LDA + 2-chlorobutane -%%fi%9 Ilebutene + cis-butene +- trans-butene 69% 20% 11% 96 Table XII. Product Ratios for Dehydrohalogenation of 2-Bromo-6-hep- tene with Secondary Lithium Amides in THF at 0°C. //’\\\//,\\\’/)E;\ amide / flM-I-Wa- 5.9 A 2 CH 3 H3 W+ 0 E .12 13. Rle A(%) B(%) C(%) B/C D + E Yield R1=R2= Rt 81 4 12 3.0 1% - R1=R2= i-Propyl 87 3 9 3.0 2% 89 tetramethylpiperidine 96 1 3 3.0 1% 81 R1= R2= EtMeZC- 93 2 4 2.0 1% - £9. R1=EtMe2C-;R2=Et2MeC— 92 2 5 2.5 1% _ 2§_ Rl= R2= EtzMeC- 85 4 10 2.5 1% - 2§_ R1=Et3C-;R2=Et2MeC- 74 6 18 3.0 2% - .21 Rl= R2= Et3C- 76 6 17 2.8 1% 90 21 All of the amides were prepared using 1 equivalent of TMEDA in hexane with n-butyllithium. Analyses were performed by GLC. Butene values are precise to within 2%. 97 Reactions of mOre hindered amides with 2-chlorobutane in THF failed to give better than 10% total yields of butenes. Further study of 2-chlo— robutane and 2—bromobutane elimination reactions with lithium amides in diethyl ether was abandoned when LDA failed to give better than a 10% yield of butenes after 24 h. None of the previous experiments yielded information which could be unequivocally interpreted by a simple concerted elimination mecha— nism for all the amides and haloalkanes. As a last recourse, a study of stereoisomeric 3-deutero—2—bromobutanes was initiated to investigate whether the dehydrohalogenation reactions of haloalkanes with lithium amides is a_§yn or antifelimination process. If the elimination reac- tion is indeed concerted, an isotope effect should be observed. There- fore, thrggf3jdf2—bromobutane_41 and erythro-3fd72-bromobutane'4§_were synthesized and reacted with a number of hindered lithium amides. Equa- tion 76 and 77 outline the synthesis of 41.and_4§.63 H CH CH3 C 3\ _ / 3 DBr A (76) C-C , CH / \ 3 H H H hv Br cis-2butene D threo-defZ-bromobutane 31 CH H H\. ’CH3 DBr 4; (77) /C=C\ / CH CH H hv 3 trans-2-butene D erythro-3-df2-bromobutane 48 98 The purity of each isomer was determined by reacting 41 and 48 with 1 molar potassium tert-butoxide in dimethylsulfoxide at 300C and col- lecting the butenes evolved from solution. The three butene isomers were separated by preparative GLC and the deuterium content of the gig and trans-Z- butene'isomers determined by mass Spectral analysis. This dehydrohalogenation reaction is known to proceed with angifstereo— chemistry. Isomers_4z and 48 should therefore yield the butene pro- ducts predicted in eqs 78 and 79, respectively. - + CH (CH ) 0 K H H D CH CH 3 3 3 ~> CH =CDCH CH + ‘C=C’ + \C=C’ 3 3 H 2 2 3 CH’ ‘CH CH’ ‘H r (CH ) SO 3 3 3 3 2 D d -l-butene cis-d trans-d 1 — o —— 1 ‘41: threo (78) - + H H (CH3)30 K D H H CH 3 \ _ \ / \ I 3 H 7 CH -CDCH CH + C=C + C=C H3 2 2 3 CH’ ‘CH CH’ \H Br (CH3)ZSO 3 3 3 D d -l—butene cis-d trans-d 48f erythro 1 “" 1 '——_—' 0 (79) Each butene isomer was separated by GLC and analyzed by mass spectroscopy at 10.7 eV, just enough energy to ionize the molecule yet not enough energy to cause ion fragmentation. The deuterium content of each isomer was determined. From this data the actual purity of the two deuterated bromobutanes was calculated (see the Experimental section for details of the calculation), and the results are reported below in Figure 4. 99 DBr, hv . _ _ _ , _ _ a _ cis_2-butene ______) 89.7 A fll_r_e_0_ 3 g 4.2/. erythro 3 6.1% 2 100°C 2-bromobutane .de-bromobutane bromobutane DBr, hv _ _ 3 1 . 2.7% threo-3-d- 88. l % erythro-B- 10. 7 % 2- -££EE§ 2 butene -lOO°C 2-bromobutane ng-bromobutane + bromobutane Figure 4. Isotopic Purity of Threo and Erythro 3jd72-Bromobutane Dehydrohalogenation of 41_and 4§_was accomplished under experimental conditions identical with those used for the dehydrohalogenation of 2-bromobutane. The results of the elimination reactions with both 41 and 48 are presented in Tables XIII and XIV, respectively. There are several clearly noticeable effects on butene product ratios for 41_and 48 when compared to undeuterated 2—bromobutane. The relative proportion of l-butene formed in the amide elimination reaction of either 41 or 4§_is greater than that observed for 2-bromobutane using the same amides. Secondly, the relative £332§%gi§—2-butene ratio from .Ehrggf3jdf2—bromobutane.41, is greater than the t£§n§723§72-butene ratio obtained from the elimination reaction of 2-bromobutane with the corresponding lithium amides. Finally, the relative ££§n§[gi§-2-butene ratio from erythro-Bjde-bromobuta e 4§,is smaller than the transfigig— 2-butene ratio obtained from the elimination reaction of 2-bromobutane with the corresponding lithium amides. .Cis and transz-butene, isolated from the reaction of diethylamide and glfLi with 41 and 48, were sepa- rated by GLC and analyzed by mass spectroscopy at 10.7 eV. The deuterium 100 Table XIII. Product Ratios 1km: the Dehydrohalogenation of Threo- 3jdf2-bromobutane with Lithium Amides in THF at 0°C. (H)D /CH (D)H\ [H R R NLi _ \ _ 3 _ CH3CHBrCHDCH3 ———3>3———) CHz'CHCHDCH3 + CH/C‘C\H + CH/C‘C\CH THF 3 3 3 i7— :1 _ _ RlR2 A(%) B(%) C(%) B/C R1=R2= Et 63 32.9 3.8 8.6 R1=R2= i-Propyl 71.5 25 3.5 7.3 tetramethylpiperidine 92 6.1 1.8 3.4 R1=R2= EtMeZC— 85.5 13 1.5 8.8 _24 Rl=EtMe2C-;R2=Et2MeC- 87.5 11 1.4 9.8 22 R1=R2= EtZMeC- 78 20 2.0 13.5 26_ R1=Et2MeC—;R2=Et3C— 62 35 2.5 14.2 .27 R1=R2= Et3C— 65 33 1.7 17. 3 21 All amides were prepared using 1 equivalent of TMEDA in hexane with n-butyllithium. 101 Table XIV. Product Ratios for the Dehydrohalogenation of Erythro— 3—dr2-bromobutane with Lithium Amides in THF at 0°C. RleNLi (D)H /CH3 (H)D /H CH CHBrCHDCH ) CH =CHCHDCH + C=C + ‘C=c 3 3 2 3 / \ / \ THF CH H CH 3 3 CH 3 .£§ .5 .E .9 RlR2 A(%) B(%) C(%) B/C R1=R2= Et 77.3 6.4 16.3 0.39 R1=R2= i-Propyl 83.6 4.0 12.4 0.33 tetramethylpiperidine 90.8 1.2 7.9 0.15 R1=R2= EtMezc‘ 89.3 3.1 7.9 0.39 .21 R1=EtMe2C—;R2=Et2MeC- 91. 6 2.4 5.9 0.41 2; R1=R2= EtZMeC— 86.4 6.2 7.2 0.86 .29 R1=Et2MeC-;R2=Et3C- 79.6 8.8 11.6 0.75 31, R1=R2= Et3C— 79.0 8.7 12.3 0.71 1 All amides were prepared using 1 equivalent of TMEDA in hexane with n-butyllithium. The butene values are known with a precision of i0.3% 102 isotope jratio was determined for each isomer and the results are shown in Table XV. Table XV. Deuterium Content in Cis and Trans-Z-Butene from the Dehydrohalogenation of Threo and Erythro-B-d—Z—bromobutane. (Et2)NLi [(Et3)C]2NLi 7 trans—d- 7cis—d- ,trans-d- 7cis-d- ° 2-butene °2-butene °2-butene °2-butene threo-3-d—2 bromobutane 87 13 89 11 erythro-B-d- 2—bromobutane 10 90 12 88 Butene isomers separated by GLC. Mass spectral analysis at 13.5 eV Another set of experiments was performed with 2-bromobutane and the amides in Table IX. These were designed to determine the exact orientational and geometrical effect of 1,4,7,lO-tetraoxacyclododecane (12—Crown—4) on the dehydrohalogenation of 2-bromobutane with lithium amides. One equivalent of 12—Crown-4, a chelating ligand for li- thium ions,was added to a 1 molar solution of 2—bromobutane, THF and a secondary lithium amide at 00C. Yields of butenes were very low for most of the amides studied. Low temperature addition (-780C) of 103 12-Crown-4 to the amide solutions followed by 2-bromobutane, gave the best yields of butenes. The reaction solutions were stirred for 15 min at -780C, warmed to 00C and stirred for an additional 15 min. GLC ana- lysis of the solution products was performed as before, without