March 20, 1963
82 1
COMMUNICATIONS TO THE EDITOR
being carried out. Studies into the generality of halogen and pseudo-halogen migration and their migratory apptitudes in this and other systems are also presently under investigation. The authors wish to thank the Research Corporation for financial assistance through a Frederick Gardner Cottrel grant-in-aid. (11) A portion of a dissertation t o be presented by P. A. Schwab t o the Graduate School of Kansas State University in partial fulfillment of t h e requirements f o r t h e degree of Doctor of Philosophy.
DEPARTMENT OF CHEMISTRY N. MCDONALD KANSASSTATEUNIVERSITY RICHARD MANHATTAN, KANSAS PETER4. S C H W A B ~ ~ RECEIVEDDCCEMBER 7 , 1962
1086 ID26
7.73 7.59
6 Y 553
PROTON CONTACT SHIFTS IN DYNAMICAL NICKEL(I1) COMPLEXES AND THE DETECTION OF DIASTEREOISOMERS BY NUCLEAR RESONANCE
Sir: We have shown recently that bis(N-sec-alkylsalicyla1dimine)-Ni(I1) complexes (I, R = i-Pr, sec-Bu)
undergo a conformational equilibrium in solution involving planar (diamagnetic) and tetrahedral (paramagnetic) species. Susceptibility measurements indi40-F~O7~ paramagnetic form for most complexes cate examined in chloroform a t 30'. This situation suggests that, like the bis(N,N'-disubstituted-aminotroponeimine)-Ni(I1) c o m p l e x e ~ , ~TI-' - ~ and/or Te-l might be sufficiently large to render isotropic hyperfine contact shifts observable5 in the proton resonance spectra of these complexes. We are currently investigating n.m.r. spectra of complexes of the type I and have observed such contact shifts, similar in sign but not in magnitude to those detected6 in the related bis(N-arylsalicyla1dimine)-Ni(I1) complexes, which attain paramagnetism by a different means a t room temperature.? X large variety of ring-substituted N-isopropyl and -set-butyl complexes I have been examined, but discussion is confined to the X = 5-Me complexes. For R = i-Pr the following contact shiftss were observed (31°, CDCl,, -0.15 M ) : +825(3-H), -624(4-H), - 138(6-H), -597(5-Me), -6059(C-H, i-Pr)j - 11570 (HC=N) .9 The solution equilibria are demonstrably temperature and solvent dependent. In particular, the nature of the inert solvent has a marked effect on the position of equilibrium. In CS2 the equilibrium is displaced loward the diamagnetic species and first order spin-spin splitting is clearly resolved, verifying the above assignments. Under the above conditions, these shifts were observed in CSp: +314(doublet, 3-H), -234(doublet, 4-H), -55.3(singlet, 6-H), -234 (5-Me), -2224(C-H, i-Pr), -4228(HC=N). The
-
(1) R. H. Holm and K . Swaminathan, Inarc. Chem., 2, 181 (1963); similar conclusions have also been reached by others: cf. L. Sacconi, P . L. Orioli P. Paoletti and M. Ciampolini, PYOC. Chem. Soc., 255 (1962). (2) D. R. Eaton, A. D. Josey, W. D . Phillips and R. E. Benson, J. Chem. P h y s . . S T . 347 (1962). ( 3 ) D. R. Eaton, A. L). Josey, W. D. Phillips and R E Benson, Mol. P h y s . , 6, 407 (1962). (4) D. R. Eaton, A. D. Josey, R. E . Benson, W. D. Phillips and T. L. Cairns, J. A m . Chem. Sac., 84, 4100 (1962). ( 5 ) H. .M. McConnell and D. B. Chesnut, J . Cheni. Phys., 18, 107 (1958). (6) E. A. LaLancette, D . R. Eaton, R. E. Benson a n d W. D. Phillips, J . A m . Chem. Sac., 84, 3 9 6 8 (1962). (7) R. H. Holm and K. Swaminathan, Inors. Chem., 1, 5 9 9 (1962). (8) Contact shifts, AHi, are proportional t o g o i [ T e A F / R T 31-1 and are given as vcomPiex - viigsnd in C.P.S.a t 60 M c . Both complex and ligand were referenced internally t o tetramethylsilane. (9) Assignment verified b y deuteration,
+
OPTICALLY ACTIVE
Fig. 1.-Proton resonance spectra of bis( 5-methyl-S-secbutylsalicyla1dimine)-Ni(I1) complexes in CSz a t 31" and 60 Mc. Chemical shifts are in p,p,rn; assignments are as shown; ( a ) optically inactive mixture of complexes: (+, +) and ( - , -), (+, - ) ; additional pairs of signals not shown are found at 35.3 and 41.8 p.p.m. ( C - H , sec-Bu) and at 67.5 and 78.7 p.p.rn. (HC=N); ( b ) optically active complex: ( + , +) or ( - , - 1; additional signals not shown are a t 41.8 and 78.7 p . p m
