Thiosteroids. 111. The Transannular Addition of Hydrogen Disulfide to

A nuinher of AIz'-8-oxosteroids have becii converted to lol,jol-epidithio-3-o~05teruids2 by the addition of hydrogen di-. The uroof of structure of th...
0 downloads 0 Views 542KB Size
XDDITIONOF HYDROGEN DISULFIDETO

Aug. 20, 1959

[CONTRIBUTION FROM

Thiosteroids.

111.

THE

THE

STEROIDA RING

4409

CHEMICAL RESEARCH DIVISION OF G D. SEARLE 81 C O . ]

The Transannular Addition of Hydrogen Disulfide to the Steroid A Ring BY ROBERTC. TWEITAND K . 51. DODSOK RECEIVED FEBRUARY 20, 1959

A nuinher of AIz'-8-oxosteroids have becii converted to lol,jol-epidithio-3-o~05teruids2b y the addition o f hydrogen disulfide. T h e uroof of structure of these steroid disulfides and the mechanism of the addition of hydrogen disulfide are discussed

I n recent years a number of transannular reactions and interactions involving non-adjacent atoms have been d i s ~ o v e r e d . ~The larger rings, where various conformations can enable atoms a t opposite sides of the ring to approach each other closely, seem especially prone to this type of reaction. I n the reactions which have given good yields of anomalous products in the larger rings, transannular effects in six-membered rings have been observed only as very minor side reactions. However, other interactions involving non-adjacent atoms have been found in the cyclohexane system; for instance, the formation of 1,4epoxycyclohexane in the alkaline hydrolysis of trans-4-ch1orocyc1ohexanol,* the rearrangement of cis-3-methoxy-cyclohexanecarboxylic acid on treatment with acetic anhydride and acid to a mixture of 3- and 4-acetoxy esters,j and a number of others6 In the steroid series the 3a,Sc~-oxides,~ the photoisopyrocalciferols8 and the i-isteroidsga are examples of transannular interactions. The addition of oxygen across the a2.4, a j ,7 and As ( 1 4 ) . 9 (1')-systemslO,11,gb, 12 Seems to be analogous to transannular Diels-Alder reactions. During our investigations of additions to double bonds in the A ring of steroids' a new type of transannular reaction was discovered. When an attempt was made to add hydrogen sulfide to the Al-bond of l,4-androstadiene-3,17-dione7 a new compound (11),

Cl9H26O2S2,was isolated. The yield of I1 was greatly improved by the addition of an atomic equivalent of sulfur to the pyridine solution of the steroid before the solution was saturated with hydrogen sulfide. The infrared and ultraviolet spectra of the new compound indicated a saturated six-membered ring ketone and a saturated five-membered ring ketone. The analysis, the failure of the compound to decolorize iodine solution, and the absence of a band in the 4 region in the infrared spectrum showed the absence of free sulfhydryl groups. A Rast molecular weight determination indicated that the molecule contained only oiie steroid nucleus. Reduction of the dione-disulfide I1 with sodium borohydride gave two substances, a diol-disulfide I11 and a diol-dithiol IV, which could be oxidized to the diol-disulfide 111 with iodine. The diol-dithiol I V formed a tetraacetate. The diol-disulfide I11 was desulfurized with Raney nickel to 3-androstene3a,17/3-diol. In contrast the dione-disulfide I1 was converted to l,Landrostadiene-3,17-dione, probably by the base present in the Raney nickel used. hfethanolic sodium hydroxide alone converted the dione-disulfide I1 to l,~-androstadiciie-3,17-dione.

