Ecological and methodological drivers of non-stationarity in tree growth response to climate.

Radial tree growth is sensitive to environmental conditions, making observed growth increments an important indicator of climate change effects on forest growth. However, unprecedented climate variability could lead to non-stationarity, that is, a decoupling of tree growth responses from climate over time, potentially inducing biases in climate reconstructions and forest growth projections. Little is known about whether and to what extent environmental conditions, species, and model type and resolution affect the occurrence and magnitude of non-stationarity. To systematically assess potential drivers of non-stationarity, we compiled tree-ring width chronologies of two conifer species, Picea abies and Pinus sylvestris, distributed across cold, dry, and mixed climates. We analyzed 147 sites across the Europe including the distribution margins of these species as well as moderate sites. We calibrated four numerical models (linear vs. non-linear, daily vs. monthly resolution) to simulate growth chronologies based on temperature and soil moisture data. Climate-growth models were tested in independent verification periods to quantify their non-stationarity, which was assessed based on bootstrapped transfer function stability tests. The degree of non-stationarity varied between species, site climatic conditions, and models. Chronologies of P. sylvestris showed stronger non-stationarity compared with Picea abies stands with a high degree of stationarity. Sites with mixed climatic signals were most affected by non-stationarity compared with sites sampled at cold and dry species distribution margins. Moreover, linear models with daily resolution exhibited greater non-stationarity compared with monthly-resolved non-linear models. We conclude that non-stationarity in climate-growth responses is a multifactorial phenomenon driven by the interaction of site climatic conditions, tree species, and methodological features of the modeling approach. Given the existence of multiple drivers and the frequent occurrence of non-stationarity, we recommend that temporal non-stationarity rather than stationarity should be considered as the baseline model of climate-growth response for temperate forests.

Bayesian Learning of Adatom Interactions from Atomically Resolved Imaging Data.

Atomic structures and adatom geometries of surfaces encode information about the thermodynamics and kinetics of the processes that lead to their formation, and which can be captured by a generative physical model. Here we develop a workflow based on a machine-learning-based analysis of scanning tunneling microscopy images to reconstruct the atomic and adatom positions, and a Bayesian optimization procedure to minimize statistical distance between the chosen physical models and experimental observations. We optimize the parameters of a 2- and 3-parameter Ising model describing surface ordering and use the derived generative model to make predictions across the parameter space. For concentration dependence, we compare the predicted morphologies at different adatom concentrations with the dissimilar regions on the sample surfaces that serendipitously had different adatom concentrations. The proposed workflow can be used to reconstruct the thermodynamic models and associated uncertainties from the experimental observations of materials microstructures. The code used in the manuscript is available at https://github.com/saimani5/Adatom_interactions.

Learning from Imperfections: Predicting Structure and Thermodynamics from Atomic Imaging of Fluctuations.

In materials characterization, traditionally a single experimental sample is used to derive information about a single point in the composition space, while the imperfections, impurities, and stochastic details of material structure are deemed irrelevant or complicating factors in the analysis. Here we demonstrate that atomic-scale studies of a single nominal composition can provide information about microstructures and thermodynamic response over a finite area of chemical space. Using the principles of statistical inference, we develop a framework for incorporating structural fluctuations into statistical mechanical models and use it to solve the inverse problem of deriving effective interatomic interactions responsible for elemental segregation in a La5/8Ca3/8MnO3 thin film. The results are further analyzed by a variational autoencoder to detect anomalous behavior in the composition phase diagram. This study provides a framework for creating generative models from a combination of multiple experimental data and provides direct insight into the driving forces for cation segregation in manganites.

Materials Science in the AI age: high-throughput library generation, machine learning and a pathway from correlations to the underpinning physics.

The use of advanced data analytics and applications of statistical and machine learning approaches ('AI') to materials science is experiencing explosive growth recently. In this prospective, we review recent work focusing on generation and application of libraries from both experiment and theoretical tools, across length scales. The available library data both enables classical correlative machine learning, and also opens the pathway for exploration of underlying causative physical behaviors. We highlight the key advances facilitated by this approach, and illustrate how modeling, macroscopic experiments and atomic-scale imaging can be combined to dramatically accelerate understanding and development of new material systems via a statistical physics framework. These developments point towards a data driven future wherein knowledge can be aggregated and used collectively, accelerating the advancement of materials science.