signs of the contact shifts are in accord with a valence bond model for a non-alternant system inasmuch as structures can be written6 for the paramagnetic form placing an unpaired spin a t carbons 3 and 5 and nitrogen (positive spin densities), Remaining position acquire negative spin densities through electron correlation effects.2 The spectra of R = sec-Bu complexes are very similar to those of the corresponding i-Pr complexes in signs and magnitudes of contact shifts, but one striking difference is evident upon inspection of Figure l a , viz., that all signals related to a given proton occur in pairs. This effect is especially evident from the doubled signal of the 5-Me group. These doubled features imply two distinct averages of chemical shift over the planar and tetrahedral forms. The independence of the two sets of resonances was first established by proton double resonancelo whereby it was possible (in CS2) to decouple the outer set of doublets a t 1.32 and 10.86 p.p.m. without disturbing the inner set at 2.03 and 10.26 p.p.m. The magnitude of contact shift differences for a given proton and for the 5-Me group strongly imply that these differences arise from two distinct paramagnetic tetrahedral configurations over which the chemical shifts are separately averaged, l 1 Because contact shifts in equilibrium systems of this type can be a t least in part dependent upon a free energy difference between the two conformersj8the results indicate two A F values for the planar t tetrahedral equilibrium correlated with two tetrahedral conformations of slightly different (IO) We thank Drs. J. D . Baldeschwieler and K. Kuhlmann of this department for these measurements and for helpful discussion. (11) Evidence t h a t conformation effects of t h e scc-butyl group in diamagnetic planar and tetrahedral configurations produce negligible chemical shift difference is afforded by t h e very similar spectra of t h e inactive ( v i d e s u p m ) bis(5-methyl-N-sec-butylsalicylaldimine) Pd(I1) and Zn(I1) complexes which show no evidence of doubling.
COMMUNICATIONS TO THE EDITOR
822
VOl. 85
free energies.12 These two conformations are due to the a n e s 4 He was prevented from making any meaningful presence of diastereosiomers. The complex (m.p. estimate of silicon involvement in electron delocaliza153-154.5’) prepared from optically inactive sec-butyltion by the nature of arylsilane ions studied. amine contains equimolar amounts of a racemic [(+, +) We have found that ions produced by the potassium and (-,-)I and a meso [(+,-)I form. The two reduction of bis(2,2’-biphenylene) silane, a compound optically active complexes (m.p. 192-193’) prepared first prepared by Gilman and Gorsichj6 possess magfrom (+)-amine ([a]% 7.53’) or (-)-amine-( [cy]: netic and chemical properties which afford direct evi-7.50’)13 are accordingly (+,+) or ( - j - ) and give dence of silicon participation in electron delocalization. identical spectra corresponding to the outer set of When a dilute solution of bis(2,2’-biphenylene) silane lines (Fig. lb). These active complexes therefore in 1-methyltetrahydrofuran or in 1,2-dimethoxyethane possess a slightly more stable tetrahedral form than is treated briefly at -60’ with pure potassium metal, does the meso complex, which produces the smaller a wine red solution is obtained. This solution appears set of contact shifts. The intrinsically dissimilar to be reasonably stable in the absence of excess potastetrahedral forms are possibly further differentiated by sium metal. At -60” the wine red solution exhibits a lack of free rotation of the sec-butyl groups about the an intense paramagnetic resonance absorption centered C-N bond. Two distinct sets of interactions between a t approximately the position for a free electron. At these groups are possible through the agencies of direct this temperature the absorption is resolved into nine steric interference or a correlation type solvent effect. lines. At higher temperatures, near room temperature, Evidence that solvent does not interact equivalently a five line spectrum is obtained. From the nature of with the two diastereoisomers is given by the inequality the hyperfine structure we deduce that this uninegative of the following contact shift ratios for the (+,+) and ion radical probably involves complete delocalization (+, -) complexes, respectively: ( A H i ) c ~ c ~ ~ I ( A H i ) c s 2 of the electron spin over both biphenyl ring systems. = 2.37 f 0.02, 2.52 f 0.04. In the case of the uninegative ion radical of the carbon Facile ligand exchange is uniquely demonstrated by analog, bis (2,2’-biphenylene) methane, the hyperfine use of the active complexes. Equimolar solutions of structure has been completely resolved and the results the (+,+) and (-,-) species in CS2 or CDCl, when would support this conclusion.6 mixed develop within 7 min. the spectrum of Fig. la. Of particular