Finding the atomic configuration with a required physical property in multi-atom structures

Effective optimization of atomic decoration in giant and superstructurally ordered crystals with machine learning

Crystals with complicated geometry are often observed with mixed chemical occupancy among Wyckoff sites, presenting a unique challenge for accurate atomic modeling. Similar systems possessing exact occupancy on all the sites can exhibit superstructural ordering, dramatically inflating the unit cell size. In this work, a crystal graph convolutional neural network (CGCNN) is used to predict optimal atomic decorations on fixed crystalline geometries. This is achieved with a site permutation search (SPS) optimization algorithm based on Monte Carlo moves combined with simulated annealing and basin-hopping techniques. Our approach relies on the evidence that, for a given chemical composition, a CGCNN estimates the correct energetic ordering of different atomic decorations, as predicted by electronic structure calculations. This provides a suitable energy landscape that can be optimized according to site occupation, allowing the prediction of chemical decoration in crystals exhibiting mixed or disordered occupancy, or superstructural ordering. Verification of the procedure is carried out on several known compounds, including the superstructurally ordered clathrate compound Rb8Ga27Sb16 and vacancy-ordered perovskite Cs2SnI6, neither of which was previously seen during the neural network training. In addition, the critical temperature of an order-disorder phase transition in solid solution CuZn is probed with our SPS routines by sampling site configuration trajectories in the canonical ensemble. This strategy provides an accurate method for determining favorable decoration in complex crystals and analyzing site occupation at unprecedented speed and scale.

Predicting Phase Stability at Interfaces

We present the RAFFLE methodology for structural prediction of the interface between two materials and demonstrate its effectiveness by applying it to MgO encapsulated by two layers of graphene. To address the challenge of interface structure prediction, our methodology combines physical insights derived from morphological features observed in related systems with an iterative machine learning technique. This employs physical-based methods, including void-filling and n-body distribution functions to predict interface structures. For the carbon-MgO encapsulated system, we have shown the rocksalt and hexagonal phases of MgO to be the two most energetically stable in the few-layer regime. We demonstrate that monolayer rocksalt is heavily stabilized by interfacing with graphene, becoming more energetically favorable than the graphenelike monolayer hexagonal MgO. The RAFFLE methodology provides valuable insights into interface behavior, and a route to finding new materials at interfaces.

Optimization of General Molecular Properties in the Equilibrium Geometry Using Quantum Alchemy: An Inverse Molecular Design Approach

Inverse molecular design allows the optimization of molecules in chemical space and is promising for accelerating the development of functional molecules and materials. To design realistic molecules, it is necessary to consider geometric stability during optimization. In this work, we introduce an inverse design method that optimizes molecular properties by changing the chemical composition in the equilibrium geometry. The optimization algorithm of our recently developed molecular design method has been modified to allow molecular design for general properties at a low computational cost. The proposed method is based on quantum alchemy and does not require empirical data. We demonstrate the applicability and limitations of the present method in the optimization of the electric dipole moment and atomization energy in small chemical spaces for (BF, CO), (N₂, CO), BN-doped benzene derivatives, and BN-doped butane derivatives. It was found that the optimality criteria scheme adopted for updating the molecular species yields faster convergence of the optimization and requires a less computational cost. Moreover, we also investigate and discuss the applicability of quantum alchemy to the electric dipole moment.

Exploring chemical space with discrete, gradient, and hybrid optimization methods

Discrete, gradient, and hybrid optimization methods are applied to the challenge of discovering molecules with optimized properties. The cost and performance of the approaches were studied using a tight-binding model to maximize the static first electronic hyperpolarizability of molecules. Our analysis shows that discrete branch and bound methods provide robust strategies for inverse chemical design involving diverse chemical structures. Based on the linear combination of atomic potentials, a hybrid discrete-gradient optimization strategy significantly improves the performance of the gradient methods. The hybrid method performs better than dead-end elimination and competes with branch and bound and genetic algorithms. The branch and bound methods for these model Hamiltonians are more cost effective than genetic algorithms for moderate-sized molecular optimization.

Band-gap design of quaternary (In,Ga)(As,Sb) semiconductors via the inverse-band-structure approach

Quaternary systems illustrated by (Ga,In)(As,Sb) manifest a huge configurational space, offering in principle the possibility of designing structures that are lattice matched to a given substrate and have given electronic properties (e.g., band gap) at more than one composition. Such specific configurations were however, hitherto, unidentified. We show here that using a genetic-algorithm search with a pseudopotential "Inverse-band-structure (IBS) approach it is possible to identify those configurations that are naturally lattice matching (to GaSb) and have a specific band gap (310 meV) at more than one composition. This is done by deviating from randomness, allowing the IBS to find a partial atomic ordering. This illustrates multitarget design of the electronic structure of multinary systems.