Search

Isotopic substitution, which can be realized both in experiment and computer simulations, is a direct approach to assess the role of nuclear quantum effects on the structure and dynamics of matter. Yet, the impact of nuclear quantum effects on the structure of liquid water as probed in experiment by comparing normal to heavy water has remained controversial. To settle this issue, we employ a highly accurate machine-learned high-dimensional neural network potential to perform converged coupled cluster-quality path integral simulations of liquid H$_2$O versus D$_2$O at ambient conditions. We find substantial H/D quantum effects on the rotational and translational dynamics of water, in close agreement with the experimental benchmarks. However, in stark contrast to the role for dynamics, H/D quantum effects turn out to be unexpectedly small, on the order of 1/1000 \r{A}, on both intramolecular and H-bonding structure of water. The most probable structure of water remains nearly unaffected by nuclear quantum effects, but effects on fluctuations away from average are appreciable, rendering H$_2$O substantially more "liquid" than D$_2$O.

Training accurate machine learning potentials requires electronic structure data comprehensively covering the configurational space of the system of interest. As the construction of this data is computationally demanding, many schemes for identifying the most important structures have been proposed. Here, we compare the performance of high-dimensional neural network potentials (HDNNPs) for quantum liquid water at ambient conditions trained to data sets constructed using random sampling as well as various flavors of active learning based on query by committee. Contrary to the common understanding of active learning, we find that for a given data set size, random sampling leads to smaller test errors for structures not included in the training process. In our analysis we show that this can be related to small energy offsets caused by a bias in structures added in active learning, which can be overcome by using instead energy correlations as an error measure that is invariant to such shifts. Still, all HDNNPs yield very similar and accurate structural properties of quantum liquid water, which demonstrates the robustness of the training procedure with respect to the training set construction algorithm even when trained to as few as 200 structures. However, we find that for active learning based on preliminary potentials, a reasonable initial data set is important to avoid an unnecessary extension of the covered configuration space to less relevant regions.

The infrared (IR) spectra of protonated water clusters encode precise information on the dynamics and structure of the hydrated proton. However, the strong anharmonic coupling and quantum effects of these elusive species remain puzzling up to the present day. Here, we report unequivocal evidence that the interplay between the proton transfer and the water wagging motions in the protonated water dimer (Zundel ion) giving rise to the characteristic doublet peak is both more complex and more sensitive to subtle energetic changes than previously thought. In particular, hitherto overlooked low-intensity satellite peaks in the experimental spectrum are now unveiled and mechanistically assigned. Our findings rely on the comparison of IR spectra obtained using two highly accurate potential energy surfaces in conjunction with highly accurate state-resolved quantum simulations. We demonstrate that these high-accuracy simulations are important for providing definite assignments of the complex IR signals of fluxional molecules.

The study of molecular impurities in $para$-hydrogen ($p$H$\rm_2$) clusters is key to push forward our understanding of intra- and intermolecular interactions including their impact on the superfluid response of this bosonic quantum solvent. This includes tagging with one or very few $p$H$\rm_2$, the microsolvation regime, and matrix isolation. However, the fundamental coupling between the bosonic $p$H$\rm_2$ environment and the (ro-)vibrational motion of molecular impurities remains poorly understood. Quantum simulations can in provide the necessary atomistic insight, but very accurate descriptions of the involved interactions are required. Here, we present a data-driven approach for the generation of $impurity\cdots p$H$\rm_2$ interaction potentials based on machine learning techniques which retain the full flexibility of the impurity. We employ the well-established adiabatic hindered rotor (AHR) averaging technique to include the impact of the nuclear spin statistics on the symmetry-allowed rotational quantum numbers of $p$H$\rm_2$. Embedding this averaging procedure within the high-dimensional neural network potential (NNP) framework enables the generation of highly-accurate AHR-averaged NNPs at coupled cluster accuracy, namely CCSD(T$^*$)-F12a/aVTZcp in an automated manner. We apply this methodology to the water and protonated water molecules, as representative cases for quasi-rigid and highly-flexible molecules respectively, and obtain AHR-averaged NNPs that reliably describe the H$\rm _2$O$\cdots p$H$\rm_2$ and H$\rm _3$O$^+\cdots p$H$\rm_2$ interactions. Using path integral simulations we show for the hydronium cation that umbrella-like tunneling inversion has a strong impact on the first and second $p$H$\rm_2$ microsolvation shells. The data-driven nature of our protocol opens the door to the study of bosonic $p$H$\rm_2$ quantum solvation for a wide range of embedded impurities., Comment: The following article has been submitted to the Journal of Chemical Physics