Decoding Oxyanion Aqueous Solvation Structure: A Potassium Nitrate Example at Saturation.

The ability to probe the structure of a salt solution at the atomic scale is fundamentally important for our understanding of many chemical reactions and their mechanisms. The capability of neutron diffraction to "see" hydrogen (or deuterium) and other light isotopes is exceptional for resolving the structural complexity around the dissolved solutes in aqueous electrolytes. We have made measurements using oxygen isotopes on aqueous nitrate to reveal a small hydrogen-bonded water coordination number (3.9 ± 1.2) around a nitrate oxyanion. This is compared to estimates made using the existing method of nitrogen isotope substitution and those of computational simulations (>5-6 water molecules). The low water coordination number, combined with a comparison to classical molecular dynamics simulations, suggests that ion-pair formation is significant. This insight demonstrates the utility of experimental diffraction data for benchmarking atomistic computer simulations, enabling the development of more accurate intermolecular potentials.

Combining configurational energies and forces for molecular force field optimization.

While quantum chemical simulations have been increasingly used as an invaluable source of information for atomistic model development, the high computational expenses typically associated with these techniques often limit thorough sampling of the systems of interest. It is therefore of great practical importance to use all available information as efficiently as possible, and in a way that allows for consistent addition of constraints that may be provided by macroscopic experiments. Here we propose a simple approach that combines information from configurational energies and forces generated in a molecular dynamics simulation to increase the effective number of samples. Subsequently, this information is used to optimize a molecular force field by minimizing the statistical distance similarity metric. We illustrate the methodology on an example of a trajectory of configurations generated in equilibrium molecular dynamics simulations of argon and water and compare the results with those based on the force matching method.

Multimodality of Structural, Electrical, and Gravimetric Responses of Intercalated MXenes to Water.

Understanding of structural, electrical, and gravimetric peculiarities of water vapor interaction with ion-intercalated MXenes led to design of a multimodal humidity sensor. Neutron scattering coupled to molecular dynamics and ab initio calculations showed that a small amount of hydration results in a significant increase in the spacing between MXene layers in the presence of K and Mg intercalants between the layers. Films of K- and Mg-intercalated MXenes exhibited relative humidity (RH) detection thresholds of ∼0.8% RH and showed monotonic RH response in the 0-85% RH range. We found that MXene gravimetric response to water is 10 times faster than their electrical response, suggesting that H2O-induced swelling/contraction of channels between MXene sheets results in trapping of H2O molecules that act as charge-depleting dopants. The results demonstrate the use of MXenes as humidity sensors and infer potential impact of water on structural and electrical performance of MXene-based devices.

Knowledge Extraction from Atomically Resolved Images.

Tremendous strides in experimental capabilities of scanning transmission electron microscopy and scanning tunneling microscopy (STM) over the past 30 years made atomically resolved imaging routine. However, consistent integration and use of atomically resolved data with generative models is unavailable, so information on local thermodynamics and other microscopic driving forces encoded in the observed atomic configurations remains hidden. Here, we present a framework based on statistical distance minimization to consistently utilize the information available from atomic configurations obtained from an atomically resolved image and extract meaningful physical interaction parameters. We illustrate the applicability of the framework on an STM image of a FeSexTe1-x superconductor, with the segregation of the chalcogen atoms investigated using a nonideal interacting solid solution model. This universal method makes full use of the microscopic degrees of freedom sampled in an atomically resolved image and can be extended via Bayesian inference toward unbiased model selection with uncertainty quantification.

Consistent Integration of Experimental and Ab Initio Data into Effective Physical Models.

We describe and test theoretical principles for consistent integration of experimental and ab initio data from diverse sources into a single statistical mechanical model. The approach is based on the recently introduced concept of statistical distance between partition functions, uses a simple vector algebra formalism to describe measurement outcomes and coarse-graining operations, and takes advantage of thermodynamic perturbation expressions for fast exploration of the model parameter space. The methodology is demonstrated on a combination of thermodynamic, structural, spectroscopic, and imaging pseudoexperimental data along with ab initio-type trajectories, which are incorporated into models describing the behavior of a near-critical fluid, liquid water, thin-film mixed oxides, and binary alloys. We evaluate how different target data constrain the model parameters and how the uncertainty associated with incomplete target information and limited sampling of the system's phase space might influence the choice of optimal parameters.

