Nitrogen- 14 Nuclear Quadrupole Resonance in Nitrogen-Silicon

(17) S. G. Bishop and P. C. Taylor, Solid State Commun., 4, 1323 (1972); K. Bergmann and K. ... D. Hadzi, Nature(London), 212, 1307 (1966). (1969). (2...
0 downloads 0 Views 333KB Size
193

14N Nuclear Quadrupole Resonance in N-SI Compounds (15) H. A. Resing, J. fhys. Chem.. 76, 1279 (1974). (16) J. K. Thompson, J. J. Krebs, and H. A. Resing, J. Chem. fhys., 43, 669 (1965). (17) S.G. Bishop and P. C. Taylor, Solid State Commun., 4, 1323 (1972); K . Bergmann and K. Demmler, Colloid folym. Sci., 252, 193 (1974); M. A. Butler and H. J. Guggenheim, Phys. Rev. B., 10, 1778 (1974); S. K. Garg, S.R. Gough, and D. W. Davidson. J. Chem. fhys., in press; H. A. Resing, Can. J. Phys., in press. (18) H. A. Resing, J. Chem. fhys., 43, 669 (1965). (19) The fitting procedure was discussed previously in ref 3. (20) C. H. Townes and B. P. Dailey, J. Chem. fhys., 17, 783 (1949). (21) T. Chiba, J. Chem. Phys., 34, 947 (1963); 41, 1352 (1964); R. Blinc and

D. Hadzi, Nature(London),212, 1307 (1966). (22) L. J. Lynch, K. H. Marsden, and E. P.George, J. Chem. fhys., 51, 5673 (1969). (23) For example, H. A. Resing, Mol. Cryst. Liq. Cryst., 9, 101 (1969). (24) D. E. D'Reilley and E. M. Peterson, J. Chem. Phys., 56, 5536 (1972). (25) H. A. Resing and R. A. Neihof, J. Colloid interface Sci., 34, 480 (1974). (26) J. Y. Wei and A. J. Maeland, J. Chem. fhys., 60, 3718 (1974). (27) H. W. Dodgen and J. L. Ragle, J. Chem. fhys., 25, 376 (1956). (28) W. R. Busing and H. A. Levy, Acta Crystaliogr,,11, 798 (1958). (29) D. H. Olson and E. Dempsey, J. Catal., 13, 221 (1969). (30) J. J. Fripiat estimates a pK. of about +5; private communication. (31) P. Salvador and J. J. Fripiat, J. fhys. Chem., 79, 1842 (1975).

Nitrogen- 14 Nuclear Quadrupole Resonance in Nitrogen-Silicon Compounds Ellory Schempp' and Mlng Chao Department of Crystallography, University of Pittsburgh, Pittsburgh, Pennsylvania 15260 (Received September 8, 1975) fublicatlon costs assisted by the National Science Foundation

Nuclear quadrupole resonance frequencies of 14N in several compounds containing nitrogen-silicon bonds have been measured. The data show quadrupole coupling constants considerably larger than would be expected on the basis of the 35Cl NQR frequencies in Si-C1 bonds. The results are compared with other amines and are interpreted in terms of a planar configuration a t the nitrogen in silyl amines. The NQR data indicate considerable ionic character in the N-Si bonds, but only a modest degree of pr-dr bonding.

Nuclear quadrupole coupling constants (NQCC) in molecules are closely correlated with molecular electronic structure and are particularly sensitive to variations in the electronegativity of the atomic species directly bonded to the atom containing the quadrupolar nucleus. In using coupling constants to extract chemical bonding information, it is useful to have data from a wide range of bonding configurations; however, nuclear quadrupole resonance (NQR) spectroscopy of I4N has been largely concerned with N-C, N-H, N-N, and N-0 b0nds.l We report here the first 14N NQR results in several compounds containing the nitrogensilicon bond. Since silicon has no natural isotopes with spin greater than one-half, there is no possibility of NQR with Si and the nature of silicon bonds must be inferred from its bonds to other species. The results of NQR investigations in Si-Cl compounds have been reviewed p r e v i o u ~ l y . ~ ~ ~ The experimental data, presented in Table I, were obtained at 77 K; the compounds are liquid a t room temperature. For nuclei with spin unity, as 14N,there are two NQR frequencies, v+ and Y-, which are related to the NQCC e2qQ and the asymmetry parameter 3 of the electric field gradient tensor by1 2 e2qQ = - ( Y + 3 3(Y+ - Y-)

