Molecular Asymmetry of Olefins. I. Resolution of trans-Cyclooctene1-3

J. Am. Chem. .... Sequential Photosubstitution of Carbon Monoxide by (E)-Cyclooctene in Hexacarbonyltungsten: .... Joel F. Liebman and Arthur Greenber...
0 downloads 0 Views 613KB Size
COPE,GANELLIN, JOHNSON,

3276

JR., V A N

SOL

THE

XRTHUR

85

DEPARTMEST O F CHEMISTRY, MASSACHUSETTS INSTITUTE O F TECHNOLOGY, CAMBRIDGE 39, MASS.]

Molecular Asymmetry of Olefins. BY

VOl.

(19) W. F. Charnicki and J , B . Data, J . A m . Pharm. Assoc., 46, 65 (1956).

chim. Fraiice, 12, 10 (1945)

[CONTRIBCTIOS FROM

WINKLER

stirring for 90 min., formation of dicarbanion X was assumed to be complete. Benzylation of Dicarbanion X.-To a suspension of 0.09 mole of dicarbanion X in a mixture of liquid ammonia and tetrahydrofuran was added 35 g . (0.278 mole) of benzyl chloride and the mixture stirred for 2 h r . The ammonia was evaporated and the tetrahydrofuran removed under reduced pressure. The residue was taken up in ice water and excess benzyl chloride and any neutral products were removed by extraction with ether. The aqueous solution was acidified with acetic acid and extracted three times with ether. The ethereal extracts were combined and dried. Distillation at reduced pressure gave initially several grams of starting P-ketoaldehyde I X , which crystallized in the column, followed by 12.1 g. (51%) of 2-benzyl-5-phenyl-l,3-butanedione ( X I I a ) , b.p. 170-173' a t 0.6 m m . Anal. Calcd. for CI6Hl802: C , 81.17; H, 6.81. Found: C , 81.31; H , 6.95. Ketoaldehyde X I I a gave a violet enol test with ferric chloride and the infrared spectrum showed a n enol band a t 6.18 P . Treatment of X I I a with copper acetate solution gave a green copper chelate, m.p. 163-164" after three recrystallizations from ethanol. Anal. Calcd. for C38H3404Cu: C , 72.77; H, 5 . 7 7 ; Cu, 10.69. Found: C, 72.59; H , 5.93; Cu, 10.60. Alkaline hydrolysis of a sample of undistilled reaction product X I I a gave 1,5-diphenyl-3-pentanone ( X I I I ) , b.p. 152-154' a t 0.6 mm., in 5OC&yield. Vapor phase chromatography showed the presence of XI11 and 4-phenyl-2-butanone, which presumably arose from starting material. The semicarbazone of XI\. melted at 115-117' after three recrystallizations from 8OCGethanol, reportedLgm.p. 120-121 o . Butylation of Dicarbanion X.-To a solution of 0.1 mole of dicarbanion X in a mixture of ammonia and tetrahydrofuran was added 40.0 g. (0.29 mole) of n-butyl bromide. After 6 hr., the ammonia was evaporated and ether and water were added. The aqueous layer was separated and acidified with cold dilute hydrochloric acid. The acidified mixture was extracted with ether and the combined ether extracts dried. Most of the solvent was removed t o precipitate 7.0 g. (40%) of starting @-ketoaldehydeI X , which crystallized out and was removed by filtration. The remainder of the solvent was removed from the filtrate, and the residue distilled to give 2 . 5 g. (145;) more @-ketoaldehyde I X , b.p. 95-100" a t 0.4 mm., and 6.2 g. ( 2 T C , ) of %-benzyl-l,3octanedione ( X I I b ) , b.p. 125-140' a t 0.4 mm. Anal. Calcd. for CI5H200?: C , 77.55; H , 8.68. Found: C, 77.74; H , 8.98. Ketoaldehyde X I I b gave a violet enol test with ferric chloride solution and the infrared spectrum indicated the presence of an enol structure. Treatment of X X I b with cyanoacetamide in the same macner as above gave pyridoneamide XIV, m . p . 195-197' after recrystallization from acetone. The infrared spectrum showed n o nitrile band near 4.5 fi b u t showed amide absorption a t 3.02 and 3.18~. Anal. Calcd. for CL8H22N202: C, 72.45; H , 7.43; S , 9.39. Found: C , 72.79; H , 7.32; N , 9.58.

