Accurately predicting NMR chemical shifts of exchangeable protons in solution remains challenging because of the combined influence of solute–solvent interactions and molecular dynamics. We introduce a framework that integrates machine-learning molecular dynamics (ML-MD) with the ShiftML3 machine-learning shielding model for rapid and accurate prediction of NMR spectra in solvated molecules. Although originally developed for solids, ShiftML3 effectively captures intermolecular contributions to shielding in solution. We validate the method across a range of chemically diverse systems, including water in organic solvents, solvated alcohols, hydrogen-bonded nucleobases, glucose anomers, and alkylated acetamides. The ML-MD + ShiftML3 framework reproduces experimentally observed chemical shifts of exchangeable protons with near-quantitative accuracy, resolving subtle hydrogen-bonding and conformational effects that implicit-solvent DFT fails to capture. These results establish ML-MD + ShiftML3 as a transferable and computationally efficient way of incorporating solvation and dynamics into NMR spectroscopy, enabling realistic chemical shift predictions for flexible, hydrogen-bonded, and complex molecular systems. The authors develop a framework integrating machine-learning molecular dynamics with the ShiftML3 machine-learning shielding mode to accurately predict NMR chemical shifts of exchangeable protons in solution, outperforming DFT approaches in capturing solvation effects.
A machine-learning-assisted framework to improve quantum-chemical prediction of 19F NMR chemical shifts by using machine learning to diagnose and correct subset-dependent limitations in the shielding-shift relationship within a practical quantum-chemical workflow is developed.
Dongdong Chen, Yuan-Xiang Ye, Yijie Zhu et al.· Journal of Chemical Informat...· 0 citations
We evaluated a fully machine-learning-assisted workflow for NMR crystallography by combining the Universal Model for Atoms (UMA) interatomic potential for crystal structure optimization with the ShiftML3 prediction of solid-state NMR shieldings. Benchmarking against conventional periodic density functional theory (DFT) calculations for 1H, 13C, and 15N chemical shifts demonstrates that ML-based geometry optimization consistently improves the accuracy of 13C and 15N predictions relative to standard PBE optimization, highlighting the dominant role of structural refinement. ShiftML3 achieves DFT-level accuracy for shielding prediction and, when combined with UMA-optimized geometries, matches or surpasses periodic DFT for 13C and 15N while reducing the computational cost by orders of magnitude. We further show that hybrid PBE0 single-molecule corrections remain effective for both DFT- and ShiftML3-derived shieldings, extending their applicability to modern machine-learning models. These results establish a new computational paradigm for NMR crystallography by replacing both computational bottlenecks of the conventional DFT workflow with modern machine-learning models.
E.v.a. Chaloupecká, O. Socha, Martin Dračínský· Journal of Physical Chemistr...· 0 citations
Understanding how hydration reshapes the structure and conformational flexibility of biomolecular ions is essential for connecting gas-phase spectroscopy to behavior in aqueous environments. Glycine, the simplest amino acid, exhibits rich microsolvation behavior, with competing intra- and intermolecular hydrogen-bonding motifs that evolve with hydration and temperature. Although cryogenic ion spectroscopy has provided detailed measurements of hydrated protonated glycine (GlyH+) clusters, interpreting these spectra and relating them to molecular hydration motifs remains challenging. Here, we develop a data-driven many-body potential energy function for GlyH+-H2O interactions and combine it with replica-exchange molecular dynamics to identify isomeric equilibria, and with temperature-elevated path-integral coarse-graining simulations to model GlyH+(H2O)n clusters, accounting for nuclear quantum effects. This framework captures many-body interactions with high-level ab initio accuracy and enables direct computation of infrared spectra for comparison with experiment. By applying an inverse spectral reconstruction of isomeric ensembles, we quantitatively decompose the experimental spectra into contributions from competing hydration motifs and extract their relative populations. Our results characterize the sequential formation of the first and second solvation shells, quantify the competition between intramolecular and water-mediated hydrogen bonds, and reveal temperature dependence and nuclear quantum effects. Overall, this study provides a transferable approach to understanding the hydration of biomolecular systems across scales, from gas-phase clusters to bulk aqueous solutions.