Infrared spectroscopy is key to elucidate molecular structures, monitor reactions and observe conformational changes, while providing information on both structural and dynamical properties. This makes the accurate prediction of infrared spectra based on first-principle theories a highly desirable pursuit. Molecular dynamics simulations have proven to be a particularly powerful approach for this task, albeit requiring the computation of energies, forces and dipole moments for a large number of molecular configurations as a function of time. This explains why highly accurate first principles methods, such as coupled cluster theory, have so far been inapplicable for the prediction of fully anharmonic vibrational spectra of large systems at finite temperatures. Here, we push cutting-edge machine learning techniques forward by using neural network representations of energies, forces and in particular dipoles to predict such infrared spectra fully at "gold standard" coupled cluster accuracy as demonstrated for protonated water clusters as large as the protonated water hexamer, in its extended Zundel configuration. Furthermore, we show that this methodology can be used beyond the scope of the data considered during the development of the neural network models, allowing for the computation of finite-temperature infrared spectra of large systems inaccessible to explicit coupled cluster calculations. This substantially expands the hitherto existing limits of accuracy, speed and system size for theoretical spectroscopy and opens up a multitude of avenues for the prediction of vibrational spectra and the understanding of complex intra- and intermolecular couplings.

Molecular dynamics simulations of supercritical water reveal distinctly different distance-dependent modulations of dipolar response and correlations in particle motion compared to ambient conditions. The strongly perturbed H-bond network of water at supercritical conditions allows for considerable translational and rotational freedom of individual molecules. These changes give rise to substantially different infrared spectra and vibrational density of states at THz frequencies for densities above and below the Widom line that separates percolating liquid-like and clustered gas-like supercritical water.

The mechanochemistry of ring-opening reactions of cyclopropane derivatives turns out to be unexpectedly rich and puzzling. After showing that a rare so-called uphill bifurcation in the case of trans-gem-difluorocyclopropane turns into a downhill bifurcation upon substitution of fluorine by chlorine, bromine and iodine in the thermal activation limit, the dichloro derivative is studied systematically in the realm of mechanochemical activation. Detailed exploration of the force-transformed potential energy surface of trans-gem-dichlorocyclopropane in terms of Dijkstra path analysis unveils a hitherto unknown topological catastrophe where the global shape of the energy landscape is fundamentally changed. From thermal activation up to moderately large forces, it is an uphill bifurcation that decides about dis-versus conrotatory ring-opening followed by separate transition states along both pathways. Above a critical force, the two distinct transition states merge to yield a single transition state such that the decision about the dis-versus conrotatory ring-opening process is taken at a newly established downhill bifurcation. The discovery of a force-induced qualitative change of the topology of a reaction network vastly transcends the previous understanding of the ring-opening reaction of this species. It would be astonishing to not discover a wealth of such catastrophes for mechanochemically activated reactions which will greatly extend the known opportunities to manipulate chemical reaction networks.

The design of accurate helium-solute interaction potentials for the simulation of chemically complex molecules solvated in superfluid helium has long been a cumbersome task due to the rather weak but strongly anisotropic nature of the interactions. We show that this challenge can be met by using a combination of an effective pair potential for the He-He interactions and a flexible high-dimensional neural network potential (NNP) for describing the complex interaction between helium and the solute in a pairwise additive manner. This approach yields an excellent agreement with a mean absolute deviation as small as 0.04 kJ/mol for the interaction energy between helium and both, hydronium and Zundel cations compared to CCSD(T) reference calculations with an energetically converged basis set. The construction and improvement of the potential can be performed in a highly automated way, which opens the door for applications to a variety of reactive molecules to study the effect of solvation on the solute as well as the solute-induced structuring of the solvent. Furthermore, we show that this NNP approach yields very convincing agreement with the CCSD(T) reference for properties like many-body spatial and radial distribution functions. This holds for the microsolvation of the protonated water monomer and dimer by a few helium atoms up to their solvation in bulk helium as obtained from path integral simulations at about 1 K.

