Toward a Predictive Understanding of Cyanobacterial Harmful Algal Blooms through AI Integration of Physical, Chemical, and Biological Data.

Freshwater cyanobacterial harmful algal blooms (cyanoHABs) are a worldwide problem resulting in substantial economic losses, due to harm to drinking water supplies, commercial fishing, wildlife, property values, recreation, and tourism. Moreover, toxins produced from some cyanoHABs threaten human and animal health. Climate warming can affect the distribution of cyanoHABs, where rising temperatures facilitate more intense blooms and a greater distribution of cyanoHABs in inland freshwater. Nutrient runoff from adjacent watersheds is also a major driver of cyanoHAB formation. While some of the physicochemical factors behind cyanoHAB dynamics are known, there are still major gaps in our understanding of the conditions that trigger and sustain cyanoHABs over time. In this perspective, we suggest that sufficient data sets, as well as machine learning (ML) and artificial intelligence (AI) tools, are available to build a comprehensive model of cyanoHAB dynamics based on integrated environmental/climate, nutrient/water chemistry, and cyanoHAB microbiome and 'omics data to identify key factors contributing to HAB formation, intensity, and toxicity. By taking a holistic approach to the analysis of all available data, including the rapidly growing number of biological data sets, we can provide the foundational knowledge needed to address the increasing threat of cyanoHABs to the security of our water resources.

How inversion relates to disordering tendencies in complex oxides.

Complex oxides exhibit great functionality due to their varied chemistry and structures. They are quite flexible in terms of the ordering of cations, which can also impact their functional properties to a large extent. Thus, the propensity for a complex oxide to disorder is a key factor in optimizing and discovering new materials. Here, we show that the propensity to disorder cations in perovskites, pyrochlores, and spinels correlates with the energy to "invert" the structure - to directly swap the cations across the sublattices. This relatively simple metric, involving only two energetic calculations per compound, qualitatively captures disordering trends amongst compounds across these three families of materials and is quantitative in several cases. This provides a fast and robust metric to determine those complex oxides that are easy or hard to disorder, providing new avenues for quick screening of compounds for cation-ordering-dependent functionalities.

Biodegradation Studies of Polyhydroxybutyrate and Polyhydroxybutyrate-co-Polyhydroxyvalerate Films in Soil.

Due to increased environmental pressures, significant research has focused on finding suitable biodegradable plastics to replace ubiquitous petrochemical-derived polymers. Polyhydroxyalkanoates (PHAs) are a class of polymers that can be synthesized by microorganisms and are biodegradable, making them suitable candidates. The present study looks at the degradation properties of two PHA polymers: polyhydroxybutyrate (PHB) and polyhydroxybutyrate-co-polyhydroxyvalerate (PHBV; 8 wt.% valerate), in two different soil conditions: soil fully saturated with water (100% relative humidity, RH) and soil with 40% RH. The degradation was evaluated by observing the changes in appearance, chemical signatures, mechanical properties, and molecular weight of samples. Both PHB and PHBV were degraded completely after two weeks in 100% RH soil conditions and showed significant reductions in mechanical properties after just three days. The samples in 40% RH soil, however, showed minimal changes in mechanical properties, melting temperatures/crystallinity, and molecular weight over six weeks. By observing the degradation behavior for different soil conditions, these results can pave the way for identifying situations where the current use of plastics can be replaced with biodegradable alternatives.

Predicting the Mechanical Response of Polyhydroxyalkanoate Biopolymers Using Molecular Dynamics Simulations.

Polyhydroxyalkanoates (PHAs) have emerged as a promising class of biosynthesizable, biocompatible, and biodegradable polymers to replace petroleum-based plastics for addressing the global plastic pollution problem. Although PHAs offer a wide range of chemical diversity, the structure-property relationships in this class of polymers remain poorly established. In particular, the available experimental data on the mechanical properties is scarce. In this contribution, we have used molecular dynamics simulations employing a recently developed forcefield to predict chemical trends in mechanical properties of PHAs. Specifically, we make predictions for Young's modulus, and yield stress for a wide range of PHAs that exhibit varying lengths of backbone and side chains as well as different side chain functional groups. Deformation simulations were performed at six different strain rates and six different temperatures to elucidate their influence on the mechanical properties. Our results indicate that Young's modulus and yield stress decrease systematically with increase in the number of carbon atoms in the side chain as well as in the polymer backbone. In addition, we find that the mechanical properties were strongly correlated with the chemical nature of the functional group. The functional groups that enhance the interchain interactions lead to an enhancement in both the Young's modulus and yield stress. Finally, we applied the developed methodology to study composition-dependence of the mechanical properties for a selected set of binary and ternary copolymers. Overall, our work not only provides insights into rational design rules for tailoring mechanical properties in PHAs, but also opens up avenues for future high throughput atomistic simulation studies geared towards identifying functional PHA polymer candidates for targeted applications.

