I'Z'c This is to certify that the . are, a thesis entitled 519’s 3 ~19” Lu" 5’ S CWWL H cicticjfmcd 10m (37/ ., C: Q t , - . 3 +6 \‘I < QLCW H1 “(CLQCL Scf‘c‘thICl 0k, /) but lid? 3 \ presented by :1: 1/11; 151/145 Ko101< 191 has been accepted towards fulfillment , of the requirements for (\H“ v” 1 ‘1 ,4, degreeln \_ ;(_l I] 1/// // / ~ », 777777 /"//>)1 éu%(LL / { (Lg/C A- L / Major professor DatesL CW '3 V 0-7639 THE SYNTHESIS AND HYDROGENATION OF STERICALLY HINDERED SECONDARY AMINES BY Ihor Elias Kopka A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Chemistry 1978 ABSTRACT THE SYNTHESIS AND HYDROGENATION OF STERICALLY HINDERED SECONDARY AMINES BY Ihor Elias Kopka Tertiary propargylic chlorides react with primary propargylic amines giving hindered bispropargylic secondary amines in good yield. Thus 1,1,1',1’-tetraethyl-di-2-pro— pynylamine l was synthesized using a 1:2 molar ratio of l-chloro-3-ethyl-l-pentyne 2 with 3-amino-3-ethyl—1-pentyne % for three days at 4°C in a DMF solution containing Cl catalytic amounts of Cu and COpper bronze powder. 2 2 Hydrogenation of l to the diallyl secondary amine 1,1,1’,1' tetraethyl-di—l—propenylamine Z and the saturated secondary amine l,l,l,l',l’,l’ hexaethyl-di-methylamine 1Q was investigated. Platinum dioxide hydrogenolyzed 1 completely upon low pressure hydrogenation in ethanol. Hydrogenation of 1 with 10% palladium on charcoal gave two heterocyclic amines; 3,4-dimethy1-2,2,5,5—tetraethyl—3-pyrroline g and 3-methylene-4—methyl-2,2,5,5-tetraethyl—3-pyrrolidine 2. Semihydrogenation of l with 10% Pd/C in ligroine gave Z in fair yield. Different Raney nickel catalysts were tried in hydrogenating l. The best yield of 1g was obtained when l was hydrogenated in ethanol containing W2 Raney nickel and a 2:1 ratio of potassium hydroxide to l. To My Parents, without whose guidance and support I would not be the man I am today ii ACKNOWLEDGEMENTS The author wishes to thank Dr. Michael W. Rathke for his guidance, advice and warm personal interest in this work throughout the course of this investigation. Appreciation is extended to Dr. William Reusch for his help in interpreting spectral data and to Bruce Osterby for his invaluable help in taking a number of 13C NMR spectra. The author also extends appreciation to his parents for their continuing confidence, love and support in him. Finally, the financial support of Michigan State University and the National Science Foundation is grate— fully acknowledged. iii TABLE OF CONTENTS INTRODUCTION . . . . . . . . . . . . . . . LITERATURE REVIEW . . . . . . . . . . . . . RESULTS . DISCUSSION. . . . . . . . . . . . . . . . . . . . . . EXPERIMENTAL. . . . . . . . . . . . . . . . . . . I. Materials . . . . . . . . . . . . II. Preparation of Tertiary PrOpargylic Alcohols . . . . . . . . . . . . . . . . A. General Procedure. . . . . . B. Product Analysis . . . . . . . . . III. Preparation of Tertiary Propargylic Chlorides. . . . . . . . . . . . . A. General Procedure. . . . . . . . . . . . B. Product Analysis . . . . . . . IV. Preparation of Primary Propargylic Amines . . . . . . . . . . . . . A. General Procedure (Hennion's Sodamide Me thOd) o o o o o o o o o o o o o B. General Procedure (DMF with Ammonia Method) . . . . . . . . . . . . . . . C. Product Analysis . . . . . . . . . . . . V. Preparation of Primary Allylic Amines . . . A. General Procedure. . . . . . . . . . B. Product Analysis . . . . . . . . . . . . iv Page 18 32 57 57 58 58 60 61 61 62 62 62 64 65 65 65 66 TABLE OF CONTENTS (Cont.) VI. Preparation of Hindered Bis-Propargylic Secondary Amines . . . . . A. Product Analysis VII. Hydrogenation of 1,1,1’,l'-Tetraethyl- di-2-propynylamine ‘1 in Absolute Ethanol With 10% Palladium on Charcoal . . . . . . A. General Procedure. B. Product Analysis VIII. Hydrogenation of l,l,l',1’-Tetraethy1— di-2-propyny1amine 1-in Absolute Ethanol With Platinum Oxide. . A. General Procedure- IX. Hydrogenation of 1,1,1’,1’-Tetraethyl- di-2-propynylamine 1 to Bis(1,1- diethylallyljamine Z in Ligroine With 10% Palladium on Charcoal . A. General Procedure. B. Product Analysis X. Hydrogenation of l l,1’,1’-Tetraethy1- di-2—propynylamine in Absolute Ethanol With Raney Nickel Catalyst A. Preparation of Raney Nickel Catalysts- 1. W2 Raney Nickel. 2. W4 Raney Nickel. 3. W6 Raney Nickel. . . . . . . B. General Procedure. . . . . . .7. C. Product Analysis BIBLIOGRAPHY. Page 67 68 69 69 7O 70 7O 71 71 72 72 72 72 73 74 74 76 77 LIST OF TABLES TABLE Page I. Yields of 2° Propargylic Amines Under Various Reaction Conditions. . . . . . . . . 26 vi INTRODUCTION Organic bases which are efficient proton abstractors but poor nucleophiles are generally valuable in synthetic organic chemistry. Metal salts of sterically hindered 2° amines, particularly the lithio derivatives of di—tert— butylamine , 2 , 2 , 6 , 6-tetramethylpiperidine and other com- pounds of the general structure LiNR2, where R is a bulky aliphatic group, are good non-nucleophilic bases. The most hindered secondary amines are reagents with the nitrogen atom attached to two tertiary carbons. They are also the most difficult amine bases to synthesize in a convenient and efficient manner. There are very few Viable general synthe- tic routes for making secondary amines which are more hindered than di—tert-butylamine. Therefore, efforts are still being made to find convenient synthetic methods for preparing secondary amines more hindered than di—tert-butyl— amine. In this thesis we will present a very simple and gener- al method for the synthesis of saturated secondary amines of unprecedented steric bulk from tertiary acetylenic chlorides and the corresponding primary acetylenic amines. Also discussed is a method for greatly reducing hydrogeno- lysis of the bulky secondary acetylenic amines, precursors 2 to the saturated secondary amines, by catalytically hydro- genating them in alkaline alcoholic suspension of Raney nickel. Since less hindered secondary amines have been made previously by coupling propargylic chlorides with satura- ted primary amines, experiments were performed with satur— ated and unsaturated primary amines to determine what effect the degree of unsaturation of the amines had on the formation of hindered secondary amines. LITERATURE REVIEW A. The Synthesis of Hindered Secondary Amines There have been many reactions developed for synthe— sizing secondary amines but only a handful are amenable to the synthesis of amines containing two tertiary carbon atoms attached to a secondary nitrogen center as, for example, di-tert-butylamine 11. Primary amines, when reacted with tertiary alkyl halides, give the elimination product almost exclusively rather than the coupled amine. (CH3)3-C-NH—C—(CH3) - _ _ ll (CH3)3 CNH2 + (CH3)3 C C1 __ CH3 H c=é Eq° 1 2 I CH3 A generally useful method for the synthesis of amines is treatment of an aldehyde or ketone with ammonia or a pri- mary or a secondary amine to form an intermediate imine which may be subsequently reduced to an amine by a number of reducing agents. By carefully choosing the starting aldehyde or ketone and amine, one can obtain good yields of secondary amine. In general, secondary amines cannot 4 be usefully prepared with aldehydes of less than five car- bon atoms. The best yields are reported for aromatic alde- hydes, presumably because of the greater ease of imine (Schiff base) formation. Secondary amines can be prepared by two possible procedures: 2 moles of ammonia and 1 mole of aldehyde or ketone (Eq. 2), or 1 mole of primary amine and 1 mole of carbonyl compound (Eq. 3 ), the latter method being better for all but aromatic aldehydes. u.) 9 Red — — (R=H,alkyl)»NH2 CHR2 OH NH3 + O=CR2 3.2 CR q ' 2 Red .__9 = + o=CR2—->(___NH(CR2) (c|:R2)<__, N CR2 ——>NH(CR2) 2 OH NHz-CHR2 Red + O=CR2 (R=H, alkyl) (:2 RNH-CRZ <_---_:>_RN=CR2 ———> OH RNH2 Eq. 3 ____> RNHCR2 Many Of these general observations were taken from two reviews of the chemistry of the amine group.26 The formation of secondary amines works well only for long chain aldehydes, aromatic aldehydes and ketones,and ali— phatic ketones that are not hindered about the carbonyl carbon. The intermediate imine rapidly decomposes or dis- prOportionates with hindered ketones to give polymers when there is not at least one phenyl group attached to the 5 nitrOgen atom. Aromatic imines are much more stable than purely aliphatic imines because conjugation increases the thermodynamic stability of the azomethine linkage. Imines derived from aliphatic primary amines and enolizable or non-enolizab1e ketones and aldehydes will add organolithium compound across the azomethine linkage to give hindered secondary amines.l’5 1) n-BuLi (9H2)3CH3 (CH3) 2CHCH=N-CH (CH3) 2 “aw—9 (CH3) Z-CH-CHNH-CH- (CH3) 2 2) H 0 2 12 13 (60%) 7“ Eq. 4 1) n-BuLi (C6H5)(CH3) C=N-CH2-R 2) (IE—gm“) (C6H5)(CH3)-C-NHCH2-R 2 (CH2)3CH3 Eq. 5 14a R=(C6H5) 15a R=(C6H5) (34%) 14b R=i-propyl 15b R=i-propyl (0%) However, if an aliphatic R group is substituted for an aromatic substituent in Eq. 5, there is no addition across the (:>C=N-) bond.2 Another complication arising from the reaction of organolithium compounds with aldimines and ketimines is LiH elimination of the aminolithium compounds; the imines react with organolithium compounds to give secon- dary amines with branched alkyl groups. 