For a long time, performing converged path integral simulations at ultra-low, but finite temperatures of a few Kelvin has been a nearly impossible task. However, recent developments in advanced colored noise thermostatting schemes for path integral simulations, namely the Path Integral Generalized Langevin Equation Thermostat (PIGLET) and the Path Integral Quantum Thermal Bath (PIQTB), have been able to greatly reduce the computational cost of these simulations, thus making the ultra-low temperature regime accessible in practice. In this work, we investigate the influence of these two thermostatting schemes on the description of hydrogen-bonded systems at temperatures down to a few Kelvin as encountered, for example, in helium nanodroplet isolation or tagging photodissociation spectroscopy experiments. For this purpose, we analyze the prototypical hydrogen bond in the Zundel cation (H$_5$O$_{2}^{+}$) as a function of both, oxygen-oxygen distance and temperature in order to elucidate how the anisotropic quantum delocalization and, thus, the shape of the shared proton adapts depending on the donor-acceptor distance. The underlying electronic structure of the Zundel cation is described in terms of Behler's Neural Network Potentials of essentially converged Coupled Cluster accuracy, CCSD(T*)-F12a/AVTZ. In addition, the performances of the PIQTB and PIGLET methods for energetic, structural, and quantum delocalization properties are assessed and directly compared. Overall, our results emphasize the validity and practical usefulness of these two modern thermostatting approaches for path integral simulations of hydrogen-bonded systems even at ultra-low temperatures.

A frozen, solid helium core, dubbed snowball, is typically observed around cations in liquid helium. Here we discover, using path integral simulations, that around a cationic molecular impurity, protonated methane, the $^4$He atoms are indeed strongly localized akin to snowballs but still participate in vivid bosonic exchange induced by the ro-vibrational motion of the impurity. Such combination of solid-like order with pronounced superfluid response in the first helium shell indicates that manifestations of local supersolid behavior of $^4$He can be induced -- and probed experimentally -- by charged molecules.

A previously published neural network potential for the description of protonated water clusters up to the protonated water tetramer, H$^+$(H$_2$O)$_4$, at essentially converged coupled cluster accuracy (J. Chem. Theory Comput. 16, 88 (2020)) is applied to the protonated water hexamer, H$^+$(H$_2$O)$_6$ -- a system that the neural network has never seen before. Although being in the extrapolation regime, it is shown that the potential not only allows for quantum simulations from ultra-low temperatures $\sim$ 1 K up to 300 K, but that it is able to describe the new system very accurately compared to explicit coupled cluster calculations. This transferability of the model is rationalized by the similarity of the atomic environments encountered for the larger cluster compared to the environments in the training set of the model. Compared to the interpolation regime the quality of the model is reduced by roughly one order of magnitude, but most of the difference to the coupled cluster reference comes from global shifts of the potential energy surface, while local energy fluctuations are well recovered. These results suggest that the application of neural network potentials in extrapolation regimes can provide useful results and might be more general than usually thought., Comment: The following article has been submitted to the Journal of Chemical Physics

We employ the k-th nearest-neighbor estimator of configurational entropy in order to decode within a parameter-free numerical approach the complex high-order structural correlations in fluxional molecules going beyond the usual linear, bivariate correlations. This generic entropy-based scheme for determining many-body correlations is applied to the complex configurational ensemble of protonated acetylene, a prototype for fluxional molecules featuring large-amplitude motion. After revealing the importance of high-order correlations beyond the simple two-coordinate picture for this molecule, we analyze in detail the evolution of the relevant correlations with temperature as well as the impact of nuclear quantum effects down to the ultra-low temperature regime of 1 K. We find that quantum delocalization and zero-point vibrations significantly reduce all correlations in protonated acetylene in the deep quantum regime. At intermediate temperatures of approximately 100 to 800 K, a quantum-to-classical cross-over regime is found where classical mechanics starts to correctly describe trends in the correlations whereas it even qualitatively fails below 100 K. Finally, a classical description of the nuclei provides correlations that are in quantitative agreement with the quantum ones only at temperatures exceeding 1000 K. This data-intensive analysis has been made possible due to recent developments of machine learning techniques based on neural network potentials in full dimensionality that allow us to exhaustively sample both, the classical and quantum ensemble of protonated acetylene at essentially converged coupled cluster accuracy. The presented non-parametric analysis of correlations is expected to complement and guide the analysis of experimental measurements, in particular multi-dimensional vibrational spectroscopy, by revealing the complex coupling between various degrees of freedom.

Many experimental techniques such as tagging photodissociation and helium nanodroplet isolation spectroscopy operate at very low temperatures in order to investigate hydrogen bonding. To elucidate the differences between such ultra-cold and usual ambient conditions, different hydrogen bonded systems are studied systematically from 300 K down to about 1 K using path integral simulations that explicitly consider both, the quantum nature of the nuclei and thermal fluctuations. For that purpose, finite sized water clusters, specifically the water dimer and hexamer, protonated water clusters including the Zundel and Eigen complexes, as well as hexagonal ice as a condensed phase representative are compared directly as a function of temperature. While weaker hydrogen bonds, as present in the neutral systems, show distinct structural differences between ambient conditions and the ultra-cold regime, the stronger hydrogen bonds of the protonated water clusters are less perturbed by temperature compared to their quantum ground state. In all studied systems, the quantum delocalization of the nuclei is found to vary drastically with temperature. Interestingly, upon reaching temperatures of about 1 K, the spatial quantum delocalization of the heavy oxygens approaches that of the protons for relatively weak spatial constraints, and even significantly exceeds the latter in case of the centered hydrogen bond in the Zundel complex. These findings are relevant for comparisons between experiments on hydrogen bonding carried out at ultra-cold versus ambient conditions as well as to understand quantum delocalization phenomena of nuclei by seamlessly extending our insights into noncovalent interactions down to ultra-low temperatures.