quench- ing. Quenching studies at -780C with H20 showed that for all the amides studied, except LiTMP, there was no dehydrohalogenation of 2-bromobutane within 30 min. LiTMP reacted very quickly (<10 min) with 2-bromobutane in the 12-Crown—4 THF solution at -78OC. The results obtained for the dehydrohalogenation of 2-bromobutane in lZ-Crown-4 amide solutions are given in Table XVI, and show the same general trends which were observed in all the other lithium amide-haloalkane studies. There is an initial rise, then a fall in the l-butene/Z-butene ratio with increasing amide bulk. Concomitantly, one observes a fall, then a rise in the £333§%gi§- 2-butene ratio with increasing amide bulk. The major difference appar- ent between the presence and absence of crown ether in these elimina— tions is that crown ether leads to poorer regioselectivity for 1-butene with the more hindered amides. A very notable exception occurs, how- ever, with LiTMP, 12-Crown-4 (1 equivalent) and 2-bromobutane. This reaction shows excellent regioselectivity for l-butene formation. The only other amide which showed a better regioselectivity with crown ether is LDA. The trans/gisz-butene ratio variation with increasing amide bulk followed quite closely the results obtained without crown ether. The only exception occurs with very hindered amides, where the .££§2§[gi§f2—butene ratio is lower than that observed without any crown ether. Experiments performed with LiTMP, 2-bromobutane and varying a- mounts of lithium crown ether showed that as less lZ—Crown-4 ether is added 104 Table XVI. Product Ratios for the Dehydrohalogenation of 2-Brnmobutane with Lithium Amides Using 12-Crown-4 in THF at —78°C. R R NLi CH H CH CH CH3CHBrCH2CH3 1 2 >=CH2=CHCH20H3 + 33:3: + 3p=él 3 12-Crown-4 H CH3 H H THF .A .2 .2 Equiv Yield RlRZ 12-Crown-4 A B C B/C (%)GLC R _R _ Et 1.0 53 29 18 1.6 -- 1 2 R1=R2= i-Propyl 1.0 71 19 10 1.9 74 tetramethylpiperidine 1.0 99 <0.5 NO.5 --- 102 " no TMEDA 1.0 99 <0.5 m0.5 —-- 99 " 0 5 95 2.5 3 5 1 4 —- " 0 25 92 3 4 5 1 5 -- ” 0 1 89 4 7 1 8 -- R =R2= EtMeZC— 1.0 88 5 7 1 4 75 Rl=EtMe2C-;R2=Et2MeC- 1.0 81 13 6 2.2 40 R1=R2= EtZMeC— 1.0 53 32 15 2 1 33 = _. = _ , 1 2. R1 EtzMeC ,R2 Et3C 1 0 46 38 5 5 33 R1=R2=Et3C- 1.0 40 45 15 3.0 36 All amides prepared with 1 equivalent of TMEQA. Initial reaction temp- erature at -78 C for 15 min, then warmed to O C. 105 to the reaction solution, the regioselectivity of the dehydrohalogena- tion reaction is diminished. The presence or absence of TMEDA in the 12-Crown-4 dehydrohalogenation reaction of 2-bromobutane with LiTMP had no noticeable effect on the product butene ratios. 106 DISCUSSION The E2 reaction of alkyl halides with secondary lithium amides has not been well studied, and there are few references to synthetic or physical organic investigations of dialkyl amide initiated E2 reac- . 59,64,65 tions. A study was therefore initiated to investigate the stereo and regiochemical influence of secondary lithium amide base pro- moted dehydrohalogenation reactions. Table IX lists the results of 2-bromobutane elimination reactions with dialkyl lithium amides. The rise in the l-butene/Z—butene ratio and concomitant fall in the Eran§[gi§72—butene ratio is what one might expect for the E2 reaction of increasingly hindered base species. Diethylamide, diisopropylamide, tetramethylpiperdylamide and the amide of 24 show progressively better terminal olefin regioselectivity. The Hoffmann product preference can be rationalized using a simple steric model, as- suming an antifelimination transition state (eq 80). This model is sim— ilar to one proposed by Bartsch for the E2 reaction using aggregated O I I 4 alkoxide bases of small to moderate steric dimentions. 11% HN// §N//// 1'1 / RCH2 CH3 H H / CH\3 (80) H R . R H x X x A C We have no a priori knowledge of the extent of amide base aggre- lw gation in THF solution. Whether or not the amide bases are associated, the simple steric model proposed in eq 80 seems adequate to explain the experimental results, at least for the less hindered amides. However, 107 as the amide bases become bulkier, transition states_A and_§ for the formation of l-alkene and ging-alkene, respectively, are less af- fected than that for transz-alkene formation,_§, because of the pos- sibility of tilting the base in A_and §_to relieve the steric inter— action between the bulky base and the 0 and/or B-alkyl groups. Since these destabilizing steric interactions should increase in the order A D’ ‘H H/ \CH3 dl—l—butene cis-d -2- trans-d —2 Br butene butene syn Erythro-Bjd—Z- bromobutane é§. syn H CH3 CH [CH CH D . 3 3 3. 1’ CH antl CH CHDCH=CH + C=C\ + =C 3 D ——-9 3 2 H’ H H/C \CH3 dl-l-butene cis-dO-Z- trans-d —2 Br butene butene Threo-3jd52- bromobutane 47 Figure 5. Syn and Anti Elimination Product Scheme for 41 and 48, The magnitude of the isotope effect for an E2 elimination reaction depends on the bonding of the B—hydrogen atom in the transition state. If the hydrogen lies midway between the base and the B-carbon (central E2 like), the kH/kD ratio will be a maximum. Either less (E1 like) or more (E1C like) C-H bond stretching will result in a lower isotope B effect. A comparison of the data for the E2 reaction of 2-bromobutane (Table IX), threo-3jdf2-bromobutane‘41 (Table XIII) and erythro-3jdf2— bromobutane (Table XIV) provides strong evidence favoring 112 anti-configuration with the hindered amides. For threof3fgf2-bromobutane, one would predict that an anti- transition state in the E2 elimination would look like the following (eq 81). ,‘N / ’N—fl X’N—fi D, 1’ 1:14 1‘ 1 CH CH3 H CH3CHD H (81) O O H . CH3 1 H D ' CH3 1 H o 7 I Br Br Br A —9 cis--dO B —-) trans-dl E —-9 1-d1- butene Assuming that the C-H(D) bond breaking step is rate limiting, retarda— tion of the elimination step represented by A_should be observed while the steps represented by.§ and 9 should proceed at a rate very similar to undeuterated 2-bromobutane. The retardation of step A (gi§72—do— butene formation) would result in an apparent rate increase in trans: dl—Z—butene and l-dl-butene formation. Thus an EEEiTEZ reaction with Ehrgngfd-Z-bromobutane should result in an increase in the l-butene/ 2-butene and E£§n§7gi§72-butene ratios when compared with undeuterated 2-bromobutane. This is exactly what is observed in Table XIII. Both of these trends are observed with both moderately and very hindered a- mides. In fact, the very hindered amide E2 reactions show considerably higher trans/Eisz-butene ratios when compared with the same results from moderately hindered amide E2 reactions. Product analysis for deu— terium content from the E2 reaction of diethylamide and glfLi with 41 113 confirms, within experimental error, that the E2 reaction proceeds by an anti-elimination mechanism (eq 82). % dO-cis-butene % dl-trans-butene Diethylamide-Li + 47 ---9 13 87 21-Li + 47 ——> 11 89 (82) The contamination by undeuterated 2—butenes is a result of the 4% and 6% contamination of 4§_(erythro isomer) and undeuterated bromobutane, respectively. Undeuterated 2-bromobutane results from the inability to remove all adsorbed H20 from the red phosphorus starting material. An identical amide elimination study for erythro-B-d-Z-bromo- butane was performed. The results (Table XIV) indicate that the E2 reaction also proceeds by an anti-elimination mechanism for all the a- mides studied. Equation 83 illustrates the transition states for anti- elimination of the erythro-3jd72-bromobutane isomer. 