Aug. 20, 1961
COMB.IUNICATIONS TO r m EDITOR
directly within the microwave cavity by reduction of millimolar solutions of the neutral parent molecule using acetonitrile as solvent. Detailed e.s.r. and polarographic data for compounds in the table as well as several related compounds currently being studied will be presented in detail elsewhere. Financial support of the U. S. Army Research Office (Durham) and the U. S. Air Force Directorate of Solid State Sciences is gratefully acknowledged.
3333
The lack of observable coupling between the electronic spin in the biradicals here reported as well as in others previously described is a matter of some theoretical interest. Lye are continuing experimental studies of radicals bound together by various metal ions, including tripositive ones, in order to gather further information concerning spin couplings in many electron systems. (4) H hl McConnell, J Chon Phys , 33, 115 ( l Y b O ) , 186s (i96n).
zbid , 3 3 ,
WASHIXGTON UNIVERSITY NOBORU HIROTA DEPARTMENT OF CHEMISTRY Lours, MISSOURI S I \~-EISSMAN CORNELL UNIVERSITY DAVIDH. GESKE SAINT JOHX L. RAGLE ITHACA, A-EW YORK RECEIVED JULY 1, l N l RECEIVED J U L Y 11, 1961
BIRADICALS IN THE KETYL SERIES‘
Sir:
THE STEREOCHEMISTRY O F AN
ETHYLENEDIAMINETETRAACETATO COMPLEX OF MANGANESE(I1)’
Sir:
We have reported previously that the paraIn further elucidation of the frequently unmagnetic molecules formed on reduction of hexaof ethylenediaminemethylacetone and pentamethylacetone by sodium orthodox stereochemi~try?,~ or lithium are characterized by hyperfine splittings tetraacetato (EDTX, Xr-&) chelates, we have in which each unpaired electron spin is coupled to utilized spectrometrically measured X-ray dif the nuclei in one ketone molecule and two equiva- fraction data from the acid salt, Mn [Mn(OH:)lent alkali metal nuclei.2 Dimeric molecules in HYI2.8H20, to give the remarkable structural which the two organic radicals are joined through features noted below. The monoclinic cell contwo equivalent alkali metal ions could account for taining 2Mn[Mn(OHz)HY]2.8Hz0has a = 9.21, the observations. Such molecules are biradicals b = 16.10,c = 11.88A , , p = 90.60’; the space with each electron spin belonging predominantly group is PBl/’n. Intensity counts were taken with to one-half of the dimer but sharing the two alkali MoKa radiation for all forms. { hkl) in the range, metal nuclei with the other electron spin. The 0 < (sin f?)/A < 0.96. Results given herein are sharing refers to a time average and does not imply based upon the data for (sin e)/A < 0.67, comprissimultaneous presence of the electron spins a t a ing 4440 forms of which 90y0 are recordable above background. Objective analysis by Patterson particular site. We have carried out new experiments with and Fourier methods, with subsequent partial reketyls of the alkaline earths which suggest the for- finement by difference syntheses, yield structural mation of biradicals in which the two radicals are configurations with rather accurate bond paramjoined by a single dipositive alkaline earth ion. eters. Thermal parameters range from 1.50 for oxyReduction of xanthone or benzophenone by mag- for manganese in the anion to above 3 nesium, calcium, or barium in ethereal solvents gen in some water molecules. The discrepancy leads in each case to a single paramagnetic species index for all 4440 forms is 0.10. Manganous ions in fourfold positions of P21,!n with hyperfine couplings t o the protons in one molecule of the original ketone. The proton coupling are centered in sexadentate seven-coordinate ayuo constants are slightly dependent on the positive ion, complexes, [Mn(OHs)HT]-. ‘The geometry of varying monotonically with size of the positive ion. the inner coordination group is not that reported” N o trace of the dissociated ions is observed. The for [Fe(OHz)lT]-, but is roughly that of the configuration’, (cJ. electron transfer reactions between the dimeric sterically superior NbFi ketyls, both aliphatic and aromatic, and their ref. 4 f p diagram). Bond lengths, $0 thc nearest parent ketones are too slow (k < l o 5 h - l sec.