Predicting the phase diagram of titanium dioxide with random search and pattern recognition

Reinhardt, Aleks; Pickard, Chris J.; Cheng, Bingqing

doi:10.1039/d0cp02513e

Cited by 29 publications

(28 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A clustering algorithm (like the k-means algorithm) could identify those conformers by grouping the data based on common patterns. 373 , 374 Alternatively, a projection technique could reveal a low-dimensional representation of the data set. 375 Often data is represented in high dimension, despite being intrinsically low-dimensional.…”

Section: Machine Learning Tutorial and Intersections With Chemistrymentioning

confidence: 99%

Combining Machine Learning and Computational Chemistry for Predictive Insights Into Chemical Systems

et al. 2021

Self Cite

View full text Add to dashboard Cite

Machine learning models are poised to make a transformative impact on chemical sciences by dramatically accelerating computational algorithms and amplifying insights available from computational chemistry methods. However, achieving this requires a confluence and coaction of expertise in computer science and physical sciences. This Review is written for new and experienced researchers working at the intersection of both fields. We first provide concise tutorials of computational chemistry and machine learning methods, showing how insights involving both can be achieved. We follow with a critical review of noteworthy applications that demonstrate how computational chemistry and machine learning can be used together to provide insightful (and useful) predictions in molecular and materials modeling, retrosyntheses, catalysis, and drug design.

show abstract

Section: Machine Learning Tutorial and Intersections With Chemistrymentioning

confidence: 99%

Combining Machine Learning and Computational Chemistry for Predictive Insights Into Chemical Systems

et al. 2021

Self Cite

View full text Add to dashboard Cite

show abstract

“…It is worth mentioning that the accuracy of (semi-)local DFT functionals for TiO 2 at low pressure is subject to debate 35 , since for example rutile is not predicted to be a stable phase, although DMC shows good agreement in the ranking of static-lattice energies of phases with a number of DFT functionals 36 . In addition, three functionals, LDA, PBE and PBEsol, all give consistent results regarding the ranking of stabilities for the known polymorphs of TiO 2 30 . We employed the CASTEP ab initio simulation package 37 , and full details of the DFT set-ups and configurations can be found in the input files supplied in the Supplementary Information.…”

Section: A Dft Referencementioning

confidence: 54%

“…For problems that demand large system sizes and long simulation times, many studies utilise empirical potentials such as the Matsui-Akaogi (MA) potential 28 , which has been shown to perform poorly compared to DFT, for example the incorrect prediction of the ordering of the stable phases 29 , which contribute to the poor prediction of the TiO 2 pressure-temperature phase diagram compared to DFT or experiments 30 . Recently, some machine learning potentials of titanium dioxide have also been developed 31 .…”

Section: Introductionmentioning

confidence: 99%

High-pressure phase behaviors of titanium dioxide revealed by a $Δ$-learning potential

Lee,

Pickard,

Cheng

2021

Preprint

Self Cite

View full text Add to dashboard Cite

Titanium dioxide has been extensively studied in the rutile or anatase phases, while its high-pressure phases are less well understood, despite that many are thought to have interesting optical, mechanical and electrochemical properties. First-principles methods such as density functional theory (DFT) are often used to compute the enthalpies of TiO 2 phases at 0 K, but they are expensive and thus impractical for long time-scale and large system-size simulations at finite temperatures. On the other hand, cheap empirical potentials fail to capture the relative stablities of the various polymorphs. To model the thermodynamic behaviors of ambient and high-pressure phases of TiO 2 , we design an empirical model as a baseline, and then train a machine learning potential based on the difference between the DFT data and the empirical model. This so-called ∆-learning potential contains long-range electrostatic interactions, and predicts the 0 K enthalpies of stable TiO 2 phases that are in good agreement with DFT. We construct a pressuretemperature phase diagram of TiO 2 in the range 0 < P < 70 GPa and 100 < T < 1500 K. We then simulate dynamic phase transition processes, by compressing anatase at different temperatures. At 300 K, we observe predominantly anatase-to-baddeleyite transformation at about 20 GPa, via a martensitic two-step mechanism with highly ordered and collective atomic motion. At 2000 K, anatase can transform into cotunnite around 45-55 GPa in a thermally-activated and probabilistic manner, accompanied by diffusive movement of oxygen atoms. The pressures computed for these transitions show good agreement with experiments. Our results shed light on how to synthesize and stabilize highpressure TiO 2 phases, and our method is generally applicable to other functional materials with multiple polymorphs.

show abstract

“…Kernel PCA has been successfully used for visualization and exploration of materials databases, 123 for identifying new materials candidates, 92 and to predict phase stability of crystal structures. 124 …”

Section: Dimensionality Reduction and Manifold Learningmentioning

confidence: 99%

Unsupervised Learning Methods for Molecular Simulation Data

et al. 2021

View full text Add to dashboard Cite

Unsupervised learning is becoming an essential tool to analyze the increasingly large amounts of data produced by atomistic and molecular simulations, in material science, solid state physics, biophysics, and biochemistry. In this Review, we provide a comprehensive overview of the methods of unsupervised learning that have been most commonly used to investigate simulation data and indicate likely directions for further developments in the field. In particular, we discuss feature representation of molecular systems and present state-of-the-art algorithms of dimensionality reduction , density estimation , and clustering , and kinetic models . We divide our discussion into self-contained sections, each discussing a specific method. In each section, we briefly touch upon the mathematical and algorithmic foundations of the method, highlight its strengths and limitations, and describe the specific ways in which it has been used-or can be used-to analyze molecular simulation data.

show abstract

Predicting the phase diagram of titanium dioxide with random search and pattern recognition

Abstract: Predicting phase stabilities of crystal polymorphs is central to computational materials science and chemistry.

Cited by 29 publications

References 57 publications

Combining Machine Learning and Computational Chemistry for Predictive Insights Into Chemical Systems

Combining Machine Learning and Computational Chemistry for Predictive Insights Into Chemical Systems

High-pressure phase behaviors of titanium dioxide revealed by a $Δ$-learning potential

Unsupervised Learning Methods for Molecular Simulation Data

Contact Info

Product

Resources

About