1 ' I 1 N ’N I’N v’ \ x \ 1* \ I? 1‘ Q Q . ‘8” H CH3 . D H . CH 3 . H I I 0 Br Br Br A —-9 sis-d1 2 ---3 trans-d0 _C_ “'3 l-dl- . butene The erythro-Bfing-bromobutane isomer should show a relative increase in l-butene and cis-2-butene formation when compared with undeuterated 2-bromobutane. Thus an increase in the l-butene/Z-butene ratio and a 114 decrease in the ££33§[gi§f2-butene isomer ratio is expected when com- pared with undeuterated 2-bromobutane. This is exactly what is observed experimentally. There is a very substantial decrease in the EEEE§4El§T 2-butene ratio for 48 when compared to the results observed with undeu- terated 2-bromobutane. Product analysis of £i§_and transz-butenes for deuterium content was performed from the E2 reaction of diethylamide and _lfLi with 48. The results are listed in eq 84 and confirm, within ex- perimental error, that the E2 reaction proceeds by an anti-elimination mechanism (eq 84). % dl-cis-butene % dO-trans-butene Diethylamide-Li + ‘4§-—~——) 90 10 (84) 1-Li + .4§-—-——) 88 12 Simple steric arguments are inadequate to rationalize why l-bu- tene selectivity decreases with increasing steric bulk. The expecta- tion would be that the less hindered terminal methyl hydrogen is more accessible than the 3-methylene hydrogen of the alkyl halide. A kinetic study of the E2 reaction of 2-bromobutane with lithium amides showed that moderately hindered amides (diethylamde through gflgLi) reacted within 1.5 h at 0°C. With the very hindered amides, the elimination reaction required anywhere from 2.5 h for ggfLi to almost 4 days with ZlfLi. An explaination for the slowdown in this E2 reaction for the very hindered amides is that they are just too hindered to effectively coordinate to the B-methyl or B-methylene hydrogen atom of the alkyl ha- lide. Space filling models of very hindered amides show that the amide 115 nitrogen is effectively shielded by the alkyl sidechains, burying the anionic center within the hydrocarbon folds of the amide molecule. This shielding effect should limit the number of effective trajectories by which the amide could approach the alkyl halide to a rather narrow "window" through which the transition state geometry is exactly right to effect dehydrohalogenation by an EEEET elimination mechanism. Another plausible explanation. might be that the very hindered amides are not as kinetically basic as the less hindered amides. Evi— dence for this proposal is found in the kinetic base study of lithium amides by their attack on THF (Chapter II). The E2 reaction of 2—bromobutane and the lithium amides was in— vestigated with 12—Crown-4 ether to determine if amide base aggrega~ tion played a significant role in influencing E2 elimination regiochem— istry. The results obtained shed little light on the question of amide base aggregation in THF. The influence of 12—Crown-4 ether on amide base selectivity in the E2 reaction with 2-bromobutane was inconsistant with the models posed by Bartsch for alkoxide promoted E2 reactions in alcohol. There is an increase, then a decrease in the l-butene/Z-bu— tene ratio with increasing amide bulk in the presence of 12-Crown-4. The l-butene/2-butene ratio maximum and minimum are greater than and less than, respectively, the same ratios without added crown ether. Based on Bartsch's observations that crown ether acts to reduce effective base size by dissociating alkoxide base aggregates, one could conclude from the results in Table XVI that the moderately hindered a- mides like diethyl and diisopropylamide are dissociated monomeric bases which are not influenced by crown ether. On the other hand, the very hindered amides appear to form aggregate pairs or oligomers whose 116 effective steric bulk and, in turn, regioselectivity, is reduced by the crown ether dissociation of amide base aggregates. However, it is hard to believe that very hindered amides form aggregates while moderately hindered amides apparently remain unassociated in THF solution. An even more difficult observation to rationalize is the greater regiose- lectivity of LDA and LiTMP for l-butene formation with crown ether than without crown ether. Addition of crown ethers has never before been shown to improve regioselectivity in base promoted E2 reactions. Based on the data obtained in Table XVI, no firm conclusions can really be made as to the aggregation state of the lithium amides in THF. Further studies directed towards determining the nature of the aggregation state of lithium amides in ethereal solution remain to be performed. One other piece of information which might be used in elucidating the nature of the dichotomy in E2 reactions involving moderately and very hindered amides is the apparent deuterium isotope effect of deuter- ated and undeuterated 2-bromobutane. If we assume that the amide bases in THF are dissociated, stretching of the CB—H bond in the transition state for the amide induced eliminations may increase with very bulky bases to a point where the base's steric effect is reduced. Table XVI lists data from which an apparent deuterium isotope effect can be cal- culated between 2-bromobutane and each of the two deuterated stereoiso— mers, 41 and 48, The values of kH/kD for threo and erythro—Bjd—Z—bro- mobutane were calculated (see Experimental section) and corrected for the deuterium content of each isomer. Table XIX lists the apparent isotope effects for each amide reaction shown in Table XVI. 117 Table XIX. Apparent Primary Deuterium Isotope Effect for 2-Bromobutane 111R2 kH/kD £11339 ———9 Egg-do kH/kD erythro —-) Earls-do R1=R2= Et 3.9 7.7 R1=R2= i-Propyl 3 0 7 8 tetramethylpiperidine 3.2 5.1 R1=R2=Me2EtC— 2.8 2.2 R1=Me2EtC-;R2=MeEt2C- 2.7 3.7 R1=R2=MeEt2C- 3.7 4.9 R1=MeEt2C-;R2=Et3C- 4.7 6.4 R1=R2=Et3C- 7.0 6.8 Ratios are subject to an uncertainty of 10.5. 118 Two things are immediately apparent. First, there is a sub— stantially smaller isotope effect observed for moderately hindered a- mides using thrgngfng-bromobutane than with the erythro-isomer. Se- condly, the isotope ratios for both stereoisomers initially decrease, then increase with increasing amide bulk. The range of the deuterium isotope effects reflect changes in the transition state of the dehydrohalogenation reactions with hindered amides. For diethylamide through gfl-Li, the decrease in kH/kD can be rationalized by arguing that the transition state shifts from a ”central E2" to a more "ElC like" transition state as the kinetic B basicity of the amide bases increase. The range of the isotope effect for the threo—isomer is substantially smaller (3.9 to 2.8) than the isotope effect (7.8 to 2.2) for the erythro isomer with moderately hin- dered amides. One reason for this effect might be the steric interaction of the terminal methyl group of 2-bromobutane with the amide base in the deprotonation step of the E2 reaction. The transition state for the Ehrggfisomer leading to gisde-Z-butene has the terminal methyl group eclipsing the B-methyl group. The amide anion has an unencumbered ap- proach from one side of the bromobutane molecule. The lack of steric interference would permit the moderately hindered bases to bind tightly with the C ~deuterium atom and break the