Precise determination of water exchanges on a mineral surface.

Solvent exchanges on solid surfaces and dissolved ions are a fundamental property important for understanding chemical reactions, but the rates of fast exchanges are poorly constrained. We probed the diffusional motions of water adsorbed onto nanoparticles of the mineral barite (BaSO4) using quasi-elastic neutron scattering (QENS) and classical molecular dynamics (MD) to reveal the complex dynamics of water exchange along mineral surfaces. QENS data as a function of temperature and momentum transfer (Q) were fit using scattering functions derived from MD trajectories. The simulations reproduce the dynamics measured in the experiments at ambient temperatures, but as temperature is lowered the simulations overestimate slower motions. Decomposition of the MD-computed QENS intensity into contributions from adsorbed and unbound water shows that the majority of the signal arises from adsorbed species, although the dynamics of unbound water cannot be dismissed. The mean residence times of water on each of the four surface sites present on the barite {001} were calculated using MD: at room temperature the low barium site is 194 ps, whereas the high barium site contains two distributions of motions at 84 and 2.5 ps. These contrast to 13 ps residence time on both sulfate sites, with an additional surface diffusion exchange of 66 ps. Surface exchanges are similar to those of the aqueous ions calculated using the same force field: Baaq2+ is 208 ps and SO4aq2- is 5.8 ps. This work demonstrates how MD can be a reliable method to deconvolute solvent exchange reactions when quantitatively validated by QENS measurements.

Rigorous force field optimization principles based on statistical distance minimization.

We use the concept of statistical distance to define a measure of distinguishability between a pair of statistical mechanical systems, i.e., a model and its target, and show that its minimization leads to general convergence of the model's static measurable properties to those of the target. We exploit this feature to define a rigorous basis for the development of accurate and robust effective molecular force fields that are inherently compatible with coarse-grained experimental data. The new model optimization principles and their efficient implementation are illustrated through selected examples, whose outcome demonstrates the higher robustness and predictive accuracy of the approach compared to other currently used methods, such as force matching and relative entropy minimization. We also discuss relations between the newly developed principles and established thermodynamic concepts, which include the Gibbs-Bogoliubov inequality and the thermodynamic length.

Vapor-liquid equilibrium and polarization behavior of the GCP water model: Gaussian charge-on-spring versus dipole self-consistent field approaches to induced polarization.

We developed the Gaussian charge-on-spring (GCOS) version of the original self-consistent field implementation of the Gaussian Charge Polarizable water model and test its accuracy to represent the polarization behavior of the original model involving smeared charges and induced dipole moments. For that purpose we adapted the recently proposed multiple-particle-move (MPM) within the Gibbs and isochoric-isothermal ensembles Monte Carlo methods for the efficient simulation of polarizable fluids. We assessed the accuracy of the GCOS representation by a direct comparison of the resulting vapor-liquid phase envelope, microstructure, and relevant microscopic descriptors of water polarization along the orthobaric curve against the corresponding quantities from the actual GCP water model.

Thermodynamics of small alkali metal halide cluster ions: comparison of classical molecular simulations with experiment and quantum chemistry.

We evaluate the ability of selected classical molecular models to describe the thermodynamic and structural aspects of gas-phase hydration of alkali metal halide ions and the formation of small water clusters. To understand the effect of many-body interactions (polarization) and charge penetration effects on the accuracy of a force field, we perform Monte Carlo simulations with three rigid water models using different functional forms to account for these effects: (i) point charge nonpolarizable SPC/E, (ii) Drude point charge polarizable SWM4-DP, and (iii) Drude Gaussian charge polarizable BK3. Model predictions are compared with experimental Gibbs free energies and enthalpies of ion hydration, and with microscopic structural properties obtained from quantum DFT calculations. We find that all three models provide comparable predictions for pure water clusters and cation hydration but differ significantly in their description of anion hydration. None of the investigated classical force fields can consistently and quantitatively reproduce the experimental gas-phase hydration thermodynamics. The outcome of this study highlights the relation between the functional form that describes the effective intermolecular interactions and the accuracy of the resulting ion hydration properties.

