Predicting Emission of Heteroleptic Iridium Complexes using Artificial Chemical Intelligence.

We report a deep learning-based approach to accurately predict the emission spectra of phosphorescent heteroleptic [Ir(C^N)2(NN)]+ complexes, enabling the rapid discovery of novel Ir(III) chromophores for diverse applications including organic light-emitting diodes and solar fuel cells. The deep learning models utilize graph neural networks and other chemical features in architectures that reflect the inherent structure of the heteroleptic complexes, composed of C^N and N^N ligands, and are thus geared towards efficient training over the dataset. By leveraging experimental emission data, our models reliably predict the full emission spectra of these complexes across various emission profiles, surpassing the accuracy of conventional DFT and correlated wavefunction methods, while simultaneously achieving robustness to the presence of imperfect (noisy, low-quality) training spectra. We showcase the potential applications for these and related models for \insilico\ prediction of complexes with tailored emission properties, as well as in "design of experiment'' contexts to reduce the synthetic burden of high-throughput screening. In the latter case, we demonstrate that the models allow to exploit a limited amount of experimental data to explore a wide range of chemical space, thus leveraging a modest synthetic effort.

Stable Pentagonal Layered Palladium Diselenide Enables Rapid Electrosynthesis of Hydrogen Peroxide.

Electrosynthesis of hydrogen peroxide (H2O2) via the two-electron oxygen reduction reaction (2e- ORR) is promising for various practical applications, such as wastewater treatment. However, few electrocatalysts are active and selective for 2e- ORR yet are also resistant to catalyst leaching under realistic operating conditions. Here, a joint experimental and computational study reveals active and stable 2e- ORR catalysis in neutral media over layered PdSe2 with a unique pentagonal puckered ring structure type. Computations predict active and selective 2e- ORR on the basal plane and edge of PdSe2, but with distinct kinetic behaviors. Electrochemical measurements of hydrothermally synthesized PdSe2 nanoplates show a higher 2e- ORR activity than other Pd-Se compounds (Pd4Se and Pd17Se15). PdSe2 on a gas diffusion electrode can rapidly accumulate H2O2 in buffered neutral solution under a high current density. The electrochemical stability of PdSe2 is further confirmed by long device operational stability, elemental analysis of the catalyst and electrolyte, and synchrotron X-ray absorption spectroscopy. This work establishes a new efficient and stable 2e- ORR catalyst at practical current densities and opens catalyst designs utilizing the unique layered pentagonal structure motif.

The Impact of Electron Donating and Withdrawing Groups on Electrochemical Hydrogenolysis and Hydrogenation of Carbonyl Compounds.

The hydrogenolysis or hydrodeoxygenation of a carbonyl group, where the C═O group is converted to a CH2 group, is of significant interest in a variety of fields. A challenge in electrochemically achieving hydrogenolysis of a carbonyl group with high selectivity is that electrochemical hydrogenation of a carbonyl group, which converts the C═O group to an alcohol group (CH-OH), is demonstrated not to be the initial step of hydrogenolysis. Instead, hydrogenation and hydrogenolysis occur in parallel, and they are competing reactions. This means that although both hydrogenolysis and hydrogenation require adding H atoms to the carbonyl group, they involve different intermediates formed on the electrode surface. Thus, revealing the difference in intermediates, transition states, and kinetic barriers for hydrogenolysis and hydrogenation pathways is the key to understanding and controlling hydrogenolysis/hydrogenation selectivity of carbonyl compounds. In this study, we aimed to identify features of reactant molecules that can affect their hydrogenolysis/hydrogenation selectivity on a Zn electrode that was previously shown to promote hydrogenolysis over hydrogenation. In particular, we examined the electrochemical reduction of para-substituted benzaldehyde compounds with substituent groups having different electron donating/withdrawing abilities. Our results show a strikingly systematic impact of the substituent group where a stronger electron-donating group promotes hydrogenolysis and a stronger electron-withdrawing group promotes hydrogenation. These experimental results are presented with computational results explaining the substituent effects on the thermodynamics and kinetics of electrochemical hydrogenolysis and hydrogenation pathways, which also provide critically needed information and insights into the transition states involved with these pathways.

Halide Adsorption Enhances Electrochemical Hydrogenolysis of 5-Hydroxymethylfurfural by Suppressing Hydrogenation.

