The use of machine learning (ML) algorithms in molecular simulations has become commonplace in recent years. There now exists, for instance, a multitude of ML force field algorithms that have enabled simulations approaching ab initio level accuracy at time scales and system sizes that significantly exceed what is otherwise possible with traditional methods. Far fewer algorithms exist for predicting rotationally equivariant, tensorial properties such as the electric polarizability. Here, we introduce a kernel ridge regression algorithm for machine learning of the polarizability tensor. This algorithm is based on the bond polarizability model and allows prediction of the tensor components at the cost similar to that of scalar quantities. We subsequently show the utility of this algorithm by simulating gas phase Raman spectra of biphenyl and malonaldehyde using classical molecular dynamics simulations of these systems performed with the recently developed MACE-OFF23 potential. The calculated spectra are shown to agree very well with the experiments and thus confirm the expediency of our algorithm as well as the accuracy of the used force field. More generally, this work demonstrates the potential of physics-informed approaches to yield simple yet effective machine learning algorithms for molecular properties.

"Line tension", a concept that features in an additional term to the Young's equation, was introduced to describe the size dependence of contact angles of nanodroplets on surfaces. Although this concept describes the observations in a succinct, elegant manner, theorists have long had misgivings about the physical interpretation of the phenomenon. Papers have been published that attempt to nail down its value, which is reportedly very small (similar to 10 pN) and evidently even the sign has been uncertain. Attempts to interpret it in a mechanical manner analogous to interfacial tension, i.e., due to the curvature of the three-phase contact line, have run into conceptual problems that require invocations of ever more complex models. In this work, we have used molecular simulations to systematically relate "line tension" to the additional free energy per unit length of the three-phase line and found no direct relation. However, when we rederived the Young's equation without ignoring the interfacial molecules, we found a physically satisfying explanation for the size dependence of the contact angle of nanodroplets without invoking the curvature of the three-phase contact line. The new model does not have the elegant form of the modified Young's equation, but each parameter in it has an unambiguous physical interpretation. An approximate form of this model, linearized in the inverse droplet radius, yields a quantity that is mathematically analogous to what is conventionally called "line tension", but unpacked at the molecular level, showing that it is unrelated to a restoring force associated with the curvature of the macroscopic three-phase contact line.

The surfaces of colloidal nanocrystals are frequently passivated with carboxylate ligands that exert significant effects on their optoelectronic properties and chemical stability. The binding geometries of these ligands are often experimentally investigated using vibrational spectroscopy, but the interpretation of the IR signal is usually not trivial. Here, using machine-learning (ML) algorithms trained on DFT data, we simulate an IR spectrum of a lead-rich PbS nanocrystal passivated with butyrate ligands. We obtain good agreement with the experimental signal and demonstrate that the observed line shape stems from a very wide range of "tilted-bridge"-type geometries and does not indicate the coexistence of "bridging" and "chelating" binding modes as has been previously assumed. This work illustrates the limitations of empirical spectrum assignment and demonstrates the effectiveness of ML-driven molecular dynamics simulations in reproducing the IR spectra of nanoscopic systems.

The morphology of semiconducting polymer thin films is known to have a profound effect on their opto-electronic properties. Although considerable efforts have been made to control and understand the processes which influence the structures of these systems, it remains largely unclear what physical factors determine the arrangement of polymer chains in spin-cast films. Here, we investigate the role that the liquid-vapor interfaces in chlorobenzene solutions of poly(3-hexylthiophene) [P3HT] play in the conformational geometries adopted by the polymers. Using all-atom molecular dynamics (MD), and supported by toy-model simulations, we demonstrate that, with increasing concentration, P3HT oligomers in solution exhibit a strong propensity for the liquid-vapor interface. Due to the differential solubility of the backbone and side chains of the oligomers, in the vicinity of this interface, hexyl chains and the thiophene rings, have a clear orientational preference with respect to the liquid surface. At high concentrations, we additionally establish a substantial degree of inter-oligomer alignment and thiophene ring stacking near the interface. Our results broadly concur with the limited existing experimental evidence and we suggest that the interfacial structure can act as a template for film structure. We argue that the differences in solvent affinity of the side chain and backbone moieties are the driving force for the anisotropic orientations of the polymers near the interface. This finer grained description contrasts with the usual monolithic characterization of polymer units. Since this phenomenon can be controlled by concurrent chemical design and the choice of solvents, this work establishes a fabrication principle which may be useful to develop more highly functional polymer films.