Machine Learning for Melting Temperature Predictions and Design in Polyhydroxyalkanoate-Based Biopolymers.

Diminishing fossil fuel-based resources and ever-growing environmental concerns related to plastic pollution demand for the development of sustainable and biodegradable polymeric material alternatives. Polyhydroxyalkanoates (PHAs) represent an eco-friendly and economically viable class of polymers with a wide range of applications. However, the chemical diversity combined with tunable physical properties available within PHAs poses discovery and optimization challenges with respect to identifying optimal application-specific chemical compositions. Here we use an example of melting temperature (Tm) prediction to demonstrate the promise of machine learning (ML)-based techniques for establishing efficient structure-property mappings in PHA-based chemical space. We employ a manually curated data set of experimentally measured Tm values for a wide range of PHA homo- and copolymer chemistries along with their reported polymer molecular weights and polydispersity indices. Descriptors based on topology, shape, and charge/polarity of specific motifs forming the polymer backbone were then used to numerically represent the polymers. The ML models developed by using available data were used to rapidly predict the property of multicomponent PHA-based copolymers, while estimating uncertainties underlying the predictions. Combined with a previously developed glass transition temperature (Tg) prediction model and an evolutionary algorithm-based search strategy, the approach is demonstrated to address polymer design with multiobjective optimization challenges.

Strong Purcell enhancement at telecom wavelengths afforded by spinel Fe3O4 nanocrystals with size-tunable plasmonic properties.

Developments in the field of nanoplasmonics have the potential to advance applications from information processing and telecommunications to light-based sensing. Traditionally, nanoscale noble metals such as gold and silver have been used to achieve the targeted enhancements in light-matter interactions that result from the presence of localized surface plasmons (LSPs). However, interest has recently shifted to intrinsically doped semiconductor nanocrystals (NCs) for their ability to display LSP resonances (LSPRs) over a much broader spectral range, including the infrared (IR). Among semiconducting plasmonic NCs, spinel metal oxides (sp-MOs) are an emerging class of materials with distinct advantages in accessing the telecommunications bands in the IR and affording useful environmental stability. Here, we report the plasmonic properties of Fe3O4 sp-MO NCs, known previously only for their magnetic functionality, and demonstrate their ability to modify the light-emission properties of telecom-emitting quantum dots (QDs). We establish the synthetic conditions for tuning sp-MO NC size, composition and doping characteristics, resulting in unprecedented tunability of electronic behavior and plasmonic response over 450 nm. In particular, with diameter-dependent variations in free-electron concentration across the Fe3O4 NC series, we introduce a strong NC size dependency onto the optical response. In addition, our observation of plasmonics-enhanced decay rates from telecom-emitting QDs reveals Purcell enhancement factors for simple plasmonic-spacer-emitter sandwich structures up to 51-fold, which are comparable to values achieved previously only for emitters in the visible range coupled with conventional noble metal NCs.

A Deep Neural Network for Accurate and Robust Prediction of the Glass Transition Temperature of Polyhydroxyalkanoate Homo- and Copolymers.

The purpose of this study was to develop a data-driven machine learning model to predict the performance properties of polyhydroxyalkanoates (PHAs), a group of biosourced polyesters featuring excellent performance, to guide future design and synthesis experiments. A deep neural network (DNN) machine learning model was built for predicting the glass transition temperature, Tg, of PHA homo- and copolymers. Molecular fingerprints were used to capture the structural and atomic information of PHA monomers. The other input variables included the molecular weight, the polydispersity index, and the percentage of each monomer in the homo- and copolymers. The results indicate that the DNN model achieves high accuracy in estimation of the glass transition temperature of PHAs. In addition, the symmetry of the DNN model is ensured by incorporating symmetry data in the training process. The DNN model achieved better performance than the support vector machine (SVD), a nonlinear ML model and least absolute shrinkage and selection operator (LASSO), a sparse linear regression model. The relative importance of factors affecting the DNN model prediction were analyzed. Sensitivity of the DNN model, including strategies to deal with missing data, were also investigated. Compared with commonly used machine learning models incorporating quantitative structure-property (QSPR) relationships, it does not require an explicit descriptor selection step but shows a comparable performance. The machine learning model framework can be readily extended to predict other properties.

Band-Edge Engineering To Eliminate Radiation-Induced Defect States in Perovskite Scintillators.