20°, 24 hr. (CH3)2CH-CH=N-CH(CH3)-C6H5 + 2 n-BuLl ; ether l6 (CH3)2CH-CH-NH-CH(CH3)-C6HS + (CH3)2CH-CH-N=C(CH3)-C6HS + C4H9 C4H9 __z (59%) y; (14%) (CH3)2CH-CH-NH-C(CH3)-C6H5 C4H9 C4H9 Eq. 6 _1__9_ (10%) The Schiff bases of some a-substituted aldehydes and ketones, in addition to being difficult to make, tend to metalate at the former a-carbonyl carbon position giving an imine eno- late3 rather than adding the organolithium reagent across the carbon-nitrogen double bond. CH3 CH \ I 3 . 21 C=N-C-CH + CH L1 __ CH / ' 3 3 3 CH3 Eq. 7 CH3.\ 9H3 29 . /C=N-C-CH3 LiCH2 CH3 22 In many cases, heating 1° amines with catalytic amounts of strong base forms the corresponding secondary amines.4 7 Refluxing primary amines and NaH gives mixtures of 2° amines as well as unreduced imines. NaH 90-13o° H H H CH3 PhCH(CH )NH ————> Ph+—NH+—Ph + Ph' NH Ph 3 2 4 hr. T— xylene CH3 CH3 CH3 Ii 22 24 (15%) 25 (22%) Eq. 8 CH I 3 + PhC=NCHPh CH3 26 (14%) This procedure suffers from the limitations that yields are generally low, significant quantities of unreduced imine remain in the reaction mixture and attempted coupling of two different primary amines generally give mixtures of all possible cross-coupled secondary amines. Another useful technique for chain extension at the d—carbon of secondary amines is to convert the amine to its N-nitroso derivative. The d-alkylated N-nitroso product is easily hydrolyzed to the product amine.6 R' R' . . R' | HONO l (1—Pr) 2NLIL e l RZCH-NH -—-> R2CH-N-NO > RZC-N-NO Eq. 9 R” R! I! I R"X , . 1) H+ I} If ——9 R2-C—N--NO -—-€> RZC—NH 2) 0H Though fairly hindered amines can be made in good yield by this method, there is an obvious safety hazard when handling carcinogenic nitrosamines. Grignard reagents react with imines to form addition products which on hydrolysis give hindered secondary amines. The reaction is usually applied to Schiff bases prepared from aryl halides. The reactions with Grignard reagents provide a general synthetic method for secondary amines of the type RR’CHNHR" where R is an aryl group. Sterically hindered reactions of Grignard reagents with Schiff bases have been studied.8 N-Benzylidene-t-butyl- amine 21 reacts with allyl-magnesium bromide; however, methylmagnesium iodide does not react with 22, even under forcing conditions. CH =CH-CH MgBr _ 2 2 x _ _ _ C6H5CH—N-t-Bu , C6H5CH NH t Bu CHZ-CH=CH2 MeMgI C H CH=NCH -—————9' No Reaction 6 5 3 29 9 Various imines derived from aliphatic primary amines and enolizable aldehydes or ketones undergo complete enoli- zation with one equivalent of alkylmagnesium compound in THF.8 The resulting enamines react with alkyl halides giving addition products which hydrolyze to a-substituted aldehydes or ketones rather than the hindered secondary amine (Eq. 11a). EtMgBr F‘Bu C6H5CH2C1 (CH3) 2HCCH=N-t-Bu ———> (CH3) 2C=CHN-MgBr > THF CH I 3 H :; CH3-C-CH2N-t-Bu CH2C6H5 9H3 H o CH3-C-CH=N-t-Bu El Eq. 11a CH C H 2 6 5 Sharpless et al9 found that aza analogues of selenium dioxide effect allylic amination of reactive olefins. Unfortunately, amination of less reactive olefins gave poor yields of hindered allylic amines. 10 CH 3 I N—C-CH3 (CH3)3C-N—Se—NC(CH3)3 H CH + > 3 22 23 (62%) Eq. 11b The nitrogen insertion reaction is similar to the allylic insertion of oxygen into olefins by selenium dioxide. These aminations probably occur via the same sequence of -ene and [2,3]-sigmatropic reactions proposed for the 10 analogous oxo-process. Alkyl boranes R’BB react with organic azides, RN3, in benzene or xylene to give R’RNBR’ which is readily con— 2 verted by alkaline hydrolysis to the corresponding secondary amine R/NHR I1 The reaction becomes quite slow with steri— cally hindered azides. It fails completely when both steri- cally hindered azides and sterically hindered organoboranes are used. These results are interpreted by a mechanism involving reversible coordination of the azide with the trialkyl borane. This step is followed by loss of nitrogen from the intermediate with migration of the alkyl group from boron to nitrogen (Eq. 12). + NaOH R'NNEN + R3B‘___——"" R'NBR3-——-) R'RNBR2 + N2 -—-) RR'NH + H 0 NEN 2 Eq. 12 11 A number of dialkylchloroboranes12 and alkyldichloroboranes were prepared14 and treated with organic azides. BHCl :OEt + R-CH=CH 2 2 2 + BC13—>RCH CH BCl + BCl :OEt 2 2 2 3 2 Eq. 13a BH Cl:OEt + 2R-CH=CH —->(RCH 2 2 2 CH2)BCl + OEt 2 2 Eq. 13b These chloroboranes proved to be quite reactive relative to the trialkylborane, even when the alkyl group(s) on the boron derivative and the organic azide were both secondary. This increased reactivity of the dialkylchloro and alkyl- dichloroboranes may be attributed to decreasing steric interference of the alkyl groups on the boron atom as well as to an increase in Lewis acidity of the organohaloboranes, facilitating the coordination of the azide with the boron derivative. The most hindered amine synthesized by this procedure is N-3—hexy1cyclohexylamine (Eq. 14). 2 I NaOH CH CH - .. _ 3( 2)2 C CHZCH3 + C6Hll N3 -———9- -———> H H20 22 EB Eq. 14 [CHz-CH3 [::]//NH-cH CHZ-CHZCH3 (85%) 37 —~_ The use of alkyldichloroboranes provides a highly useful synthesis of fairly hindered secondary amines. This has some synthetic potential since a simple general synthesis of alkyldichloroboranes has been developed.14 The first efficient synthesis of an amine with two tertiary carbon centers bonded to nitrogen was that of di—tert-butylamine 44. This was achieved by the reaction of 2-methyl-2-nitr0propane 22 with sodium metal to generate the presumed unstable intermediate 22 which was hydrolyzed to di-tert-butylnitroxide 49.15 Subsequent mechanistic study established that 49 is formed by the hydrolysis of a compound formulated as sodium di-t-butylhydroxylamine oxide £4.16 44 is considered to arise by a combination of the tert-butyl radical with t-nitro butyl anion radical (Eq. 15 and 16). 2t-C4H9N02 + Na ———9 t-C4H9NO2 -—9»t-C4H9 + N02 38 Eq. 15 13 . H O —- 2 4H9NO2 ——9 (t-C4H9)2NO2 —-> t-C4H9 + t-C 39 Eq. 16 +/0H [(t-C4H9)2-N\O_] ———> (t-C4H9)2N-O' Nitroxide 42 then may be converted to 22 by a reducing mixture of sodium sulfide nonahydrate, elemental sulfur and N,N,dimethy1formamide in the presence of light.17 Direct reduction of the intermediate 22 by this procedure gives di-tert-butylamine in improved yield.18 However, there is a major limitation in the application of nitro compounds as precursors in the synthesis of hindered secon- dary amines. Subsequent work designed to define the scope of the transformation of nitro compounds to secondary amines showed that the reductive conversion of nitro compounds to the corresponding nitroxides is BEE general. The formation of nitroxide in substantial yields was successful only with 22 among the large number of nitro compounds studied.19 l4 CH CH l 3 / 2 CH ’CH -C-N0 + Na° -————9’CH CH -C + 3 2 g 2 1 me 3 2 \\ CH3 9 y CH3 42 EE- Eq. 17 //CH3 CH 'CH -CH 3 2 \CH 3 22 CH I 3 IH3 $H3 CH3-C-N02 + Na° > CHB-C—-N-—CF-CH3 C'2H glyme CH 6 CH 3 3. 3 38 40 Eq. 18 This result severely restricts the usefulness of this reaction as a general method for the synthesis of steri— cally hindered secondary amines. Probably the most thorough investigation of a generally applicable method for the synthesis of sterically hindered amines was initiated by Hennion in the 1950's. He and his co-workers discovered that aliphatic tertiary propargyl chlorides and bromides successfully alkylate alcohols and amines of virtually all classes to produce propargylic ethers and amines, respectively.20 These reac- tions proceed under very mild conditions in alkaline, par— tially aqueous solution and generally give good yields of product. The utility of these reactions is that they 15 achieve nucleophilic substitution at tertiary aliphatic carbon centers, a reaction which is ordinarily difficult. It is generally accepted that the t-propargylic halide reaction involves an intermediate zwitterion-allene- carbene which is stable to proton elimination and quite electrophilic at the tertiary carbon.21 Cl $1 base I slow _ _ RZC-CECH .