interest is the narrowness of the total Integration shows the ratio of active and inactive splitting in the uninegative bis(2,2’-biphenylene) silane products is the same from both direct reaction and ion radical, the line breadth being 11.5 gauss compared ligand exchange and is 1.10 f 0.03 in CS2 solution. with a line breadth of 24.2 gauss observed for the biAlthough attempts to detect geometrical isomers of phen ylyl. metal complexes by nuclear resonance have recently A comparison of these two line breadths indicates been successful, these results represent the first dethat approximately 50% of the spin resides in the two 5 tection by such means of diastereosiomeric complexes or, membered hetero rings. For the analogous uninegato our knowledge, of any diastereoisomeric pair. Full tive bis(2,2’-biphenylene) methane the line breadth is details of resonance studies of type I complexes will be 21 gauss. Thus silicon in the spiro position has a very reported subsequently. large effect in increasing spin density in the five membered rings. Acknowledgment.-Financial support by the NaWhen the wine-red solution of the uninegative bistional Institutes of Health is gratefully acknowledged. (2,2’-biphenylene) silane ion radical is treated with exThe authors are indebted to Dr. W. D. Phillips for cess potassium metal for sweral minutes, even at stimulating discussion. temperatures as low as -79”, a royal purple solution (12) The possibility t h a t the two tetrahedral forms could also differ in of the dinegative ion is obtained. This species is diatheir intrinsic g-values and thereby contribute partially t o t h e distinguishamagnetic in contrast with the dinegative ion of bis(2,2’hility of the two sets of resonances cannot be excluded a t present. biphenylene) methane which is a triplet in its ground (13) T h e amines were resolved with (+) and (-)-tartaric acids according t o Thome14 and Bruck, e l al.16 The latter obtained [ a l z 2 D +7.48’. -7.04’. state.6 (14) L . G. Thome, B e r . , 86, 582 (1903). The dimagnetism of the dinegative bis(2,2’-biphenyl(1.5) P. Bruck, I. N . Denton and A. H. Lamberton, J. C h m . Soc., 921 ene) silane ion can be explained by assuniing a 26 (1956). electron closed shell aromatic x structure. Such a (16) R . C. F a y and T. S . Piper, J . A m . Chem. Soc., 84, 2303 (1962). DEPARTMEST OF CHEMISTRY R. H. HOLM structure would involve conjugation of the two biphenyl A . CHAKRAVORTY ring systems through the silicon atom. Such conjugaHARVARD UNIVERSITY CAMBRIDGE 38, MASS. G . 0. DUDEX tion does not require coplanarity of the molecule since RECEIVED JANUARY 26, 1963 in the D,d symmetry of the parent compound the x functions transform as A?, B1, and E and the silicon d orbitals transform as AI, B1, B,? and E. Thus the p x ELECTRON DELOCALIZATION INVOLVING SILICON I N d x B1 and E m.o.’safford conjugation for this symmetry. IONS DERIVED FROM BIS(2,2‘ BIPHENYLENE) SILANE’ The presence of the low lying d orbitals of the silicon Sir: would account for the observed difference in ground The participation of the 8d orbitals of silicon to effect states for the dinegative bis(2,2’-biphenylene) silane enhanced bonding in various covalently bonded silicon and its carbon analog. compounds has long been a subject of speculation. We are continuing our study of these and closely Participation of this kind has been used t o explain such related ions and ion radicals. diverse anomalies as the trigonal planar structure of triDEPARTMEST OF CHEMISTRY RICHARD D . COWELL PURDUE UNIVERSITY GRANT URRY silylaniine? and the unexpectedly low +I effect of a LAFAYETTE, INDIANA trimethylsilyl substituent in aromatic molecules. S. I. WEISSMAN OF CHEMISTRY DEPARTMENT Recently, Townsend has reported electron spin resWASHINGTON UTIVERSITY onance results for ions derived from various arylsilST. LOUIS5 , MISSOURI
+
(1) The research herein reported was supported by funds from t h e Air Force Office of Scientific Research, the Kational Science Foundation and t h e Petroleum Research Fund. t o whom xrateful acknowledgment is made. (2) A. G. MacDiarmid and A . C. Maddock, J. Inorg. A’rrcl. Chem., 1, 411 (195.5 1. (3) J. C h a t t and A. A. Williams, J. Chem. Soc., 4403 (1964).
RECEIVED JANUARY 14, 1963 (4) M. G. Townsend, i b i d . , 5 1 (1962).
( 5 ) H. Gilman a n d R . Gorsich, J . A m . Chem. Sor., 80, 1884 (19.58). (6) K. D. Cowell, G. Urry and S . I. Weissman, J. Chem. P h y s . , in press. (7) E. deBoer, Ph.D. Thesis, Free University, Amsterdam, 1957.