Materials that undergo singlet fission are of interest for their use in light-harvesting, photocatalysis, and quantum information science, but their ability to undergo fission can be sensitive to local variations in molecular packing. Herein we employ transient absorption microscopy, molecular dynamics simulations, and electronic structure calculations to interrogate how structures found at the edges of orthorhombic rubrene crystals impact singlet fission. Within a micrometer-scale spatial region at the edges of rubrene crystals, we find that the rate of singlet fission increases nearly 4-fold. This observation is consistent with formation of a region at crystal edges with reduced order that accelerates singlet fission by disrupting the symmetry found in rubrene's orthorhombic crystal structure. Our work demonstrates that structural distortions of singlet fission materials can be used to control fission in time and in space, potentially offering a means of controlling this process in light harvesting and quantum information applications.

Solvated electrons are powerful reducing agents capable of driving some of the most energetically expensive reduction reactions. Their generation under mild and sustainable conditions remains challenging though. Using near-ultraviolet irradiation under low-intensity one-photon conditions coupled with electrochemical and optical detection, we show that the yield of solvated electrons in water is increased more than 10 times for nanoparticle-decorated electrodes compared to smooth silver electrodes. Based on the simulations of electric fields and hot carrier distributions, we determine that hot electrons generated by plasmons are injected into water to form solvated electrons. Both yield enhancement and hot carrier production spectrally follow the plasmonic near-field. The ability to enhance solvated electron yields in a controlled manner by tailoring nanoparticle plasmons opens up a promising strategy for exploiting solvated electrons in chemical reactions.

Semiconducting nanocrystals passivated with organic ligands have emerged as a powerful platform for light harvesting, light-driven chemical reactions, and sensing. Due to their complexity and size, little structural information is available from experiments, making these systems challenging to model computationally. Here, we develop a machine-learned force field trained on DFT data and use it to investigate the surface chemistry of a PbS nanocrystal interfaced with acetate ligands. In doing so, we go beyond considering individual local minimum energy geometries and, importantly, circumvent a precarious issue associated with the assumption of a single assigned atomic partial charge for each element in a nanocrystal, independent of its structural position. We demonstrate that the carboxylate ligands passivate the metal-rich surfaces by adopting a very wide range of “tilted-bridge” and “bridge” geometries and investigate the corresponding ligand IR spectrum. This work illustrates the potential of machine-learned force fields to transform computational modeling of these materials.

Understanding molecular transport in polyelectrolyte brushes (PEBs) is crucial for applications such as separations, drug delivery, anti-fouling, and biosensors, where structural features of the polymer control intermolecular interactions. The complex structure and local heterogeneity of PEBs, while theoretically predicted, are not easily accessed with conventional experimental methods. In this work, we use 3D single-molecule tracking to understand transport behavior within a cationic poly(2-(N,N-dimethylamino)ethyl acrylate) (PDMAEA) brush using an anionic dye, Alexa Fluor 546, as the probe. The analysis is done by a parallelized, unbiased 3D tracking algorithm. Our results explicitly demonstrate that spatial heterogeneity within the brush manifests as heterogeneity of single-molecule displacements. Two distinct populations of probe motion are identified, with anticorrelated axial and lateral transport confinement, which we believe to correspond to intra- vs inter-chain probe motion.

