An Accurate Tensorial Model for Prediction of Full Zeolite NMR Spectra

Abstract

Solid state nuclear magnetic resonance (ss-NMR) is one of the most sensitive and popular techniques for structure elucidation in geometrically complex crystalline materials, such as zeolites. Synergistic support from computational modelling is vital to interpret experimental spectra, and relate ss-NMR to atomistic models. Nevertheless, computational predictions are hindered by the high expense of calculating magnetic shielding (MS) and electric field gradient (EFG) tensors from first principles. In this work, we leverage a novel tensorial machine learning approach to train a general model for predicting complete NMR tensors. We demonstrate the utility of the approach for a diverse dataset of zeolitic materials and NMR-active nuclei ($^{27}$Al, $^{29}$Si, $^{17}$O, $^{23}$Na and $^{1}$H), predicting all NMR observables to a high degree of precision. These observables are then translated into predictions of the full $^{27}$Al and $^{29}$Si ss-nMR spectra for the exemplary zeolite RTH. Thus, this work opens a pathway to accurate, high-throughput NMR simulation for large-scale and realistic models of chemically complex zeolites.

Figures (24)

• Figure 1: Parity plots comparing DFT-calculated and ML-predicted irreducible spherical tensor components ($\sigma^{(0)}$, $\sigma^{(1)}$, and $\sigma^{(2)}$) of the magnetic shielding tensor for $^{1}$H, $^{17}$O, $^{23}$Na, $^{27}$Al, and $^{29}$Si on the test set. Marginal histograms show the distribution of DFT (gray) and ML (colored) values. All values are in ppm.
• Figure 2: Parity plots comparing DFT-calculated and ML-predicted $V^{(2)}$ components of the electric field gradient tensor for $^{2}$H, $^{17}$O, $^{23}$Na, $^{27}$Al, and $^{29}$Si on the test set. The EFG tensor is symmetric and traceless and therefore contains only $V^{(2)}$ components. All values are in atomic units.
• Figure 3: Parity plots comparing DFT-calculated and ML-predicted magnetic shielding tensor observables for $^{1}$H, $^{17}$O, $^{23}$Na, $^{27}$Al, and $^{29}$Si on the test set. The isotropic chemical shift ($\delta_\mathrm{iso}$), span ($\Omega$), and skew ($\kappa$) are shown. All values are in ppm.
• Figure 4: Parity plots comparing DFT-calculated and ML-predicted EFG tensor observables for $^{2}$H, $^{17}$O, $^{23}$Na, and $^{27}$Al on the test set. The absolute value of the quadrupolar coupling constant ($|C_Q|$, in MHz) and the asymmetry parameter ($\eta_Q$, dimensionless) are shown. $^{29}$Si is omitted as it is a spin-$1/2$ nucleus.
• Figure 5: Simulated $^{29}$Si NMR spectrum of a) pure-silica RTH and b) Al-containing RTH zeolites, $^{27}$Al NMR spectra of Al-containing RTH zeolite in c) hydrated and d) dehydrated form.