____ R2—Cczc- __+19 4, fast -Cl “ R2CCEC- RZ—C=C=C: Eq. 19 R2C=C=CHNu and/or RZC—CECH A, r I ~ Nu V Nucleophile It is obvious that the zwitterion-allenecarbene should be an ambident electrophile capable of yielding both propargylic and allenic products; both products have been seen in a number of cases.21d Good evidence for the above mechanism 20g,21 but comes not only from kinetic and product studies also from the fact that the allenecarbene has been trapped by a stereospecific reaction with olefins, a reaction typi— cal of ordinary carbenes.22 There are a number of observations however, not readily reconcilable with the zwitterion-allenecarbene mechanism. Methanol and ethanol give good yields of propargylic ethers 16 whereas alkaline aqueous alcohol solutions containing 50 mole % water produce much more of the propargylic ether than the carbinol. These reactions follow second order kinetics, first order each in t-propargyl halide and base. tert-Butyl alcohol, however, does not give any ether pro- duct despite the fact that steric inhibition of tert-butyl allenyl ether should not be serious.20e This is in sharp contrast to reactions with t-alkyl and other sterically hindered amines which give N-tert-propargylic amines in 20d120f'23 It was later discovered reasonably good yields. that some primary and secondary saturated aliphatic amines could be used in place of strong base to give hindered sec- 200,20d,23 ondary acetylenic amines. It is not clear why all reactions with amines are markedly catalyzed by trace amounts of cuprous salts.20d Some good nucleophiles give no substitution products, even though steric effects cannot explain the failure. Thus, the reaction of 42 with excess KCN in aqueous methanol yields only the solvent derived methyl ether 42 and no nitrile (Eq. 20). MeOH (CH ) -C(Cl)-CECH + KCN ——> (CH ) -C(OCH )—CECH 3 2 3 2 3 H 0 2 45 46 Eq. 20 u"— #— Presently, no satisfactory correlation exists between steric features of the nucleophilic reagent, its basicity, nucleophilicity, polarizability, solvent employed and the l7 outcome of the substitution reaction. Hennion synthesized the following compound 42 based on the observation that hindered primary amines react with tert-propargylic chlorides to give hindered N-tert- propargylic secondary amines. $H3 ?H3 40% KOH, H20 FH3 9H3 HCEC-C-Cl + HCEC-C-NH2 > H-CEC—C-NH-C—CECH I l I CH CH COpper Bronze CH CH 3 3 Cu C1 3 3 2 2 41 42 8 days, 30°C 42 (47%) Eq. 21 Compound 42 was semihydrogenated to 22 using 10% palla- dium on charcoal catalyst and then hydrogenated to the saturated amine 22 using Raney nickel in ethanol. Compound 22 appears to be the most hindered 2° amine obtained by any method reported to date. Pd/Char. $H3 FHB RaNickel 42 —————-—> CH2=CH—{}1§-—C——CH=CH2 > I l Pet Ether CH3 CH3 EtOH i9 Eq. 22 CH3 CH3 7 CH3-CH2—C—NH—C—CH2-CH3 CH3 CH3 51 —— 18 RESULTS Tertiary propargylic chlorides were synthesized from the corresponding propargylic alcohols: 3-Methyl-l—bu- tyne-3-ol 23, 3-ethyl-l-pentyne-3-ol 22 and 4-methyl—3- isopropyl-l-pentyne-B-ol 23. The alcohols were made from acetone, 3-pentanone and 2,5-dimethyl—3-pentanone, respec— tively, by reacting the ketones with sodium acetylide in anhydrous liquid ammonia (Eq. 23). 0 9H R—g-R + NaCECH .——————> R-c-R Eq. 23 NH3(1) éECH -33°C 52; R: Methyl (55%) 53; R: Ethyl (93%) 2g; R== Isopropyl (89%) Bubbling acetylene gas slowly through the ammonia solu— tion for several hours once NaCECH was formed,35 then adding the ketone to the solution, increased the carbinol yields substantially over reported literature values.24 19 The alcohols were worked up and then purified by vacuum distillation. The tertiary acetylenic chlorides were prepared from the corresponding propargylic alcohols by reacting the alcohols with excess cold hydrochloric acid containing copper bronze powder, calcium chloride and cuprous chloride.25 When R is methyl or ethyl, the chloride may be prepared with concentrated HCl and CaCl2 alone. When R is isopropyl, rearrangement products are obtained at the expense of the desired tert-prOpargylic chloride (Eq. 24). HCl OH cu2C12 cl ! R-C-R CaClZ, Copper Bronze\ R—C-R Eq. 24 ‘ f 7 ) CECH 0°C, 1 hr. CECH 47; R== Methyl (65%) ; R: Ethyl (73%) 6; R== Isopropyl (70%) All of the propargylic chlorides are sensitive to heat and were used without further purification. The proportion of chloride in each reaction mixture was determined by GLC and the value calculated was used in determining the yield of propargylic amine. Tertiary propargylic chlorides 2, £1 and 22 were converted to the corresponding primary propargylic amines 20 by dropwise addition of the chloride to a liquid ammonia suspension of sodamide, prepared 22 situ (Eq. 25). Cl NH , NaNH2 . 2 R-g-R ——————+> Rec—R + NaCl CECH NH3(1) CECH Eq. 25 32; R== Methyl (20%) 2; R== Ethyl (73%) 22; R== Isopropyl (40%) Ordinarily, sodamide in liquid ammonia reacts with ali— phatic halides by eliminating HX.29 Sodium acetylide may be substituted for sodamide in the propargylic amine syn- thesis with no difference seen in the results.24 Cl NH2 I ( (C2H5)2-C-C:CH + NaC:CH + NH3-—-€>(C2H5)2C-C:CH 2 2 Eq. 26 An attempt was made to synthesize 2 by applying Hennion's procedure (Eq. 21) for synthesizing l,l,l',l’- tetramethyl-di—2-propyny1 amine £2. 21 5 mg. Cu2C12 14 ml. 40% KOH $1 §H2 5 mg. Cu Bronze (C H ) CCECH + (C H ) CCECH > 2 5 2 2 5 2 8 days, 350 2 .3. Eq. 27 OH I (C2H5)2CC:CH 53 No coupled amine was seen by GLC, only the solvolysis pro- duct 22 and low boiling impurities. The experiment was repeated with the di—isopropyl propargylamine and chloride, 22 and 22, with similar results. Low boiling impurities and the carbinol 23 were the only products seen after reaction for two weeks at 30°. Another approach was used to prevent solvolysis of the chloride. Conversion of the propargylic chloride to the acetylide with the strong base NaH (oil dispersion) should be complete. The reaction's progress could be followed by monitoring H2 evolution with a gas buret. Thus the chloride 2 was added to 1 mlmxle of 2 and l lnmole of NaH in 2 ml of tetrahydrofuran (THF) (Eq. 28). 22 c1 l .. ___) _ _ z ——> _ ___ = . 2 + NaH 9. (c2145)2 c c_c + H2 (C2H5)2 c c c. THF 22° Eq. 28 $H2_ + (C2H5)2C-C:CH —->_1_ Only 0.8 Inmole of H2 was evolved over a 5 hour period. The solution was quenched with H20 and the brown mixture was analyzed by GLC. No coupled amine 2 was detected. Analy- sis for recovered starting material was impossible, since the starting amine and chloride were inseparable by GLC. The same slow, incomplete H evolution was observed (0.55 2 n1mole) with KH (mineral oil suspension) when 1 Inmole of the base was used with 2 and 2 under identical experimental conditions. Again, no coupled amine 2 was detected by GLC analysis. The above experiment with KH as base was repeated, but this time a catalytic quantity of Cu2C1 and c0pper bronze 2 was added. Evolution of H was very slow (1 Inmole in 21 2 hrs.) and GLC analysis of the mixture showed no coupled amine 2. The same experiment was repeated using 2 Inmoles each of 2 and 2 in THF along with 10 mg. of copper bronze, Cu2Cl2 and 4 Inmoles of potassium tert-butoxide as base. A thick black tar was the only material obtained. 23 Easton 23 in his investigation of hindered amines noted that dimethylformamide, when used as a solvent, gave much better yields of hindered secondary prOpargylic amines than either ethyl ether or THF. We thought that it might be possible to synthesize primary propargylic amines directly from an ammonia saturated dimethylformamide solu- tion of Cu Cl 2 2 chlorides directly to the mixture. The experiment was and copper bronze by adding the propargylic performed by adding 2, 22_and 22 dropwise to a well stirred DMF solution of ammonia, Cu2C12 and copper bronze at 0°C. The DMF solution was kept saturated with ammonia by rapidly bubbling the gas through the solution. In every case we obtained the corresponding primary propargylic amine in only fair yield (Eq. 29). Cu Cl 2 2 $1 Cu Bronze sz R-C-R + NH3 % R-C-R ‘_ DMF !_ C:CH 1 hru,O°C C:CH Eq. 29 47; R: Methyl 512; R: Methyl (-) 2 ; R= Ethyl _3_ ; R= Ethyl (40%) 56; R== Isopropyl 22; R== Isopropyl (36%) No attempt was made to determine the yield of 22 because on closer investigation by GLC, two products were seen; the primary amine 22 and an unidentified high boiling product. 