Reductive upgrading of 5-hydroxymethylfurfural (HMF), a biomass-derived platform molecule, to 2,5-dimethylfuran (DMF), a biofuel with an energy density 40% greater than that of ethanol, involves hydrogenolysis of both the aldehyde (C═O) and the alcohol (C-OH) groups of HMF. It is known that when hydrogenation of the aldehyde occurs to form 2,5-bis(hydroxymethyl)furan (BHMF), BHMF cannot be further reduced to DMF. Thus, aldehyde hydrogenation must be suppressed to increase the selectivity for DMF production. Previously, it was shown that on a Cu electrode hydrogenolysis occurs mainly through proton-coupled electron transfer (PCET), where a proton from the solution and an electron from the electrode are transferred to the organic species. In contrast, hydrogenation occurs not only through PCET but also through hydrogen atom transfer (HAT), where a surface-adsorbed hydrogen atom (H*) is transferred to the organic species. This study shows that halide adsorption on Cu can effectively suppress HAT by decreasing the steady-state H* coverage on Cu during HMF reduction. As HAT enables only aldehyde hydrogenation, a striking suppression of BHMF is observed, thereby enhancing DMF production. We discuss how the identity and concentration of the halide, along with the reduction conditions (i.e., potential and pH), affect halide adsorption on Cu and identify when optimal halide coverages are achieved to maximize DMF selectivity. Our experimental results are presented alongside computational results that elucidate how halide adsorption affects the adsorption energy of hydrogen and the steady-state H* coverage on Cu, which provide an atomic-level understanding of all experimentally observed effects.

Development and Implementation of Atomically Anisotropic First-Principles Force Fields: A Benzene Case Study.

π-interactions are an important motif in chemical and biochemical systems. However, due to their anisotropic electron densities and complex balance of intermolecular interactions, aromatic molecules represent an ongoing challenge for accurate and transferable force field development. Historically, ab initio force fields for aromatics have not exhibited good accuracy with respect to bulk properties or have only been used to study gas-phase dimers. Using benzene as a proof of concept, herein we show how our own ab initio MASTIFF force field incorporates an atomically anisotropic description of intermolecular interactions to yield an accurate and robust model for aromatic interactions irrespective of phase. Compared to existing models, the MASTIFF benzene force field not only is accurate for liquid phase properties but also offers transferability to the gas and solid phases. Additionally, we introduce a computationally efficient OpenMM plugin which enables customizable anisotropic intermolecular functional forms and which can be generically used in any MD simulation where a model for nonspherical atomic features is required. Overall, our results demonstrate the importance of atomic-level anisotropy in enabling next-generation ab initio force field development.

Guiding epitaxial crystallization of amorphous solids at the nanoscale: Interfaces, stress, and precrystalline order.

The crystallization of amorphous solids impacts fields ranging from inorganic crystal growth to biophysics. Promoting or inhibiting nanoscale epitaxial crystallization and selecting its final products underpin applications in cryopreservation, semiconductor devices, oxide electronics, quantum electronics, structural and functional ceramics, and advanced glasses. As precursors for crystallization, amorphous solids are distinguished from liquids and gases by the comparatively long relaxation times for perturbations of the mechanical stress and for variations in composition or bonding. These factors allow experimentally controllable parameters to influence crystallization processes and to drive materials toward specific outcomes. For example, amorphous precursors can be employed to form crystalline phases, such as polymorphs of Al2O3, VO2, and other complex oxides, that are not readily accessible via crystallization from a liquid or through vapor-phase epitaxy. Crystallization of amorphous solids can further be guided to produce a desired polymorph, nanoscale shape, microstructure, or orientation of the resulting crystals. These effects enable advances in applications in electronics, magnetic devices, optics, and catalysis. Directions for the future development of the chemical physics of crystallization from amorphous solids can be drawn from the structurally complex and nonequilibrium atomic arrangements in liquids and the atomic-scale structure of liquid-solid interfaces.

Operando Elucidation of Electrocatalytic and Redox Mechanisms on a 2D Metal Organic Framework Catalyst for Efficient Electrosynthesis of Hydrogen Peroxide in Neutral Media.