(1) Paper I, R.. hf. Dodson and R . C. Tweit, THISJ O U R N A L , 81, 1224 (1459); Paper 11, R . C. Tweit and R . M. Dodson, J . Ovg. Chem., 24, 2 i 7 (1959). ( 2 ) This nomenclature was suggested t o us b y Dr. Leonard T. Capell uf "Chemical Abstracts." ( 3 ) For r e v i e w see: Airiz. K e p i s . o n Progv. C h e m (Chent. SOC. L o n d o n ) , 64, 181, 210 (1957), and Ii. Huisgen, A??gew.Chem., 69, 341 (1057). ( 4 ) H. TV, Heine, TITISJ O U R N A L , 79, ti268 (11157). m d H. I. Weingarten, i b i d . , 79, 3008 (lY57). (6) E. I,. Beiiiil:tt and C. S i e m a n n , zbid.,74, 5076 (1052); \V. R . Hatchard and A. K. Schneider, i b i d . , 79, 6261 (1957); D. S. Noyce and H . I. Weingaiten, ibid., 79, 3093 (1957); D. S. Noyce and B. R . Thomas, i b i d . , 79, 755 (19571, S. Archer, M. R. Bell, T. R. Lewis, The positions (1 atid 5 ) , beta to the ketone, for the J. W. Schulenberg and M. J. Unser, ibid., 79, 6337 (1957); W. G. disulfide linkage are consistent with the ease of reDauhen. R . C. T n e i t and R.1,. Maclean, ibid., 77, 48 (1955); N. A . Nelson and G. A. Mortinier, J . 01,s.C k e m . 22, 1140 (1957); L. S . versal of the addition, and the isolation of 3-androOwen and P. il. Rubins. J . Ciieln. Soc., 3 2 0 , 32(i (1940); E. F. Ullman, stene-3a,l7P-diol froni desulfurization is further evC / i c w i ~ l v .&~ ' I ~ d u t ~ 1173 y , (1958); H . K . Hall, Jr., THISJ O U R N A L , 80, idence for the attachment of a sulfur at C-3. Thc 0.112 (1!)38); and rrtlier references. relatively easy acetylation of the tertiary thiol at C-3 ( 7 ) 1'. R.h l a t t o x . R . B. Turner, 1,. 1,. Itngel, B. I;. McKenzie, W. F. AlcGuchin and E. C. Kendali, J . Bio/. C J ~ C ~164, ~ Z . , 5tjO (1946). is an interesting contrast to the more vigorous con(6) W . G. Daubeii and G. J . Fonken, T m s J O U R N A L , 79, 2971 ditions usually required for a C-3cr-hydroxyl.' (1!)37). The difference can probably be best ascribed to the (9) 1.. 1'. Fieser a n d Irl. Fieser, "Xatural Products Related t o greater length of the carbon-sulfur bond and the Phenanthrene," 3rd Edition, Reinhold Publishing Corp., New York, N. I,.,1949; (a) pp. 2 . 5 6 2 6 1 , (b) pp. I(j3-164. greater ionization of the sulfur-hydrogen bond in (IO) H . B. Henbest and R . A . L . Wilson, Chemislry & I n d u s f v y , 86 pyridine compared to a 5-hydroxyl group. (1956). The disulfide bridge was placed on the a-side of (11) U'.Bergmann, F. Hirschmann and E. L. Skau, J . O r g . Chem., 4, the molecule by analogy to other additions to C-1 of 29 (1939). (12) G. D . Laubach. E. C. Schreiher, E. J. Agnello and K. J. Brun(13) P A Plattner, T Petrzilka and W Lang, H t l e Chrin Actu 27, i n s , THISJ O U X X A 78, L , 474U (195ti). 513 (1911)

4-110

x 13 I1 Ii O= 0-01-1

AMD =

ROBERT

Y

z

RZ.p., dcc.,

cc.

c. ' h ' E I T

AND

[a]D

R.hI. DODSON

-Carbon, Calcd.

A.lluL"

I4 321-222 - 4O -397" OAc 4-4 188-190 -412 OAc 222-223 - 10 -309 OH OAc 235-237 - 4 - 403 OH OAc 222-224 + 1 -274 MD (3C=O,la,5oc-~pidithio) - MD(3C=O,Scu-pregiiane). 1-1 11 0€1

