Feb. 5, 1958
COMMUNICATIONS TO THE EDITOR
for 45 min. completely destroyed the biological activity. After decolorizing, the solution was taken to a small volume and methanol and ethanol were added, yielding crystalline 1,3-diamino-4,5,6trihydroxycyclohexane dihydrochloride (desoxystreptamine) ? Anal. Calcd. for CsH14N203.2HCl: C, 30.66; H, 6.86; N, 11.93; C1, 30.17; neut. eq., 117.5. Found: C, 30.76; H, 6.99; N, 11.36; C1, 29.6; neut. eq., 118.7. The compound showed no optical activity in acid or alkaline aqueous solution. The dihydrochloride was converted to desoxystreptamine by dissol_ving in water, treating with Amberlite I R 410 (OH),3 concentrating, and crystallizing with ethanol. Anal. Calcd. for C6H14N203: C, 44.4; H, 8.70. Found: C, 44.66; H, 8.86. A portion was converted to the dihydrobromide by crystallization from dilute aqueous HBr with methanol. It gave an infrared spectrum identical with that of an authentic sample of 1,3-diamino4,5,6-trihydroxycyclohexanedihydrobromide isolated from neomycin. lo The desoxystreptamine was additionally characterized by acetylation in pyridine and acetic anhydride to give pentaacetyl1,3-diamino-4,5,6-trihydroxycyclohexane. Anal. Calcd. for C16H24N208: C, 51.6; H, 6.48; N, 7.53. Found: C, 51.7; H, 6.98; N, 7.30. Selective de0-acetylation with Amberlite I R 410 (OR)%in water yielded N, N ’-diacetyl-1,3-diamino-4,5,6-trihydroxycyclohexane. Anal. Calcd. for CloHlsNz0 6 : C, 48.77; H, 7.37. Found: C, 48.82; H, 7.73. Other products of the acid hydrolysis shown to be present by paper chromatography include two ninhydrin-positive reducing materials, accounting for the major portion of the material left after removal of desoxystreptamine, and several minor components. These data allow speculation that kanamycin is a trisaccharide-like molecule, composed of two aminosugar moieties glycosidically linked to desoxystreptamine.
753
Consequently, i t appears highly probable that the cation of I should have an octahedral configurationj3s4 which leads t o the conclusion that i t should be asymmetric. A resolution of racemic [Si(acac),]+ into its optical enantiomers6 would offer conclusive evidence for its octahedral configuration. Optical Resolution Studies.-Racemic [Si(acac)3]Cl*HClwas prepared by a modification of the method of Dilthey6 and its resolution was attempted by treatment with sodium (-)-dibenzoyld-tartrate,?+ according to the equationlo 2D,L-[Si(acac)a]Cl.HCl-I- d-(-)-Na&HlzOg + ~-[Si(acac)a]d-( -)-HC18H12O8 ~-[Si(acac)~]Cl.HCl -I2NaCI (1)
+
The diastereoisomer L- [Si(acac)s]d-(-)-HC18HI208is less soluble than the other diastereoisomer and precipitates from solution. For L- [Si(acac)r]d-( -)-HC1~H1~08,a 0.008870 solution showed sobs to be -0.068 + 0.003’; [ r u I z 3 ~-773’. The optical rotation of the diastereoisomer decreased (became more positive) with time as the asymmetric silicon complex ion racemized, and, after six hours, the observed rotation decayed to a constant value ( -O.OOSO) due to the asymmetric monohydrogen dibenzoyl-d-(-)tartrate