C—D bond before a substantial B stretching of the CB-D bond could occur. This would in effect decrease the observed deuterium isotope effect because the transition state would become more "EICB like". The erythro isomer on the other hand has the B-methyl group trans to the terminal methyl group in the transition state leading to trans- dO-Z-butene. This transition state blocks 119 the approach of the incoming amide anion in the deprotonation step. The steric interaction would not permit moderately hindered and kine— tically weak bases like diethyl and diisopropylamide to bind as tightly with the B-deuterium atom in the trans-erythro transition state as in the EEETEEEEE transition state. Thus, substantial stretching of the CB-D bond would occur in the transtion state, giving rise to a more "E2-like" transition state with the erythro isomer. As the kinetic ba— sicity of the amide anions increase (LiTMP, géfLi and ggfLi), this should offset the steric interference of the B-methyl group. The kin- etically powerful amide bases would break the C-D bond before any sub— stantial stretching of the C—D could occur and this would explain the drop in the apparent isotope effect from 7.7 to 2.2 for diethylamide through_24—Li, respectively. For the very hindered amides géfLi through_21-Li, there is a gradual increase in the isotope effect for both deuterated 2-bromobu- tane isomers. This increase in the isotOpe effect would seem to in— dicate a steady increase in the stretching of the C —H bond, reflecting B a shift towards a more symmetrical transition state. If the CB-H bond stretching were to go too far, the transition state would become more Ech—like. This is apparently not observed with the hindered amides. This increased stretching of the C -H bond probably reflects a 8 combination of both steric and kinetic base effects. It was previously shown in the amide base stability study and the E2 reaction rate study that as the amide bases become more hindered, their kinetic base strength decreases. The elimination reactions with kinetically slow, hindered amides probably proceeds with a central E2 transition state. This would permit the B-alkyl group to lower the energy of the very 120 hindered amide transition state by electron release to the developing double bond, so that 2—butene formation is favored in the elimination step. Also, one would expect that a more central or symmetrical E2 transition state would give a larger trans[gi§—2—butene ratio when com— pared to the Ech—like transition state, owing to the eclipsing effect between two alkyl groups in the transition state leading to Eis—Z—butene. This is also observed as the very hindered amide bases increase in bulk. 121 EXPERIMENTAL 2-Bromobutane, 2-iodobutane and 2—bromo—6-heptene were all at least 98% pure and stored in amber bottles over copper foil under re- frigeration. N',N',N ,N-Teteramethylethylenediamine, tetramethylpiper— idine, diisopropylamine and diethylamine were purchased from Aldrich and distilled over calcium hydride. Hindered amines 24) 253 26) 21_and __1 were prepared and purified as described in Chapter I. Tetrahydro- furan and diethyl ether were dried over lithium aluminum hydride and stored over sieves under argon. n—Butyllithium was purchased from Al— drich and standardized as described in Chapter II. The lithium chelat- ing macrocycle 12—Crown-4 was purchased from Aldrich and used without purification. Deuterium bromide was synthesized by in_§i£u formation of phosphorus tribromide with bromine and red phosphorus and subsequent reaction with 99.8% pure deuterium oxide (Stohler Isotopes). All re- ference butenes were obtained from Matheson and were at least 99% pure. IR spectra were taken as neat films on NaCl plates with a Perkin—Elmer 237—B spectrophotometer. NMR spectra were taken on Varian T—6O and Bruker WR-250 spectrometer using tetramethylsilane (TMS) as the reference standard. Mass spectra were taken with a Finnigan EI—CI gas chromato— graph-—mass spectrometer. Analytical gas-—liquid chromatography was performed with a Varian 920 gas chromatograph using a 40' x 1/8" alumi— num column packed with 80/100 mesh Chromosorb w DMCS—AW treated with 20% SE—30 liquid phase. All preparative GLC analyses were performed on the same chromatograph using a 20' x 1/4” stainless steel column packed with the same support at —200C. 122 Dehydrohalogenation of 2-Bromobutane with Lithium Diisopropylamide in THF. The following dehydrohalogenation procedure is representative of all the dehydrohalogenation reactions of 2-bromobutane, 2-iodobutane, erythro and thrggf3fd72-bromobutane and 2-bromo-6-heptene with hindered alkyl lithium amides in THF or diethyl ether. A 10 mL round bottomed-flask, fitted with a rubber septum glass sidearm, Teflon stirring bar and gas inlet valve, was connected to a mer— cury bubbler and flame dried under argon. After cooling, 1.90 mL of n- butyllithium (3.0 mmol, 1.56 M) was injected by syringe into the reaction flask, followed by 0.45 mL (3.0 mmol) of TMEDA. The flask was cooled to 00C with an ice bath. Then 0.42 mL (3.0 mmol) of diiSOpropylamine was slowly injected into the reaction mixture. The cooling bath was removed and the solution stirred for 10 min. A warm water bath (40-500C) was placed beneath the reaction flask and the hexane solvent removed under vacuum. After the hexane was removed, the flask was cooled to 00C and 3 mL of THF solvent was added to the flask, followed by 0.33 mL (3.0 mmol) of 2-bromopentane and 0.345 mL pentane internal standard (3.0 mmol). After 1 h, a 10 microliter sample aliquot was withdrawn with a micro- liter syringe and injected into the GLC (500C). The three butenes elu- ted sequentially as l-butene, transz-butene and Sing-butene. Coinjec- tion of authentic butene samples dissolved in THF confirmed the product identities. Elution of the product butenes and pentane standard required about 1.5 h. The same experimental procedure was followed using 3 mmole of di- ethylamine (0.31 mL), diisopropylamine (0.42 mL), tetramethylpiperidine (0.51 mL), 24_(0.60 mL), 2§_(0.64 mL), 26 (0.68 mL), 21 (0.72 mL) and 21_(0.76 mL). The reaction times for the dehydrohalogenation reactions 123 are reported in Table X. None of the reaction mixtures were quenched with water before GLC analysis. Metallation of THF and Attempted Reaction of Metallated THF with 2-Bromobutane. LiTMP (3 mmol) was prepared by the procedure outlined in Chapter II using TMEDA. After hexane solvent was removed from the amide solu— tion, THF was added (3 mL) and the mixture stirred at 240C for 2 days. 2—Bromobutane (3 mmol) was added to the metallated THF solution and the mixture was stirred for another 3 days at 240C. Pentane standard (3 m- mol, 0.345 mL) was added to the solution and a sample aliquot was ana— lysed by GLC. Only trace amounts (<2%) of butenes were obtained. Reaction of 2—Iodobutane with Hindered Lithium Amides in THF. The procedure described for the dehydrohalogenation reaction of 2—bromobutane was applied to the same reaction with 2-iodobutane. All the amides were prepared (3 mmol scale) as described previously in Cha— pter II. Iodobutane (3 mmol,.37 mL) was added to the appropriate amide in THF solution at 00C. The reaction times for the dehydrohalogenations ranged from 5 min for diethylamide to l h for ZéfLi. Reaction rates for the very hindered amides were not determined. However, all the very hindered amides had completely reacted within 10 h of 2-iodobutane ad- dition. Yields and relative butene ratios were determined by GLC as described previously with 3 mmol of pentane. 