24 Upon isolation of the high boiling component by GLC, we identified it as l,l,l’,l’ tetramethyl-di-2-propyny1amine 22, the same amine which Hennion synthesized (Eq. 21) only after reacting 22 and 22 for 8 days in a 40% aqueous solu- tion of KOH. Cu C1 2 2 Cl I Cu Bronze §H2 (CH ) C-CECH + NH 2_;. (CH ) -C-CECH 3 2 3 DMF 3 2 1 hr.,0°C .42 .42 Eq. 30 $H3 9H3 + HC EC-C—NH—C—C ECH | I CH3 CH3 49 The amine 32 was identified by its physical and spectral properties; NMR (CDC13) 6 1.23 (S, 1H), 1.5 (S, 12H), 2.23 (s, 2H), m.p. 34-35°, Lit?0f 32-35°. This is the first method reported for preparing primary propargylic amines directly from ammonia. Hennion attempted to prepare 2 by reacting the chloride 2 with aqueous ammonia at 100°C in an autoclave.20C The only products obtained were 3-ethy1—3- pentyne (30%), produced by HCl elimination, and the hydrol— ysis product 22. Based on the observation that DMF greatly increases the reaction rate between propargylic chlorides and propargylic 25 amines, a number of experiments were performed to determine the limits of this method for obtaining hindered secondary amines. Table I lists the results of a number of these experiments. Amine 2 was isolated by quenching the reaction mixture containing amine hydrochloride with NaOH. DMF was removed by extracting the solution with water. The remaining solution was steam distilled to remove the primary and secondary amine, and the two amines were then separated by distillation. A minimum 2:1 ratio of propargylic amine to chloride was maintained so that the extra equivalent of amine would act as an HCl accepUns Adding either triethylamine or di-isopropylamine: as an HCl acceptor greatly reduces the yield of 2 (<:7%). Compound 2 was identified by its spec- tral properties; NMR (CDC13) 6 0.9 (t, 13H, J=6Hz), 1.72 (q, 8H, J=6Hz), 2.25(S, 2H); mass spec (parent peak m/e 205). Several interesting trends appear in Table I. As the temperature of the reaction mixture increases, the overall yield of coupled amine decreases. The ratio of amine to chloride has a greater influence on the yield of coupled amine formed at higher temperatures than it does for amines formed at lower temperatures. As the size of the alkyl groups on the chloride and amine increase, the yield of coupled amine decreases. Finally, as the degree of unsatur— ation of the primary amine decreases, the yield of coupled amine decreases drastically. 26 mm she a oem use e.m e.H m.m mesnum fisnum ma sot H oem mze o.H e.H e.H messum asnpm om amp H ome mzo e.m e.H m ocsgum assum me msee m oem mze e.m e.H m wasnpm Henge Nv map H ovm NmZoU: w.m v.H m mcwnpm H>£um 0v how H ovm hEQ ©.m va om mcmnum Hwnwm mm .u: H oo mzo m.m nm em mssgum Hanumz pamew maee .dsme pem>aom moeeoHno meeeoHno maesm .m m Q m mcHE¢ mmaoarz me082: 4 m m 2 _ momouonmzuon.m ”A .muonm + momouwmm _ A. . m m mmz Ho mcofluflpcou coepommm mSOHHm> “moss mmcflE¢ oflawmummoum om mo mpamflw H mamflfi 27 .mmmmnpcmumm CH mum mpamflx UwamHOmH .mpamflm mum mum mpawflw Hamn mpflhoazommlxéha w.o me mNCOHQ Hmmmoo paw NHUNDU mo 20mm 08 H meAMpcoo mucmfiflummxw Hadm 0 .m23 m 0mm- use e.m ea om memsum H>e0220mH 0 .m23 m .4 use e.m ea om wasnpm H>e0heomH e msme m 04 use m.m m.e ma fisnum flange 5H msme m 0e mzo m.m m.e ma mamnum esnpm me mane m 04 use o.m m ea meanpm Henge Amwvam wwwp m ow mza ©.m v.a m mc%£#m Hwnum seamew mafia .esme mucm>aom wmmhoHno Mmmmmmmo hwmwmme .m m mCHE< l|||||"|||lllllll'lllllllII'Il'llIII-I'll!I'll-III|l||||||llll|ll|la|lll|lllIIIIIIIIIIIIIIIIIIIIIIII lllll'llllllIII‘IIIII'I'II'IIIIIIIII'IIIIl'l'lll'lll'l'-Ill-I'llIIIIIIIII'IIIIIII'IIIIII A.pcoov H mqm (CH ' ' EtOH CH2CH3CH2CH3 CH2) -C—NH Eq. 31 3 2 law (100 o\0 ) l 29 Attempted semihydrOgenation of 2 to 2 using 10% pal- ladium on charcoal in the aprotic solvent ligroine gave fair yield of 2 along with some hydrogenolysis. However, when ethanol was added to the solution and the hydrogenation continued, 1 was hydrogenolyzed almost immediately to 2. 10% Pd/Char.,H2 $H2CH3 9H2CH3 _l_ ‘ > H2C=CH-C-—NH—C'3- CH=CH2——> _6__ ligroine EtOH CHZCH3 CHZCH3 Z (55%) Eq. 32 The same sequence of reactions attempted with 2 and 10% palladium on charcoal in ethanol gave two unusual cycliza— tion products; 3,4 dimethy1-2,2,5,5—tetraethy1-3-pyrroline, 2 and 3—methylene-4-methy1—2,2,5,5-tetraethyl-3—pyrrolidine Pd/H CHZ Pd 1 2 > ‘ Ethanol N (CH3CH2)2 H L. CH3\ CH3 CH3 C (CH3CI'12)2 )“(CH2CI~13)2 + (CH3 H2)2 N H 2 9 in about a 3:1 molar ratio, respectively. (CH |\o CH 2CH3)2 .4 (CHZCH3)2 Eq. 33 30 Spectral data confirm the structures of 2 and 2; compound g, 13 l C NMR 20 MHz (CDC13) 6 7.2, 8.7, 29.8, 70.7, 134.2; H NMR (CDC13) 6 0.82 (t, 13, J=6Hz) 1.42 (broad multiplet, 8H) 1.45 (S, 6H); mass spectrum (parent peak m/e 209). Compound 2, 1H NMR (CDC13) 6 0.85 (t, 13H J=6Hz), 1.41 m, 11H), 2.4 (m, 1H), 4.63 (t, 2H, J=3Hz); mass spectrum (parent peak m/e 209). The reaction of 2 with 10% palladium on charcoal in ethanol appears to be the first catalytic cyclization reac— tion forming substituted pyrrolines and pyrrolidines from bispropargylic secondary amines. Attempts to reduce 2 to either 2 or 22 with a number of stochiometric reducing agents proved equally fruitless. Reduction of the bisalkynyl amine by hydroboration with borane30 prepared 22 2222 gave a number of low boiling pro- ducts, none of which corresponded to either the bisallyl amine Z or the saturated amine 22(Eq. 34)). HN-(C(CH2CH3)2-CECH)2 + 3 NaBH + 4 BF3:OEt 4 2 .1. Eq. 34 ._—————9 >- No Product diglyme l) propionic acid, 100° 4 hrs. 2) NaOH Attempts to reduce 2 with nickel boride, a reactive olefin hydrogenation catalyst, gave similar results.31 31 The saturated amine 22 was finally synthesized in about 20% yield by low pressure catalytic hydrogenation of 2 in ethanol with W2 grade Raney nickel. The hydrogena- tion took about 12 hours to complete at 22° and 30 psi H2 pressure. When more reactive grades of Raney nickel were used for hydrogenating 2 in ethanol, (W4 and W6 grades),33 hydrogenolysis was more extensive. A180: the hydrogenation generally fails to go to completion with the more active grades of Raney nickel. One can stop the hydrogenation at the diallylamine stage by monitoring the reaction by GLC. Hydrogenation of the diallylamine Z_ to the saturated amine 22 was much slower than hydrogenation of 2 to the diallyl- amine Z. If the hydrogenolysis of 2 occurs at theciiallylamine stage through a carbonium ion mechanism, a presumption sup— 28,32 perhaps ported by results of other investigators, hydrogenolysis could be minimized by hydrogenating in a strongly basic ethanolic solution of W2 Raney nickel. This hypothesis was tested and proved correct. A 71% yield of 22 was obtained by hydrogenating 2 under the same condi- tions with W2 Raney nickel, except that a 2:1 molar excess of KOH was added to the ethanolic solution of 2. Amine 22 is the most hindered secondary amine ever synthesized. The spectral data confirm its structure; 1H NMR (CDC13) 6 0.78 (t, 19, J=7Hz) 1.32 (q, 12, J=7Hz); mass spec (parent peak m/e 213). 32 DISCUSSION The reaction of metal acetylides with ketones and aldehydes has been thoroughly studied by a number of invest- igators. The most satisfactory general procedure for the synthesis of propargylic carbinols involves the condensation of sodium acetylide with aldehydes and ketones in anhydrous liquid ammonia.24 Comparatively small amounts of the glycols, formed by the condensation of two molecules of the carbonyl compound with one of acetylene, are obtained. The yields of the acetylenic carbinols 22, _2 and 22 are increased and that of the glycols decreased by passing acetylene gas into the mixture during the entire course of the reaction. Passing acetylene gas through the ammonia solution during the course of the reaction presumably suppresses the disproportionation of sodium acetylide to disodium acetylide and acetylene (Eq. 35). 2 NaCECH —————9 NaCECNa + HCECH Eq. 35 Attempted acetylation of di-tert-butylketone with sodium acetylide in liquid ammonia gave rapid evolution of acety- lene and, upon workup, starting ketone and glycol (Eq. 36a). 