ADVERTISEMENT RETURN TO ISSUEEditorialNEXTThe Age of Direct Chemical DynamicsJiali GaoJiali GaoShenzhen Bay LaboratoryUniversity of MinnesotaMore by Jiali Gaohttps://orcid.org/0000-0003-0106-7154 and Peter J. RosskyPeter J. RosskyRice UniversityMore by Peter J. Rosskyhttps://orcid.org/0000-0002-0461-4179Cite this: Acc. Chem. Res. 2022, 55, 4, 471–472Publication Date (Web):February 15, 2022Publication History Received11 January 2022Published online15 February 2022Published inissue 15 February 2022https://pubs.acs.org/doi/10.1021/acs.accounts.2c00021https://doi.org/10.1021/acs.accounts.2c00021editorialACS PublicationsCopyright © Published 2022 by American Chemical Society. This publication is available under these Terms of Use. Request reuse permissions This publication is free to access through this site. Learn MoreArticle Views2056Altmetric-Citations-LEARN ABOUT THESE METRICSArticle Views are the COUNTER-compliant sum of full text article downloads since November 2008 (both PDF and HTML) across all institutions and individuals. These metrics are regularly updated to reflect usage leading up to the last few days.Citations are the number of other articles citing this article, calculated by Crossref and updated daily. Find more information about Crossref citation counts.The Altmetric Attention Score is a quantitative measure of the attention that a research article has received online. Clicking on the donut icon will load a page at altmetric.com with additional details about the score and the social media presence for the given article. Find more information on the Altmetric Attention Score and how the score is calculated. Share Add toView InAdd Full Text with ReferenceAdd Description ExportRISCitationCitation and abstractCitation and referencesMore Options Share onFacebookTwitterWechatLinked InRedditEmail PDF (497 KB) Get e-AlertscloseSUBJECTS:Chemical reactions,Direct dynamics,Excited states,Molecular dynamics,Molecular dynamics simulations Get e-Alerts

Semiconductor nanocrystals (NCs) interfaced with molecular ligands that function as charge and energy acceptors are an emerging platform for the design of light-harvesting, photon-upconverting, and photocatalytic materials. However, NC systems explored for these applications often feature high concentrations of bound acceptor ligands, which can lead to ligand-ligand interactions that may alter each system's ability to undergo charge and energy transfer. Here, we demonstrate that aggregation of acceptor ligands impacts the rate of photoinduced NC-to-ligand charge transfer between lead(II) sulfide (PbS) NCs and perylenediimide (PDI) electron acceptors. As the concentration of PDI acceptors is increased, we find the average electron transfer rate from PbS to PDI ligands decreases by nearly an order of magnitude. The electron transfer rate slowdown with increasing PDI concentration correlates strongly with the appearance of PDI aggregates in steady-state absorption spectra. Electronic structure calculations and molecular dynamics (MD) simulations suggest PDI aggregation slows the rate of electron transfer by reducing orbital overlap between PbS charge donors and PDI charge acceptors. While we find aggregation slows electron transfer in this system, the computational models we employ predict ligand aggregation could also be used to speed electron transfer by producing delocalized states that exhibit improved NC-molecule electronic coupling and energy alignment with NC conduction band states. Our results demonstrate that ligand aggregation can alter rates of photoinduced electron transfer between NCs and organic acceptor ligands and should be considered when designing hybrid NC:molecule systems for charge separation.

Electron microscopy is often required to correlate the size and shape of plasmonic nanoparticles with their optical properties. Eliminating the need for electron microscopy is one crucial step toward in situ sensing applications, especially for complicated sample conditions such as during irreversible chemical reactions or when particles are embedded in a matrix. Here, we show that a machine learning decision tree can accurately predict gold nanorod dimensions over a wide range of sizes. The model is trained by using similar to 450 nanorod geometries and corresponding scattering spectra obtained from finite-difference time-domain simulations. We test the model using a set of experimental spectra and sizes obtained from correlated scanning electron microscopy images, resulting in predictions of the dimensions of gold nanorods within similar to 10% of their true values (root-mean-squared percentage error) over a large range of sizes. Analysis of the decision tree structure reveals that a relationship with resonance energy and line width of the localized surface plasmon resonance is sufficient to predict nanorod dimensions, notably outperforming more complicated models. Our findings illustrate the advantages of using machine learning models to infer single particle structural features from their optical spectra.