124 Preparation of 2-g72—Bromobutane from 2—Butanone. A 400 mL round—bottomed flask was fitted with a Teflon stirring bar, septum sidearm and a reflux condenser. The reflux condenser was connected to a mercury bubbler with a gas flow valve and rubber tubing. The entire system was flame dried and flushed with argon. Anhydrous ether (distilled over LiAlHA) 200 mL, was added to the reaction flask, followed by 5.0 g of LiAlDa (98% Stohler Isotopes). The solution was stirred vigorously while 38 mmol of 2—butanone (34 mL, 27.0 g) was added by syringe at such a rate as to maintain a steady reflux. After ketone addition was complete, the solution was stirred overnight at 240C. Water (9 mL) was carefully added to the solution to decompose any unre— acted deuteride. The supernatant was carefully decanted and the alumi— num hydroxide residue was washed five times with 50 mL aliquots of ether. The gel was vacuum filtered to remove any remaining traces of alcohol. The ether extracts were combined and dried over anhydrous K2C03. The ether was removed under reduced pressure and the alcohol distilled through a 20 cm Vigreux column. By GLC (5% Carbowax 20M, Chromosorb W), the product alcohol was 95% pure, the remaining 5% unreacted ketone. Recovered alcohol yield was 66% (17 g). The 2jd-2—butanol was converted to 2jd—2—bromobutane by a litera— ture procedure.66 Bromine (10 mmol, 0.51 mL) was added to a 25 mL round bottomed flask flushed with argon, fitted with a stirring bar, septum inlet and a gas inlet valve connected to a mercury bubbler. The reaction flask was cooled to -780C and 9.6 g (3.36 mL) of PBr3 (Kodak, 95%) was added dropwise by syringe over 5 min. Then 97 mmol (8.9 mL) of 2—d72— butanol was added dropwise to the vigorously stirred reaction solution. 125 After addition of the alcohol was complete, stirring was continued for 10 h at 00C. The organic layer was decanted and washed with 2 x 10 mL cold water and 2 x 10 mL cold saturated aqueous Na2C03. The organic layer was separated and distilled (bulb to bulb) to obtain 7.0 g (60% yield) of 98% pure (GLC) Zide-bromobutane. No proton resonance was ob- served corresponding to 2—H—2—bromobutane: NMR (CD013) 0 1.0 (3 H, t, J=7), 1.7 (3 H, s), 1.83 (2H, q, J=7); mass spectrum, m/e (relative in— tensity) 138 (M3), 58(100), 57(3). Reaction of ngfZ—Bromobutane with 1—Li in THF at 00C. Reaction of 2jd32-bromobutane with amide glfLi in THF was per— formed as described previously. The reaction time was 60 h at 00C. Preparative GLC was performed on a 20' x 1/4" column at -200C filled with 20% SE—30 on Chromosorb W. The three butene isomers were separated and collected at -1980C in glass collection tubes. The individual iso- mers were dissolved in D6 benzene and NMR spectra were obtained with the flame sealed tubes. Product analysis of 1-butene showed no proton re- sonance at 0 5.3 - 6.0, corresponding to a secondary olefinic hydrogen of undeuterated l-butene; NMR of 1—butene (273) 0 0.8 (3 H, t, J=6), 1.8 (2 H, q, J=6), 4.75 ( 2 H, bs). NMR of £i§_and transz-butene from 2fid72-bromobutane shows a 4:1 integrated proton signal between the ole— finic proton at 05.3 and the methyl doublet at<51.6. Undeuterated SEE and transz-butene show a 2:1 integrated proton signal at the same po- sition: mass spectrum, m/e (relative intensity) cis and trans-Z-butene 57 (MI), 100 (41). 126 Preparation of 2-Butyl p:Toluenesulfonate 2-Butanol (7 g, 95 mmol) was added to 125 mL of dry pyridine in a 250 mL Erlenmeyer flask. A one mol excess of toluene sulfonyl chloride (0.2 mol, 38.1 g) was added to the pyridine solution and the flask stOp— pered with a rubber stopper. The solution was stirred for l h at 220C with a magnetic stirring bar. The solution was then placed in a refig- erator overnight (12 h). The solution was then poured into a 2 L beaker filled with 1 L of chopped ice and water. The organic tosylate falls to the bottom of the beaker and is separated from the aqueous layer. The tosylate is washed 3 times with 200 mL of water to remove any remaining pyridine. The tosylate is recrystallized from petroleum ether at -780C. The yield of product is 41 g (90%). The product is a liquid at room tem— perature; NMR 0 0.5 (3 H, t, J=6), 0.95 (2 H, d, J=6), 1.2 (2 H, q, J=6) 1.93 (3 H, s), 4.3 (1 H, q, J=6), 6.77 (2 H, d, J=9), 7.57 (2 H, d, J=9). Preparation of 2—Butyl Trifluoroacetate. 2—Butanol (10 g, 0.14 mol) was added to 150 mL of dry pyridine in a 250 mL Erlenmeyer flask fitted with a stirring bar and rubber stopper. The solution was cooled to 0°C and 56.7 g (0.27 mol) of trifluoroacetic anhydride was added to the solution. The solution was stoppered and reacted overnight at 00C. The reaction mixture was poured into 500 mL of ice water and the aqueous phase decanted. The ester was dried over 4 A molecular sieve and distilled (bulb to bulb) under vacuum to give 14 g of 99% pure (GLC) ester. Yield is 59%: NMR 6 0.93 (3 h, t, J=6), 1.33 (2 H, d, J=6), 1.63 (2 H, q, J=6), 6.0 (1 H, sext, J=6). 127 Reaction of 2-Buty1 Tosylate and 2-Butyl Trifluoroacetate with Secondary Lithium Amides in THF. Both 2—butyl tosylate (3 mmol, 0.57 mL) and 2—butyl trifluoro— acetate (3 mmol, 0.47 mL) were reacted with lithium diethylamide and LiTMP (3 mmol in each) in THF (3 mL) at 00C as previously described. Less than a 20% yield of total butenes was achieved for each trial (GLC) af— ter 10 h reaction time. Preparation of 1-Pentene—5-ol from Allyl Bromide and Ethylene Oxide. This procedure was adapted from a previously reported synthesis?7 A solution of a few iodine crystals and 2 g of allyl bromide in 50 mL of anhydrous ether was added to 57 g of magnesium (2.34 mol) in a flame dried 3-necked round-bottomed flask fitted with a mechanical stirrer, reflux condenser and a dropping funnel (pressure equalized). The entire apparatus was under argon and connected with a gas inlet valve to a mer- cury bubbler. The Grignard reaction sets in within 10 min. Once the reaction begins to reflux vigorously, 500 mL of anhydrous ether was ad— ded to the reaction vessel. The reaction flask was placed in an ice bath. Then a solution of 250 mL of ether, 125 mL (178 g, 1.45 mol) al- lyl bromide and 46 g ethylene oxide (1.43 mol) at 00C was dripped into the reaction solution with vigorous stirring. The temperature of the reaction mixture should not rise above 100C. After addition, the so- lution is stirred for 5 h at 00C, then warmed to 500C and refluxed for 1 h. The reaction solution was quenched carefully with 50 mL ice water. The reaction solution was vacuum filtered. The solid residue was washed with 3 x 50 mL ether. The filtrates were combined and the ether was 128 removed under low pressure. The product was distilled (50 mm, 560— 580C) and gave 55 g (42% yield) of 98% pure (GLC) 1-pentene—5-ol. Preparation of 5—Chloro—15pentene from l-Pentene-S-ol. 1-Pentene—5-ol (30 g, 0.35 mol) was added to 250 mL of dry pyr- idine and cooled to -50C. Then 80 g of p-toluenesulfonyl chloride (80 g, .42 mol) was added to the solution and the mixture stirred over- night at 00C. Water (10 mL) was added in two mL aliquots at 5 min