Targeting Productive Composition Space through Machine-Learning-Directed Inorganic Synthesis

Lotfi, Sogol; Zhang, Ziyan; Viswanathan, Gayatri; Fortenberry, Kaitlyn; Tehrani, Aria Mansouri; Brgoch, Jakoah

doi:10.1016/j.matt.2020.05.002

Cited by 18 publications

(18 citation statements)

References 43 publications

(50 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The sparse number of compounds identified in the PCD also suggests it is improbable to find entirely new superhard materials using this strategy. We address this inadequacy by merging this hardness model with our recently developed formation energy and convex hull prediction tool [ 33 ] to identify more than ten previously unreported high hardness compounds. This work proves that using ensemble learning to predict load‐dependent hardness can potentially provide the next big step in the search for new superhard materials.…”

Section: Figurementioning

confidence: 99%

“…This allows the model to make general predictions of hardness for any chemical composition without knowing the crystal structure. Merging this idea with our recently developed convex‐hull phase diagram analysis, [ 33 ] we can identify regions of composition space where new, unreported compounds with a hardness >40 GPa are likely to reside.…”

Section: Figurementioning

confidence: 99%

“…Our previously created machine learning‐based phase diagram model has proven extremely useful for discovering intermetallic materials. [ 33 ] Applying the method here produces multiple compounds on the convex hull, shown by the black squares on each plot. These represent the most energetically favorable compositions that should be observed at equilibrium (at 0 K).…”

Section: Figurementioning

confidence: 99%

See 2 more Smart Citations

Finding the Next Superhard Material through Ensemble Learning

et al. 2020

Self Cite

View full text Add to dashboard Cite

An ensemble machine‐learning method is demonstrated to be capable of finding superhard materials by directly predicting the load‐dependent Vickers hardness based only on the chemical composition. A total of 1062 experimentally measured load‐dependent Vickers hardness data are extracted from the literature and used to train a supervised machine‐learning algorithm utilizing boosting, achieving excellent accuracy (R2 = 0.97). This new model is then tested by synthesizing and measuring the load‐dependent hardness of several unreported disilicides and analyzing the predicted hardness of several classic superhard materials. The trained ensemble method is then employed to screen for superhard materials by examining more than 66 000 compounds in crystal structure databases, which show that 68 known materials have a Vickers hardness ≥40 GPa at 0.5 N (applied force) and only 10 exceed this mark at 5 N. The hardness model is then combined with the data‐driven phase diagram generation tool to expand the limited number of reported high hardness compounds. Eleven ternary borocarbide phase spaces are studied, and more than ten thermodynamically favorable compositions with a hardness above 40 GPa (at 0.5 N) are identified, proving this ensemble model's ability to find previously unknown materials with outstanding mechanical properties.

show abstract

Section: Figurementioning

confidence: 99%

Section: Figurementioning

confidence: 99%

Section: Figurementioning

confidence: 99%

See 1 more Smart Citation

Finding the Next Superhard Material through Ensemble Learning

et al. 2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…This way, the discovery of materials can be refocused from traditional "exploratory" into a more targeted approach. [15][16][17][18][19][20][21] An accurate, rapid, and reliable method suitable for experimental verification of theoretically predicted compounds containing alkali or alkaline-earth elements is the synthetic hydride route. [22][23][24][25][26][27][28][29][30][31][32][33][34][35] This method utilizes salt-like hydrides (instead of ductile metals) that can be ball-milled into homogeneous mixture of precursors.…”

Section: Introductionmentioning

confidence: 99%

“…High‐throughput computations combined with an appropriate experimental method can promote an effective screening strategy for new phases. This way, the discovery of materials can be refocused from traditional “exploratory” into a more targeted approach [15–21] …”

Section: Introductionmentioning

confidence: 99%

How to Look for Compounds: Predictive Screening and in situ Studies in Na−Zn−Bi System

Gvozdetskyi

Wang

Xia

et al. 2021

Chemistry A European J

View full text Add to dashboard Cite

Here, the combination of theoretical computations followed by rapid experimental screening and in situ diffraction studies is demonstrated as a powerful strategy for novel compounds discovery. When applied for the previously "empty" NaÀ ZnÀ Bi system, such an approach led to four novel phases. The compositional space of this system was rapidly screened via the hydride route method and the theoretically predicted NaZnBi (PbClF type, P4/nmm) and Na 11 Zn 2 Bi 5 (Na 11 Cd 2 Sb 5 type, P � 1) phases were successfully synthesized, while other computationally generated compounds on the list were rejected. In addition, single crystal Xray diffraction studies of NaZnBi indicate minor deviations from the stoichiometric 1 : 1 : 1 molar ratio. As a result, two isostructural (PbClF type, P4/nmm) Zn-deficient phases with similar compositions, but distinctly different unit cell parameters were discovered. The vacancies on Zn sites and unit cell expansion were rationalized from bonding analysis using electronic structure calculations on stoichiometric "NaZnBi". In-situ synchrotron powder X-ray diffraction studies shed light on complex equilibria in the NaÀ ZnÀ Bi system at elevated temperatures. In particular, the high-temperature polymorph HT-Na 3 Bi (BiF 3 type, Fm � 3m) was obtained as a product of Na 11 Zn 2 Bi 5 decomposition above 611 K. HT-Na 3 Bi cannot be stabilized at room temperature by quenching, and this type of structure was earlier observed in the high-pressure polymorph HP-Na 3 Bi above 0.5 GPa. The aforementioned approach of predictive synthesis can be extended to other multinary systems. This method has been successfully applied to prepare several complex materials, including ternary borides, antimonides, arsenides, silicides, and germanides, where comprehensive control over composition is crucial for the targeted preparation

show abstract

Prediction of thermodynamic stability of actinide compounds by machine learning model

Qin,

Liu,

et al. 2024

Ceramics International

View full text Add to dashboard Cite

Targeting Productive Composition Space through Machine-Learning-Directed Inorganic Synthesis

Cited by 18 publications

References 43 publications

Finding the Next Superhard Material through Ensemble Learning

Finding the Next Superhard Material through Ensemble Learning

How to Look for Compounds: Predictive Screening and in situ Studies in Na−Zn−Bi System

Prediction of thermodynamic stability of actinide compounds by machine learning model

Contact Info

Product

Resources

About