the A ring.' The reduction of the 3-ketone to give a Sa-hydroxyl group, in contrast to the usual reductions of 3-ketones to give mainly the 3p-isomer, also supported the a-configuration of the disulfide bridge. The two sulfur atoms apparently hindered the approach of the borohydride ion to the 3-ketone to a greater extent than the 10-methyl group. The reduction of the disulfide bridge with sodium borohydride provided some information about the strain involved in the molecules containing it. The disulfide link in lipoic acid was cleaved by sodium borohydride at 5 ' , I 4 while only the carbonyl groups of the dione-disulfide I1 were reduced below 30". When the temperature was allowed to rise to 60') the mixture of diol-dithiol IV and diol-disulfide I11 mentioned above was obtained. Little information is available on the reduction of linear disulfides with sodium borohydride, although Brown and Subba RaoI5 reduced diphenyl disulfide with sodium borohydride and aluminum chloride in diethylene glycol dimethyl ether a t 25". Thus the stability of the steroidal disulfide bridge is apparently greater than that of the trimethylene disulfides.'6 The ultraviolet absorption spectrum of I 1 indicated that the stability of the dione-disulfide was probably closer to that of the linear disulfides than to trimethylene disulfide. Calvin's1' studies of the ultraviolet spectra of various disulfides showed that trimethylene disulfide had a maximum at 334 m p , while the iliaximum for linear disulfides was at 230 inp. The dione-disulfide I1 had a maximum a t 268 inp, much closer to the value which Calvin found for the linear disulfides, than to the valuc for triiiicthylene disulfide. The mechanism for the foriiiation of tlic disulfide bridge in I1 probably involves the addition of hydrogen disulfide, rather than separate additions of hydrogen sulfide to the Al- and A4-bonds, followed by oxidation to the disulfide bond. On chromatography of the crude product, no products corresponding to the addition of hydrogen sulfide to only one double bond were found. This conclusion was bolstered by the failure of other nionofunctional compounds, such as ethanethiolic acid' and methyl (14) I. C . Gunsalus, L. S. Barton and W. Gruber, THISJ O U R N A L , 7 8 ,

1763 (1956). (15) H. C. Brown and B. C. Subba R a o , ibid., 78, 2582 (1950). ( I F ) S. Sunner [Nature, 176, 217 (195511 has shown trimethylene disulfide t o be 4 kcal. less stable than linear disulfides. (17) J . A . Barltrop, P. M. Hayes and hl. Calvin, THISJ O U R N A L . 76, 4: 15 ( l ! I , x ) .

b

Vol. Sl

5%-

Found

66 . e2 66. 72 62.97 63.27 61.12 61.03 58.80 59.20 59.10 59. 2 g b Contains one niole of acetrine

-Hydrogen, Culcd.

RFound

7.99 8.00 7.39 7.60 7.13 7.18 6.48 e . 15 7.27 7.45 of crystallization

mercaptan, to add to the A4-bond in various steroids. Rough calculations indicate that, in a pyridine solution of hydrogen sulfide, an appreciable fraction of the sulfide is present as monohydrogen sulfide ions. Addition of sulfur should produce a mixture of monohydrogen disulfide ions and hydrogen disulfide, just as addition of sulfur to a sodium sulfide solution gives sodium disulfide. The sulfur-sulfur kond in hydrogen disulfide was reported to be 2.05 A.18 The two sulfur atoms can interact with the orbitals of the C-1 and C-5 atoms (indicated by calculation to be 2.5 A. apart) to form the disulfide linkage in either one or two steps. No choice can be made between a concerted or stepwise mechanisni on the basis of available evidence. Attempts to add hydrogen disulfide to saiitoniti, benzoquinone and 2,6-diphenyl-4-pyrone failed. The failure of santonin to react is particularly interesting, since this is a very close analog of thc steroid A and B rings. The lack of reactivity is probably due to the 4-methyl group, and the lactone ring. 178 -Hydroxy - l a , h - epidithioandrostan-3one, lcr,5a-epidithio-l'ia-oxa-D-homoa1~drostane-3, 17-dione and the other steroid adducts listed in Table I were prepared by similar methods. ,4s can be seen from the table and the experimental section AMD values for the adducts were consistent. The values for the reduction products I11 and II' are worthy of note. The AMDvalue for the disulfide-diol I11 relative to 5a-androstane-3au,17P-diol was -235'. Reduction of I11 to the dithiol IV produced a change in MD of +13(jo, evidence for a rather striking shift in the conformatioll of thc molecule and comparable in magnitude to a A J I D of -2214" for the reductio11 o f the disulfide bond in lipoic acid.14s'9 Experimenta120 I n ,Sa-Epidithioandrostane-3,17-dione(II).- .I holutioii of 46 g. of 1,4-3ndrostadienc-3,17-dlon~aild 5.2 g. of sulfur (18) D . P. Stevrnsc,n and J. Y . Reach, i b t d , 60, 2872 (l!l38). (10) E. Walton, A . 12. Wajiner. F. W. Bachelor, I,. 11. Peterson, F.\L'. IIolly and K. Folkers, ibid.,7 7 . 5144 (1955).