anion. I n order to replace the resolving anion with chloride, the diastereoisomer was treated with a n anion-exchange resin in the chloride form (Dowex 1-X8), which gave ~-[Si(acac)~]CI.HCl.A 0.0311% solution of this enantiomer showed CUobs t o be -0.300 The observed rotation de0.003’; [ r u I 2 3 ~-965’. creased (became more positive) as the compound racemized, and decayed t o zero degrees in six hours with a half-life of one hour. By extrapolation, the specific rotation, [cY]’~Da t zero time is found to be -1,300°. I n order to determine whether the anion-exchange resin removed the negative rotating monohydrogen (- )-dibenzoyl-d-tartrate anion (9) D. A. Kuehl, Jr., M. N. Bishop and K. Folkers, THISJOURNAL, 73, 881(1951). quantitatively from the diastereoisomer, an experi(10) Supplied by Dr. H. E. Carter, University of Illinois. ment was performed in which a sample of soM. J. CRON dium monohydrogen (-)-dibenzoyl-D-tartrate was D. L.JOHNSON treated with resin under similar conditions; the RESEARCH DEPT. F. M. PALERMITI BRISTOL LABORATORIES, INC. Y. PERRON filtrate showed no optical rotation, indicating that SYRACUSE I, NEW YORK H. D. TAYLOR essentially complete removal of the asymmetric D. F. WHITEHEAD anion had been achieved. I. R. HOOPER The optical rotation of the filtrate from reaction RECEIVED JANUARY 2, 1958 1 was found to be strongly dextrorotatory. It was treated with the chloride form of an anion-exchange resin (Dowex 1-XS) in order to remove the THE RESOLUTION OF A HEXACOVALENT SILICON monohydrogen (- ))-dibenzoyl-D-tartrateanion, and (IV) COMPLEX its rotation was measured. A 0.1 164y0 solution Sir: showed a0bo = 0.790 f 0.03’; [ruIz3D +679O. The possibility of six-coordinate silicon(1V) has The observed rotation of this solution of the dexbeen discussed by Stone and Seyferth‘ who de(3) A. Werner, Ber., 47,3087(1914). scribed the availability of d-orbitals for bonding (4) J. C. Bailar, Jr., Ed., “Chemistry of the Co(lrdination Comin silicon(1V) compounds. West2 described the in- pounds,” Reinhold Publishing Co., New York, N. Y., 1958. PP. 274 frared spectra of [Si(acac)3]Cl.HCI (I) and other ff. (5) acac acetylacetonate anion, CrH?Ol-. This cation Is also P-diketones of silicon(IV), and from a comparison tris-(2,4-pentanediono)-silicon(IV)cation. of these spectra with that of free acetylacetone named (8) W. Dilthey, Bn.,86, 932,1595 (1903). concluded that the ligands of I behave as bidentate (7) S.Kirschner, Dissertation, University of Illinois, 1954,p. 43. (8) C. L. Butler and L. H. Cretcher, TEIS JOURNAL, 66, 2805 chelating agents.
-
(1) F. G. A. Stone and D. Seyferth, J . Inorp. N U C ~ CChem., Q ~ 1, 112 (1955). (2) R. West, Abstr. of Papers, 132nd Meeting, Am. Chem. Soc.. 11-N(1957),THISJOURNAL, in press.
(1933). (9) F.Zetzsche and M. Hubacher, Rclu. Chim. A c f a ,9,291(1928). (10) “n” and “L” indicate dextrorotatory and levorotatory complex Ions, respectively; the “d-( -)-” prefix for dibenzoyltartrate indicates a negative rotating anion derived from dextrorotatory tartaric acid.
