Research on discharge performance prediction of magnesium anode materials assisted by graph neural networks

Abstract

In Mg anode materials design, the second phase plays a significant role in influencing the alloy's properties. This paper proposes a fusion model based on graph neural networks (GNNs) aimed at optimizing alloy design by considering the impact of the second phase in the alloy. We calculate the types and mass percentages of the thermodynamically stable second phases in the alloy. A GNN-based fusion model is then used to capture the interactions between the second phases within the alloy to predict the voltage of the magnesium anode. Our model outperforms the traditional MLP model. The average error is reduced from 0.096 to 0.075 V, representing a 21.8% decrease in error. We use Bayesian optimization and Pareto front analysis for multi-objective optimization. The Mg-2Sr alloy, which has high voltage values at 5, 10 and 20 mA cm⁻², is selected for experimental validation. The GNN fusion model improves the accuracy of the model, providing a more scientific basis for the design and optimization of magnesium alloy anodes.

Keywords

cell voltages / graph neural networks / machine learning / magnesium alloys / multi-objective optimizations

References

[1] Cheng F, Chen J. Metal–air batteries: from oxygen reduction electrochemistry to cathode catalysts. Chem Soc Rev. 2012;41(6):2172-2192.

[2] Wang N, Wang R, Peng C, Peng B, Feng Y, Hu C. Discharge behaviour of Mg-Al-Pb and Mg-Al-Pb-In alloys as anodes for Mg-air battery. Electrochimica Acta. 2014;149:193-205.

[3] Zhao J, Yu K, Hu Y, et al. Discharge behavior of Mg–4 wt% Ga–2 wt% Hg alloy as anode for seawater activated battery. Electrochimica Acta. 2011;56(24):8224-8231.

[4] Li W, Li C, Zhou C, Ma H, Chen J. Metallic magnesium nano/mesoscale structures: their shape-controlled preparation and Mg/air battery applications. Angew Chem Int Ed. 2006;45(36):6009-6012.

[5] Xiong H, Wang Z, Ye L, Wang X, Chen T, Yu H. Microstructure and discharge performance of Mg–6Al and Mg–6Al-0.5 Er alloys. Mater Chem Phys. 2022;280:125822.

[6] Deng M, Höche D, Lamaka SV, Snihirova D, Zheludkevich ML. Mg-Ca binary alloys as anodes for primary Mg-air batteries. J Power Sources. 2018;396:109-118.

[7] Bao J, Sha J, Li L, et al. Electrochemical properties and discharge performance of Mg-3Sn-xCa alloy as a novel anode for Mg-air battery. J Alloys Compd. 2023;934:167849.

[8] Yuan S, Lu H, Sun Z, Fan L, Zhu X, Zhang W. Electrochemical performance of Mg-3Al modified with Ga, in and Sn as anodes for Mg-air battery. J Electrochem Soc. 2016;163(7):A1181-A1187.

[9] White A. The materials genome initiative: one year on. MRS Bull. 2012;37(8):715-716.

[10] Ward L, Agrawal A, Choudhary A, Wolverton C. A general-purpose machine learning framework for predicting properties of inorganic materials. npj Comput Mater. 2016;2(1):1-7.

[11] Pei Z, Yin J. Machine learning as a contributor to physics: understanding Mg alloys. Mater Des. 2019;172:107759.

[12] Wen YF, Cai CZ, Liu XH, Pei JF, Zhu XJ, Xiao TT. Corrosion rate prediction of 3C steel under different seawater environment by using support vector regression. Corros Sci. 2009;51(2):349-355.

[13] Pan S, Yu J, Han J, et al. Customized development of promising Cu-Cr-Ni-Co-Si alloys enabled by integrated machine learning and characterization. Acta Mater. 2023;243:118484.

[14] Zhao Q, Yang H, Liu J, Zhou H, Wang H, Yang W. Machine learning-assisted discovery of strong and conductive Cu alloys: data mining from discarded experiments and physical features. Mater Des. 2021;197:109248.

[15] Chen Y, Tian Y, Zhou Y, et al. Machine learning assisted multi-objective optimization for materials processing parameters: a case study in Mg alloy. J Alloys Compd. 2020;844:156159.

[16] Xiang S, Chen X, Fan Z, Chen T, Lian X. A deep learning-aided prediction approach for creep rupture time of Fe–Cr–Ni heat-resistant alloys by integrating textual and visual features. J Mater Res Technol. 2022;18:268-281.