Under radiative environments such as extended hard X- or γ-rays, degradation of scintillation performance is often due to irradiation-induced defects. To overcome the effect of deleterious defects, novel design mitigation strategies are needed to identify and design more resilient materials. The potential for band-edge engineering to eliminate the effect of radiation-induced defect states in rare-earth-doped perovskite scintillators is explored, taking Ce3+-doped LuAlO3 as a model material system, using density functional theory (DFT)-based DFT + U and hybrid Heyd-Scuseria-Ernzerhof (HSE) calculations. From spin-polarized hybrid HSE calculations, the Ce3+ activator ground-state 4f position is determined to be 2.81 eV above the valence band maximum in LuAlO3. Except for the oxygen vacancies which have a deep level inside the band gap, all other radiation-induced defects in LuAlO3 have shallow defect states or are outside the band gap, that is, relatively far away from either the 5d1 or the 4f Ce3+ levels. Finally, we examine the role of Ga doping at the Al site and found that LuGaO3 has a band gap that is more than 2 eV smaller than that of LuAlO3. Specifically, the lowered conduction band edge envelopes the defect gap states, eliminating their potential impact on scintillation performance and providing direct theoretical evidence for how band-edge engineering could be applied to rare-earth-doped perovskite scintillators.

Molecular dynamics simulations for glass transition temperature predictions of polyhydroxyalkanoate biopolymers.

Polyhydroxyalkanoates (PHAs) represent an emerging class of biosynthetic and biodegradable polyesters that exhibit considerable potential to replace petroleum-based plastics towards a sustainable future. Despite the promise, general structure-property mappings within this class of polymers remain largely unexplored. An efficient exploration of this vast chemical space calls for the development and validation of predictive methods for accurate estimation of a diverse range of properties for PHA-based polymers. Towards this aim, here we present and validate the results of our molecular dynamics (MD) simulation based approach aimed at predicting glass transition temperatures (Tg) of PHA-based polymers. Since generally available and widely used polymer forcefields exhibit a relatively poor performance for Tg predictions, we have developed a new forcefield by modifying the polymer consistent force field (PCFF) via refining a selected set of torsion potentials of the polymer backbone using accurate density functional theory (DFT) computations. After carefully assessing the dependence of critical simulation parameters, such as, polymer chain length, number of polymer chains, supercell size, and thermal quenching rate used in the simulation, the applicability and transferability of the modified PCFF (mPCFF) is demonstrated by directly comparing the computed Tg predictions of various polymers with different chemistries, polymer side chain lengths and functional groups forming the polymer side chains against the respective experimentally measured values. Furthermore, the transport properties such as self-diffusion coefficient and viscosity are computationally determined and their well-known correlation with the target properties is demonstrated. Lastly, we have employed the developed approach to predict Tg values for a number of yet-to-be-synthesized PHA-based polymers with a diverse set of functional groups in the polymer side chains. The results are further rationalized by correlating the predicted Tg values with the inter-chain H-bond formation tendencies of the different side chain functional groups. This work represents an important first step towards computationally guided design of PHA-based functional polymers and opens up new directions for a systematic investigation of composition- and configuration-dependent structure-property relationships in more complex binary and ternary copolymer systems.

Machine-Learning-Based Predictive Modeling of Glass Transition Temperatures: A Case of Polyhydroxyalkanoate Homopolymers and Copolymers.

Polyhydroxyalkanoate-based polymers-being ecofriendly, biosynthesizable, and economically viable and possessing a broad range of tunable properties-are currently being actively pursued as promising alternatives for petroleum-based plastics. The vast chemical complexity accessible within this class of polymers gives rise to challenges in the rational discovery of novel polymer chemistries for specific applications. The burgeoning field of polymer informatics addresses this challenge via providing tools and strategies for accelerated property prediction and materials design via surrogate machine-learning models built on reliable past data. In this contribution, we use glass transition temperature Tg as an example target property to demonstrate promise of the data-enabled route to accelerated learning of accurate structure-property mappings in PHA-based polymers. Our analysis uses a data set of experimentally measured Tg values, polymer molecular weights, and a polydispersity index for PHA-based homo- and copolymers that was carefully assembled from the literature. A fingerprinting scheme that captures key properties based on topology, shape, and charge/polarity of specific chemical units or motifs forming the polymer backbone was devised to numerically represent the polymers. A validated statistical learning model is then developed to allow for a mapping of the polymer fingerprints onto the property space in a physically meaningful and reliable manner. Once developed, the model can not only rapidly predict the property of new PHA polymers but also provide uncertainties underlying the predictions. The model is further combined with an evolutionary-algorithm-based search strategy to efficiently identify multicomponent polymer compositions with a prespecified Tg. While the present contribution is focused specifically on Tg, the surrogate model development approach put forward here is general and can, in principle, be extended to a range of other properties.