754
COMMUNICATIONS ro THE EDITOR
Vol. 80
tro enantiomer decayed to zero degrees with a half-
and analyzed for hydrogen and deuterium by the life of one hour a t 23", which is the same half-life quantitative pyrolytic decomposition of this mateobserved for the levo isomer. The lower specific rial to boron and hydrogen-deuterium a t 900" folrotation of the dextro isomer is due to the presence lowed by mass spectrometric measurements of the of some levo isomer in a relatively large quantity of H / D ratio. dextro isomer. Table I sets forth the infrared absorptions and Resolution of the complex cation D,L- [Si(acacja]+ deuterium analyses of several samples of decabointo its optical enantiomers indicates that the sili- rane which were equilibrated with deuterium oxide con(Ilr) is six-coordinate in this complex, and that in dioxane solution for varying periods of time a t the cation is octahedral. The possibility of seven- room temperature. coordination (coordination by the chloride ion) is As seen in Table I the first detectable deuterium remote in view of the conductivity of the complex which enters the decaborane molecule enters the in anhydrous chloroform solution. Additional de- bridge positions but longer equilibration times protails of the resolution of this and other six-coordi- duce appreciable exchange a t terminal positions. nate silicon(1V) complexes will appear in a forth- Furthermore the equilibration of a sample of BIOcoming publication. H10D4 (bridge-&) in pure dioxane produces after The authors wish to express their sincere appre- one day a t room temperature appreciable amounts ciation to the Union Carbide Corporation for a re- of absorption a t 5.25 p as a sharp band charactersearch fellowship grant which contributed signifi- istic of terminal B-D bonds. Similar rearrangecantly to the progress of this investigation. ment of deuterium in B l ~ H l ~ D (bridge-&) I decaboDEPARTMENT O F CHEMISTRY S. K. DHAR rane does not occur in methylene chloride solution VERA DORON under identical conditions of time and temperature. WAYNESTATEUSIVERSITY STANLEY KIRSCHNER DETROIT2, MICHIGAN Table I1 presents the results of three exchange RECEIVED DECEMBER 19, 1957 experiments carried out with decaborane and deuterium chloride in sealed bulbs a t room temperaDEUTERIUM EXCHANGE OF DECABORANE WITH ture. It is to be noted that methylene chloride DEUTERIUM OXIDE AND DEUTERIUM CHLORIDE solvent does not produce exchange. Sir: These results clearly point to the fact that the We wish to report the results of a series of deu- four bridge hydrogens of decaborane are rapidly terium exchange experiments which illustrate the removed by bases (water, dioxane, etc.) and are exchange of as many as eight hydrogen atoms of decaborane by deuterium and which further illus- easily replaced by acids (hydronium ion, hydrogen trate either the existence of two types of exchange- chloride, etc.). Furthermore, less rapid but apable hydrogens (bridge or terminal) or the dioxane preciable exchange involving terminal positions apcatalyzed migration of bridge deuterium to ter- pears to occur only in the presence of basic solvents minal positions. such as dioxane or diethyl ether by what may be a TABLE I RESULTS O F DYO-BIOH~~ EXCHANGE EXPERIMENTS I N DIOXAKE SOLUTION AT ROOMTEMPERATURE Equilibration time, hr.
-Initial BioH14
0.15 0.15 2.00 36.0 72.0
0.0S3 0.167 0.167 0.083 0.120
concn., mole/l.-Dz0
B-D infrared absorption
Bridge Bridge Bridge Bridge Bridge
3.10 6.20 6.20 3.10 1.55
+ extremely weak terminal + very weak terminal + weak terminal + moderate terminal + strong termirial
Deiiteriuni/ mole
2.3 4 ti 5 2
0 7 7.5
TABLE I1 EXCHANGE OF BlaH,4WITH DC1 Solvent
Dioxane Dioxane hlethyleiie chloride
Exchange time, hr.
114
96 93
--Initial DCI
0.036 .032 .0384
quantities, moles-BioHii
0.0024 ,0016 .004
Perdeuterodecaborane was prepared by the pyrolysis (100") of perdeuterodiborane. This material gave B-D stretching absorptions in the infrared a t 5.25 p (terminal-dlo) and 7.30 (bridged?). The latter absorption is characterized by a broad band and weak intensity. LVith normal decaborane the corresponding absorptions are found a t 3.5 and 5.25 p , respectively. All spectra were determined in carbon disulfide solution. Partially deuterated samples of decaborane (prepared as described below) were sublimed, weighed
B-D infrared absorption
Bridge $- strong terminal Bridge strong terminal Identical with B:oHI~
+
Deuterium! molt
6.2 5.9 0
base-catalyzed internal rearrangement of bridge deuterium to terminal positions or a base-catalyzed process involving another type of BlaH13- anion formed by direct removal of a terminal proton. That this latter terminal hydrogen exchange process is not as rapid as bridge exchange is apparent from the above results. REDSTOSEARSESAL RESEARCH HAWTHORNE DIV. OF ROHMAND HAASCo. M. FREDERICK ALABAMA Jor-is J. MILLER HUNTSVILLE, RECEIVED JANUARY 2 , 1958