[17] Wang XJ, Xu DK, Wu RZ, et al. What is going on in magnesium alloys? J Mater Sci Technol. 2018;34(2):245-247.

[18] Bao J, Sha J, Li L, et al. Electrochemical properties and discharge performance of Mg-3Sn-xCa alloy as a novel anode for Mg-air battery. J Alloys Compd. 2023;934:167849.

[19] Song Y, Ma J, Li Y, Wang J, Qin C, Stock HR. Effects of second phases in anode materials on discharge performance of Mg-air batteries. Ionics. 2019;25(12):5899-5906.

[20] Yuasa M, Huang X, Suzuki K, Mabuchi M, Chino Y. Discharge properties of Mg–Al–Mn–Ca and Mg–Al–Mn alloys as anode materials for primary magnesium–air batteries. J Power Sources. 2015;297:449-456.

[21] Liu X, Xue J, Zhang D. Electrochemical behaviors and discharge performance of the as-extruded Mg-1.5 wt% Ca alloys as anode for Mg-air battery. J Alloys Compd. 2019;790:822-828.

[22] Ma Y, Li N, Li D, Zhang M, Huang X. Performance of Mg–14Li–1Al–0.1 Ce as anode for Mg-air battery. J Power Sources. 2011;196(4):2346-2350.

[23] Liu Y, Ren H, Hu WC, et al. First-principles calculations of strengthening compounds in magnesium alloy: a general review. J Mater Sci Technol. 2016;32(12):1222-1231.

[24] Wang WY, Zhang S, Li G, et al. Artificial intelligence enabled smart design and manufacturing of advanced materials: the endless frontier in AI+ era. Mat Genome Eng Adv. 2024;2(3):e56.

[25] Ghorbani M, Boley M, Nakashima PNH, Birbilis N. A machine learning approach for accelerated design of magnesium alloys. Part A: alloy data and property space. J Magnesium Alloys. 2023;11(10):3620-3633.

[26] Ward L, Liu R, Krishna A, et al. Including crystal structure attributes in machine learning models of formation energies via Voronoi tessellations. Phys Rev B. 2017;96(2):024104.

[27] Xie T, Grossman JC. Crystal graph convolutional neural networks for an accurate and interpretable prediction of material properties. Phys Rev Lett. 2018;120(14):145301.

[28] Xu G, Xue Y, Geng X, Hou X, Xu J. A machine learning-based crystal graph network and its application in development of functional materials. Mat Genome Eng Adv. 2024;2(3):e38.

[29] Chen C, Ye W, Zuo Y, Zheng C, Ong SP. Graph networks as a universal machine learning framework for molecules and crystals. Chem Mater. 2019;31(9):3564-3572.

[30] Schütt KT, Sauceda HE, Kindermans PJ, Tkatchenko A, Müller KR. Schnet–a deep learning architecture for molecules and materials. J Chem Phys. 2018;148(24).

[31] Veličković P, Cucurull G, Casanova A, Romero A, Liò P, Bengio Y. Graph Attention Networks; 2017. arXiv preprint arXiv:1710.10903.

[32] Shi B, Lookman T, Xue D. Multi-objective optimization and its application in materials science. Mat Genome Eng Adv. 2023;1(2):e14.

[33] Jain A, Ong SP, Hautier G, et al. Commentary: the Materials Project: a materials genome approach to accelerating materials innovation. APL Mater. 2013;1(1).

[34] Louis SY, Zhao Y, Nasiri A, et al. Graph convolutional neural networks with global attention for improved materials property prediction. Phys Chem Chem Phys. 2020;22(32):18141-18148.

[35] Solomou A, Zhao G, Boluki S, et al. Multi-objective Bayesian materials discovery: application on the discovery of precipitation strengthened NiTi shape memory alloys through micromechanical modeling. Mater Des. 2018;160:810-827.

[36] Deng M, Wang L, Vaghefinazari B, et al. High-energy and durable aqueous magnesium batteries: recent advances and perspectives. Energy Storage Mater. 2021;43:238-247.

RIGHTS & PERMISSIONS

2025 The Author(s). Materials Genome Engineering Advances published by Wiley-VCH GmbH on behalf of University of Science and Technology Beijing.