The practical electrosynthesis of hydrogen peroxide (H2O2) is hindered by the lack of inexpensive and efficient catalysts for the two-electron oxygen reduction reaction (2e- ORR) in neutral electrolytes. Here, we show that Ni3HAB2 (HAB = hexaaminobenzene), a two-dimensional metal organic framework (MOF), is a selective and active 2e- ORR catalyst in buffered neutral electrolytes with a linker-based redox feature that dynamically affects the ORR behaviors. Rotating ring-disk electrode measurements reveal that Ni3HAB2 has high selectivity for 2e- ORR (>80% at 0.6 V vs RHE) but lower Faradaic efficiency due to this linker redox process. Operando X-ray absorption spectroscopy measurements reveal that under argon gas the charging of the organic linkers causes a dynamic Ni oxidation state, but in O2-saturated conditions, the electronic and physical structures of Ni3HAB2 change little and oxygen-containing species strongly adsorb at potentials more cathodic than the reduction potential of the organic linker (Eredox ∼ 0.3 V vs RHE). We hypothesize that a primary 2e- ORR mechanism occurs directly on the organic linkers (rather than the Ni) when E > Eredox, but when E < Eredox, H2O2 production can also occur through Ni-mediated linker discharge. By operating the bulk electrosynthesis at a low overpotential (0.4 V vs RHE), up to 662 ppm of H2O2 can be produced in a buffered neutral solution in an H-cell due to minimized strong adsorption of oxygenates. This work demonstrates the potential of conductive MOF catalysts for 2e- ORR and the importance of understanding catalytic active sites under electrochemical operation.

Kinetic Model of Acid Gas Induced Defect Propagation in Zeolitic Imidazolate Frameworks.

Understanding the degradation of nanoporous materials under exposure to common acid gas contaminants (e.g., SO2, CO2, NO2, and H2S) is essential to elongate their lifetime and thus enable their practical applications in separations and catalysis. Previous theoretical investigations have focused on the formation of isolated point defects, which are insufficient to provide direct insights into the long-term evolution of the bulk properties of materials such as zeolitic imidazolate frameworks (ZIFs) under sustained acid gas exposure. To bridge this divide in both length and time scales, we developed a first-principles lattice-based kinetic model to simulate the defect propagation and bulk material breakdown in ZIFs. This model closely reproduces the experimentally measured macroscopic evolution of the time-dependent bulk materials proprieties and also yields important new insights regarding the autocatalytic nature of ZIF degradation and the spatial distribution of defects. Our results suggest new experimental directions to identify nascent defect clusters in degraded ZIFs and avenues to mitigate degradation under challenging conditions of acid gas exposure.

Mechanistic Differences between Electrochemical Hydrogenation and Hydrogenolysis of 5-Hydroxymethylfurfural and Their pH Dependence.

Hydrogenation and hydrogenolysis are two important reactions for electrochemical reductive valorization of biomass-derived oxygenates such as 5-hydroxymethylfurfural (HMF). In general, hydrogenolysis (which combines hydrogenation and deoxygenation) is more challenging than hydrogenation (which does not involve the cleavage of carbon-oxygen bonds). Thus, identifying factors and conditions that can promote hydrogenolysis is of great interest for reductive valorization of biomass-derived oxygenates. For the electrochemical reduction of HMF and its derivatives, it is known that aldehyde hydrogenation is not a part of aldehyde hydrogenolysis but rather a competing reaction; however, no atomic-level understanding is currently available to explain their electrochemical mechanistic differences. In this study, combined experimental and computational investigations were performed using Cu electrodes to elucidate the key mechanistic differences between electrochemical hydrogenation and hydrogenolysis of HMF. The results revealed that hydrogenation and hydrogenolysis of HMF involve the formation of different surface-adsorbed intermediates via different reduction mechanisms and that lowering the pH promoted the formation of the intermediates required for aldehyde and alcohol hydrogenolysis. This study for the first time explains the origins of the experimentally observed pH-dependent selectivities for hydrogenation and hydrogenolysis and offers a new mechanistic foundation upon which rational strategies to control electrochemical hydrogenation and hydrogenolysis can be developed.

Surface diffusion of a glassy discotic organic semiconductor and the surface mobility gradient of molecular glasses.