inter— vals to the reaction solution at 00C. Then 100 mL of ice water was ad— ded to the reaction mixture. The entire solution mixture was extracted with 3 x 100 mL of chloroform. The chloroform layers were combined and extracted twice successively with l M H2804, H20 and Na2C03. The to- sylate was dried over anhydrous Na2804. The dried product was a liquid at 240C and was used without further purification: NMR 0 1.5 - 2.2 (4 H, m), 2.33 (3 H, s), 3.93 (2H, t, J=6), 4.7 (1 H, m), 4.9 (l H, m) 5.2-5.9 (l H, bm), 7.16 (2 H, d, J=7), 7.67 (2 H, d, J=7). Anhydrous lithium chloride (.32 mol, 13.4 g) was added to dry dimethylsulfoxide (130 mL) in a 300 mL round-bottomed flask equipped with a magnetic stirring bar. The mixture was stirred until all of the lithium chloride was dissolved. Then 63 g (0.26 mol) of l—pentene- 5-01 tosyl ester was poured into the reaction flask. The flask was stop- pered and stirred for 2 days at 500. Water (200 mL) was added to the .solution and the solution cooled to 00C. The l-chloro-S-pentene separated :frrmlthe solution and was decanted. The chloroolefin was distilled tllrough a short path distillation column (bp 105—1070C, 750mm) and the reCovered yield (19 g) was 70% of 98% pure product (GLC): NMR 6 1.7-2.4 129 (4 H, m), 3.47 (2 H, t, J=6), 4.9 (l H, m), 5.1 (1 H, m), 5.4 - 6.0 (1 H, m). Synthesis of 1—Heptene-6-ol. To a 100 mL round-bottomed flask fitted with a stirring bar, septum sidearm and reflux condenser, was added 4.6 g of magnesium sha- vings ( 0.19 mol). The flask was flamed dried and flushed with argon. Ether (20 mL) was added to the reaction flask, followed by 16.1 g of 5—chloro-1—pentene in 20 mL of ether at a rate where the solution re— fluxed gently. After the addition was completed, the solution was heated at reflux for 2 h. The solution was cooled to -100C and 6.4 g of acetaldehyde in 20 mL of ether was added dropwise. The addition must be done slowly enough to insure that the solution never exceeds 00C. The mixture was stirred for 2 h at 00C after acetaldehyde addition was completed. Ice (50 g) was added to the solution and the mixture acidified with 10% H SO The reaction mixture was extracted with 3 x 2 4' 10 mL of ether. The ether layers were combined and the ether was re- moved under reduced pressure. The residue was distilled through a short path distillation apparatus (20 mm, 62-63OC). About 13.8 g of 1—heptene— 2-ol (76% yield) was recovered: NMR 0 1.17 (3 h, d, J=7), 1.43 (4 h, bm), 1.8 — 2.2 (2 H, bm), 2.5 (1 H, bs, -0H), 3.7 (1 H, bm), 4.78 (l H, m), 5.0 (1 H, m), 5.4 - 6.1 (1 H, m). Synthesis of 2—Bromo—6—heptene from 1-Heptene—6-ol. A solution of 14.1 g l—heptene-6-ol and 30 mL dry ether was added by syringe into an ice bath cooled 50 mL round-bottomed flask containing 130 a stirring bar, septum side arm and 15.8 g (5.5 mL) of 95% phoshorus tri— bromide (56 mmol). After the alcohol was added, the solution was stir- red at 00C overnight. Then 20 mL of cold staurated Na2003 was added slowly (vigorous CO evolution) to the reaction mixture. The mixture was 2 extracted with 3 x 30 mL of cold ether. The extracts were pooled and dried over anhydrous K2CO3. The ether was removed under reduced pressure and the residue distilled by short path distillation (50 mm, 46-470C). The yield was 58% (9.6 g, 98% pure by GLC): NMR 6 1.95 (2 H, sep, J=7), 4.0 (1 H, sex, J=6), 4.73 (l H, m), 4.95 (l H, m), 5.35 - 5.95 (l H, bm), 13C NMR (0 = ppm relative to TMS) 0 26.47, 27.0, 33.05, 40.56, 51.37, 114.94, 138.19; mass spectrum, m/e (relative intensity) 179 (M3), 177 (M3), 97 (82), 81 (57), 69 (41), 55 (100), 54 (70), 41 (78), 32 (63). Reaction of 2-Bromo-6-heptene (46) with With Hindered Lithium Amides. The reaction of 3 mmol of_46 (0.44 mL) with all the lithium amides in Table IX was carried out as described previously with 2-bromobutane. Analysis for heptadienes and either 1,2-dimethylcyclopentane or methyl— cyclohexane Was carried out by GLC (40' x 1/8H aluminum column filled with 100 mesh Chromosorb W AW—DMSC coated with 20% SE-30). Product ana— lysis was performed by coinjection of authentic heptadiene samples from Chemical Samples Co. Product analysis of 1,2-dimethylcyclopentane and methylcyclohexane was by mass spectrometry and sample coinjection: NMR analysis of sample gig and Eran§§1,5-heptadiene mixture; 6 1.6 (3 H, t, J=2), 2.03 (4 H, s), 2.06 (2 H, bm superimposed over s), 4.75 (l H, m), 5.0 ( l H, m), 5.3 (2 H, m), 5.4 — 6.1 (l H, m); mass spectrum m/e (re— lative intensity) 96 (M+), 81 (31), 67 (43), 55 (76), 54 (100), 41 (29). 131 NMR 1,2-dimethy1cyclopentane 6 0.8 (3 H, s), 0.93 (3 H, s), 1.0 — 2.0 (8 H, bm); mass spectrum,m/e (relative intensity) 98 (M+), 83 (19), 70 (90), 69 (36), 56 (100), 55 (77), 42 (35), 41, (78). Synthesis of Deuterium Bromide. In a 100 mL flask fitted with a stopcock septum side arm was placed 2.5 g of red phosphorus and sufficient D20 to thoroughly wet the flask walls. The flask was evacuated to less than one micron and then 5.1 mL of 99.8% D20 was added, followed by 30 g (0.37 mol) 0f Brz, added dropwise. The bromine reacted with considerable violence; flashes of light occured as it hit the solution. Addition required about 2 h. The DBr was collected in a dry ice—acetone cooled trap and then transfered to a vacuum line system and distilled through a —1100C trap (pentane, liquid N2) into a liquid nitrogen cooled one. The material in the li- quid nitrogen flask was condensed onto about 0.5 g of 1—octene to remove traces of elemental bromine. Four more distillations through a -1100C trap gave 98 mmol of DBr which was condensed into a thick walled py— rolysis tube and sealed under vacuum. The yield was 8.6 g ( 41% yield based on 0.24 mol of PBr3). Threo and Erythro—3—gf2—Bromobutane Synthesis, Into an evacuated quartz flask fitted with a Teflon stirring bar and connected to a high vacuum line was condensed 10.9 g (119 mmol) of DBr and 5.54 g (99 mmol) of Eisfbutene using liquid nitrogen. The flask was covered with aluminum foil and the contents degased by warming to -780C, cooling to —1980C, and applying high vacuum (<1 micron). This 132 was repeated twice. The reaction mixture was maintained at between -105 and -950C using a 3 L Dewar flask filled with pentane and cooled by periodic addition of liquid nitrogen. The pentane cooling solvent was stirred using a magnetic stirring bar. The reaction mixture was ir— radiated with a Hanovia medium pressure utility mercury arc lamp at a distance of between 12 and 18 inches. Initially, no reaction occured; however, a rapid exothermic reaction soon set in as observed by a pres- sure rise in the reaction flask. The pressure was maintained below 25 mm by adusting the distance of the mercury lamp. At no time should the cooling bath temperature rise above —9OOC. After the reaction was com— plete, irradiation of the reaction mixture caused no pressure increase. The irradiation required about 15 to 25 min. The mixture was distilled through a -100 OC trap to remove excess DBr. The liquid remaining in the —1000C trap was removed from the vacuum line and carefully washed with aqueous K2C03 and distilled by a short path distillation to give 10.6 g (80% yield) of Ehrggf3—d72—bromobutane, bp 910C. GLC analysis showed the product to contain less than 1% non—bromobutane impurities. A dehydrohalogenation procedure (vide infra) gives the stereoisomeric purity of the thrggfsample. The erythro-3—d—2—bromobutane diasteriomer was prepared in a sim— ilar manner from 60 mmol of DBr and 45 mmol transz—butene. The yield of the erythro diastereomer was 4.5 g (33 mmol, 73%). The same dehydro— halogentation procedure as above was used to determinine the stereoiso- meric purtity of the erythro isomer. 