+

3=

(u+

V-)

+ v-)

The experimental details have been given previou~ly;l*~ the samples were obtained commercially and purified by distillation only if the initial search failed to yield NQR lines. In the cyclic hexamethyltrisilazane (MezSiNH)3 the Y+

lines appear with a threefold splitting, consistent with a distortion of the ring which makes the three nitrogens inequivalent. Although other crystallographic possibilities are not ruled out, nearly axial distortion is indicated by the fact that the lines appear as a close doublet separated from another line about 50 kHz higher; such a configuration has been observed in other c y c l ~ s i l a z a n e s .The ~ ~ ~ lines were paired using pulsed double resonance techniques. Azidotrimethylsilane, Me3SiN3, shows similar splittings due to inequivalent sites. We attribute the NQR lines to the nitrogen in the Si-N bond of Me3SiNNN on the basis of comparison with the microwave results in HN3.7 In addition, we find that this compound can crystallize in two different forms; when the sample is cooled rapidly from room temperature to 77 K, NQR lines are found at 3148, 3083, 2928, and 2920 kHz. However, on warming just to the melting point (ca. -63OC) and recrystallizing, the lines appear a t 3136,3096,2956, and 2935 kHz. Since no unusual broadening is observed in either case, the effect is probably not due to disorder or strain.

Discussion In the case of Si-Cl bonds, the 36ClNQR frequencies are just about one-half the frequencies seen in the corresponding C-C1 compounds,3g8 cf. 16.506 MHz in MeBSiCl vs. 31.065 MHz in Me3CC1; 18.571 MHz in HZSiC12 vs. 35.991 MHz in HzCC12. However, substitution of Si for C in analogous bonds to nitrogen does not lead to comparable reductions in the nitrogen nuclear quadrupole coupling constant; e2qQ = 4855 kHz for MesNSiMes (Table I) which is only 6.5%smaller than the 5194 kHz in Me2NMe. Although the C1-Si NQR data may be interpreted in terms of a low effecThe Journal of Physical Chemistry, Voi. 80, No. 2, 1976

194

Ellory Schempp and Ming Chao

TABLE I: I4N NQR Data in Several Compounds Containing N-Si Bondsa

Compound N,N-Dimethylaminotrimethylsilane ((3%),NSi(CH,), N,N-Diethylaminotrimethylsilane (C,H,)*NSi(CH,),

--

Frequencies vi., kHz 4184.05 3098.22 4160.38 3046.47

Hexamethyltrisilazane [(CH,),Si”I3

2463.42 2418.53 2415.52 1628.43 1548.30 1537.36

e 2 q Q , kHz

17

4854.85

0.4473

4804.57

0.4636

2669.23b

0.64526

__-__

5147.80 3083.47 4026.18b 0.09531, 2927.60 2919.67 I1 3136.34 3096.41 2956.17 4041.426 0.0844b 2935.33 aThe frequencies were measured at 77 K and are accurate to 30.05 kHz. 6 Average values, CNQR lines are found in I or I1 depending on the crystallization history; see text. Trimethylsilylazidec (CH,),SiN,

I

tive electronegativity for Si and possibly dr-pa bonding, the nitrogen results require more explanation. For example, the coupling constants in the series of substituted aminesg-ll Men” (4369 kHz), MezNSiMes (4855 kHz), MenNMe (5194 kHz), MezNCl(6333 kHz) suggest an effective electronegativity for Si with respect to N which is intermediate between hydrogen and carbon. However, if this were all, then the asymmetry parameter in the silicon compound (0.45) should also be intermediate between the value in MezNH (0.32) and MezNMe (0) (7reflects the departure from trigonal symmetry, and hence the difference in occupancy of the three bonds, about the direction defined by the lone pair). However, the nitrogen silyl compounds probably have a planar sp2 configuration at the nitrogen,5,6J2-15 whereas the alkyl amines have the pyramidal sp3 form. Since the nitrogen lone pair electrons make the dominant contribution to the field gradient, a comparison of coupling constants must take into account the fact that when the lone pair electrons are in a nearly pure p orbital, their contribution to g,, is 4/3 larger than when they are in an sp3 orbital. Referred to an sp2 basis, and assuming no consequent change in the bond polarities, MezNH would have e2qQ = 5825 kHz. If N-Si u bonds are more polarized toward nitrogen than N-H bonds, as expected on the basis of a lower electronegativity for Si, substituting the trimethylsilyl group for hydrogen could reduce this coupling constant to the 4855 kHz observed in MezNSiMes. If, in addition, there is loss of charge from the nitrogen lone pair orbital to the vacant Si 3d orbitals, e2qQ for the nitrogen will be further reduced. Evidence for such pa-da interaction comes from the planarity of the heavy atom skeleton and the short N-Si bond lengths in (H3Si)3N,14 (H,$i)zNMe,12 and C13SiNMe2,13from low values of the dipole moments16-19 which suggest a two-way charge transfer, and from NMR experiments.20-22 Overall, the NQR data in MezNSiMe3 can probably be most usefully compared with the data for the amino nitrogen in 4-dimethylamin0pyridine:~:~ e2qQ = 4799 kHz, q = 0.025, vs. 4855 kHz, 7 = 0.45 in the silyl amine. Here the The Journal of Physical Chemistry, Vol. 80, No. 2, 1976