33 2NaCECH + 3—}—c —-> HO——-CEC———-OH + c n + +c+ Eq. 36a Successful acetylation of very hindered ketones was reportedly achieved by adding n-butyl lithium to acetylene in tetrahydrofuran at —78°C. Addition of di-tert—butylket- one at —78°C, followed by warming to room temperature, gave good yield of di-tert-butylpropargyl alcohol.24 The mono— lithium acetylide is stable when maintained at low tempera- tures. Warming the solution of monolithium acetylide to 0° results in the irreversible formation of a white solid, presumably dilithium acetylide. In cases where sodium acetylide in liquid ammonia fails to monoacetylate hindered ketones, acetylation with lithium acetylide in tetrahydro- furan at —78°C appears to be the method of choice. The tertiary acetylenic chlorides were prepared from the corresponding propargylic alcohols. When R is either methyl or ethyl (water and acid soluble tert-propargylic carbinols), the chlorides 21 and 2 were prepared in good yield and acceptable purity from concentrated hydrochloric acid and calcium chloride. When R is is0propy1, a number of products, including the propargylic chloride 22, were obtained. These rearrangement products were obtained along with the desired tert-propargylic chloride, in accord with 34 published observations25 (Eq. 36b). V RCHZ-C(OH)-CECH + HCl RCHz-C(Cl)-CECH I R R (i) + RCHZ-C=C=CHC1 + RCH(Cl)-C=C=CH2 + RCH=C-C(Cl)=CH2 R R R (ii) (iii) (iv) + RCH=C~CH=CHC1 R Eq. 36b (V) + + The carbonium ion (RZC-CECH v—A-R2C=C=CH) derived from the alcohol could lead directly to 2 and 22. Dehydration of the alcohol would yield the conjugate eneyne hydrocarbon which is converted to 222 and 22 by 1,4 and 1,2 addition of hydrogen chloride, respectively. Prototropic rearrangement of 22 yields 3. It has long been recognized in individual cases that mixtures of products are encountered when tert- propargylic carbinols are converted to tert-propargylic halides.36 The most successful general procedure for converting tert-propargylic carbinols to tert-propargylic chlorides is by treatment with excess hydrochloric acid containing calcium chloride, cuprous chloride and copper bronze 35 powder.25 The combination of these reagents in cold concentrated hydrochloric acid seem to give higher yields and purer products than any other procedure reported to date. The choice of these reagents in the synthesis of tert-propargylic chlorides was arrived at. empirically. The formation of the propargylic amines 2, 8 and 58 from the corresponding propargylic chlorides 2, 2_ and 22, respectively,i11a sodamide-liquid ammonia solution is a sol— volysis reaction by ammonia and not a simple nucleophilic displacement by the amide anion. This conclusion is based on the observation that sodium acetylide may be substituted for sodamide in the amine syntheSis (Eq. 26).20b This is good evidence for the mechanism involving the propargyl zwitterion-allenecarbene species (Eq. 19), where a conse- quence of the suggested mechanism is that the reaction of tertiary propargylic chlorides with base in a suitable solvent should likewise produce the solvolytic product. 21a,21b In 1962 Shiner et. a1 offered conclusive kinetic evidence for the intermediacy of the species 22 in the solvolysis of the propargylic halide 21 in basic aqueous ethanol (Figure l).21C 36 CH CH H I 3 _ 3\ / - CH3-C-CECH + RO C=C=C\\ + R0 ( Cl(Br) CH3// Cl(Br) 47a (b) 55a (b) CH3 _ CH3 \ .— CH3-C—CEC + HOR l/C=C=C-C1(Br) -t HOR Cl(Br) CH3 —C1-, Br- CH3\\+ - CH3\\ C-CEC < 9 C=C=C CH / 6 CH / 3 —— 3 60a 1 60b CH2=C(Me)CECH M82C(OH)CECH + Me2C(OEt)CECH 9.1. .53 22. Figure 1. Reaction scheme for the base promoted solvolysis of isomeric allenyl and tertiary propargylic halides. He showed that 212 exchanged the acetylenic hydrogen in basic 80% ethanol-d-deuterium oxide solution much faster than it solvolyzed and that the rate of the second-order solvolysis in the non-deuterated medium was depressed by adding sodium salts in the order: Br->Cl->NO3-~C104—. 37 These results are consistent with Hennion's mechanism only if the rate determining step is the ionization of the halide from the conjugate base of 222 in Figure 1. Gas chromatographic analysis of products from the reaction mixture of 222 in basic ethanol indicated that the propargylic ether 22 was the predominant product (90% relative yield) accompanied by small amounts of the propar— gylic alcohol 22 (7%) and olefin 22 (3%).21a In comparison, the first order, initially neutral, solvolysis of 1b, 21a,21b gave an entirely characterized as an SNl process, different distribution of these same propargylic products: 62, (43%); 52, (22%), and 22 (35%). Shiner concluded that "The difference in product proportions is apparently dic— tated by the different reactivity, and therefore selectivity, of the two intermediates, the carbonium ion and the zwitterion-carbene. The latter is more stable and more selective because the allene—carbene resonance contribution form [222] contains no formal charges and therefore contributes more importantly to the structure [22] than the allene carbonium ion VIIb does to the structure VII."21a + + [MeZC-CECH <————~> Me2c=c=CHJ VIIa VII VIIb With the zwitterion-allene carbene species in Figure 1. identified as the reactive intermediate in what is 38 essentially a base catalyzed solvolysis reaction with ammonia and alcohols, tertiary propargylic chlorides were shown to undergo alkylation with primary and secondary amines to give the coupled secondary and tertiary propar- gylic amines, respectively.20d Experimental results confirm that the coupling reaction of 21 with primary amines is notably insensitive to steric features of the amine, except for rate (Eq. 37). Cl ' _ 1 2 l 2 (CH3)2-C—C:CH + 2 R R NH ———> (CH3)2C(NR R )CECH Eq. 37 47 + RlRZNHoHCl Thus tert-butylamine reacts with 21 essentially as well as ethylamine to give coupled secondary propargylic amine in \ 52% and 44% yield, respectively, in aqueous solution within one day.20d The same reaction is catalyzed by copper and by cuprous salts. When the amine subjected to alkylation is a strong base, catalysis is not necessary. With weakly basic compounds (aromatic and prOpargylic amines) cuprous salt catalysis is necessary in order to obtain the products in good yield within a reasonable reaction time. Thus the sterically crowded secondary amine 22, l,l,l’,l’-tetramethyl— di-2-propynylamine:h5 prepared (Eq. 21) from the correspond— ing propargylic chloride 21 and propargylic amine 22 in 47% yield in 8 days by using cuprous chloride as a catalyst. 39 An attempt to synthesize 1,1,1',l'_tetraethyl-di-2— propynylamine 2; using the same conditions as in Eq. 21 failed to give the desired product. Only the propargylic carbinol 22 and starting propargylic amine 2 were recovered. Apparently, solvolysis of the propargylic chloride 2 is more facile than coupling with the prOpargylic amine 2. This is probably due to the severe steric crowding between the reactive allenecarbene intermediate and the primary amine 2. It appears that when both the propargylic amine and chloride are crowded about the reaction centers, solvol— ysis by the aqueous solution is faster than the coupling reaction. Attempts to couple the propargylic amine 2 with the propargylic chloride 2 in the polar aprotic solvent dimethylformamide by using KH or NaH as the base with a cuprous chloride catalyst were equally unsuccessful. The slow, incomplete H evolution seems to indicate that the 2 propargylic chloride 2 is not being converted to the acetylide. This observation is not unexpected. Jacobs et al.37 reported that lithium aluminum hydride dehalogenates tertiary propargylic halides to give allenic hydrocarbons mixed in most instances with some of the corresponding acetylenic hydrocarbon (Eq. 38). LiAlH4 RR'CXCECH -——--—> RR'C=C=CH2 + RR'CHCECH Eq. 38 40 The results can be explained by a combination of 8N2 and SN2' attack by hydride ion as pictured in equations 39 and 40, suggested by Wotiz.38 MH R l C\-CECH ——-> RR'CHCECH Eq. 39 R (x MH :B/a HCEC-C-X ————> H C=C=CRR' Eq. 40 \I; 2 RI This seems to be a reasonable explanation for the lack of H2 evolution, though no attempt was made to actually identify the low boiling products of the reactions with the metal hydride bases. Amination of the propargylic chlorides 2, 21 and 22 to the amines 2, 22 and 22, respectively, by adding the propargylic chlorides to an ammonia saturated solution of dimethylformamide and cuprous chloride was somewhat unexpected. Ammonia is not a particularly strong base and it seems unlikely that any appreciable amount of acetylide could be present. Detailed examination39a of reactions of terminal alkynes with cuprous salts have shown that cuprous alkyne derivatives are the reactive intermediates in coupling reactions with l-haloalkynes (Eq. 41), known as the 41 Cadiot-Chodkiewicz coupling.38a + R—CECH ——C—u——> RCECCu + H+ Eq. 41 + .