133 Determination of the Stereoisomeric Purity of Threo and Erythro-3jdf 2-Bromobutane, To two mL of dimethyl sulfoxide (reagent grade), containing 400 mg potassium Efbutoxide was added 200 pl (1.84 mmol) threo-Bjde-bromo- butane. The reaction mixture was kept at 300C during the reaction. Cases were evolved and were collected in a liquid N trap. The product 2 butenes were condensed into an evacuated 50 mL Schlenk tube containing 1 mL of THF. The butenes dissoved in the THF and the solvent mixture was injected into the GLC. The butene isomers were separated and col— lected in glass capillary tubes. The prOportion of butenes consisted of 36.2% l-butene, 57.8% transz-butene and 5.9% gist-butene. The yield was not determined. The same dehydrohalogenation reaction was repeated with the er - _Ehrg isomer. The proportion of butenes consisted of 54.0% l-butene, 22.4% Sing—butene and 23.6% transz-butene. The olefins were analysed with a Finnigan EI-CI GLC - mass spec- trometer at 11.4 eV. The mass spectrum of C at this electron energy H 4 8 consists solely of a peak for the parent ion; there were no P—1 or P—2 peaks. Table XX lists the results of the deuterium content analysis. Deuterium percentage was computed using the following formula. [Mass 57 - (Mass 56) (0.042)] 100 Percent Deuterium = ‘ Mass 56 + Mass 57 - Mass 56 (0.042) 134 Table XX. Mass Spectra of 2—Butenes from 3fid72-Bromobutanes Olefin Source Mass 56 Mass 57 Mass 58 Deuterium cis; Matheson 100 4.2 O 0% .EEEEEI .Ehregfsample 7.34 100 0 93.1% SEE? Ehrggfsample 100 26.72 0 18.4% Eransf erythro-sample 100 7.74 0 3.4% cis- erythro-sample 7.76 100 O 92.8% The composition of these butene samples can be used to determine the percentage of undeuterated and diastereotopic impurities in the threo and erythro—3fdf2—bromobutane samples. Assuming that the dehydrohalo— genation reactions proceed quantitatively with perfect trans-stereo- . O O 47 spec1f1c1ty, a pure erythro-sample should produce trans-2-butene—dO and cis—2—butene—d Similarly, a pure threo—sample should produce 1' transz-butene-dl and Sing-butene—do The first three rows in Table XXI give the experimental data for butene composition. These are taken from the butene compositions of the Efbutoxide mediated dehydrohalogenation of £hggg_and erythro-3fid72-bro- mobutane and from the deuterium analysis in Table XX. In Table XXI, deuterated compounds are listed in separate columns and l-butene is nor- malized to 100. NN.NH o o ass cos w mafia 664462462 AaV se.os o o m.wss sw.aw Awe mmqm. .m Aqumo .m .mqm AH messy Aomo.ov AAV mo.ms o sa.w m.mss qw.ma Ass .wmqm AWHMV mmqm. mmq. mmqm. Am 6:44 v Asaao.ov Ame m.m~ a.~ is 8.mss cos mamasmmemmm Ass mamammmemmw mung wow coaumasoamo cowuflmomaou .5 . ‘ mu H.m m.oq ow Hm.H OOH mHaEmmlowfiuNwm AmV m.ms a.N as s.wsz cos maaamm.mmmmmw Ame as men cos mamusnoaounuw Aav owlmmw. lemmm. ovlmmmmm. HwImcmuu mamusnlfi ocmusnoEomeNLwrmlomwfiH paw ousuwsm mo coaumcmwoamsowpxsmm %£ pmospoum mucousm .Hxx manmw 136 AN.NHV a.os o o Aa.moav o.msa Acosv o.nw muamumuuas muwamauoz Aomv o.: mm a $1.8 m MN. 2 2:: 338.8 8: mi 0 Tm 8.9: 8.8 a: 8.: m Ame ad. a 164m 5 9:: 338.8 C: MEAN mi m.~ : 6.9: 2: s as: G: coHuHmanoomemmwmm.pmuomuuoo wcwms .maaemmlomunu wow :owuwmanoo wumasuamomm 8.3 3.: o 2: .2 6:: 326.562 3: o 93 8.3 o 98 3: 3.2 34.3 .oJm «1.91 m a 3 9:: A8298 8: o.m was 3 o 2.8 a: 3 m m Gland Ham 8 9:: 328.8 2: mam: in new 3 2; 2: maaemmuflflflmm 8: madamm MMMMNMM mung HOW coHumHSUHmo coaufimanou cam 66-4...an Heumlm 1 6613mm“: Hana 6:323 A.Dcoov Hxx magma 137 Since cis—Z—butene—d should only be produced from the dehydro- l halogenation of erythro-3—df2—bromobutane, any present in the three sample must be due to erythro impurity. The erythro sample experimen— tal data (row 3, Table XXI) was multiplied by a factor (0.0716) that was chosen so that substraction of these products from the threo sample experimental data (row 4) would leave no ging-butene-dl. Row 6 gives the difference. transz—Butene-do must arise from either erythro-3—df 2—bromobutane or undeuterated 2-bromobutane. Since all the erythro— contribution has already been removed from line 6, the remainder of the transz—butene must have come from undeuterated Z—bromobutane. Multiplication of the 2-bromobutane data (row 1) by (0.05) and subtra— tion from row 6 removes all transz—butene dO contribution. The re— sults (row 8) relates the composition that would be obtained by dehy- drohalogenation of a diastereomerically pure sample of threo—3jd—2—bro- mobutane. In line this composition is normalized such that l—butene is 100. A similar process was used on the erythro experimental data. The process was completely analogous to that used for the threo data, except that the corrected threo data (line 9) was used instead of the raw data. (Line 10) - (0.00775) (line 9) gives line 12 in which all three contribution is removed. (Line 12) — (0.06383) (line 1) gives line 14, the composition that would come from a pure erythro-sample. Line 15 is this composition normalized such that l-butene is 100. Assuming all dehydrohalogenations are quantitative, the compo— sition of the erythro sample can now be determined. Let the sum of all butenes in any line of Table XXI be denoted by the symbol Sbn where n is the line under consideration. Sb10 represents all material obtained _K— 138 from the erythro sample. Sbll is the threo contribution, hence the percentage of threo present in this sample is 1008b11/Sb10° Similarly. the percentage undeuterated 2-bromobutane is lOOSb13/Sb When nu- 10° merical values from Table XXI are substituted into these expressions, it is found that this sample contained 1.2% £2329 and 10.7% undeuterated 2—bromobutane with the remainder consisting of erythro. The data in lines 4 through 8 cannot be used for the determina- tion of the thggg sample composition because the erythro data that was subtracted contained contributions due to both £h£32_and undeuterated 2—bromobutane. However, when the corrected erythro composition (line 15) was used instead of the experimental data (line 3), the composition could be determined. Lines 16 through 20 are analogous to lines 4 through 8, except corrected erythro data was used in this computation. The percentage erythro present is lOOSb17/Sb16 and the percentage undeuter— ated 2-bromobutane is given by the expression lOOSb19/8b16. When numer- ical values are substituted into these expressions, it is found that this sample contained 4.2% erythro and 6.1 Z undeuterated 2-bromobutane with the remainder consisting of threo. Determination of the Kinetic Isotope Effect for Dehydrohalogenation of Threo and Erythro-3jdf2-Bromobutane with Hindered Lithium Amides in THF. The primary kinetic isotope effect may be computed using the assumption that l—butene is formed at the same rate in 2-bromobutane and its 3—deuterated counterparts. A second assumption is that the E2 reaction goes to completion with the deuterated isomers. The composi- tion of the butene products