(20) All melting points mere taken on a Fisher-Johns melting poiill apparatus. We are indebted t o Dr. R . T. Dillon and his staff of t h e Analytical Division, G. D. Searle and Co., for t h e analytical and optical d a t a reported. T h e rotations were taken in chloroform a t 24 =k 2', and t h e infrared spectra were taken in potassium bromide disks. T h e methods of preparation of starting materials were discussed in a previo u s paper.' T h e AMD values listed refer to: AMu = AJD ( I a , Z a el,idithiosteroid)-Mn (1 ,sa-unsubstituted steroid).

Aug. 20, 1959

ADDITIONOF HYDROGEN DISULFIDE TO

THE

STEROID A RING

441 1

in 1.0 1. of pyridine was saturated with hydrogen sulfide. 10, p&hamol 238 mp, E 9750. The infrared spectrum had The last of the sulfur dissolved rapidly as the hydrogen sul- bands a t 5.73, 5.91,S.OO-8.03 and 9.00 p and no sulfhydryl bands in the 4 p region. fide was added. The reaction stood overnight and was then concentrated under vacuum. Toluene was added to the Anal. Calcd. for C Z ~ H ~ ~ OC, ~ S61.80; Z: H, 7.68; S, residue and removed under reduced pressure. The residue 12.22. Found: C, 62.13; H, 8.00; S, 12.35. was crystallized from methylene chloride-acetone to yield Reduction of la ,Sa-Epidithioandrostane-3a, 178-diol.39 g. of la,5a-epidithioandrostane-3,17-dione,m.p. 210The steroid, 1.9 g., was mixed with 40 ml. of methanol and 214" dec. to a red melt. This material was contaminated a solution of 2 g. of sodium borohydride in 20 ml. of water with a trace of sulfur. The filtrates were concentrated was added. A heavy precipitate formed and 40 ml. of methunder vacuum and chromatographed on 900 g. of silica gel. anol was added. The mixture was heated on the steamThe column was washed with benzene and 570 ethyl acetate bath, gas was evolved and the solid dissolved. About 50 in benzene and then eluted with 10% ethyl acetate in ben- rnl. of methanol was distilled and on cooling a solid formed. zene. The first 10% ethyl acetate in benzene eluates con- The material was separated by filtration and recrystallized tained la,5a-epidithioandrostane-3,17-dioneand were con- from methanol to yield 1.1 g. of 3a,l7p-dihydroxyandrocentrated under vacuum. After crystallization from methyl- stane-la,5a-dithiol, m.p. 213-216'. The infrared spectrum ene chloride-acetone, the product, 2, g., melted a t 214-218' identified the compound. dec., turned red, [ a ] -21 ~ zk 2 , AMD -400°, hz:ihano' Raney Nickel Desulfurization of la,Sa-Epidithioandro268 mp E 240. The infrared spectrum had carbonyl bands stane-3a,[email protected] steroid, 5.25 g., was dissolved in a t 5.76 and 5.82 and no sulfhydryl bands around 4 p . 100 ml. of 95% ethanol and three teaspoons of Raney The spectrum was identical with that of the main fraction of nickel from a suspension in water was added. The mixture material. was heated under reflux for 7 hours and then filtered through Anal. Calcd. for ClsH2602S~: C, 65.10; H , 7.48; S, charcoal. The solid on the filter was washed twice with 18.29; mol. wt., 350.5. Found: C, 64.77, 65.28; H, 7.47, hot methanol and the filtrates were concentrated. The 7.67; S,18.51; mol. wt. (Rast), 350. residue was dissolved in benzene and chromatographed on Reduction of 1~,5~-Epidithioandrostane-3,17-dione.--200 g. of silica gel. The column was washed with benzene The steroid, 10.00 g., was mixed with 800 cc. of methanol and from the 5oJ, ethyl acetate-benzene eluents a solid was obtained in low yield. The analysis was close to that of a and a solution of 10.0 g. of sodium borohydride in 100 ml. of water was added. The temperature of the mixture rose monohydroxyandrostene, but since the amount of material to about 60°, gas was evolved and the solid dissolved. The was small, it was not characterized further. The 15% ethyl acetate-benzene eluates were concentrated resulting solution was cooled to room temperature and allowed to stand for 0.5 hour. Then, the solution was con- and the residue was crystallized from methanol to yield a (20%) and 0.10 centrated uncler vacuum to one-third liter, cooled, and the total of 0.94 g. of 5-androstene-3~~,17@-diol solid which had formed was separated by filtration and g. of recovered starting material. The dihydroxy compound washed with water. Another fraction of material was ob- was recrystallized twice from methanol-methyl ethyl ketone tained on concentration of the filtrate. This solid had the to yield the analytical sample, m.p. 196-199' (a methy: same infrared spectrum as the first fraction. Finally, ethyl ketone solvate), [ a ] -52' ~ (ethanol) and -71 two more crops, whose infrared spectra did not match those (chloroform, both values calculated to non-solvated mateof the first two fractions, were separated by filtration. rial). The first two fractions were combined and partially disAnal. Calcd. for C I ~ H ~ O O Z . C ~C, H ~76.19; O: H , 10.56. solved in 100 ml. of methanol-acetone. The undissolved Found: C, 75.95; H, 10.31. material was separated and the solution was concentrated RuzickaZ1reported that the ethyl acetate solvate of 5to yield 2.25 g. of 3a,l7p-dihydroxyandrostane-la,5a-di- androstene-3a,l78-diol melts unsharply about 200" and tha: thiol (IV),m.p. 216-218', [ a ]+67 ~ f l o , AMD+201'. a sublimed sample has a m.p. of 207-208', [a]D -56 Anal. Calcd. for ClgHazOzS~:C, 64.00; H , 9.05. Found: (ethanol). The infrared spectrum of our compound has an C, 64.07; H, 9.21. hydroxyl band a t 2.98 p and a weak ketone band a t 5.81 p . The infrared spectrum of the undissolved material, 2.7 Our compound did not form a precipitate with digitonin. g., m.p. 211-213", was identical with that of the analytical A diacetate prepared by the usual method melted a t 154sample. The spectrum had a broad hydroxyl band a t 2.94 155", lit.21 m.p. 155-155.5'. The infrared spectrum had p and a sulfhydryl doublet a t 3.90 and 3.97 p . There were no hydroxyl band in 3 p region, and ester bands at 5.77, no bands in the infrared between 5 and 6.7 p (the carbonyl 7.87 and 8.03 p . region). When the dihydroxv compound was oxidized with chroThe last two fractions from the original solution were mium trioxide in pyridine; 4-androstene-3,17-dione was dissolved in the methanol-acetone filtrate from the lol,5a- obtained. dithiol. On concentration, a solid was obtained which was Basic Treatment of la,5a-Epidithioandrostane-3,17-dione. recrystallized from methylene chloride-methanol to yield -The steroid, 2.00 g., was dissolved in 100 nil. of methanol 2.20 g. of la,5a-epidithioandrostane-3a,l78-diol (111), containing 5 ml. of 8 N aqueous sodium hydroxide. The m.p. 238-239" dec., [ a ]-56 ~ f l o ,A M D -235". solution was allowed to stand overnight; some solid formed. Anal. Calcd. for C19H3b02SZ: C, 64.36; H, 8.53. The solvent was partially evaporated under nitrogen and water was added. The precipitate was separated by filtraVound: C, 64.52; H , 8.66. The infrared spectrum had a hydroxyl doublet at 2.92 tion to Xield 1.38 g. of 1,4-androstadiene-3,17-dione,m.p. 142-144 . The infrared spectrum was identical with that arid 3.00 p and neither Sulfhydryl bands around 4.0 p nor of an authentic sample. carbonyl bands in the 5 to 6.5 p region. When the reaction 17@-Hydroxy-h ,Sa-epidithioandrostan-3-0ne .-17@-Hywas repeated with the sodium borohydride added slowly and the temperature kept below 30°, only the epidithiodiol 111 droxy-l,4-androstadien-3-one,2.00 g., was dissolved in 60 ml. of pyridine, and the solution was saturated with hydrowas obtained in 72% yield. Oxidation of 3~,17p-Dihydroxyandrostane-la,5a-dithiol. gen sulfide. Then 2 drops of piperidine was added and air -The steroid, 0.05 g., was dissolved in 3 ml. of methanol was bubbled through the solution for one minute. After 3 and a solution of iodine in ethanol was added dropwise days, the reaction mixture was poured into one-half 1. of until the color persisted. On evaporation of the solution, water and extracted three times with ether. The ether crystals of la,5a-epidithioandrostane-30.,178-diol formed solution was washed five times with water; then petroleum ether, b.p. 35-40', was added, and the solution was conand were separated by filtration to yield 0.02 g., m.p. 235237" dec. A mixture of the compound with authentic centrated on the steam-bath. When the residue was trimaterial melted 235.5-237.5' dec.; the infrared spectrum turated with ether, a solid formed. I t was crystallized from acetoneether and then from acetone to give 0.24 g. of also confirmed the identity. m.p. 2173a,17~-Diacetoxy-la,5~-diacetylthioandrostae.-3~, 178- 17~-hydroxy-la,5a-epidithioandrostan-3-one, ~ f 2', AMD -301". The Dihydroxyandrostane-la,5a-dithiol,1.4 g., was dissolved 219" dec., turned red, [ a ]-81 in 10 cc. of pyridine and 10 cc. of acetic anhydride and infrared spectrum had bands a t 2.93 and 5.83 p . allowed to stand at room temperature for one day. Then Anal. Calcd. for ClgHBO&: C, 64.73; H , 8.00. the solution was poured into water; the mixturewas filtered; Found: C, 64.62; H , 7.55. and the solid was crystallized twice from acetone-petroleum ether (b.p. 60-70') to yield 1.08 g. of 3a,17P-diacetoxy-lor,( 2 1 ) L. Ruzicka. M. W. Goldberg and W. Bosshard, Helo. Chirn. Sa-diacetylthioandrostane,m.p. 235-238", [ a ] D -71 f A d a , 20, 641 (1937).