acetic acid. T h e resulting oil was separated by decantation, and gradually crystallized on standing. Recrystallization from hexane gave 6-amyl-3-cyano-2(1)-pyridone (Vc), m . p . 99-100.5" (reported" m . p . 95-96'), Preparation of Triacylbenzenes VIa-b from Monocarbanions IIIa-b.--An aqueous solution of monocarbanion I I I a , obtained from 0.052 mole of dicarbanion I1 and 0.052 mole of benzyl chloride, was acidified with hydrochloric acid. The liberated 0-ketoaldehyde was taken up in ether and 10 ml. of glacial acetic acid was added. The acidified ethereal solution was warmed and then allowed to stand t o room temperature for 24 h r . T h e solvent was evaporated and the residue was recrystallized from methanol t o give 2. 4 g. (29% based on I ) of 1,3,5-tris-(3-phenylpropiony1)-benzene, m . p . 114-117". Another recrystallization from methanol raised the m.p. t o 121.5-122.5', reported's m.p. 122O. A solution of monocarbanion I I b , obtained from 0.12 mole of dicarbanion I1 and 0.13 mole of methyl iodide, was acidified with acetic acid, heated 12 hr. a t BO", and then allowed to stand a t room temperature for 3 days. .4 precipitate formed which was removed by filtration and recrystallized from hexane t o give 2.7 g. (279; based on I ) of 1,3,5-tripropionylbenzene( V I b ) , m.p. 6669". Chromatography of aopentane solution of VIb on alumina raised the m . p . t o 71-73 , reportedL5m.p. 74-75', Independent Synthesis of Monocarbanion IIIa, Copper Chelate IVa, and Cyanopyridone Va.-To a stirred suspension of 2.70 g . (0.05 mole) of sodium methoxide iri 200 ml. of anhydrous ether was added a mixture of 7.40 g. (0.05 mole) of 4-phenyl-2-butanone and 3.70 g. (0.05 mole) of ethyl formate.7 The reaction mixture was allowed to stand a t room temperature with occasional shaking for 6 hr., during which a thick suspension of I I I a appeared. Copper chelate 1l.a was isolated in the same manner as above. The crude chelate was blue-green, m . p . 164-170', indicating the presence of some of the chelate of I7I1I. One recrystallization from chloroform-ethanol gave 4.35 g. (42',;) of chelate I l - a , m.p. 173-175' and 173-176' after two more recrystallizations. Although the recrystallized chelate was blue-green, while t h a t prepared r,iu the dicarbanion was blue, a mixture melting point of the two samples was undepressed and their infrared spectra were essentially identical. A portion of the purified chelate IVa, m.p. 173-176", was decomposed with dilute hydrochloric acid, and the liberated @-ketoaldehyde was taken up in ether. The ethereal solution was extracted with cold dilute sodium hydroxide solution, and the alkaline extract was neutralized to p H 8. This solution was treated with cyanoacetamide t o give cyanopyridone Va, m . p . 193-198" after one recrystallization from water. Further recrystallization from ethanol gave m . p . 204-205", reported' m . p . 198". h mixture m . p . with \-a ( m . p . 204-205') prepared above was undepressed. 2-Benzyl-1 ,3-butanedione ( I X ) .-4-Phenyl-2-butanone was acylated with ethyl formate in ethanol solution b,y the method of Roch* to give @-ketoaldehydeI X , m . p . 99-101 , reported m.p. 100-101 O . Preparation of Dicarbanion X.-To a stirred solution of 0.24 mole of potassium amide, prepared from 9.4 g. of potassium, in 250 ml. of liquid ammonia was added 15.8 g. (0.09 mole) of 2benzyl-1 ,a-butanedione in 100 ml. of tetrahydrofuran. After (18) h l . Roch, Bull.

AUKEN,A N D

I. Resolution of tvans-Cycl~octene~-~

c. C O P E , c. R.G A N E L L I N , H. w,JOHNSON,

JR.,

T. v. V A N

AUKEN,4 AND

HANSJ . s. % ' I N K L E R

RECEIVED J U N E 12, 1963 Resolution of trans-cyclooctene has been accomplished through platinum complexes containing optically active a-methylbenzylamine. Separate fractional crystallizations of trans-dichloro-(trans-cpc1ooctene)-[( or - )a-methylbenzylaniine]-platinum(11) were carried out t o constant rotation. The complexes were destroyed with aqueous potassium cyanide, yielding optically active ( - )- and (+)-trans-cyclooctene. This novel type of asymmetry demonstrates the inability of the trans-olefinic hnkage to rotate with respect to the rest of the molecule.

