High resolution 17O NMR spectra of siliceous ferrierite (Sil-FER) have been collected and the 29Si and 17O isotropic chemical shifts and the electric field gradients of oxygen have been calculated from first principles. The theoretical 29Si MAS NMR spectrum is found to be in excellent quantitative agreement with the experimentally determined spectrum, and is extremely sensitive to the accuracy of the structure used for the calculations, thus providing a method for assessing the quality of the structure determination. However, theoretical predictions of the chemical shifts, quadrupolar coupling constants and asymmetry parameters show only qualitative agreement with the experimental 17O NMR spectra obtained by Double Rotation (DOR) and multiple quantum magic angle spinning (MQMAS) as the spectra are much more complex (10 peaks within a shift range of less than 15 ppm, and the quadrupolar coupling constants only differ by 0.4 MHz) and hence higher accuracy is required from the shift calculations (>0.5 ppm), which is currently not possible. These findings also demonstrate the current limitations of the experimental techniques and show that no simple correlation appears to exist between the zeolite structure, such as the Si−O−Si bond angles or lengths, and the 17O NMR parameters.

state NMR have already played a vital role in the characterization of technologically important zeolites, 1 but relatively few 17 O NMR studies have been performed. 2 17 O is only 0.037% in natural abundance, it is expensive to isotopically enrich, it has a small magnetic moment, and the electric field gradient is relatively large so the static and magic-angle spinning (MAS) line shapes are often broad and featureless. Nevertheless, using high magnetic field strengths (11.7-16.9 T) and state-ofthe-art double rotation probes, we have obtained high-resolution 17 O NMR spectra in which, for the first time, all crystallographically inequivalent oxygen sites in a zeolite have been resolved. In addition, we present an assignment of the spectrum based upon quantum mechanical chemical shift and field gradient calculations. 17 O has a nuclear spin of I ) 5 / 2 so the observed NMR line shape of the central transition and its relative peak position are determined by the quadrupole coupling constant, C q , the asymmetry parameter of the electric field gradient tensor, η, and the quadrupolar and chemical shift tensors, all of which are dependent upon the coordination of the oxygen and the type of bonding in which it is participating. 3 The potential of 17 O NMR to study silicates 4 and zeolites 2 has been demonstrated using MAS and variable angle spinning techniques. Chemically distinct environments such as Si-O-Si and Si-O-Al were distinguished in several zeolites, but individual sites could not be resolved. Highresolution 17 O NMR spectra of condensed silicates 5,6 have been obtained by techniques such as double rotation (DOR) 3 and dynamic angle spinning (DAS). 3 Assignments of 17 O spectra have been made by using correlations between C q and η and the Si-O-Si bond angle. 4,6 These semiempirical correlations, based on either quantum mechanical calculations 7 or the Townes-Dailey model, 8 have predicted relationships where C q is expected to increase, and η to decrease, with Si-O-Si bond angle. However, with the limited amount of data in the literature currently it is uncertain whether these correlations are quantitatively reliable. There have also been numerous attempts to calculate field gradients in various materials. 9 Ab initio density functional theory calculations have been performed on the mineral forsterite 10 where electric field gradients within about 5% of the absolute experimental values were obtained for 25 Mg and 17 O.Sil-Y was chosen for investigation by 17 O NMR for several reasons: (i) it adopts the same topology (FAU) as zeolites X and Y, which are used widely as acid catalysts and separating media, (ii) the structure has been determined from neutron diffraction with high accuracy, 11 (iii) the framework does not contain aluminum or charge compensating cations that could cause chemical shift dispersion, and (iv) due to its structural simplicity (one crystallographically distinct Si site and four crystallographically distinct oxygen sites), it is an excellent sample for calculating the 17 O NMR chemical...

No abstract