Conformational changes of antibodies and other biologics can decrease the effectiveness of pharmaceutical separations. Hence, a detailed mechanistic picture of antibody-stationary phase interactions that occur during ion-exchange chromatography (IEX) can provide critical insights. This work examines antibody conformational changes and how they perturb antibody motion and affect ensemble elution profiles. We combine IEX, three-dimensional single-protein tracking, and circular dichroism spectroscopy to investigate conformational changes of a model antibody, immunoglobulin G (IgG), as it interacts with the stationary phase as a function of salt conditions. The results indicate that the absence of salt enhances electrostatic attraction between IgG and the stationary phase, promotes surface-induced unfolding, slows IgG motion, and decreases elution from the column. Our results reveal previously unreported details of antibody structural changes and their influence on macroscale elution profiles.

ADVERTISEMENT RETURN TO ISSUEPREVSpecial Issue Prefac...Special Issue PrefaceNEXTAn Autobiographical History of Peter J. RosskyPeter J. RosskyPeter J. RosskyMore by Peter J. Rosskyhttp://orcid.org/0000-0002-0461-4179Cite this: J. Phys. Chem. B 2020, 124, 47, 10594–10597Publication Date (Web):November 25, 2020Publication History Published online25 November 2020Published inissue 25 November 2020https://pubs.acs.org/doi/10.1021/acs.jpcb.0c09811https://doi.org/10.1021/acs.jpcb.0c09811introductionACS PublicationsCopyright © 2020 American Chemical Society. This publication is available under these Terms of Use. Request reuse permissions This publication is free to access through this site. Learn MoreArticle Views1460Altmetric-Citations-LEARN ABOUT THESE METRICSArticle Views are the COUNTER-compliant sum of full text article downloads since November 2008 (both PDF and HTML) across all institutions and individuals. These metrics are regularly updated to reflect usage leading up to the last few days.Citations are the number of other articles citing this article, calculated by Crossref and updated daily. Find more information about Crossref citation counts.The Altmetric Attention Score is a quantitative measure of the attention that a research article has received online. Clicking on the donut icon will load a page at altmetric.com with additional details about the score and the social media presence for the given article. Find more information on the Altmetric Attention Score and how the score is calculated. Share Add toView InAdd Full Text with ReferenceAdd Description ExportRISCitationCitation and abstractCitation and referencesMore Options Share onFacebookTwitterWechatLinked InRedditEmail PDF (435 KB) Get e-AlertscloseSupporting Info (2)»Supporting Information Supporting Information SUBJECTS:Chemical engineering and industrial chemistry,Quality management,Solution chemistry,Students Get e-Alerts

Coarse-grained modeling of conjugated polymers has become an increasingly popular route to investigate the physics of organic optoelectronic materials. While ultraviolet (UV)-vis spectroscopy remains one of the key experimental methods for the interrogation of these materials, a rigorous bridge between simulated coarse-grained structures and spectroscopy has not been established. Here, we address this challenge by developing a method that can predict spectra of conjugated polymers directly from coarse-grained representations while avoiding repetitive procedures such as ad hoc back-mapping from coarse-grained to atomistic representations followed by spectral computation using quantum chemistry. Our approach is based on a generative deep-learning model: the long-short-term memory recurrent neural network (LSTM-RNN). The latter is suggested by the apparent similarity between natural languages and the mathematical structure of perturbative expansions of, in our case, excited-state energies perturbed by conformational fluctuations. We also use this model to explore the level of sensitivity of spectra to the coarse-grained representation back-mapping protocol. Our approach presents a tool uniquely suited for improving postsimulation analysis protocols, as well as, potentially, for including spectral data as input in the refinement of coarse-grained potentials.