29Si and 13C magic angle spinning n.m.r. spectra of silicon carbide polymorphs

Abstract

Hexagonal and cubic silicon carbide polymorphs give distinctive 29Si and 1% magic angle spinning n.m.r. spectra that can be related to the number of nonequivalent lattice sites in the structure and variations in the non-nearest-neig h bour environment.Silicon carbide exists in many crystalline modifications based on hexagonal a-Sic (wurtzite-type ZnS), and there is also a cubic p-Sic structure (diamond or zinc blende type). The complexity arises from the numerous stacking sequences in the crystal.' A search of the Inorganic Crystal Structure Database2 shows that crystal structures of many polytypes have been determined. They differ in the number of crystallographically nonequivalent silicon and carbon atoms, although the silicon-carbon bond lengths are similar in all the forms. Hence high-resolution solid-state n .m.r. spectroscopy3 has potential for the study of such systems. We now report that 29Si and 13C magic angle spinning n.m.r. (m.a.s. n.m.r.) is highly sensitive to fine points of lattice structure.

Long-range shielding and NMR chemical shifts in silicon carbide polytypes

“…If these are assigned to the three peaks in the 2 9~i and I3c NMR spectra of 6H S i c in such a way that the chemical shift is to low field as G, increases, least-squares solution of eq. [4] yields values of 97.8 x m3/bond (7.78 x lo3 ppm A~/ bond) for AX, and 0.79 and 40.13 ppm for the intrinsic chemical shifts of silicon and carbon, respectively. The intrinsic shifts are in reasonable agreement with the 2 9~i shift in tetramethylsilane (0 ppm) and the I3c shift of the central carbon in (Me3Si),C (77 ppm relative to TMS) (15).…”

“…The sum is over all bonds n out to a certain radius. Equation [4] allows a value of AX to be calculated explicitly and compared to experimental values.…”

“…The sum is over all bonds n out to a certain radius. Equation [4] allows a value of AX to be calculated explicitly and compared to experimental values.Bonds to nearest neighbours were eliminated from the sum, both because the point dipole approximation is not valid at such short ranges (12) and because the nearest-neighbour effects are assumed to be accounted for in the intrinsic shift terms. Each crystallographically independent atom in the structure produces an equation.…”

“…There are more than 200 known polytypes, which are characterized by having identical Si-C double layers arranged in different stacking sequences (2,3). Since our initial work on the distinctive 2 9~i and I3c Magic Angle Spinning (MAS) NMR spectra of the 6H, 3C, and 15R polytypes (4,5), NMR has become an important characterization tool for silicon carbide (6). All polytypes so far studied other than 3C and 2H, several peaks in their MAS NMR spectra (4,5,(7)(8)(9) despite the fact that the immediate-neighbour tetrahedron in S i c is essentially identical for all sites (10).…”

“…Since our initial work on the distinctive 2 9~i and I3c Magic Angle Spinning (MAS) NMR spectra of the 6H, 3C, and 15R polytypes (4,5), NMR has become an important characterization tool for silicon carbide (6). All polytypes so far studied other than 3C and 2H, several peaks in their MAS NMR spectra (4,5,(7)(8)(9) despite the fact that the immediate-neighbour tetrahedron in S i c is essentially identical for all sites (10). We (4, 5) and others (7,11) showed that the number of peaks could be correlated with the number and geometrical arrangement of higher neighbours.…”