from deuterated bromobutanes and 2-bromo 139 butane were normalized to l—butene as 100. The kinetic isotope effect was calculated for each lithium amide dehydrohalogenation reaction as els-Z-butene(undeuterated)/ClS-2-do erzthro) and trans—Z—butene(undeu_ /trans—2—butene—d The kH/kD value calculated was terated) o (threo)' corrected for deuterium content; 12% d0 for the erythro isomer and 10% for the threo isomer. Beaction of Lithium Amides with Threo and Ervthro 3-d—2-Bromobutane in THF at 0°C. The alkyl lithium amides were prepared as previously described. For the threo study, the reactions were run on a 3 mmol scale while the erythro-study was performed on a 2 mmol scale. Threo (325 pl) and ery— thro (216 pl) 3jd72—bromobutane reaction were run for 1 day using the less hindered amides ( diethylamide through géfLi). Amides 26 through _lfLi were reacted for periods described in Table X. At least 3 injec- tions of each sample were made and the cis/trans—Z—butene ratios were performed with at least one peak adjusted to full scale deflection on the strip chart recorder. This was done to minimize the signal to noise ratio. Reaction of Alkyl Lithium Amides with 2-Bromobutane in THF — lZ—Crown—4 Ether Solution at —78°c. The alkyl lithium amides were prepared as before (3 mmol). After hexane solvent was removed, the reaction flask was cooled to 00C and 1.5 mL of THF was added to the amide. After the amide dissolved (about 1 min), the flask was cooled to -780C. Then 1.5 mL of standard 2.0 M 140 solution of 12-Crown-4 in THF was added by syringe within 30 sec. The solution was stirred at -780C for 1 min and immediately 0.33 ml of 2- bromobutane was added by syringe. The mixture was stirred at -7800 for 30 min, then warmed to Doc for 2 h. The sample aliquots were ana— lyzed by GLC using pentane internal standard. Care should by taken not to add pure 12-Crown-4 to the amide solution at 00C, since the crown ether greatly accelerates metallation of THF. Pure 12-Crown-4 added to THF at -780C will not dissolve, even after 2 h. BIBLIOGRAPHY l) 2) 3) 4) 5) 6) 8) 9) 10) 11) 12) 13) 14) 15) 16) 17) BIBLIOGRAPHY G. Scott, Pure and App. Chem., 52, 365 (1980). M. W. Rathke and R. Kow, J. Am. Chem. Soc., 24, 6854 (1974). a. R. A. Olofson and C. M. Dougherty, J. Am. Chem. Soc., 25} 581 (1973). b. R. A. Olofson and C. M. Dougherty, J. Am. Chem. Soc., 25, 582 (1973). G. F. Hennion and R. S. Hanzel, J. Am. Chem. Soc., 82, 4908 (1960). N. C. Deno, R. Fishbein, J. C. Wyckoff, J. Am. Chem. Soc., 23, 2065 (1971). Y. L. Chow, W. C. Danen, S. F. Nelsen, D. H. Rosenblatt, Chem. Rev., 18, 243 (1978). C. A. Brown, J. Chem. Soc., Chem. Commun., 680 (1974). G. F. Hennion, C. V. DiGiovanna, J. Org. Chem., 30, 2645 (1965). I. E. Kopka, "The Synthesis and Hydrogenation of Sterically Hindered Secondary Amines," M. S. Thesis, Michigan State University (1978). T. H. Vaughn, R. R. Vogt, J. A. Hieuland, J. Org. Chem., 2, l (1937). M. M. Midland, J. Org Chem., 49, 2250 (1975). N. R. Easton, R. D. Dillard, W. J. Doran, M. Livezy and D. E. Mor— rison, J. Org. Chem., 26, 3772 (1961). Similar reactions have been observed for sodium/NH reduction of related propargylamines: G. F. Hennion, R. H. Ode, J. Org. Chem., _31, 1975 (1966). M. Calvin and K. W. Wilson, J. Am. Chem. Soc., 61, 2003 (1945). G. G. Vexlearschi, Bull. Soc. Chim. Fr., 589 (1956). a. N. F. Hall and M. R. Sprinke, J. Am. Chem. Soc., 54, 3469 (1932). b. H. K. Hall, Jr., J. Am. Chem. Soc., 12, 5441 (1957). H. 2. Zimmer, H. I. Lipp and L. L. Jackson, J. Org. Chem., 36, 824 (1971). 141 18) 19) 20) 21) 22) 23) 24) 25) 26) 27) 28) 29) 30) 31) 32) 33) 34) 35) 36) 37) 38) 142 Z. Fataftah, "The Synthesis and Reaction of a Hindered Secondary Amine," Ph.D. Dissertation, Michigan State Univerity(1980). N. C. Deno and D. G. Pohl, J. Am. Chem. Soc., 26, 6680 (1974). N. C. Deno, D. G. Pohl and H. J. Spinelli, Bioorg. Chem., 31, 1977 (1974). G. F. Hennion and C. V. DiGiovanna, J. Org. Chem., 31, 1977 (1966). H. C. Brown and B. J. Kanner, J. Am. Chem. Soc., 88, 986 (1966). P. D. Bartlett, M. Roha and R. M. Stiles, J. Am. Chem. Soc., 16, 2349 (1954). H. K. Hall, Jr., J. Am. Chem. Soc., 22, 5441, (1957). F. Klages and H. Sitz, Chem. Ber., 26, 2394 (1963). P. G. Gassman, Acc. Chem. Res., 3, 26 (1970). R. M. Keller and H. D. Wycoff, Inorg. Synth., 2, 1 (1946). G. F. Hennion, E. G. Teach, J. Am. Chem. Soc., 15, 1653 (1953). G. F. Hennion, A. P. Boisselle, J. Org. Chem., 26, 725 (1961). R. L. Augustine, "Catalytic Hydrogenation"; Marcel Dekkar, New York 1965. a. M. Hamell and R. Levine, J. Org. Chem., 15, 162 (1950). b. R. Levine, Chem. Rev., 54, 467 (1954). . S. Raynold and R. Levine, J. Am. Chem. Soc.,_§2, 472 (1960). O M. W. Rathke and A. Lindert, J. Am. Chem. Soc., 23, 2318, 4605 (1971). H. 0. House, M. Call and H. D. Olmstead, J. Org. Chem., 36, 2361 (1971). M. W. Rathke and R. Kow, J. Am. Chem. Soc., 24, 6854 (1972). R. A. Olofson, K. D. Lotts and G. N. Barber, Tetrahedron Lett., 3779 (1976). U. Schollkopf, F. Gerhart, I. Hoppe, R. Harms, K. Hantke, K. H. Scheunemann, E. Eilers and E. Blume, Ann., 183 (1976). C.M. Dougherty and R. A. Olofson, J. Am. Chem. Soc., 25, 581, 582 (1973). I. Fleming and T. Mah, J. C. S. Perkin 1., 964 (1975). 39) 40) 41) 42) 43) 44) 45) 46) 47) 48) 49) 50) 51) 52) 53) 54) 55) 56) 57) 58) 59) 60) 61) 143 R. West and P. C. Jones, J. Organometal. Chem., 2, 165 (1967). C. A. Brown, J. Org. Chem., 32, 3913 (1974). T. F. Crimmins and E. M. Rather, J. Org. Chem., 43, 2170 (1978). H. L. Lewis and T. L. Brown, J. Am. Chem. Soc., 22, 4664 (1970). J. Mallan and R. Bebb, Chem. Rev.t§2, 693 (1969). J. F. Nobis and L. F. Moormeier, Ind. Eng. Chem., 539 (1954). F. G. Bordwell, Acc. Chem. Res., 5, 374 (1972). W. H. Saunders, Jr., Acc. Chem. Res., 8, 19 (1975). R. A. Bartsch, Acc. Chem, Res., 8, 248 (1975). E. Baciocchi, Acc. Chem. Res., 12, 430 (1979) R. A. Bartsch and J. Zavada, Chem. Rev., 89, 453 (1980). J. F. Bunnett, Angew. Chem. Int. Ed.,1, 225 (1962). D. J. McLennan, Quart. Rev.,_21, 490 (1967). R. W. Alder, R. Baker and J. M. Brown, "Mechanism in Organic Chemistry"; Wiley-Interscience, London 1971. C. H. DePuy, G. F. Morris, J. S. Smith and R. J. Smat, J. Am. Chem. Soc., 81, 2421 (1965). H. Kwart, T. Takeshita and J. L. Nyce, J. Am. Chem. Soc., 86, 2606 (1964). R. A. Bartsch, G. M. Pruss, B. A. Bushaw and K. E. Wiegers, J. Am. Chem. Soc., 25, 3405 (1973). R. A. Bartsch and J. F. Bunnett, J. Am. Chem. Soc., 29, 408 (1968). M. Pankova and J. Zavada, J. Collect. Czech. Chem. Commun. 42, 3439 (1977). R. A. Bartsch, R. Read, D.T. Larson, D. K. Roberts, K. J. Scott, B. R. Cho, J. Am. Chem. Soc., 101, 1176 (1979). R. A. Bartsch, B. R. Cho and J. C. Day, J. Org. Chem., 45, 4057 (1980). R. A. Bartsch, D. K. Roberts and B. R. Cho, J. Org. Chem.,fifi, 4105 (1979). R. S. Tipson, M. A. Clapp and L. H. Cretcher, J. Org, Chem., 12, 133 (1947). 62) 63) 64) 65) 66) 67) 144 M. Julia, Acc. Chem. Res., 4, 386 (1971). R. R. Pavlis, "Evidence for Bridged Haloalkyl Radicals and Ions," Ph. D. Dissertation, Pennsylvania State University, (1969). R. A. Bartsch and T. A. Shelly, J. Org. Chem., 88, 2911 (1973). C. L. Bumgardner, J. R. Leves and S. T. Purrington, J. Org. Chem., 48, 748 (1980). J. Jullien and A. Paillous, Bull. Soc. Chem. Fr., 8, 2236 (1965). M. P. Dreyfuss, J. ng. Chem., 28, 3269 (1963). "1111111111111111111111“ 31239