Zoe A. Solomon, R. Rashmi, Ruihan Zhou et al.· Journal of Physical Chemistr...· 0 citations
Fast and accurate chemical shielding estimators are essential for shielding-driven Nuclear Magnetic Resonance (NMR) crystallography. Machine-learning models for shielding predictions have matured significantly and today are primarily limited by the electronic structure reference data they are trained on. Here, we introduce ShiftML4, a shielding-tensor model trained directly on monomer-corrected calculations that approximate PBE0, rather than the PBE reference targeted by earlier ShiftML models. ShiftML4 is trained on a diverse set of structures containing 12 of the most common NMR nuclei in molecular organic solids. On experimental benchmark sets, the 13C isotropic RMSE against experiment is 1.67 ppm, compared with 2.34 ppm for GIPAW-PBE on the same geometries. ShiftML4 gives a similar 1H prediction RMSE to ShiftML3 (0.5 ppm) and improves the 15N RMSE from 7.24 to 6.08 ppm. The model also reduces errors in the shielding-tensor anisotropy, with an RMSE of 4.63 ppm on 13C CSA principal components against 5.85 ppm for GIPAW. The improvements in prediction accuracy are retained on better geometries. Basing shift predictions on structures relaxed with PET-MOLS, a recent machine-learned interatomic potential that reaches approximate hybrid-DFT geometries in seconds, lowers the ShiftML4 errors further to 0.48 ppm (1H), 1.49 ppm (13C) and 3.66 ppm (15N).
Matthias Kellner, Ruben Rodriguez-Madrid, Jacob B. Holmes et al.· 0 citations
Through-hydrogen-bond scalar couplings are attractive NMR observables because they connect high-precision spectroscopy with local hydrogen-bond structure. It is less clear whether they can also report hydrogen-bond covalency in amorphous ice and other frozen or heterogeneous aqueous environments. Here, we combine ab initio molecular dynamics configurations of water, density functional response calculations of indirect nuclear spin–spin couplings, and absolutely localized molecular orbital (ALMO) energy decomposition analysis. Benchmark calculations against SOPPA(CCSD) water-dimer references validate BLYP/pcJ-1 for the through-hydrogen-bond 1h J O–H coupling. The coupling is dominated by the Fermi contact term and therefore follows an approximately exponential distance dependence, but ensemble and vibrational averaging prevent a transferable one-dimensional distance ruler. Extending earlier NMR/ALMO work on liquid water, 1h J O–H correlates with ALMO charge-transfer stabilization and charge-transfer amount. Thus, 1h J O–H is an experimentally accessible, covalency-sensitive fingerprint of hydrogen bonds, provided that geometry and ensemble effects are included explicitly.
Hossam Elgabarty, T. D. Kühne· Journal of Physical Chemistr...· 0 citations
We provide design principles for predicting bond exchange kinetics in acylsemicarbazide (ASC)-based systems that can be applied to tuning the properties of dynamic networks. Because of their capability of dynamic and reversible bond dissociation, ASCs are promising motifs in the design of tunable dynamic covalent networks that combine mechanical robustness with reprocessability and stability. We elucidate the factors that determine the rate of bond dissociation using density functional theory in combination with detailed kinetic studies on the mechanism of the ASC dissociation. Several ASC compounds were investigated, R1─C(═O)(H)N─N(H)─C(═O)NH─R2, with R1 methyl or phenyl, and R2 methyl, phenyl, or benzoyl, that dissociate into hydrazide and isocyanate parts. The experimentally measured dissociation rate correlates with the proton affinity of the N─H bond next to R2, which could also be used to predict relative dissociation rates a priori. Proton-transfer assistance is required for efficient bond exchange. A water molecule, but also neighboring ASCs (reactant) and hydrazides (product), lowers the activation barrier for bond dissociation considerably, likely facilitating autocatalysis that can occur in polymeric ASC networks. These findings can aid in the rational design of reversible polymers based on ASC motifs and can also be generalized for other dynamic covalent networks.
Siebe Lekanne Deprez, Stefan J D Maessen, A. V. van Dam et al.· Chemistry· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
