Composite decoupling in CPMAS of moderate-susceptibility materials

Abstract

The limiting resolution of CPMAS has been discussed extensively (I, 2), with the conclusion that one of the factors affecting the resolution of 13C CPMAS spectra is the effect of the bulk magnetic susceptibility of the sample in distorting the magnetic field. We have investigated an indirect effect of this, that of the field distortion in spreading the proton resonance and thereby reducing the efficiency of decoupling. This effect will not be significant for diamagnetic materials, which have small negative values of the susceptibility x ( m 1 X lo-* m 3 / kg); however, paramagnetic constituents of heterogeneous samples, particularly those found in soils, will have positive values of x of the order of lo-100 X 10e8 m3/kg (3). These have been observed to broaden the resonance of free water in soil preparations to 5-10 kHz (4) and to severely distort magnetic resonance images in soils (5). Recently it has been shown that magnetic sieving will improve the quality of CPMAS spectra of soils (6). While this technique removes highly paramagnetic particles, it also is likely to change the organic composition of the soil. To investigate the effect moderate values of x will have on CPMAS spectra, we placed a standard material (hexamethylbenzene, or HMB) in two different hosts: 99.9% alumina (Johnson-Matthey, Hartfordshire, England) and vermiculite (Ernst Home Centers, Seattle, Washington). Vermiculite was chosen because it has a x typical of soil minerals ( = 15 X 1 O-* m 3 / kg (3)) and is readily available. The samples were prepared by adding 5 wt% HMB to each host material and mixing thoroughly. For HMB/alumina, this was done in a ceramic mortar, while for HMB/vermiculite the vermiculite was ground in a Wiley mill and passed through a 60 mesh screen prior to mixing, because the layered structure of vermiculite prevents grinding in a mortar. To each of these samples we applied several methods of decoupling at moderate RF field strength ( yl ): CW, COMARO-2 ( 7, 8) ALPHA-3 (9), and TREV (IO). For comparison purposes, we also applied the liquid heteronuclear decoupling sequence WALTZ16 ( II ) and solid-state homonuclear decoupling sequences MREV-8 (semi-windowless) and BLEW-48 ( 12) to the sample in vermiculite. All experiments were performed on a CXP-200 NMR spectrometer using an external pulse programmer (13) interfaced to the Aspect 2000 and a Bruker MAS probe with a 7 mm Doty Scientific (Columbia, South Carolina) spinner assembly; the spinning rate was about 2600 Hz in all cases. For both 13C and ‘H, the RF field strength was set to 4 1.7 kHz (6 ps 90” pulse). The pulse lengths and phases of the proton pulses were carefully adjusted using a small bulb of hexamethyldisiloxane, and the proton offset was set to