-I) 0.005 A., ar? Mn-0, 2.210-2.260 A , , ayerage of to be measured by e.s.r. spectroscopy, while the five, 2.235 A . ; Mn-N, 2.350, 2.395 A. Apart reaction of the monomeric aromatic ketyls of the from those angles and boiids which involve the cen~ atoms, the geometry of ring systetns is very alkali metals are very rapid (k >lo8 M-’ ~ e c . - ~ )tral Use of magnesium enriched to 9370 of &g yields like that expected from earlier studies5,6of CoYfurther evidence bearing on the constitution of the and Ni(OHz)H2Y. hfanganous ions in twofold ketyls. In the case of benzophenone magnesium (7) positions of P&/n display octahedral cobrdiriaketyl the expected splitting by one &Ig nucleus tion; each such ion is bonded to four water mole(I = s/2) is observed. The interval between adja- cules ’ and two not otherwise complexed oxygen cent components is 0.3 gauss. The other ketyls (1) P a r t of a program supported b y t h e National Science Foundawith &fg exhibit a broadening of all lines, but no tion. We t h a n k also t h e U. S. Army Research Office ( D u r h a m ) a n d t h e Advanced Research Projects Agency for support of t h e work rewell-resolved ZsMg splitting, (1) T h i s work h a s been supported in p a r t by t h e U. S . Air Force u n d e r C o n t r a c t a n d i n p a r t by an E q u i p m e n t Loan C o n t r a c t with 0 . N. R. Grateful acknowledgment is also m a d e t o t h e donors of t h e Petroleum Research F u n d administered by t h e American Chemical Society for partial support of this research. (2) N. H i r o t a a n d S. I. Weissman, J. Ani. Chem. Sac., 82, 4424 (1960). (3) F. Adam and S. Weissman, ibid., 80, I R l X (1958).
ported herein. ( 2 ) Cf.J. I,. H o a r d , G. S. Smith and 11. I.ind in “Advances i n t h e Chemistry of Coordination Compounds,” C. Stanley Kirschner, T h e Macmillan ComDanv. - - New York. N. Y . .”Tulv. _ .1961. (3) J. L. Hoard, hf. Lind a n d J. V. Silverton, J . A m . Chrm. S O ~83, , 2770 (1961). (4) ‘J. L. Hoard. ibid , 61, 1252 (1939). ( 5 ) H . A. Weakliem a n d J. I. H o a r d , ibid., 81. 5 4 9 il95!?) ( 6 ) Ccxflion S . S m i t h and J I.. H o a r d . ibiJ , 81,5 5 t i (I95$))
3534
COMMUNICATIONS TO THE EDITOR
Vol. 83
atoms from glycinate rings of neighboring anions : Mn-0 = 2.140-2.175, average, 2.165 A. The acid hydrogens are used in extraordinarily strong hydrogen bonds, 2.46OaA. (with an estimated probable error of 0.015 A.), which tie the complex anions into infinite strings parallel to b. The remaining eight water molecules of the cell are involved in the general pattern of hydrogen bonding which characterizes the crystal. The over-all structural pattern of moderately strong bonding in three-dimensions leads to relatively subdued thermal motions as compared with NHaCoY.2H205 or RbFe(OH2)Y~H20.6Indeed, we anticipate that extension of the Fourier synthesis to include all experimental amplitudes out to (sin B ) / X = 0.96 will give atomic positions requiring little or no “back-shift” corrections. The machine computations of this paper were carried out a t the Cornel1 Computing Center through the courtesy of Mr. Richard C. Lesser, Director. Preparation and preliminary X-ray examination of the crystals were carried out by Dr. Gordon S. Smith.
