Phase identification and quantification in a devitrified glass using homo- and heteronuclear solid-state NMR.

A complex mixture resulting from the devitrification of an aluminophosphate glass has been studied for the first time using a combination of homo- and heteronuclear solid-state NMR sequences that offers the advantage of subsequent quantification.

Combined 17O NMR and 11B-31P double resonance NMR studies of sodium borophosphate glasses.

17O enriched sodium borophosphate glasses were prepared from isotopically enriched NaPO3 and H3BO3. These glasses have been studied by 17O, 11B and 31P NMR including 17O and 11B multiple quantum magic angle sample spinning (MQMAS), 11B-31P heteronuclear correlation (HETCOR) NMR and 11B{31P} rotational echo double resonance (REDOR). For comparison, the crystalline borophosphates BPO4 and Na5B2P3O13 were included in the investigations. The latter compound shows three sharp 31P resonances at -0.2, -2 and -8 ppm and two BO4 sites that can only be resolved by MQMAS. The 17O NMR spectra were recorded using both the static echo method at medium magnetic field (9.4 T) as well as MAS and MQMAS methods at high field (17.6 T). In total, five oxygen sites were identified in these borophosphate glasses: P-O-P, Na...O-P, P-O-B, B-O-B, Na...O-B. However, these five sites are not present simultaneously in any of the glasses. The 17O MQMAS spectra prove that P-O-B links play a major role in borophosphate glasses. These results are confirmed by the complementary 11B MAS spectra that show the presence of asymmetric and symmetric trigonal groups BO3a and BO3s and two tetrahedral BO4 units. 11B{31P} REDOR NMR is used to give independent information to assign the 11B lines to structural units present in the glasses. These REDOR measurements reveal that B-O-P bonds are present for each borate unit, including the BO3 groups. Particularly, a structural proposal for the two different BO4 resonances is given in terms of a different number of bonded phosphate tetrahedra. The 31P MAS spectra are usually broad and not well resolved. It is shown by 11B-31P HETCOR NMR that a possible structural assignment of a 31P signal at about -20 ppm to Q2 units as in binary sodium phosphate glasses is wrong and that the phosphate tetrahedron belonging to this resonance must be connected to borate groups.