— RCECCu + BrCECR' —————e~ RCEC-CECR' + Cu + Br A number of observations by Cadiot et a1. pertaining to the unsymmetrical coupling reactions of alkynes seem appli- cable to coupling of ammonia and primary propargylic amines 38a,38b Firstly, ammonia facili— with propargylic chlorides. tates greatly the formation of very reactive cuprous derivatives. Secondly, solvents like dimethylformamide and n-methyl phosphoramide are generally the best solvents for coupling cuprous acetylides with terminal acetylenes.38b Hennion was unable to couple tertiary propargylic chlorides with an aqueous ammonia solution to form primary propargylic amines. This is not unexpected, since the Cu+ ion lifetime in water is generally very short and it readily diSpropor— tionates to Cu0 and Cu+2.40 Thirdly, ammonia facilitates the oxidation of Cu+ to Cu2+ in aqueous solution to form 1+2 ammine complexes of the form [Cu(H20)6_n(NH3)n n==1 to 5, depending on the relative ammonia concentration.40 It may be that in dimethylformamide, the dipolar inter- mediate 2 is made more reactive in the form of the acetylide structure 2 or 2. Alternatively, the t—acetylenic chloride used may form the acetylide 2, subsequently leading to 2, 2 or 2 as the species responsible for alkylation.20d 42 R c-csc' R C-CEC-Cu I? Iw R C=C=C: é———9 R C=C=C: R2C(Cl)-CEC-Cu Cu-Cl Cu—Cl l0 IO Whatever the exact nature of the cuprous acetylide, it reacts with sterically hindered propargylic and allylic amines to give remarkably hindered secondary bispropargylic and allylpropargylic amines, respectively, in good yields. Though very little is known about the reaction of cuprous acetylides with propargylic and allylic amines, the experi- mental data pose a number of interesting questions. The bulky saturated analogues of the ethynyl and ethenyl primary amines used in our coupling experiments fail to give hindered secondary amines. Below is a summary of 20f (Eq. 42) and by us (Eq. 43) results obtained by Hennion on the coupling reaction of saturated and unsaturated primary propargylic amines with tertiary propargylic chlorides. 43 3:1/amine: chloride NH CH CH ; 2 40% KOH, H20 i 3 ; 3 £1 + CH3-C-CH3 HCEC-C-NH-C-R I s : R Copper Bronze CH3 CH3 Cu2Cl2 8 days, 30°C Eq. 42 4g; R==Ethynyl 42; R==Ethynyl (47%) E1; R:=Ethenyl Q4; R==Ethenyl (47%) £2; R==Ethyl fig; R==Ethyl (-) NH2 3:1/amine: chloride (Iii C H ' Dimethylformamide ,2 5 .2 5 2 + C H -C-C H _ — 2 5 fi 2 5 COpper Bronze HCZC—1}—Ifl+—$-R C H C H Cu2Cl2 2 5 2 5 3 days, 4 C Eq. 43 3; R==Ethynyl l; R==Ethynyl (63%) 4; R==Ethenyl g; R==Ethenyl (17%) _g; R=Ethyl 66; R=Ethyl (0%) - It seems unlikely that a cuprous salt derived from the saturated and unsaturated amines is the active agent in the coupling reaction with the cuprous acetylides derived from the propargylic chlorides. The terminal hydrogens on the ethenyl and ethyl groups of the secondary amines are not sufficiently acidic to form the cuprous salt under the relatively mild reaction conditions used in the experi- ments.40 It seems that all of these coupling reactions 44 would involve formation of similar intermediary copper derivatives (in particular, cuprous acetylide). This common factor, which is very little examined as yet, seems of primary importance for understanding why the hindered secondary amines are formed so easily. A polymeric structure has been proposed for cuprous derivatives of alkynes.41 A trimethyl phosphine cuprous etc. R T e R-CEC-Cu +—— ”' C R ' Eq 44 u ‘ ' $ c R-CEC—Cu<——g I Cu Ab ! etc derivative of phenylacetylene has been isolated as a crystallized product and its structure has been determined by an X—ray study.42 Though it is hazardous to extrapolate from a solid phosphinous derivative to amino complexes formed in the reaction media, experimental evidence does not preclude the possibility of the formation of amino cuprous complexes where copper valence variation and the existence of two (or more) different copper species could be used for the coupling reaction to occur. A number of experimental observations may be rational— ized by some kind of copper stabilized complex for cuprous derivatives of propargylic chlorides and unsaturated amines (Eq. 45) . 45 RI, | C1 NHZ—Cf—R’, | R’—C-CEC-Cu <——— R R = CECH, CH=CH2 Eq. 45 I R! As the reaction mixture temperature increases for a given ratio of propargylic chloride to unsaturated amine in the dimethylformamide solution, the yield of coupled amine decreases. This result is in accord with the proposal that at higher temperatures, an unsaturated amine-propargylic chloride complex would tend to dissociate and destroy the reactive intermediate which is responsible for the coupling reaction. This proposal is reinforced by the observation that the temperature, rather than the ratio of propargylic chloride to unsaturated amine (provided that the ratio is at least 1:2, respectively), is the most important factor in maximizing yield. Reaction of propargylic chloride 2 with alkynyl amine 3 in a 1:2 and 1:3.6 ratio at 24° for 1 day gave a 33% and 43% yield, respectively, of 1. Repeating the same two experiments at 4°C for 3 days gave a 61% and 63% yield, respectively, of 1 (Table 1). For the relatively unhindered propargylic chloride 41 (R’=CH3) and propargylic amine 4g (R”=CH3) in Eq. 45, the Cué——R complex formation is strong enough to overcome the repulsive forces of steric crowding between the chloride and amine alkyl groups. This would explain why R= ? R = CHZCECH, Eq. 48 CHZCHZCECH, CHZCHZCHZCECH, etc. The terminal acetylenic amines could be synthesized by treating the propargylic amine with sodamide and the appro- 32 priate alkyl halide and then isomerizing the internal alkyne with potassium—3—aminopropylamide43 to the terminal alkyne (Eq. 49a). gHz 1) NH Na §H2 RI ICRI I 2 7} RI ICRII l ' CECH 2) CH3CH2I CEC-CHZCHB Eq. 49a NH2 H :; R”CR” NH2- (CH2) 3-NH KNH-(CH2)3-N I 2 CHZCH2C2CH This approach would be especially interesting for investi- gating the product distribution of the coupling reaction when either the propargylic chloride (Eq. 46) or the 49 propargylic amine (Eq. 47) has very bulky alkyl groups. In this manner, a better understanding might be gained into the nature of the reactive intermediate involved in the coupling reaction. When triethylamine is used asan1HCl acceptor in the coupling reaction of 2 and 2 in dimethylformamide, Cu2C12 and copper bronze, the yield of the coupled amine 2 is reduced greatly. Adding triethylamine to a reaction mixture of 2 and 2 in dimethylformamide causes a heavy white precipitate to form immediately. Addition of aqueous sodium hydroxide to the solution mixture causes the preci- pitate to disappear. Analysis by GLC shows that an extremely small amount of 2 is formed and the triethylamine is recov- ered. Hennion studied the reaction of trimethylamine with 209’27 He found tertiary propargylic chlorides in acetone. that the reaction produces quaternary ammonium chlorides that have the propargylic structure when R or R' is CH3. When R 366 R’ are larger than CH3, the products are allenic (Eq. 49b). He also found that trace amounts of copper bronze or cuprous chloride catalyze both reactions, as in the case of primary and secondary amine alkylations. RR'C(NMe3)-CECH(C1_) (R or R’ small) RR’C(Cl)-CECH + (CH3)3N Eq. 49b b RR’C=C=CH-NMe3(Cl') (R=R'=C H 2 5 or larger) 50 It may be that the propargylic chloride reacts preferentially with the tertiary amine rather than with the propargylic amines. This would explain the low yield of coupled secondary amine 2. However, since no attempt was made to identify the precipitate, any conclusions are specu- lation. Hydrogenation of 2 in ethanol with platinum oxide gave the hydrogenolysis product 6 almost exclusively (Eq. 31). The same result‘was observed with the hydrochloride of 2 under identical conditions. There are many examples in the literature28'32 illustrating that hindered allyl amines, intermediates in the hydrogenation of propargylic amines, undergo hydrogenolysis readily with heterogeneous catalysts in polar protic solvents.43 It appears that hindered amine hydrochlorides are susceptible to solvolysis as well as hydrogenolysis. This is illustrated by the fact that two tertiary amine hydrochlorides, prepared from N,N-diallyl- isobutylamine 62 and N-allyl-N-benzyl-t-butylamine 26, suffer rapid loss of an allyl group when recrystallization was attempted from mixed solvents containing ethanol.28 The products are allyl-isobutylamine hydrochloride Z2 and benzyl- t—butylamine hydrochloride 22, respectively. The latter case is particularly interesting, since 16 has three different groups presumably liable to cleavage as carbonium ions (Eq. 50,51).28’32 51 + + - HCl EtOH HZCl (CH2=CH-CH2)Z-N-CHZCH(CH3)2 ————4 CH2=CHCH2-N-CH2CH(CH3)2 £2 ~ 12 Eq. 50 + + _ HCl— EtOH HZCl CH2=CHCH2-W-CH2C6H5 ————9 (CH3)3c-N-CH2c6H5 C(CH3)3 70 72 Eq. 51 The susceptibility of hindered allyl amines towards hydrogenolytic and solvolytic cleavage in protic solvents might explain why the bisdiallyl amine 2 in Eq. 32 was hydrogenolyzed to 6 when ethanol was added to the ligroine solution containing palladium on charcoal as the hydrogena- tion catalyst. When 2 was hydrogenated with 10% palladium on char- coal in ethanol, the unsaturated heterocycles 6 and g_were obtained in good yield (Eq. 33). The formation of substi- tuted pyrrolines and pyrrolidines from unsaturated alicyclic amines is not unprecidented. Hennion synthesized 3,4- dimethyl-2,2,5,5-tetramethyl-3 pyrroline 22 and 3—methylene- 4-methy1-2,2,5,5-tetramethyl-3-pyrrolidine 26 from the corresponding bispropargylic and propargylic-allylic amines 22 and 22, respectively, with sodium in liquid 52 ammonia (Eq. 52, 53).20f H H Na NH4+ CH N CH3 CH C-N-C CH -——————9 CECHCECH 3 -- c 3 H3 22 73 + H 3 NarNH4 CH N CH (CH3)2<|:-N-c(CH3)2 ——————9 3 3 Eq. 53 I CECHCH=CH2 NH3 CH3 H3 CH3 CH2 74 75 Hennion investigated the steric and conformational effects of the sodium-ammonia reduction of bispropargylic and propargylallyl amines.28 From the data, he concluded that cyclization occurs as the major reaction only when the unsaturated centers are conformationally restrained to close proximity. Below is a scheme drawn up by Hennion showing the possible reaction pathways. 