Multifidelity Information Fusion with Machine Learning: A Case Study of Dopant Formation Energies in Hafnia.

Cost versus accuracy trade-offs are frequently encountered in materials science and engineering, where a particular property of interest can be measured/computed at different levels of accuracy or fidelity. Naturally, the most accurate measurement is also the most resource and time intensive, while the inexpensive quicker alternatives tend to be noisy. In such situations, a number of machine learning (ML) based multifidelity information fusion (MFIF) strategies can be employed to fuse information accessible from varying sources of fidelity and make predictions at the highest level of accuracy. In this work, we perform a comparative study on traditionally employed single-fidelity and three MFIF strategies, namely, (1) Δ-learning, (2) low-fidelity as a feature, and (3) multifidelity cokriging (CK) to compare their relative prediction accuracies and efficiencies for accelerated property predictions and high throughput chemical space explorations. We perform our analysis using a dopant formation energy data set for hafnia, which is a well-known high-k material and is being extensively studied for its promising ferroelectric, piezoelectric, and pyroelectric properties. We use a dopant formation energy data set of 42 dopants in hafnia-each studied in six different hafnia phases-computed at two levels of fidelities to find merits and limitations of these ML strategies. The findings of this work indicate that the MFIF based learning schemes outperform the traditional SF machine learning methods, such as Gaussian process regression and CK provides an accurate, inexpensive and flexible alternative to other MFIF strategies. While the results presented here are for the case study of hafnia, they are expected to be general. Therefore, materials discovery problems that involve huge chemical space explorations can be studied efficiently (or even made feasible in some situations) through a combination of a large number of low-fidelity and a few high-fidelity measurements/computations, in conjunction with the CK approach.

The impact of chemical order on defect transport in mixed pyrochlores.

Using temperature accelerated dynamics, an accelerated molecular dynamics method, we examine the relationship between composition and cation ordering and defect transport in the mixed pyrochlore Gd2(Ti1-xZrx)2O7, using the oxygen vacancy as a representative defect structure. We find that the nature of transport is very sensitive to the cation structure, transitioning, as a function of composition, from three-dimensional migration to two-dimensional to pseudo-one-dimensional to becoming essentially immobile before reverting back to three-dimensional as the Zr content is increased. The rates of migration are also affected by the cation structure in the various compositions. This behavior is driven by the connectivity of Ti polyhedra in the material, with more extensive networks of Ti ions leading to a greater ability of the vacancy to traverse the material. Our results indicate that the nature of transport is dictated by the cation structure of the material and that, conversely, the cation structure could be used to control transport and potentially other functionalities in mixed pyrochlores.

Computational screening of organic polymer dielectrics for novel accelerator technologies.

The use of infrared lasers to power accelerating dielectric structures is a developing area of research. Within this technology, the choice of the dielectric material forming the accelerating structures, such as the photonic band gap (PBG) structures, is dictated by a range of interrelated factors including their dielectric and optical properties, amenability to photo-polymerization, thermochemical stability and other target performance metrics of the particle accelerator. In this direction, electronic structure theory aided computational screening and design of dielectric materials can play a key role in identifying potential candidate materials with the targeted functionalities to guide experimental synthetic efforts. In an attempt to systematically understand the role of chemistry in controlling the electronic structure and dielectric properties of organic polymeric materials, here we employ empirical screening and density functional theory (DFT) computations, as a part of our multi-step hierarchal screening strategy. Our DFT based analysis focused on the bandgap, dielectric permittivity, and frequency-dependent dielectric losses due to lattice absorption as key properties to down-select promising polymer motifs. In addition to the specific application of dielectric laser acceleration, the general methodology presented here is deemed to be valuable in the design of new insulators with an attractive combination of dielectric properties.

A polymer dataset for accelerated property prediction and design.

Emerging computation- and data-driven approaches are particularly useful for rationally designing materials with targeted properties. Generally, these approaches rely on identifying structure-property relationships by learning from a dataset of sufficiently large number of relevant materials. The learned information can then be used to predict the properties of materials not already in the dataset, thus accelerating the materials design. Herein, we develop a dataset of 1,073 polymers and related materials and make it available at http://khazana.uconn.edu/. This dataset is uniformly prepared using first-principles calculations with structures obtained either from other sources or by using structure search methods. Because the immediate target of this work is to assist the design of high dielectric constant polymers, it is initially designed to include the optimized structures, atomization energies, band gaps, and dielectric constants. It will be progressively expanded by accumulating new materials and including additional properties calculated for the optimized structures provided.