Surface diffusion has been measured in the glass of an organic semiconductor, MTDATA, using the method of surface grating decay. The decay rate was measured as a function of temperature and grating wavelength, and the results indicate that the decay mechanism is viscous flow at high temperatures and surface diffusion at low temperatures. Surface diffusion in MTDATA is enhanced by 4 orders of magnitude relative to bulk diffusion when compared at the glass transition temperature Tg. The result on MTDATA has been analyzed along with the results on other molecular glasses without extensive hydrogen bonds. In total, these systems cover a wide range of molecular geometries from rod-like to quasi-spherical to discotic and their surface diffusion coefficients vary by 9 orders of magnitude. We find that the variation is well explained by the existence of a steep surface mobility gradient and the anchoring of surface molecules at different depths. Quantitative analysis of these results supports a recently proposed double-exponential form for the mobility gradient: log D(T, z) = log Dv(T) + [log D0 - log Dv(T)]exp(-z/ξ), where D(T, z) is the depth-dependent diffusion coefficient, Dv(T) is the bulk diffusion coefficient, D0 ≈ 10-8 m2/s, and ξ ≈ 1.5 nm. Assuming representative bulk diffusion coefficients for these fragile glass formers, the model reproduces the presently known surface diffusion rates within 0.6 decade. Our result provides a general way to predict the surface diffusion rates in molecular glasses.

Origins of Acid-Gas Stability Behavior in Zeolitic Imidazolate Frameworks: The Unique High Stability of ZIF-71.

Zeolitic imidazolate frameworks (ZIFs) are promising materials for industrial process separations, but recent literature reports have highlighted their vulnerability to acid gases (e.g., SO2, CO2, NO2, H2S), often present in practical applications. While previous work has documented the widely varying stability behavior of many ZIFs under varying (humid and dry) acid gas environments, efforts to explain or correlate these experimental observations via empirical descriptors have not succeeded. A key observation is that ZIF-71 (RHO topology) is an extraordinarily stable ZIF material, retaining both structure and porosity under prolonged humid SO2 exposure whereas many other well-known ZIFs with different linkers and topologies (such as ZIF-8) were shown to degrade. Through a combination of hybrid quantum mechanics/molecular mechanics (QM/MM) based methods and statistical mechanical models, we successfully explain this important experimental observation via atomistic investigations of the reaction mechanism. Our holistic approach reveals an ∼9 times lower average defect formation rate in ZIF-71 RHO compared to ZIF-8 SOD, leading to the conclusion that the observed experimental stability of this material rises from kinetic effects. Moreover, our analysis reveals that differing stability of the two materials is determined by the distributions of acid gas molecules, which is difficult to capture using empirical descriptors. Our results suggest wider applicability of the present approach, toward identifying tuned functional groups and topologies that move the acid gas distributions away from more reactive sites and thus allow enhanced kinetic stability.

Defect-mediated ferromagnetism in correlated two-dimensional transition metal phosphorus trisulfides.

Controlling the magnetic spin states of two-dimensional (2D) van der Waals (vdW) materials with strong electronic or magnetic correlation is important for spintronic applications but challenging. Crystal defects that are often present in 2D materials such as transition metal phosphorus trisulfides (MPS3) could influence their physical properties. Here, we report the effect of sulfur vacancies on the magnetic exchange interactions and spin ordering of few-layered vdW magnetic Ni1−xCoxPS3 nanosheets. Magnetic and structural characterization in corroboration with theoretical calculations reveal that sulfur vacancies effectively suppress the strong intralayer antiferromagnetic correlation, giving rise to a weak ferromagnetic ground state in Ni1−xCoxPS3 nanosheets. Notably, the magnetic field required to tune this ferromagnetic state (<300 Oe) is much lower than the value needed to tune a typical vdW antiferromagnet (> several thousand oersted). These findings provide a previously unexplored route for controlling competing correlated states and magnetic ordering by defect engineering in vdW materials.

The Impact of 5-Hydroxymethylfurfural (HMF)-Metal Interactions on the Electrochemical Reduction Pathways of HMF on Various Metal Electrodes.

5-Hydroxymethylfurfural (HMF), which can be derived from lignocellulosic biomass, is an important platform molecule that can be used to produce valuable biofuels and polymeric materials. Electrochemical reduction of HMF is of great interest as it uses water as the hydrogen source and achieves desired reduction reactions at room temperature and ambient pressure. Hydrogenation and hydrogenolysis are two important reactions for reductive HMF conversion. Therefore, elucidating key characteristics of electrocatalysts that govern the selectivity for hydrogenation and hydrogenolysis is critical in rationally developing efficient and selective electrocatalysts. In this study, combined experimental and computational investigations are used to demonstrate how the adsorption energy of HMF on metal surfaces and the resulting changes in the intramolecular bond lengths of adsorbed HMF directly impact the reduction pathways of HMF. These results make it possible to rationally understand a general trend in the behaviors observed when using various metal electrodes for HMF reduction.