53 Na(NH3g Rgf\\*EE Na(NH§) R§/~\\”é7 \/\ \/\ § E Na(NH3) Na(NH3) V ) r ._ Na(NH3) ¢¢ RN -—————9 RN RN \J E .12 He determined that for the unhindered bispropargylic amine 2 (R:CH3), the reaction product proceeded almost exclusively by the pathway 2—92—22. If bispropargylic amine 2 was more hindered (R=C(CH3)3), the product mixture contained a 2:1 ratio of 2 to 2. And in instances where there were alkyl groups a to the nitrogen, as in 22 , the product mixture consisted predominantly of substituted pyrrolines analogous to g. Apparently, considerable steric assistance is required for such reactions to occur in good yield and reactions B—aD and A—aB—aD require less steric crowding than does the reaction A—éC. 54 The catalytic cyclization reaction of 2 using palla- dium on charcoal in ethanol gives us the substituted pyrro- line 2 and the substituted pyrrolidine 2 in good yield. It appears, at least qualitatively, that the products formed by catalytic cyclization of 2 with palladium on charcoal seem to follow the same general scheme as the cyclization reaction achieved with sodium in liquid ammonia. Mechanistically, cyclizations induced by sodium- ammonia probably involve the union of a radical center derived from one ethynyl group with the appropriate allylic or propargylic carbon atom in the other unsaturated group.28 Mechanistically, little can be said about the catalytic cyclization reaction of 2, except that it is probably an oxidative addition-reductive elimination oligomerization reaction involving some change in the oxidation state of palladium. It may be a heterogeneous oligomerization reaction involving Pd(O) or it may be a homogeneous catalysis reaction involving traces of PdCl2 left unreduced when palladium metal was deposited on charcoal by reduction of a PdCl -charcoa1 solution with hydrogen gas. 2 Whatever the active agent is, amine 2 must have the unsaturated centers restrained much closer to one another than in the case of 22 ,since 4g may be reduced to 62 with palladium on charcoal in ethanol with no traces of the heterocyclic amines 12 or 22 seen.20f This is in contrast with results obtained with 2 under identical experimental conditions, where we obtained the diallyl amine 2 along with 55 substantial amounts of the heterocyclic compounds 2 and 2. An interesting and potentially useful application of substituted 3-pyrrolines, formed from the sodium-ammonia reduction of bispropargylic secondary amines, might be in their use as synthons for substituted dienes. Lemal and MCGregor44 reported that dienes are generated in high yield from 3-pyrrolines by treatment with nitrohydroxylamine. The availability of 3-pyrrolines from substituted bis- propargylic amines makes this reaction potentially useful from a synthetic viewpoint, particularly since it proceeds with complete stereospecificity via a disrotatory thermo— lytic cleavage (Eq. 54, 55).44 ___ Na2N203 / \ > CH , Eq. 54 ,/ . 3 CH N C 1311. HCl,A CH 3 3 3 H ___ NaZN203 ~> ._Jfl——T%L_ Eq. 55 Dil. HC1,A CH3 ‘ CH3 CH N 3 H 3 One might start with the appropriate aldehyde or ketone:and form the propargylic chloride and propargylic amine, then couple them to form the substituted bispropargylic secondary amine. Cyclization of the bispropargylic amine 56 and resolution of the cis—trans isomers (if any) should give the substituted 3—pyrroline, a potential source of the desired diene. The saturated amine 22 was successfully synthesized by low pressure catalytic hydrogenation of 2 in ethanol with W2 grade Raney nickel. More reactive grades of Raney nickel generally failed to completely hydrogenate 2. One explana- tion for the apparent lack of reactivity of the more active grades of Raney nickel might be that amines are effective catalyst poisons for many platinum, palladium and nickel catalysts.33 As the activity of the Raney nickel became greater, the amount of hydrogenolysis increased. Since the hydrogenolyzed amine was less hindered than the starting secondary amine, it might act as a more effective poison than the unhydrogenolyzed amine. This would tend to explain why increased hydrogenolysis seemed to go hand in hand with the catalysts' failure to completely hydrogenate 2. Hydrogenation of 2 with W2 Raney nickel in a strongly basic ethanolic solution reduced considerably the competing hydrogenolysis reaction. This observation tends to support the observation that hydrogenolysis of hindered unsaturated amines occurs via a carbonium ion mechanism and that strongly basic reaction media suppress this undesired sidereaction. 57 EXPERIMENTAL I. Materials Propargylic Alcohols 3-Methyl-l-butyne-3—ol was commercially available (Aldrich). All other propargylic alcohols were synthesized from the corresponding ketones. Propargylic Chlorides All propargylic chlorides were prepared from the corresponding alcohols and used without further purifica— tion. All were stored over anhydrous potassium carbonate. Propargylic Amines The propargylic amines were made from the correspond— ing chlorides. All were distilled under reduced pressure and stored over molecular sieve. 2olvents Tetrahydrofuran was dried over sodium benzophenone ketyl, distilled and stored under argon over molecular sieve. Anhydrous diethyl ether was used without purifica- tion. Dimethylformamide was dried over calcium hydride, distilled under vacuum and stored over molecular sieve. U.S.P. grade absolute ethanol was used for all hydrogena- tions. 58 Inorganic Chemicals Commercially available Cu2Cl2 (Alpha Co. 95%) was used for preparing the propargylic chlorides. Freshly prepared Cu2Cl2 was used in catalytic coupling reactions.45 Copper bronze powder and all other inorganic reagents were obtained commercially and used without further purification. II. Preparation of Tertiary Propargylic Alcohols A. General Procedure 3-Ethyl-l—pentyne-3-ol 22 is a representative example for the preparation of tertiary propargylic alcohols. GLC analyses used a 1/4" by 6‘ stainless steel column packed with 10% Carbowax ZO-M liquid phase on Chromasorb G support. A 5-liter three-neck round bottom flask was fitted with an efficient mechanical stirrer mounted through a short glass bushing and two gas inlet tubes for acetylene and ammonia which dipped below the surface of the liquid ammonia. The third neck of the flask was fitted with a large Dry Ice condenser which was connected to a KOH drying tower by rubber tubing. The flask was charged with about 4 l. of anhydrous liquid ammonia (Matheson Gas Co.), the stirrer started, and a rapid stream of acetylene gas was passed in for about 30 minutes to saturate the solution. Welding grade acetylene was sufficiently purified by passage through two sulfuric acid gas wash bottles. Additional ammonia gas was condensed from time to time to keep the solution level at 59 about 4 1. Sodium (115 g., 5g}.atoms) was cut into strips so that they could be inserted through the side neck of the flask. The Dry Ice condenser was removed and replaced with a short piece of 12-15 mm. wide glass tubing through which was passed a long piece of flexible iron wire. The lower end of the wire was bent into a hook. One of the pieces of sodium was attached to the wire hook and was gradually lowered into the liquid ammonia while a rapid stream of acetylene was bubbled in. The sodium was added at such a rate that the entire solution did not turn blue. If it did, the sodium was raised above the level of the ammonia until the color was discharged. The rest of the sodium was added in a similar manner. The addition required about one hour, depending on the rate of passage of the acetylene. After the sodium was added, stirring was continued for 1 hour while still bubbling acetylene through the solution. The acetylene was then shut off. The gas inlet tubes and the glass tube with the iron wire were removed and replaced with a large addition funnel and the Dry Ice con- denser. Then 430.6 g. of 3-pentanone (4.95 moles, 