Highly accurate potential energy surfaces are of key interest for the detailed understanding and predictive modeling of chemical systems. In recent years, several new types of force fields, which are based on machine learning algorithms and fitted to ab initio reference calculations, have been introduced to meet this requirement. Here we show how high-dimensional neural network potentials can be employed to automatically generate the potential energy surface of finite sized clusters at coupled cluster accuracy, namely CCSD(T*)-F12a/aug-cc-pVTZ. The developed automated procedure utilizes the established intrinsic properties of the model such that the configurations for the training set are selected in an unbiased and efficient way to minimize the computational effort of expensive reference calculations. These ideas are applied to protonated water clusters from the hydronium cation, H$_3$O$^+$, up to the tetramer, H$_9$O$_{4}^{+}$, and lead to a single potential energy surface that describes all these systems at essentially converged coupled cluster accuracy with a fitting error of 0.06 kJ/mol per atom. The fit is validated in detail for all clusters up to the tetramer and yields reliable results not only for stationary points, but also for reaction pathways, intermediate configurations, as well as different sampling techniques. Per design the NNPs constructed in this fashion can handle very different conditions including the quantum nature of the nuclei and enhanced sampling techniques covering very low as well as high temperatures. This enables fast and exhaustive exploration of the targeted protonated water clusters with essentially converged interactions. In addition, the automated process will allow one to tackle finite systems much beyond the present case.

Water, the ubiquitous solvent, is also prominent in forming liquid solid interfaces with catalytically active surfaces, in particular with promoted oxides. We study the complex interface of a gold nanocatalyst, pinned by an F center on titania support, and water. The ab initio simulations uncover the microscopic details of solvent-induced charge rearrangements at the metal particle. Water is found to stabilize charge states differently from the gas phase as a result of structure specific charge transfer from to the solvent, thus altering surface reactivity. The metal cluster is shown to feature both cationic and anionic solvation, depending on fluctuation and polarization effects in the liquid, which creates novel active sites. These observations open up an avenue toward solvent engineering in liquid-phase heterogeneous catalysis.

Extensive first principles calculations are carried out to investigate gold-promoted TiO2(110) surfaces in terms of structure optimizations, electronic structure analyses, ab initio thermodynamics calculations of surface phase diagrams, and ab initio molecular dynamics simulations. All computations rely on density functional theory in the generalized gradient approximation (PBE) and account for on-site Coulomb interactions via inclusion of a Hubbard correction, PBE+U, where U is computed from linear response theory. This approach is validated by investigating the interaction between TiO2(110) surfaces and typical probe species (H, H2O, CO). Relaxed structures and binding energies are compared to both data from the literature and plain PBE results. The main focus of the study is on the properties of gold-promoted titania surfaces and their interactions with CO. Both PBE+U and PBE optimized structures of Au adatoms adsorbed on stoichiometric and reduced TiO2 surfaces are computed, along with their electronic structure. The charge rearrangement induced by the adsorbates at the metal/oxide contact are also analyzed and discussed. By performing PBE+U ab initio molecular dynamics simulations, it is demonstrated that the diffusion of Au adatoms on the stoichiometric surface is highly anisotropic. The metal atoms migrate either along the top of the bridging oxygen rows, or around the area between these rows, from one bridging position to the next along the [001] direction. Approximate ab initio thermodynamics predicts that under O-rich conditions, structures obtained by substituting a Ti5c atom with an Au atom are thermodynamically stable over a wide range of temperatures and pressures., Comment: 20 pages, 12 figures, accepted for publication in Phys. Rev. B

The dynamics of an F--center created by an oxygen vacancy on the $\mathrm{TiO_{2}(110)}$ rutile surface has been investigated using {\it ab initio} molecular dynamics. These simulations uncover a truly complex, time-dependent behavior of fluctuating electron localization topologies in the vicinity of the oxygen vacancy. Although the two excess electrons are found to populate preferentially the second subsurface layer, they occasionally visit surface sites and also the third subsurface layer. This dynamical behavior of the excess charge explains hitherto conflicting interpretations of both theoretical findings and experimental data., Comment: 4 pages, 2 figures, accepted for publication in Phys. Rev. Letters

