No abstract

No abstract

Surface reaction networks involving hydrocarbons exhibit enormous complexity with thousands of species and reactions for all but the very simplest of chemistries. We present a framework for optimization under uncertainty for heterogeneous catalysis reaction networks using surrogate models that are trained on the fly. The surrogate model is constructed by teaching a Gaussian process adsorption energies based on group additivity fingerprints, combined with transition-state scaling relations and a simple classifier for determining the rate-limiting step. The surrogate model is iteratively used to predict the most important reaction step to be calculated explicitly with computationally demanding electronic structure theory. Applying these methods to the reaction of syngas on rhodium(111), we identify the most likely reaction mechanism. Propagating uncertainty throughout this process yields the likelihood that the final mechanism is complete given measurements on only a subset of the entire network and uncertainty in the underlying density functional theory calculations.

Catalyst discovery and optimization is key to solving many societal and energy challenges including solar fuel synthesis, long-term energy storage, and renewable fertilizer production. Despite considerable effort by the catalysis community to apply machine learning models to the computational catalyst discovery process, it remains an open challenge to build models that can generalize across both elemental compositions of surfaces and adsorbate identity/configurations, perhaps because datasets have been smaller in catalysis than in related fields. To address this, we developed the OC20 dataset, consisting of 1,281,040 density functional theory (DFT) relaxations (∼264,890,000 single-point evaluations) across a wide swath of materials, surfaces, and adsorbates (nitrogen, carbon, and oxygen chemistries). We supplemented this dataset with randomly perturbed structures, short timescale molecular dynamics, and electronic structure analyses. The dataset comprises three central tasks indicative of day-to-day catalyst modeling and comes with predefined train/validation/test splits to facilitate direct comparisons with future model development efforts. We applied three state-of-the-art graph neural network models (CGCNN, SchNet, and DimeNet++) to each of these tasks as baseline demonstrations for the community to build on. In almost every task, no upper limit on model size was identified, suggesting that even larger models are likely to improve on initial results. The dataset and baseline models are both provided as open resources as well as a public leader board to encourage community contributions to solve these important tasks.

Bimetallic catalysts are promising for the most difficult thermal and electrochemical reactions but modeling the many diverse active sites on polycrystalline samples is an open challenge. We present a general framework for addressing this complexity in a systematic and predictive fashion. Active sites for every stable low-index facet of a 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 sites. The activity of these sites is explored in parallel using a neural-network based surrogate model to share information between the many Density Functional Theory (DFT) relaxations, resulting in activity estimates with an order of magnitude fewer explicit DFT calculations. Sites with interesting activity were found and provide targets for follow-up calculations. This process was applied to the electrochemical reduction of CO 2 on nickel gallium bimetallics and indicated that most facets had similar activity to Ni surfaces, but a few exposed Ni sites with a very favorable on-top CO configuration. This motif emerged naturally from the predictive modeling and represents a class of intermetallic CO 2 reduction catalysts. These sites rationalize recent experimental reports of nickel gallium activity and why previous materials screens missed this exciting material. Most importantly these methods suggest that bimetallic catalysts will be discovered by studying facet reactivity and diversity of active sites more systematically.

Molecular recognition is central to the design of therapeutics, chemical catalysis and sensors. Motifs for doing so most commonly involve biological structures such as antibodies and aptamers. The key to such biological recognition consists of a folded and constrained heteropolymer that, via intra-molecular forces, forms a unique three dimensional structure that creates a binding pocket or an interface able to recognize a specific molecule. In this work, we demonstrate that synthetic heteropolymers can be alternatively constrained by adsorption around a nanoparticle, and specifically a single walled carbon nanotube (SWNT), forming a corona phase and resulting in a new form of molecular recognition of specific molecules. The phenomenon is shown to be generic, with new heteropolymer recognition complexes demonstrated for three distinct examples: Riboflavin, l-thyroxine, and estradiol, each predicted using a 2D thermodynamic model of surface interactions. The dissociation constants are continuously tunable by perturbing the chemical structure of the heteropolymer. Moreover, these complexes can be used as new types of spatial-temporal sensors based on modulation of SWNT photoemission in the near-infrared, as we show by tracking riboflavin diffusion in murine macrophages.

Nanopores that approach molecular dimensions demonstrate exotic transport behaviour and are theoretically predicted to display discontinuities in the diameter dependence of interior ion transport because of structuring of the internal fluid. No experimental study has been able to probe this diameter dependence in the 0.5-2 nm diameter regime. Here we observe a surprising fivefold enhancement of stochastic ion transport rates for single-walled carbon nanotube centered at a diameter of approximately 1.6 nm. An electrochemical transport model informed from literature simulations is used to understand the phenomenon. We also observe rates that scale with cation type as Li þ 4K þ 4Cs þ 4Na þ and pore blocking extent as K þ 4Cs þ 4Na þ 4Li þ potentially reflecting changes in hydration shell size. Across several ion types, the pore-blocking current and inverse dwell time are shown to scale linearly at low electric field. This work opens up new avenues in the study of transport effects at the nanoscale.

Developing active and stable oxygen evolution catalysts is a key to enabling various future energy technologies and the state-of-the-art catalyst is Ir-containing oxide materials. Understanding oxygen chemistry on oxide materials is significantly more complicated than studying transition metal catalysts for two reasons: the most stable surface coverage under reaction conditions is extremely important but difficult to understand without many detailed calculations, and there are many possible active sites and configurations on O* or OH* covered surfaces. We have developed an automated and high-throughput approach to solve this problem and predict OER overpotentials for arbitrary oxide surfaces. We demonstrate this for a number of previously-unstudied IrO2 and IrO3 polymorphs and their facets. We discovered that low index surfaces of IrO2 other than rutile (110) are more active than the most stable rutile (110), and we identified promising active sites of IrO2 and IrO3 that outperform rutile (110) by 0.2 V in theoretical overpotential. Based on findings from DFT calculations, we pro-vide catalyst design strategies to improve catalytic activity of Ir based catalysts and demonstrate a machine learning model capable of predicting surface coverages and site activity. This work highlights the importance of investigating unexplored chemical space to design promising catalysts. File list (2)download file view on ChemRxiv oxide_manuscript.pdf (16.03 MiB) download file view on ChemRxiv oxide_supporting_information.pdf (7.81 MiB)