+

5lolecular asymmetry due t o restriction of rotation about a pivot bond has been well documented in bi(1) or preceding communications on this subject see. ia) A. C . cope,

phenyl compounds suitably substituted with bulky substituent^.^ In the cyclophane series, molecular asymmetry has been demonstrated more recetltly.'

C. F. Hi>wcll, and A K n o w l e s , J . A n i . C h e m SOL, 84, 3190 (1902); ( b ) A C Coi)e. C. R . Ganellin. and H. W Johnson, rr , ibzd , 04, 3191 (19G2). ( 2 ) This work has appeared in part in a brieI communication, ref. l b . (31 S u p p m t e d in part by the Army Research Office (Durham) under G r a a t S o . I)A-ARO-r)-31-124-G240.

14) Sational Science Foundation Postdoctoral Fellow, 1901-19li2. ( 5 ) (a) R . Adams and H C . Yuan, Chem. R e v , l a , 2(31 ( 1 9 3 1 ) ; (b) R . L. Shriner and R . Adam5 i n H Gilman, Ed., "Organic Chemistry,' John WileY and Sons, I n c . , N e w Y o r k , N V , 1943, pp. 343-382.

Oct. 20, 1963

3277

RESOLUTIOX O F trUns-CYCLOOCTENE

This type of asymmetry arises from the inability of part of a ring, a substituted benzene nucleus in all cases, to rotate with respect to the rest of the molecule.' I t has been suggested that trans-cyclic olefins of intermediate size (8-10 membered rings) should be capable of existence in stable enantiomorphic conformations.k T h e molecular asymmetry of trans-cyclic olefins depends on the inability of the trans-olefinic bond to rotate with respect to the remainder of the molecule; see l a and lb. Presupposing a fixation of

taining the olefin and optically active (+ or - ) a-methylbenzylamine. The difference in solubility of the diastereoisomers makes a separation by fractional crystallization possible. The olefin is recovered by the ready decomposition of the platinum complex with aqueous potassium cyanide under conditions which do not cause cis-trans isomerization of the transcyclooctene. This resolution scheme does not intrinsically depend on the geometry of the platinum complex, square-planar dsp2-hybridization, but rather on the convenience of attaching both the olefin and the amine to the platinum atom and the subsequent readiness with which these can be removed. The coordination chemistry involved in this resolution scheme is outlined below. KzPtC14

la

+ C2H1 K[PtCla(CzH4)] + KCI + Am* -+ [PtCL(C2H4)Am*]+ KCI --f

K[PtC13(CzHd)]

2a,b,c (+)-CsH5CH(CHa)SHs; ( b ) Am* = ( - ) - C s H j C H ( C H 3 ) N H s ; ( c ) Am* = ( + ) - C ~ H ~ C H Z C H ( C H I ) N H Z ( a ) Am*

=

+ t r e n s - C s H ~--+ ~

[PtC12(CzH4)Am*]