NO3(-) coordination in aqueous solutions by (15)N/(14)N and (18)O/(nat)O isotopic substitution: what can we learn from molecular simulation?

We explore the deconvolution of water-nitrate correlations by the first-order difference approach involving neutron diffraction of heavy- and null-aqueous solutions of KNO3 under (14)N/(15)N and (nat)O(N)/(18)O(N) substitutions to achieve a full characterization of the first water coordination around the nitrate ion. For that purpose we performed isobaric-isothermal simulations of 3.5 m KNO3 aqueous solutions at ambient conditions to generate the relevant radial distribution functions required in the analysis (a) to identify the individual partial contributions to the total neutron-weighted distribution function, (b) to isolate and assess the contribution of NO3(-)···K(+) pair formation, (c) to test the accuracy of the neutron diffraction with isotope substitution based coordination calculations and X-ray diffraction based assumptions, and (d) to describe the water coordination around both the nitrogen and oxygen sites of the nitrate ion.

Ewald summation approach to potential models of aqueous electrolytes involving gaussian charges and induced dipoles: formal and simulation results.

We present a detailed derivation of the complete set of expressions required for the implementation of an Ewald summation approach to handle the long-range electrostatic interactions of polar and ionic model systems involving Gaussian charges and induced dipole moments with a particular application to the isobaric-isothermal molecular dynamics (NPT-MD) simulation of our Gaussian charge polarizable (GCP) water model and its extension to aqueous electrolyte solutions. The set is comprised of the individual components of the potential energy, electrostatic potential, electrostatic field and gradient, electrostatic force, and corresponding virial. Moreover, we show how the derived expressions converge to known point-based electrostatic counterparts when the parameters, defining the Gaussian charge and induced dipole distributions, are extrapolated to their limiting point values. Finally, we test the simulation outcomes from the Ewald implementation against the corresponding reaction-field (RF) approach at three contrasting hydrogen-bonded water environments, including thermodynamic quantities, polarization behavior, and microstructural properties, where the simulated microstructures are compared with the available neutron scattering and X-ray diffraction data.

Correspondence between cluster-ion and bulk solution thermodynamic properties: on the validity of the cluster-pair-based approximation.

Since the single-ion thermodynamic properties of bulk solutions are not directly accessible from experiments, extrapolations have been devised to estimate them from experimental measurements on small-clusters. Extrapolations based on the cluster-pair-based approximation (CPA) technique (Tissandier et al. J. Phys. Chem. A 1998, 102, 7787-7794) and its variants are currently considered one of the most reliable source of single-ion hydration thermodynamic data and have been used as a benchmark for the development of molecular and continuum solvation models. Despite its importance, the CPA has not been thoroughly tested and recent studies have indicated inconsistencies with molecular simulations. The present work challenges the key CPA assumptions that the hydration properties of single cations and anions in growing clusters rapidly converge to each other following a monotonous trend. Using a combination of simulation techniques to study the transition between alkali halide ions in small clusters and bulk solution, we show that this convergence is rather slow and involves a surprising change in trends, which can result in significant errors in the original estimated single-ion properties. When these cluster-size-dependent effects are taken into account, the inconsistencies between molecular models and experimental predictions disappear, and the value of the proton hydration enthalpy based on the CPA aligns with estimates based on other principles.

Dynamic and structural properties of room-temperature ionic liquids near silica and carbon surfaces.