the carbon dioxide pressure increases the rate. The reaction rate is fastest in the pH range 5-7, which suggests that unprotonated amino groups and molecular carbon dioxide are the primary reactants. When the reaction is completed, as indicated by a negative hydroxylamine assay for p-lactams, Compound IV can be isolated in SO-SO% yields as the crystalline disodium salt, m.p. 250-251’ (dec.), [ o ( I z 5 ~4-277’ (c 1, HzO) (calcd. for C ~ H I O K205SNa2: C, 33.53; H , 3.31; N, 9.21. Found: C, 35.56; H, 3.76; N, 9.02). The microcrystalline free acid, m.p. 136-138’, decarboxylates readily. Analytical and spectral data were consistent with Structure IV, which then was confirmed by conversion with metlianolic HC1 to the known l L dimethyl ester (V),m.p. 170-171’, [ C Y ] ? ~ D4-238’ ( C 1, CH30H) (Calcd. for C11Hl&206S: c, 45.82; H , 5.59; N,9.72. Found: C, 46.01;H , 5.67; h’, 9.85). Identity with an authentic sample, prepared by reaction of methyl ~-~~-4-carbomethoxy5,5-dimethyl-c~-amino-2-thiazolidineacetate hydrochlorideL2 with phosgene, was established by comparison of optical rotations, melting points, (7) Fellow of t h e John Simon Guggenheim Memorial Foundation, mixed melting point, infrared spectra, and ele1960. DEPARTMENT OF CHEMISTRY J. L. HOARD’ mental analyses. CORNELL UNIVERSITY BERITPEDERSEN The ease with which this transformation occurs STEPHANIE RICHARDS is surprising, but it is more easily understood if the ITHACA, NEW YORK J. V. SILVERTON course of the reaction is considered to involve RECEIVED JULS 6, 1961 intermediates I1 and 111. REACTION OF 6-AMINOPENICILLANIC ACID WITH CARBON DIOXIDE
Sir: 6-Aminopenicillanic acid has becri prepared by fermentation of precursor-free penicillin broth, by enzymatic cleavage of natural penicillins, 2--5 and by total synthesk6 Its availability has inatlc possible the synthesis of inany new, iniproved penicillins which are not amenable to preparation via fermentation. a-Phenoxyethylpenicillin7~sand 2,6-dimethoxyphenylpeni~illin~~~~ are clinically iiiiportant examples. G-Aminopenicillanic acid (I) has now been fouiitl to react rapidly with carbon dioxide in neutral aqueous solution a t room temperature to form 3,3dimethyl - S - oxo - 4 - thia - 1,7- diazabicyclo[3.3.0]octane-2,6-dicarboxylicacid (IV). Solutions of 0.01 to 1 Jl 6-AP9 dissolved in 2 equivalents of sodium bicarbonate and placed under 1 atmosphere of carbon dioxide will react completely in 4 to 6 hours a t room temperature. Increasing (1) F. R. Batchelor, 12. P . Doyle, J. 11. C. Naylor a n d G. N . Rolinson, Nalurc, 188, 257 (1969). (2) G. N. Rolinson, et ol., N a f u r e , 187, 236 (1960). (3) C.A. Claridge, A. Gourevitch a n d J. Lein, i b i d . , 187, 237 (19GO). (4) H. T. Huang, A. R . English, T. A . Seto, G. R I . Sbull a n d B. A. Sobin, J. A m . Chem. Soc., 82, 3790 (1960). ( 5 ) W. K a u f m a n n a n d K. Bauer, Natirvwissenschaften, 47, 474 (1960). ( 6 ) J. C. Sheehan and K . R. Henery-Logan, J . A m . Chem. SOL.,81, 5838 (1959). (7) Y.G. Perron, el d . , i b i d . , 82, 3934 (1960). (8) T h e t r a d e n a m e of Bristol Laboratories for potassium aphenoxyethylpenicillin is Syncillin. (9) E. T. Knudsen a n d G . N. Rolinson, B n l . ,Wed. J., ii, 700
(1960). (IO) T h e t r a d e name of Rristol Laboratories for sodium Z,G-(Iimethoxyphenylpenicillin is Staphcillin.
111
1V, R = II V,R=C€Ia
Tlle scquencc is malogous to that advanced for the penillic acid rearrangement of the natural penic i l l i n ~ . ’ ~Indeed, by writing the carbonyl group a t position S in the enolic form, the complete nucleus of the penillic acids is obtained. Therefore, in order to simplify the nomenclature, Compound IV has been named ti-hydroxypenillic acid. Since the reaction of 6-aminopenicillanic acid with carbon dioxide takes place so readily, i t seemed probable that it might be occurring in P. chrysogenurn fermentations in which 6-aminopenicillanic acid is an intermediate or end-product. Indeed, (11) H . T. Clarke, J. R. Johnson a n d R. Robinson, Editors, “ T h e Chemistry of Penicillin,“ Princeton University Press, Princeton, N.J., 1949, pp. 869, 891. (12) J. C. Sheehan a n d J. P. Ferris, J . A m . Chem. Soc., 81, 2912
(1939). (13) 1%.T. Clarke, J , R . Jnhnson and R. Robinson, Editors, “ T h e C h r m i s t r y o f Penicillin,” Princcton University Press, Princeton, N. J . 19l!J, p . 433.