[PtCMC&L4)Am*l

lb

the double bond (roughly with the plane of the sp2hybridized bonds perpendicular to the "plane" of the methylene groups) it may be seen from the figure and confirmed by inspection of models that the mirror images l a and l b are nonsuperimposable. Conversion of one enantiomorph of a trans-cycloalkene into its mirror image involves rotation b y lSOo of the plane containing the sp2-hybridized carbon bonds. The ease with which this rotation can occur depends on the ease with which the hydrogen atoms bonded to the olefinic linkage can pass through the loop of the methylene groups. I t is evident that the ease of racemization will increase as the nonclassical strain (nonbonded interactions) decreases, or the larger the methylene bridge is. The particular difficulty in attempting resolution of trans-cycloalkenes is due to the absence of groups suitable for the direct formation of a diastereoisomeric pair of compounds by reaction of such a group with an optically active compound. The ease with which strained trans-cycloalkenes isomerize to cis-cycloalkenesg likewise discourages schemes of resolution requiring attachment and subsequent removal from the ring of such a group. I t has IIOW been found t h a t the strong complexing tendency of trans-cycloalkenes with platinum may be used in a method of resolution that should be general. Olefins are known to form coordination compounds with platinum. The resolution of trans-cyclooctene has been accomplished through the formation of a diastereoisomeric pair of platinum( I I ) complexes con(6) For a general discussion of molecular asymmetry see' E. I,. Eliel, "Stereochemistry of Carbon Compounds,'' McGraw-Hill Book Co., Inc., P;ew York. N . Y . , 1909, pp 156-179. (7) A. T. Blomquist, R. E. Stahl, Y . C. Meinwald, and B. H. Smith, J . Org, Chem Z6, 1687 (1961), and references quoted therein. (8) ( a ) A. T. Blomquist, L. H. I i u , and J . C. Bohrer, J A m . C h e m . SOL., 74, 3643 (19.32); ( b ) V. Prelog in Sir A. Todd, E d , "Perspectives in Organic Chemistry," Interscience Publ., Inc., New York, N . Y..1950, p. 129. (9) A C Cope, P . T. Moore, and W. R . Moore, J. A m . Chem. Soc., 82, 1744 (1900). (10) For reviews of such complexes see: (a) R. N. Keller, Chem. Rev , 28, 229 (1911), (b) J C h a t t , A n n . Repl. PYORT. Chem. (Chem Sac London), 43, 120 (1940); (c) J C h a t t in P H. PI-sch, Ed., "Cationic Polymerization," W Heffer and Sons, Cambridge, Eng , 1953, p. 4 0 ; (d) B. E. Douglas in J. C. Bailar, Jr., Ed , "The Chemistry of CoGrdination Compounds," John Wiley and Sons, Inc., New York. N . Y . , 19.56, pp. 487-508. (11) For discussions of olefin-platinum complexes see. A . A Babushkin, L. A. Gribov, and A D. Gelman, R K S SJ. . Inorg. Chem., 4, 695 (1959); D . M . Adams and J C h a t t , Chem. I n d . [London), 149 (1960). I

3a,b,c l 2

+ CPH~

+ 4KCX + K2Pt(Ch1)1+ 2KCl + Am* + trans-CsH14

[PtCI2(CsHl4)Am*]

Zeise's salt,I3 potassium trichloro-(ethylene)-platinate(11) was prepared from ethylene and potassium tetrachloroplatinate(I1). This salt on reaction with (+)a-methylbenzylamine in aqueous medium yielded 2a, (+)-trans-dichloro- (ethylene) - ( a- methylbenzy1amine)platinum(I1). This compound was generally obtained as a viscous oil, but the infrared spectrum and specific rotation were identical with those of a crystalline sample. Purification of the compound by chromatography has been shown to give crystalline material.14 An olefin exchange reaction is a well-known route to olefin-platinum c0mplexes.~5 In this case the ethylene was readily displaced b y trans-cyclooctene to give 3a, 3b, or 3c.I6 The crude complex of type 3a is a mixture of the diastereoisomers trans-dichloro- [(-)-transcyclooc tenel- [ (+)-a-methylbenzylamine ]-platinum(11)l7 (3a'), which is dextrorotatory, and transdichloro- [ (+)-trans-cyclooctenel- [ (+)-a-methylbenzylaminel-platinum(I1) (3a"), which is levorotatory with an angle of smaller absolute magnitude. The less soluble isomer 3a' can readily be obtained by fractional crystallization to constant rotation, [ a] ? 3 ~ +67", from carbon tetrachloride. I * However. since it proved difficult to obtain the more soluble isomer 3a" from the mother liquors, the enantiomorph of the s~ was prepared less soluble fraction, 3b', [ a ] 2 1 .-B8.2', (12) In this discussion the least soluble fraction of t h e diastereoisomeric pairs Sa, Sb, and 3c will be designated Sa', 3b', and 3c' and the more sl. Nagy of this department. Infrared spectra were recorded on a Perkin-Elmer model 2 1 spectrophotometer by hlrs S a n c y Alvord. Optical rotations were measured with Rudolph or Zeiss polarimeters. Concentrations are given in g.,'100 ml., lengths of polarimeter tubes in d m . Gas chromatographic analyses were carried o u t using 180 X 0.8-cm Pyrex tubes packed with 48-100 mesh firebrick (JohnsLlanville) which was coated with the appropriate stationary phase. T h e samples =-ere eluted with helium at 15 p s i a n d thermal conductivity cells were used as detectors.