4412

BARBARA W.Low, F. L. HIRSHFELDAND F. A i . RICHARDS

la,5o-Epidithio-17a-oxa-D. homoandrostane-3,17-dione.6D-Homo-l7a-oxa-l,4-androstandiene-3,17-dione, 2.00 g., was dissolved in 60 ml. of pyridine and hydrogen sulfide wab bubbled into the solution for two hours, Then 2 drops of piperidine was added and the reaction was allowed t o stand for 3 days. After this time 200 ml. of water was added and the solution was extracted three times with ether. A solid

VOl. 81

formed in the aqueous layer and was separated by filtration. It was crystallized from ethyl acetate t o give 0.34 g. of la,5a-epidithio-17a-oxa-D-homoandrostane-3,17-dione, m .p. 232-233" dec., red melt, [ a ] D -140 3~ 2', AMD -457'. Anal. Calcd. for C19H26O3S2: C, 62.26; H , 7.15. Found: C, 62.06, 62.00; H , 6.76, 7.00. CHICAGO 80, ILL.

[COSTRIBGTIOX FROM THE DEPARTMEST OF BIOCHEMISTRY, COLLEGE O F P H Y S I C I A V S A N D S U R G E O N S , COLUMBIA UAIVERSITY ]

Glycinate Complexes of Zinc and Cadmium

mr.

B Y BARBARA LOX'.,' F. L. HIRSHFELD.4ND F. A I . KICHARDS' RECEIVED MARCH13, 1959 Crystals of Zn( XHpCHrCO2)2.HaOand Cd( SHaCH2C02)2,H20 are approximately isomorphous. The latter, being of higher crystallographic symmetry, has been studied in somewhat greater detail although only one projection has been analyzed. In both complexes the metal coordination is octahedral, two glycine ligands chelating with the metal in trans planar array while the other two coordination positions are occupied by carboxyl oxygens of neighboring glycine ligands. The amino groups and the water are fully hydrogen bonded. The difference in symmetry of the two structures is not explained.