Machine Learning Strategy for Accelerated Design of Polymer Dielectrics.

The ability to efficiently design new and advanced dielectric polymers is hampered by the lack of sufficient, reliable data on wide polymer chemical spaces, and the difficulty of generating such data given time and computational/experimental constraints. Here, we address the issue of accelerating polymer dielectrics design by extracting learning models from data generated by accurate state-of-the-art first principles computations for polymers occupying an important part of the chemical subspace. The polymers are 'fingerprinted' as simple, easily attainable numerical representations, which are mapped to the properties of interest using a machine learning algorithm to develop an on-demand property prediction model. Further, a genetic algorithm is utilised to optimise polymer constituent blocks in an evolutionary manner, thus directly leading to the design of polymers with given target properties. While this philosophy of learning to make instant predictions and design is demonstrated here for the example of polymer dielectrics, it is equally applicable to other classes of materials as well.

Termination chemistry-driven dislocation structure at SrTiO3/MgO heterointerfaces.

Exploiting the promise of nanocomposite oxides necessitates a detailed understanding of the dislocation structure at the interfaces, which governs diverse and technologically relevant properties. Here we report atomistic simulations demonstrating a strong dependence of the dislocation structure on the termination chemistry at the SrTiO3/MgO heterointerface. The SrO- and TiO2-terminated interfaces exhibit distinct nearest neighbour arrangements between cations and anions, leading to variations in local electrostatic interactions across the interface that ultimately dictate the dislocation structure. Networks of dislocations with different Burgers vectors and dislocation spacing characterize the two interfaces. These networks in turn influence the overall stability of and the behaviour of oxygen vacancies at the heterointerface, which will dictate vital properties such as mass transport at the interface. To date, the observed correlation between the dislocation structure and the termination chemistry at the interface has not been recognized, and offers novel avenues for fine-tuning oxide nanocomposites with enhanced functionalities.

Rational design of all organic polymer dielectrics.

To date, trial and error strategies guided by intuition have dominated the identification of materials suitable for a specific application. We are entering a data-rich, modelling-driven era where such Edisonian approaches are gradually being replaced by rational strategies, which couple predictions from advanced computational screening with targeted experimental synthesis and validation. Here, consistent with this emerging paradigm, we propose a strategy of hierarchical modelling with successive downselection stages to accelerate the identification of polymer dielectrics that have the potential to surpass 'standard' materials for a given application. Successful synthesis and testing of some of the most promising identified polymers and the measured attractive dielectric properties (which are in quantitative agreement with predictions) strongly supports the proposed approach to material selection.

Revisiting the Al/Al2O3 interface: coherent interfaces and misfit accommodation.

We study the coherent and semi-coherent Al/α-Al2O3 interfaces using molecular dynamics simulations with a mixed, metallic-ionic atomistic model. For the coherent interfaces, both Al-terminated and O-terminated nonstoichiometric interfaces have been studied and their relative stability has been established. To understand the misfit accommodation at the semi-coherent interface, a 1-dimensional (1D) misfit dislocation model and a 2-dimensional (2D) dislocation network model have been studied. For the latter case, our analysis reveals an interface dislocation structure with a network of three sets of parallel dislocations, each with pure-edge character, giving rise to a pattern of coherent and stacking-fault-like regions at the interface. Structural relaxation at elevated temperatures leads to a further change of the dislocation pattern, which can be understood in terms of a competition between the stacking fault energy and the dislocation interaction energy at the interface. Our results are expected to serve as an input for the subsequent dislocation dynamics models to understand and predict the macroscopic mechanical behavior of Al/α-Al2O3 composite heterostructures.

Accelerating materials property predictions using machine learning.

The materials discovery process can be significantly expedited and simplified if we can learn effectively from available knowledge and data. In the present contribution, we show that efficient and accurate prediction of a diverse set of properties of material systems is possible by employing machine (or statistical) learning methods trained on quantum mechanical computations in combination with the notions of chemical similarity. Using a family of one-dimensional chain systems, we present a general formalism that allows us to discover decision rules that establish a mapping between easily accessible attributes of a system and its properties. It is shown that fingerprints based on either chemo-structural (compositional and configurational information) or the electronic charge density distribution can be used to make ultra-fast, yet accurate, property predictions. Harnessing such learning paradigms extends recent efforts to systematically explore and mine vast chemical spaces, and can significantly accelerate the discovery of new application-specific materials.