( 2 0 ) J C h a t t and M L. Searle, Inovg. S m . , 6, 210 (1957). (21) A W Ingersoll, "Organic Syntheses," Coll. Vol 11, John Wiley and Sons, Inc., S e w York, N . Y , 1943, p, 506. (22) A. C. Cope, R. A . Pike, and C. F. Spencer, J . A m . Chem. S o l , 76, 3212 (19531.. (23) T h e large variation in specific rotation observed for the same sample is the result of the small volumes (usually 2 ml.) and volatile solvent which it was necessary t o use. T e n measurements of t h e rotation of fully resolved material gave ID 1-87.5 3z l . O o , assuming no concentration and temperature effects in the ranges used.

RESOLUTION OF

Oct. 20. 1963

with three 30-ml. portions of pentane. T h e organic layers were combined, washed with l o c i hydrochloric acid, aqueous saturated bicarbonate solution, and water. The solution was dried over magnesium sulfate and concentrated without heating a t 400 inm. in a 46-cm. spinning-band column. Distillation of the residue without heating a t 0.3 mm. into a t r a p cooled with Dry Ice gave 357 mg. (77'x) of (-)-trans-cyclooctene, [ a I z 5 D -458' ( n e a t ) ; [ c Y ] ~ ~-426', D -423' (c 0.41, 0.49, methylene chloride). The compound had an infrared spectrum and gas chromatographic retention times [ X M P S (30 and 75') and T C E P f63') columns'g] identical with those of authentic dltrans-cyclooctene. Anal. Calcd. for CSH14: C, 86.97; H, 12.91. Found: C , 86.98; H , 12.97. ( +)-trans-Cyc1ooctene.-From 1.94 g. (3.90 mmoles) of [ - ):trans-dichloro-( trans-cyc1ooctene)-(a-methylbenzy1amine)platinum(I1) (3b') there was obtained 288 mg. (679:) of ( +)-trans-cpclooctene using the procedure described for ( - 1trans-cyclooctene. The ($- )-trans-cyclooctene had [ a ]2 5 +440" ~ (neat), [ ~ ] % S D f41-L' ( c 0.55, methylene chloride), and an identical infrared spectrum and gas chromatographic retention time [ T C E P column (68")] with authentic dl-tvans-cyclooctene. An analytical sample was obtained by gas chromatography ( S M P N ,

75'). Anal. Calcd. for C ~ H I L :C, 87.19; H , 12.81. Found: C, 86.88; H , 12.80. .A sample of this enantiomer was also obtained from the more soluble fraction 3a" by decomposition with aqueous potassium cyanide as previously described; this sample of (+)-transcyclooctene had [ c z ] ~ ~ D+382" ( c 2.01 in methylene chloride, 2 I ) , [a]265161 462" ( c 1.52 in methylene chloride, 1 2 ) and an identical retention time ( N M P N column, 65") with authentic dltrans-cyclooctene. The sample contained no cis-cyclooctene. Anal. Calcd. for C8Hl4: C , 87.19; H , 12.81. Found: C , 86.97; H , 12.91. Hydrogenation of ( - )-trans-Cyclooctene .-( - )-trans-Cyclooctene (83.0 mg.) in 8 mi. of glacial acetic acid was hydrogenated a t atmospheric pressure using 92 mg. of platinum oxide catalyst. Uptake of hydrogen was quantitative. After the catalyst was removed by filtration, the reaction mixture was diluted with water and extracted with pentane. The pentane extracts were then washed with saturated aqueous bicarbonate solution and with water and dried over magnesium sulfate. The pentane solution, after concentration to 0.3 ml. using a 46-cm. spinning-band column a t 400 mm., showed no rotation (1 1 d m . ) . Examination of the acetic acid solution ( I 2 d m . ) and of the bicarbonate and water washes combined ( 1 2 dm.) also showed no rotation. Cyclooctane, obtained from the pentane solution by gas chromatography, had a n infrared spectrum and gas chromatographic retention times [ N M P X (26 or 75') and T C E P (68') columns] identical with an authentic sample of cyclooctane. A solution of ( - )-trans-cyclooctene in glacial acetic acid showed no change in rotation after 25 hr. Hydrogenation of ( )-trans-Cyclooctene .-( )-trans-Cyclooctene (130 m g . ) in 9.5 ml. of glacial acetic acid was hydrogenated a t atmospheric pressure using 71.3 mg. of platinum oxide catalyst. The product was isolated in the manner described for the hydrogenation of ( - )-trans-cyclooctene. Neither the concentrated pentane solution ( 3 ml., 1 1 dm.), the acetic acid solution (1 2 d m . ) , nor the combined bicarbonate and water washes ( 1 2 d m . ) showed optical activity. Cyclooctane, obtained by gas chromatography, had a n identical infrared spectrum and identical gas chromatographic retention time [ T C E P column (68")] with authentic cyclooctane.