Introduction The marked effect of zinc ion on protein solubilitiesZ suggested that a study of the coordination of this metal with amino acids would be of interest (for a review of other work in this field, see Gurd and IYilcox),3 =In X-ray crystallographic investigation of the zinc glycine complex was undertaken in an attempt to establish the nature of the metal coordination in this structure. At the same time, glycine complexes with several other metals were prepared, of which the cadmium complex proved to be crystallographically similar to the zinc complex. X-Ray diffraction photographs showed the two crystals to be nearly isomorphous. The cadmium complex, of much higher symmetry, was far the more amenable to crystallographic analysis. Accordingly, when work on the zinc complex had proceeded far enough to show the general nature of the metal coordination, attention was temporarily diverted t o the cadiniuni complex, which offered greater promise of a more detailed structure determination. Experimental

groups). Most crystals show { 100); other forms commonly observed include {llO] and { 170)of the zinc complex (designated in terms of the triclinic holosymmetric class, as discussed below), { 210 } and { 110)for the cadmium complex. The crystals frequently show etched lines parallel to [ O l O ] on the main face (001) with principal cleavage (100). Crystals of the zinc complex tend to grow radially in clumps so that single crystals with unbroken ends are somewhat rare. I n well-developed specimens both like and unlike termination has been observed.

The zinc and cadmium coniplexes wcrc both prcparcd iii the same way. The metallic oxide (1 mole) and rccrystallizcd glycine (2 moles) were dissolved together in boiling water. After filtration, the solution was allowed t u cool slowly to rooiii temperature. Large crystals were obtained by recrystallization from \rater and from water-alcohol ( 1 : 1) mixtures. hlicro-lijcldahl aualyses and dry-weight deterininations (air, 110') of both complexes establishcc! the crystalline composition: M( SH2CH2COO)~~I-I20.

-;.

Morphology and Optics.-The approximate isoiuorphism of the two preparations was immediately evident from the morphological and optical examination. Both the zinc and the cadmium glycinate monohydrate crystallize as colorless plates elongated along [ O l O ] and usually lying on (001) (see Table I for cell dimensions and space ( I ) P a r t 01 this work w a s carried r u t while ( B . W . L . , anrlF.RI.R i were members of the Department of Biophysical Chemistry, Harvard IIedical School (2) E. J. Cohn, F R . N . Gurd, D . S I . Surgenor. B. A . Barnes. R . E;. Browm, G. Derouaux, J . 11, Gillespie, F. X'. K a h n t , \V. F. I.ever, C . H. Liu, D. Mittelman, R . F.Mouton, K . Schmid and E. 4 .Uroma. THISJ O U R N A L , 72, 465 (1950). (3) F. R. h'.Gurd and P. E. W-ilcox, Adsaizces in Proleiit C h i n . , 11, 311 (1956).

TABLE I CRYSTAL LATTICE COSSTANTS Lattice cliniensicins measured oil a Ge~icr:tl Elcctric Spectrogciniometer fitted with a Eulerian cradle, Goniostat for single-crystal studies. Copper radiation employed : X K a i = 1.54051, X Kcr. = 1.54433. ,'L

Cadmium glycinate monohydrate (I,

.A.

11,

A.

Zinc glycinate monohydrate

14 862 =k 0 ,00.5 15 0.51 zk (i,( N K i ;i. 20; f 005 10.411 zk , oox 10.006 f , 003 10.693 =t , 012 XR.89 i .or, 9(J,-lil f ,i)3 S 7 1 3 zk ,115 !iii,:3!) f ii.5 ,

A'l. CY, degree5 @, degrecs tlcgrccs Space grc t i l ) c,

.\I o r .\i I ii 1 !)!IO _t 11 O O I

12/:1 .1

z

I)ciiiity(calctl.),g ;1111. Ihibity (fh~tatii)ii), 6.1

?,;L-I(~i- o . c i i l l

2.338 0.010 1 .!J!l:j ~ l , O l l t - ) ' & T ,Ik-nas aucl U . l-farkcr, Rcu. S c i . f j z s t i . . , 26, 4&