The dynamic and structural properties of a room-temperature ionic liquid (RTIL) 1-butyl-3-methyl-imidazolium(trifluoromethanesulfonimide) ([C4mim][Tf2N]) confined in silica and carbon mesopores were investigated by molecular dynamics (MD) simulations and nuclear magnetic resonance (NMR) experiments. The complex interfacial microstructures of confined [C4mim][Tf2N] are attributed to the distinctive surface features of the silica mesopore. The temperature-dependent diffusion coefficients of [C4mim][Tf2N] confined in the silica or carbon mesopore exhibit divergent behavior. The loading fraction (f = 1.0, 0.5, and 0.25) has a large effect on the magnitude of the diffusion coefficient in the silica pore and displays weaker temperature dependence as the loading fraction decreases. The diffusion coefficients of mesoporous carbon-confined [C4mim][Tf2N] are relatively insensitive to the loading faction and exhibit a temperature dependence that is similar to the bulk dependence at all loading levels. Such phenomena can be attributed to the unique surface heterogeneity, dissimilar interfacial microstructures, and interaction potential profile of RTILs near silica and carbon walls.

Structure and stability of SnO2 nanocrystals and surface-bound water species.

The structure of SnO2 nanoparticles (avg. 5 nm) with a few layers of water on the surface has been elucidated by atomic pair distribution function (PDF) methods using in situ neutron total scattering data and molecular dynamics (MD) simulations. Analysis of PDF, neutron prompt gamma, and thermogravimetric data, coupled with MD-generated surface D2O/OD configurations demonstrates that the minimum concentration of OD groups required to prevent rapid growth of nanoparticles during thermal dehydration corresponds to ~0.7 monolayer coverage. Surface hydration layers not only stabilize the SnO2 nanoparticles but also induce particle-size-dependent structural modifications and are likely to promote interfacial reactions through hydrogen bonds between adjacent particles. Upon heating/dehydration under vacuum above 250 °C, nanoparticles start to grow with low activation energies, rapid increase of nanoparticle size, and a reduction in the a lattice dimension. This study underscores the value of neutron diffraction and prompt-gamma analysis, coupled with molecular modeling, in elucidating the influence of surface hydration on the structure and metastable persistence of oxide nanomaterials.

Modeling the interaction between integrin-binding peptide (RGD) and rutile surface: the effect of cation mediation on Asp adsorption.

The binding of a negatively charged residue, aspartic acid (Asp) in tripeptide arginine-glycine-aspartic acid, onto a negatively charged hydroxylated rutile (110) surface in aqueous solution, containing divalent (Mg(2+), Ca(2+), or Sr(2+)) or monovalent (Na(+), K(+), or Rb(+)) cations, was studied by molecular dynamics (MD) simulations. The results indicate that ionic radii and charges will significantly affect the hydration, adsorption geometry, and distance of cations from the rutile surface, thereby regulating the Asp/rutile binding mode. The adsorption strength of monovalent cations on the rutile surface in the order Na(+) > K(+) > Rb(+) shows a "reverse" lyotropic trend, while the divalent cations on the same surface exhibit a "regular" lyotropic behavior with decreasing crystallographic radii (the adsorption strength of divalent cations: Sr(2+) > Ca(2+) > Mg(2+)). The Asp side chain in NaCl, KCl, and RbCl solutions remains stably H-bonded to the surface hydroxyls and the inner-sphere adsorbed compensating monovalent cations act as a bridge between the COO(-) group and the rutile, helping to "trap" the negatively charged Asp side chain on the negatively charged surface. In contrast, the mediating divalent cations actively participate in linking the COO(-) group to the rutile surface; thus the Asp side chain can remain stably on the rutile (110) surface, even if it is not involved in any hydrogen bonds with the surface hydroxyls. Inner- and outer-sphere geometries are all possible mediation modes for divalent cations in bridging the peptide to the rutile surface.

Optimized unlike-pair interactions for water-carbon dioxide mixtures described by the SPC/E and EPM2 models.

The unlike-pair interaction parameters for the SPC/E-EPM2 models have been optimized to reproduce the mutual solubility of water and carbon dioxide at the conditions of liquid-supercritical fluid phase equilibria. An efficient global optimization of the parameters is achieved through an implementation of the coupling parameter approach, adapted to phase equilibria calculations in the Gibbs ensemble, that explicitly corrects for the overpolarization of the SPC/E water molecule in the nonpolar CO(2) environments. The resulting H(2)O-CO(2) force field accurately reproduces the available experimental solubilities at the two fluid phases in equilibria as well as the corresponding species tracer diffusion coefficients.