configuration is close to and owing to conjugation with the ring there is some loss of charge from the amino lone pair. Both cases, however, display very much less p a delocalization than occurs in N-methylpyrrole,l where e2qQ = 2393.1 kHz and 7 = 0.171. There are, thus, two parts to the explanation of the observed NQR results in the silylamines: the polarity of the N-Si a bond and the degree of pa-da bonding. Another question concerns the relative u bond polarity in Si-C1 bonds compared with Si-N bonds. In spite of several inadequacies, the Townes and Dailey model remains the most generally reliable scheme for the interpretation of quadrupole coupling constants in terms of bond populations. Accordingly, if the nitrogen bonds are written as sp2 hybrid orbitals and electron occupancies b and b’ are assigned to the N-C and N-Si u bonds, the NQR parameters can be expressed as1

b’ = b

+ aq

where a is the nitrogen p a occupancy and a is the reduced coupling constant, ( e 2 q Q e x p / e 2 q p Qthe ) ; coupling constant per p electron, e2qpQ,for nitrogen is taken as 8.4 MHz.’ In order to estimate b’ and a , the N-C rn population b must be known and, unfortunately, no reliable value for an sp2 configuration is available: for sp3 Me3N, b N 1.17, whereas for the nitrogen in pyridine, b N 1.29 has been found. In the first case, eq 2 give for the N-Si bond in Me2NSiMe:c b‘ N 1.43, and for the a charge a = 1.84, or a transfer of 0.16 e from the nitrogen p a to the silicon d orbitals. In the second case for b = 1.29, b‘ N 1.55, a = 1.95, suggesting a much smaller pa-da interaction. By comparison, in 4-dimethylaminopyridine for b = 1.17, we find a = 1.74, and for b = 1.29, a = 1.86 for the amino a-charge density. This contrasts with a = 1.5 in the highly conjugated N-methylpyrrole. The N-Si bonding thus appears to be characterized by strong polarization of the N-Si u bond and rather weak

I4N

Nuclear QuadrupoleResonance in N-SI Compounds

p?r-d?r bonding, weaker a t least than the PT-PT bonding in aminopyridines. This conclusion is borne out by the results in hexamethylcyclotrisilazane, where e2qQ (2.67 MHz) is only about half that in MezNSiMea, owing to the presence of three readily polarizable bonds. Using the previous model, and taking the N-Si occupancy b‘ = 1.5 e gives b” = 1.3 for the N-H u bond, close to the value in Me2NH, and a = 1.75. The T loss from the nitrogen is seen to be somewhat greater in this case where each nitrogen is bonded to two silicon atoms, and suggests a slight degree of aromatic character for the silazane ring with a silicon d occupancy of about 0.25 e. The bond population discussion above indicates an approximately 20% increase in the nitrogen u occupancy in going from N-C to N-Si. It is difficult to estimate an analogous change in u polarity in the case of C-C1 and Si-C1 bonds since the coupling constant is related to three unknown parameters,l the u charge b, the T occupancy, and the s character of the chlorine bonding orbital, a: = (2 b)(l - s) - T . If there is no dT-p?r bonding, and if s crl 0.2 for both the carbon and silicon cases, we find bc 1.29 and bsj = 1.62 in Me3CC1 and Me3SiC1, or an increase of about 25%. However, Kaplansky and Whitehead29 have pointed out that the C1 s character probably increases in the Si bonds, to perhaps 0.45; they find bc = 1.26 and bsi = 1.45, or an increase of only 15%.If T # 0, which Kaplansky and Whitehead doubt, bsi would be still smaller. The most that can be concluded, therefore, is that the u polarization toward nitrogen in N-Si bonds is comparable to the u polarization toward chlorine in Si-Cl bonds. The 14N quadrupole coupling constants in the silyl amines can thus be qualitatively accounted for on the basis of considerable N-Si u polarization in a planar sp2 configuration with a relatively small degree of pr-dr bonding.