+

+

t?'U?ZS-CYCLOOCTENE

3279

Thermal Stability of ( - )-trans-Cyc1ooctene.-Samples of ( - )-trans-cyclooctene in sealed tubes were heated a t the temperature of refluxing chloroform. After 7 days there was no significant change in rotation. trans-Dichloro-( ethylene)-( 1-phenyl-2 -aminopropane)-platinum(11) (2c).-The procedure used for t h e preparation of transdichloro-(ethylene)-(p-toluidine)-platinurn( 11) was used . t 4 A solution of 3.30 g. of Dexedrine in 45 ml. of water containing just enough hydrochloric acid to dissolve the amine mas added dropwise with stirring t o a solution containing 7.42 g. of di-rchloro-l,3-dichloro-2,4-bis-(ethylene)-d~plat~num( 11) and 1.50 g. of sodium chloride in 45 ml. of water. An ice bath was used t o keep the temperature below 5", and the addition required 1 . 5 hr. T h e pH of the solution was raked t o 6 by the dropwisp addition of 3m, sodium hydroxide solution with cooling and stirring. .4 small amount of tar formed which was removed by careful decantation. The product was separated by filtration and dried over phosphorus pentoxide for 3 days a t 0.05 m m . Crystallization from cyclohexane gave 9.1 g. (86c/,) of yellow needles, m . p . 1 2 G 1 2 l o , [ a I z 8+20.5" ~ ( c 2, methylene chloride). Anal. Calcd. for C I I H I ~ X P ~ C C I ~, :30.76; H , 3.99; ?;, 3.27; Pt, 45.74. Found: C, 30.54; H , 4.00; S , 3.31; P t , 45.74. trans-Dichloro-( trans-cyclooctenei-( 1-phenyl-2-aminopropane)platinum(I1) (3c).--A solution of 5 g . of trans-dichloro-(ethylene)(l-ph,enyl-2-aminopropane)-platinum( 11) in 15 ml. of methylene chloride was treated with a solution of 1.35 g. of trans-cyclooctene Effervescence occurred during in 15 ml. of methylene chloride. the addition. After the addition Nas complete the solution was refluxed for 2 min. The solvent was removed under reduced pressure t o yield 6 g. of a viscouc oil still containing a small amount of methylene chloride. The viscous oil was fractionally crystallized from hexane. The amount of solvent was adjusted t o give approximately 50Vc of the solute in the precipitate a n d the remainder in the solvent. I t was necessary t o reflux the solvent t o dissolve the gum, a n d by cooling t o -20' a solid t h a t appeared t o be crystalline separated. q-n warming t o room temperature, the yellow solid melted t o a very viscous oil. Using low temperatures, apparently satisfactory separation occurred. After ten steps of fractional crystallization the more soluble fraction 3c" showed [ ~ ] Z @ D 4-4'' ( c 2.5 in methylene chloride); the less soluble fraction 3c' showed [ a ] " D +24" ( c 2.5, methylene chloride). The yield of 3c" was 0.75 g. and of 3c' 0.70 g. Anal. Cdlcd. for C17H27SPtCI2: C , 39.91; H , 5.32; N , 2.75; P t , 38.17. Found for 3c": C, 39.90; H , 5.26; S , 2.82; Pt, 38.04. Found for 3c': C, 40.28; H , 5.41; S , 3.03; P t , 37.95. Partially Resolved trans-Cyc1ooctene.-The complex 3c" was dissolved in 20 ml. of methylene chloride and then shaken with 10 ml. of 207, potassium cyanide solution. The color of the methylene chloride layer changed from yellow t o colorless. The methylene chloride layer was washed with three 5-ml. portions of water, three 5 m l . portions of 554 hydrochloric acid, water, 5"/L.sodium bicarbonate, and water. After drying the organic phase ovcr sodium sulfate, the solvent was removed on a steam bath, and thP residue was distilled in a short path still, yi-lding 120 mg. (75YG) of trans-cyclooctene, [a]*'D +18" ( c 2.5, pentane). T h e complex 3c' was treated similarly and gave 108 mg. of tmnscyclooctene (7056) from the complex, [ a ] " D -21.5" ( ( 2.5, pentane). The observed rotations were of the order of 0 50'. (24) J. Chatt,

J. Chem. Soc., 3340 (1949).