Phase Selection and Structure of Low-Defect-Density γ-Al2O3 Created by Epitaxial Crystallization of Amorphous Al2O3.

A multistep phase sequence following the crystallization of amorphous Al2O3 via solid-phase epitaxy (SPE) points to methods to create low-defect-density thin films of the metastable cubic γ-Al2O3 polymorph. An amorphous Al2O3 thin film on a (0001) α-Al2O3 sapphire substrate initially transforms upon heating to form epitaxial γ-Al2O3, followed by a transformation to monoclinic θ-Al2O3, and eventually to α-Al2O3. Epitaxial γ-Al2O3 layers with low mosaic widths in X-ray rocking curves can be formed via SPE by crystallizing the γ-Al2O3 phase from amorphous Al2O3 and avoiding the microstructural inhomogeneity arising from the spatially inhomogeneous transformation to θ-Al2O3. A complementary molecular dynamics (MD) simulation indicates that the amorphous layer and γ-Al2O3 have similar Al coordination geometry, suggesting that γ-Al2O3 forms in part because it involves the minimum rearrangement of the initially amorphous configuration. The lattice parameters of γ-Al2O3 are consistent with a structure in which the majority of the Al vacancies in the spinel structure occupy sites with tetrahedral coordination, consistent with the MD results. The formation of Al vacancies at tetrahedral spinel sites in epitaxial γ-Al2O3 can minimize the epitaxial elastic deformation of γ-Al2O3 during crystallization.

Solvation Induced Ring Puckering Effect in Fluorinated Prolines and Its Inclusion in Classical Force Fields.

Strategic incorporation of fluorinated prolines can accelerate folding and increase thermal stability of proteins. It has been suggested that this behavior emerges from puckering effects induced by fluorination of the proline ring. We use electronic structure calculations to characterize the potential energy surface (PES) along puckering coordinates for a simple dipeptide model of proline and its fluorinated derivatives. Significant shifts in puckering trends between gas phase and implicit solvent calculations shed light on the effect of solvation on electronic structure and conformational preferences of the ring. This solvation induced puckering effect is previously unknown in the context of prolines. The PES based on implicit solvent is then utilized to construct a correction for a classical force field. The corrected force field accurately captures the experimental conformational equilibrium including the coupling between ring puckering and cis-trans isomerism in fluorinated prolines. This method can be extended to other rings and substituents besides fluorine.

Modeling the Nucleation of Weak Electrolytes via Hybrid GCMC/MD Simulation.

We have developed a hybrid grand canonical Monte Carlo/molecular dynamics (GCMC/MD) method for simulating the nucleation of weak electrolytes in an explicit solvent. In contrast to brute-force MD simulation, the approach is capable of efficiently simulating the nucleation of dilute solutions while including the atomistic influence of the surrounding solvent and provides access to the full nucleation free energy surface and associated nucleation free energy barrier. After validating the method against a simple model system, we applied the approach to the nucleation of a low-solubility rock-salt structure in liquid water. We find that the calculated nucleation barriers, in conjunction with analytic rate theories, yield predicted nucleation rates that are in excellent agreement with brute-force MD simulations of the supersaturated solution. We anticipate possible applications of this approach to a wide variety of related weak electrolytes, including CaCO3, zeolites, and metal-organic frameworks.

Influence of Charge Scaling on the Solvation Properties of Ionic Liquid Solutions.

Scaled-charge force fields (FFs) are widely employed in the simulation of neat ionic liquids (ILs), where the charges on the ions are empirically scaled to approximately account for electronic polarization and/or charge transfer. Such charge scaling has been found to yield significant improvement in liquid-state thermodynamic and dynamic properties (when compared to experiment). However, the mean field approximation inherent in charge scaling becomes suspect when applied to IL mixtures or solutions. In this work, we simulate solutions of IL with various nonpolar and polar gas solutes and compare results of charge-scaled and polarizable FFs to experiment. Our results demonstrate that scaling of the Coulomb interaction inherent in scaled-charge FFs leads to an underestimation of the solute-solvent electrostatic interaction and thus also the enthalpy and free energy of solvation; this effect is particularly pronounced for polar solutes. In some cases, we find that this artificial reduction in the solute-solvent interaction can also alter the apparent phase behavior of the resulting solution. Overall, the totality of our results suggests that explicit polarization (rather than charge scaling) is likely necessary to provide high transferability to both neat IL and IL mixtures and solutions.