I95

Acknowledgment. Supported by the National Science Foundation, University Science Development Grant No. GU-3184. References and Notes (1) E. Schempp and P. J. Bray in “Physical Chemistry”, Vol. IV, D. Henderson, Ed., Academic Press, New York, N.Y., 1970. (2) E. A. C. Lucken, “Nuclear Quadrupole Coupling Constants”, Academic Press, London, 1969. (3) E. A. C. Lucken in “Structure and Bonding”, Vol. 6, C. K. Jdrgensen, et ai., Ed., Springer-Verlag, New York, N.Y., 1969. (4) E. Schempp and P. J. Bray, J. Msgn. Reson., 5 , 78 (1971). (5) G. W. Adamson and J. J. Daly, J. Chem. SOC.A, 2724 (1970). (6) G. S.Smith and L. E. Alexander, Acta Crystallogr,, 16, 1015 (1963). (7) R. A. Forman and D. R. Lide, J. Chem. Phys., 39, 1133 (1963). (8) I. P. Biryukov, M. G. Voronkov, and I. A. Safin, “Tables of NQR Frequencies”, Chemistry Publishing House, Leningrad, 1968. (9) P. J. Haigh, P. C. Canepa, G. A. Matzkanin, and T. A. Scott, J. Chem. Phys., 48, 4234 (1968). (IO) C. T. O’KonskiandT. J. Flautt, J. Chem. Phys., 27, 815 (1957). (11) E. Schempp, Chem. Phys. Lett., 8, 562 (1971). (12) C. Glidewell. D. W. H. Rankin, A. G. Robiette, and G. M. Sheldrick, J. Mol. Struct., 4, 215 (1969). (13) W. Airey. C. Glidewell, A. G. Robiette, G. M. Sheldrick, and J. M. Freeman, J. Mol. Struct., 8, 423 (1971). (14) K. Hedberg, J. Am. Chem. Soc., 77, 6491 (1955). (15) M. Yokoi and K. Yamasaki, J. Am. Chem. Soc., 75, 4139 (1953); M. Yokoi, Bull. Chem. SOC.Jpn., 30, 100 (1957). (16) T. Moeller, Ed., “inorganic Synthesis”, Vol. V, McGraw-Hill, New York, N.Y., 1957, pp 58-59. (17) R. L. Cook and A. P. Mills, J. Phys. Chem., 65, 252 (1961). (18) K. Schaarschmidt, Z.Anorg. Allg. Chem., 310, 78 (1961). (19) I. Yu. Kokoreva, Ya. K. Syrkin, E. D. Babich, and V. N. Vdovin, Zh. Struct. Khim., 8, 1102 (1967). (20) E. W.Randall, C. H. Yoder, and J. J. Zuckerman, Inorg. Chem., 6, 744 (1967). (21) J. Mack and C. Yoder, Inorg. Chem., 8, 278 (1969). (22) H. Vahrenkamp and H. Noeth, J. Organomefal. Chem., 12, 281 (1968). (23) R . A. Marino, L. GuibB, and P. J. Bray, J. Chem. Phys.. 49, 5104 (1968). (24) T. C. W. Mak and J. Trotter, Acta Crystallogr., 18, 68 (1965). (25) J. Tanaka and N. Sakabe, Acta Crystallogr., Sect. B,24, 1345 (1968). (26) G. J. Bullen, D. J. Corney, and F. S . Stephens, J. Chem. Soc.. Perkin Trans. 2,642 (1972). (27) J. L. DeBoer and A. Vos, Acta Crystallogr,, Sect. E, 24, 720 (1968). (28) M. Chao, E. Schempp, and R . D. Rosentein, Acta Crystallogr.,in press. (29) M. Kaplansky and M. A. Whitehead, Mol. Phys., 16, 481 (1969).

The Journal of Physical Chemistry. Vol. 80, NO. 2, 1976