A comprehensive phase diagram of lowest-energy structures and compositions of the rutile TiO_2(110) surface in equilibrium with a surrounding gas phase at finite temperatures and pressures has been determined using density functional theory in combination with a thermodynamic formalism. The exchange of oxygen, hydrogen, and water molecules with the gas phase is considered. Particular attention is given to the convergence of all calculations with respect to lateral system size and slab thickness. In addition, the reliability of semilocal density functionals to describing the energetics of the reduced surfaces is critically evaluated. For ambient conditions the surface is found to be fully covered by molecularly adsorbed water. At low coverages, in the limit of single, isolated water molecules, molecular and dissociative adsorption become energetically degenerate. Oxygen vacancies form in strongly reducing, oxygen-poor environments. However, already at slightly more moderate conditions it is shown that removing full TiO_2 units from the surface is thermodynamically preferred. In agreement with recent experimental observations it is furthermore confirmed that even under extremely hydrogen-rich environments the surface cannot be fully hydroxylated, but only a maximum coverage with hydrogen of about 0.6-0.7 monolayer can be reached. Finally, calculations of migration paths strongly suggest that hydrogen prefers to diffuse into the bulk over desorbing from the surface into the gas phase., Comment: 17 pages, 11 figures, to appear in PRB

A joint experimental / theoretical investigation of the elastin-like octapeptide GVG(VPGVG) was carried out. In this paper a comprehensive molecular dynamics study of the temperature dependent folding and unfolding of the octapeptide is presented. The current study, as well as its experimental counterpart find that this peptide undergoes an "inverse temperature transition", ITT, leading to a folding at about 310-330 K. In addition, an unfolding transition is identified at unusually high temperatures approaching the boiling point of water. Due to the small size of the system two broad temperature regimes are found: the "ITT regime" (at about 280-320 K) and the "unfolding regime" at about T > 330 K, where the peptide has a maximum probability of being folded at approximately 330 K. A detailed molecular picture involving a thermodynamic order parameter, or reaction coordinate, for this process is presented along with a time-correlation function analysis of the hydrogen bond dynamics within the peptide as well as between the peptide and solvating water molecules. Correlation with experimental evidence and ramifications on the properties of elastin are discussed., Comment: 15 pages, 1 table, 8 figures

The ammonia dimer (NH3)2 has been investigated using high--level ab initio quantum chemistry methods and density functional theory (DFT). The structure and energetics of important isomers is obtained to unprecedented accuracy without resorting to experiment. The global minimum of eclipsed C_s symmetry is characterized by a significantly bent hydrogen bond which deviates from linearity by about 20 degrees. In addition, the so-called cyclic C_{2h} structure is extremely close in energy on an overall flat potential energy surface. It is demonstrated that none of the currently available (GGA, meta--GGA, and hybrid) density functionals satisfactorily describe the structure and relative energies of this nonlinear hydrogen bond. We present a novel density functional, HCTH/407+, designed to describe this sort of hydrogen bond quantitatively on the level of the dimer, contrary to e.g. the widely used BLYP functional. This improved functional is employed in Car-Parrinello ab initio molecular dynamics simulations of liquid ammonia to judge its performance in describing the associated liquid. Both the HCTH/407+ and BLYP functionals describe the properties of the liquid well as judged by analysis of radial distribution functions, hydrogen bonding structure and dynamics, translational diffusion, and orientational relaxation processes. It is demonstrated that the solvation shell of the ammonia molecule in the liquid phase is dominated by steric packing effects and not so much by directional hydrogen bonding interactions. In addition, the propensity of ammonia molecules to form bifurcated and multifurcated hydrogen bonds in the liquid phase is found to be negligibly small., Comment: Journal of Chemical Physics, in press (305335JCP)

An extensive theoretical investigation of the nonpolar (10$\bar{1}$0) and (11$\bar{2}$0) surfaces as well as the polar zinc terminated (0001)--Zn and oxygen terminated (000$\bar{1}$)--O surfaces of ZnO is presented. Particular attention is given to the convergence properties of various parameters such as basis set, k--point mesh, slab thickness, or relaxation constraints within LDA and PBE pseudopotential calculations using both plane wave and mixed basis sets. The pros and cons of different approaches to deal with the stability problem of the polar surfaces are discussed. Reliable results for the structural relaxations and the energetics of these surfaces are presented and compared to previous theoretical and experimental data, which are also concisely reviewed and commented., Comment: 12 pages, with 9 postscript figures embedded. Uses REVTEX and epsf macros