98% pure) was added drOpwise over an hour period to the ammonia solu- tion. After the addition, the dropping funnel was removed and the neck stopped up. The flask was insulated with a 5]” heating mantle and glass wool. The solution was allowed to stir overnight. The heating mantle was removed, the stirring stopped and the reaction mixture was allowed 61 III. Preparation of Tertiary Propargylic Chlorides. A. General Procedure The following procedure for the conversion of 3-ethyl- l-pentyne-B-ol 22 to 3-chloro-3-ethyl-l-pentyne 2 is representative for preparing the chlorides. A ll” 3-neck flask provided with a magnetic stirrer, thermometer and dropping funnel was charged with 56 g. (0.5 mole) calcium chloride, 40 g. (0.4 mole) Cu2Cl2 chloride (95% brown powder), 400 mg. copper bronze powder (Illinois Bronze Powder Co.) and 430 ml. (5 moles) of cold concentrated hydrochloric acid. The mixture was flushed with argon and cooled (ice bath) with stirring. One mole of 3-ethyl-l- pentyne-3-ol 22 (112.2 g.) was added dropwise within 30 minutes. Stirring was continued for 1 hour. (0-5° solu- tion temperature). The upper organic layer was separated and washed immediately with three 100 m1. portions of cold concentrated hydrochloric acid, then with two 100 ml. por- tions of water and once with 100 ml. of saturated aqueous sodium carbonate. The colorless product was dried super- ficially with anhydrous potassium carbonate and then thor- oughly with fresh potassium carbonate. Analysis of the sample by GLC with a 10% Carbowax 20-M on Chromasorb—G column showed the sample to be 96% pure. The chloride was used without further purification. Total isolated yield of pure chloride was 73%, n25D 1.4387. 62 B. Product Analysis 3-Chloro-3-methyl-l-butyne 22 NMR (CDC13): 6 1.82 (S, 6H), 6 2.57 (S, 1H). IR 1 (CEC-H), 2110 cm"1 (-CsC-). n25D 1.4156. (neat): 3390 cm- Yield: 65%. 3-Chloro-3-ethyl-lfipentyne 2 NMR (CDC13): 6 1.47 (t, 6H, J=7Hz), 6 1.92 (q, 4H, J=7Hz), 6 2.58 (s, 1H). IR (neat): 3390 cm“1 (CEC-H), 2115 cm'1 (—CsC—). n25D 1.4387. Yield: 73%. 3-Chloro-4-methyl-3-isopropyl-l-pentyne 26 NMR (CDC13): 6 1.10 (d, 12H, J=6Hz), 6 2.13 (m, 2H, J=6Hz), 6 2.55 (s, 1H). IR (neat): 3395 cm"1 (CEC—H), 1390-1375 cm'1 (isopropyl doublet). n25D = 1.4560. Yield: 70%. IV. Preparation of Primary Propargylic Amines A. General Procedure (Hennion's Sodamide Method) The following procedure for the conversion of 3-chloro-3— ethyl-l-pentyne 2 to 3-amino-3-ethyl-l-pentyne 2 is representative. Twenty four grams of sodium (1.04 g. atom) was converted to the amide in 1]" of anhydrous liquid ammonia within a 3].. three neck round bottom flask provided with a mechanical stirrer, Dry Ice condenser and a long stem gas inlet tube for introducing ammonia into the flask. To the mechanically stirred ammonia solution was added 0.39% of finely powdered anhydrous ferric (III) 63 chloride and 16% of sodium. Dry air was bubbled through the solution until the blue color was discharged. The remaining 2395% pure. Determination of the amount of hydrogenolysis pro- duct by GLC was not possible, since the hydrogenolyzed amine eluted with the solvent. The same experiment was performed using W2, W4 and W6 Raney nickel under identical conditions. Tridecane was added as an internal standard in each case. Analysis of the products of each experiment by GLC showed that as the reactivity of the catalyst increased, the degree of hydro- genation of the bispropargylic secondary amine decreased. Thus W2 Raney nickel was the most satisfactory catalyst for the hydrogenation of 2 to the saturated secondary amine. The same experiment was performed under identical con- ditions, except that 20 mmoles (1.12 g.) of potassium hydroxide was dissolved in the ethanolic solution of the bispropargylic secondary amine before adding the W2 Raney nickel catalyst. The catalyst was filtered after the hydro- genation and the ethanol was removed under reduced pressure. 76 Twenty ml. each of water and ether were added to the Viscous residue. The solution was transferred to a small separatory funnel, the ether layer removed and the aqueous layer extracted with two 20 ml. portions of ether. The ether layers were pooled, dried over anhydrous potassium carbonate and then evaporated under reduced pressure. The remaining amine was distilled (bulb to bulb) under reduced pressure. About 1.52 g. (71%) of the saturated amine 22 was isolated, n24D 1.4653. C. Product Analysis l,l,l,1’,l’,l’-Hexaethyl-di-methylamine‘22 NMR (CDC13): 6 0.78 (t, 19H, J=6Hz), 6 1.4 (q, 12H, J=6Hz). IR (neat): no distinguishing bands. Mass Spec: parent ion m/e 213. Yield: 71%. BIBLIOGRAPHY l) 2) 3) 4) 5) 6) 7) 8) 9) 10) ll) 12) l3) 14) 15) 77 BIBLIOGRAPHY Jean Huet, Bull. Soc. Chim. Fr., 952 (1964). L. Hu, B. Mauze and L. Miginiac, C. R. Hebd. Seances, Acad. Sci., Ser. C, 284, 195 (1977). G. Wittig and H. Reiff, Aggew. Chem. Int. Ed., 2, 7 (1968). - H. G. Richey, Jr. and W. F. Erickson, Tetrahedron Lett., 2807 (1972). W. F. Erickson and H. G. Richey, Jr., Tetrahedron Lett., 2811 (1972). D. Seebach and D. Enders, Chem. Ber., 108, 1293 (1975). P. Beltrame, C. Veglio and M. Simonetta, Chem. Commun., 996 (1967). G. Stork and S. R. Dowd, J. Am. Chem. Soc., 22, 2178 (1963). __ K. B. Sharpless, T. Hori, L. K. Truesdale, C. O. Dietrich, ibid., 22, 269 (1976). H. P. Jensen and K. B. Sharpless, J. Org. Chem., 22, 264 (1975). _— A. Suzuki, S. Sono, M. Itoh, H. C. Brown and M. Mark Midland, J. Am. Chem. Soc., 93, 4329 (1971). H. C. Brown, M. Mark Midland and A. B. Levy, 2222., 22 2144 (1972). _— H. C. Brown, J. Mark Midland and A. B. Levy, ibid., 22 2394 (1973). ‘“ H. C. Brown, M. Mark Midland and A. B. Levy, ibid., 22 2396 (1973). A. K. Hoffmann, and A. T. Henderson, ibid, 22J 4671 (1961). _’ 78 16) A. K. Hoffmann,A~ M. Feldman, E. Gelblum and W. G. Hodgson, ibid., 22, 639 (1964). 17) N. Kornblum and H. W. Pinnick, J. Org. Chem., 37, 2050 (1972). ___ 18) T. G. Back and D. H. R. Barton, J. Chem. Soc., Per. Trans. I, 924 (1977). 19) A. K. Hoffmann, W. G. Hodgson, D. L. Maricle and W. H. Jura, J. Am. Chem. Soc., 22, 631 (1964). 20a) G. F. Hennion and D. E. Maloney, ibid., 22, 4735 (1951). 20b) G. F. Hennion and E. G. Teach, ibid., 22, 1653 (1953). 20c) G. F. Hennion and K. W. Nelson, ibid., 12, 2142 (1957). 20d) G. F. Hennion and R. S. Hanzel, ibid., 22, 4908 (1960). 20e) G. F. Hennion and A. P. Boissele, J. Org. Chem., 26, 2677 (1961). :: 20f) G. F. Hennion and C. V. DiGiovanna, ibid., 30, 2645 (1965). ___ 209) G. F. Hennion and C. V. DiGiovanna, ibid., 31, 1977 (1966). _— 21a) V. J. Shiner, Jr. and J. W. Wilson, J. Am. Chem. Soc., 24, 2402 (1962). 21b) V. J. Shiner, Jr., J. W. Wilson, G. Heinemann and N. Solliday, ibid., 24, 2408 (1962). 21c) V. J. Shiner, Jr. and J. S. Humphrey, Jr., ibid, 22, 622 (1967). _— 21d) A. F. Bramwell, L. Crombie and M. H. Knight, Chem. Ind., (London), 1265 (1965). 22) H. D. Hartzler, J. Org. Chem., 22, 1311 (1964). 23) N. R. Easton, R. D. Dillard, W. J. Doran, M. Livezey, and D. B. Morrison, ibidu,g§, 3772 (1961). 24) K. N. Cambell, B. K. Campbell and L. T. Eby, J. Am. Chem. Soc., 62, 2882 (1938). For a more recent preparation scheme of highly hindered propargylic alcohols, see M. M. Midland, J. Org. Chem., 40, 2250 (1975). ‘ 25) 26a) 26b) 27) 28) 29) 30) 31) 32) 33) 34) 35) 36) 37) 38) 39a) 39b) 40) 79 G. F. Hennion and A. P. Boisselle,.J. Org. Chem., 264 725 (1961). ‘— For a review, see S. Dayagi and Y. Degani in Patai, "The Chemistry of the Carbon-Nitrogen Double Bond," pp. 61-147, Interscience Publishers, Inc., New York, 1970. For a review, see B. C. Challis and Anthony R. Butler in Patai, "The Chemistry of the Amino Group," pp. 298- 302, Interscience Publishers, Inc., New York, 1968. G. F. Hennion and C. V. DiGiovanna, J. Org. Chem., 39, 3696 (1965). _’ G. F. Hennion and R. H. Ode, J. Org. Chem., 31, 1975 (1966). _— F. W. Bergstrom and W. C. Fernelius, Chem. Revs., 20, 433 (1937). ‘— H. C. Brown, "Organic Synthesis Via Boranes,‘ pg. 9, Wiley-Interscience, Inc., New York, 1975. C. A. Brown and H. C. Brown, J. Am. Chem. Soc., 85, 1003, 1005 (1963). _— G. F. Hennion and E. G. Teach, ibid., 22} 4297 (1953). R. L. Augustine, "Catalytic Hydrogenation,’ Marcel Dekker, Inc., New York, 1965. pg. 27, T. H. Vaughn, R. R. Vogt and J. A. Nieuwland, ibid., §§. 2120 (1934). T. H. Vaughn, R. R. Vogt and J. A. Nieuwland, J. Org. Chem., 2, 1 (1937). For leading references, see 22 and the references cited therein. T. L. Jacobs, E. G. Teach and D. Weiss, J. Am. Chem. Soc., 22, 6254 (1955). J. H. Wotiz, ibid., 12, 693 (1951). W. Chodkiewicz and P. Cadiot, Compt. Rend., 41, 1055 (1955). “’ W. Chodkiewicz, Ann. Chim., 2, 819 (1957). A. E. Jukes, Adv. Organomet. Chem., 22, 215 (1974). _u—