Álvaro Vázquez-Mayagoitia, PhD

I am a Computational Scientist working at Argonne National Laboratory in the Computational Science Division, with a joint appointment as Senior Scientist at the University of Chicago Consortium for Advanced Science and Engineering (UChicago CASE).

My expertise is in Computational Chemistry, Materials Science, and Data Science on extreme-scale computing platforms. I have 23+ years of experience in large-scale scientific simulations on leadership-class DOE supercomputers.

Interests

I am interested in applications and development of scientific codes for chemistry and materials to solve challenging problems, particularly in cases where such problems can only be solved using leadership-class supercomputers such as Polaris and Aurora (DOE's first exascale computer).

I am constantly seeking projects in chemistry, materials science, and artificial intelligence that could leverage ALCF resources at capability scale to pursue high-impact scientific outcomes. I dedicate a significant portion of my time to supporting such projects through proposal preparation, review, and technical guidance.

I conduct research to study materials for solar cells, energy storage, CO₂ capture and conversion, and electrocatalysis. My recent work applies artificial intelligence and machine learning to predict molecular properties including redox potentials, toxicity of PFAS compounds, and thermal properties of 2D materials.

I am passionate about developing machine-learned interatomic potentials (MLIPs) for materials discovery, including amorphous and liquid metal oxides, molten salts, and refractory materials.

Research

• Development of High Performance Computing codes for atomistic simulations

  • Porting parallel codes to accelerator-based architectures (SYCL, OpenMP, CUDA)
  • Benchmarking and tuning parallel algorithms on Leadership-class DOE computers (Polaris, Aurora)
  • GPU-accelerated Monte Carlo and molecular dynamics simulations

  References:

  • Li, Z.; Shi, K.; Dubbeldam, D.; Dewing, M.; Knight, C.; Vázquez-Mayagoitia, Á.; Snurr, R. Q. Efficient Implementation of Monte Carlo Algorithms on Graphical Processing Units for Simulation of Adsorption in Porous Materials. J. Chem. Theory Comput. 2024, 20, 10649-10666.
  • Williams-Young, D. B.; Bagusetty, A.; de Jong, W. A.; Doerfler, D.; van Dam, H. J. J.; Vázquez-Mayagoitia, Á.; Windus, T. L.; Yang, C. Achieving Performance Portability in Gaussian Basis Set Density Functional Theory on Accelerator Based Architectures in NWChemEx. Parallel Comput. 2021, 108, 102829.

• AI/ML applied to chemical and materials properties

  • Machine-learned interatomic potentials (MLIPs) for amorphous/liquid metal oxides and molten salts
  • Graph neural networks for redox potential prediction of metal complexes
  • Deep transfer learning for PFAS toxicity prediction
  • Active learning workflows for automated materials discovery

  References:

  • Bhuiyan, F.; Harb, H.; Assary, R.; Vázquez-Mayagoitia, Á. Redox Potential Prediction of Fe(II)/Fe(III) Complexes: A Density Functional Theory and Graph Neural Network Approach. Digital Discovery 2026, in press.
  • Sivaraman, G.; Gallington, L.; Krishnamoorthy, A. N.; Stan, M.; Csányi, G.; Vázquez-Mayagoitia, Á.; Benmore, C. Experimentally Driven Automated Machine-Learned Interatomic Potential for a Refractory Oxide. Phys. Rev. Lett. 2021, 126, 156002.

• Molecular weak interactions and chemical reactivity

  • BSSE corrections of energies and geometries
  • Semiempirical and Double Hybrid Functionals
  • Local, global and condensed reactivity indexes using DFT

  References:

  • Medrano Sandonas, L.; Hoja, J.; Ernst, B. G.; Vazquez-Mayagoitia, A.; DiStasio Jr., R. A.; Tkatchenko, A. "Freedom of Design" in Chemical Compound Space: Towards Rational In Silico Design of Molecules with Targeted Quantum-Mechanical Properties. Chem. Sci. 2023, DOI: 10.1039/d3sc03598k.
  • Kocabaş, T.; Keçeli, M.; Vázquez-Mayagoitia, Á.; Sevik, C. Gaussian Approximation Potentials for Accurate Thermal Properties of Two-Dimensional Materials. Nanoscale 2023, 15, 8772-8780.

• Dye-sensitized solar cells and photovoltaic materials

  • Computational design of organic dyes and co-sensitizers
  • Hybrid organic-inorganic perovskites for tunable semiconductors
  • In-silico device performance prediction

  References:

  • Devereux, L. R.; Vázquez-Mayagoitia, Á.; Sternberg, M. G.; Cole, J. M. In-Silico Device Performance Prediction of Cosensitizer Dye Pairs for Dye-Sensitized Solar Cells. Adv. Energy Mater. 2023, 13, 2203536.
  • Liu, C.; Huhn, W.; Du, K.-Z.; Vazquez-Mayagoitia, Á.; Dirkes, D.; You, W.; Kanai, Y.; Mitzi, D. B.; Blum, V. Tunable Semiconductors: Control over Carrier States and Excitations in Layered Hybrid Organic-Inorganic Perovskites. Phys. Rev. Lett. 2018, 121, 146401.

Biography

Álvaro Vázquez-Mayagoitia is an expert in computational chemistry and data science. His experience spans both methods and applications of electronic structure theory with high-performance computing. He has authored and co-authored more than 60 refereed papers, 4 book chapters, and technical reports, with over 5,800 citations (h-index=27).

He joined Argonne National Laboratory at the Argonne Leadership Computing Facility in 2011 as part of an Early Science Program (ALCF-2). In 2013, he accepted a position with the ALCF Computational Science team. As a team member, Álvaro actively contributed to the continuous enhancement of features and performance of quantum chemistry codes, including NWChem, BigDFT, Quantum Espresso, MADNESS, VASP, and FHI-aims. With the goal of efficiently using ALCF resources and accelerating simulations, he has optimized codes and libraries for Argonne's petascale and exascale systems, including Mira, Theta, ThetaGPU, Polaris, and Aurora.

In 2019, Álvaro joined the Computational Science Division, where he provides support, guidance, and training for scientific projects in chemistry and materials science. He holds a joint appointment as Senior Scientist at UChicago CASE since 2020, serves as AI/data-driven science capabilities lead at the CPS Division, and is currently the Point of Contact (POC) for two Early Science Program projects for Aurora, DOE's first exascale computer.

Prior to joining ANL, Álvaro held postdoctoral positions at Oak Ridge National Laboratory and the University of Tennessee, where he worked with MADNESS and NWChem codes. He developed tools for molecular spectroscopy and studied weak interactions.

Álvaro collaborates with multiple projects, deploying AI/ML models and complex workflows to accelerate chemical and materials discovery. His work has contributed to advancements in machine-learned interatomic potentials for refractory oxides, organic dye prediction for solar cells, PFAS toxicity prediction, and molecular crystal structure prediction.

Álvaro led the ALCF Postdoctoral Committee, managing recruitment and mentoring of postdoctoral researchers. He contributes to the IGEN organization, which fosters underrepresented communities in physical sciences, and serves as a training facilitator for mentoring programs promoting inclusion and diversity at Argonne.

Software

I develop and contribute to the following open-source scientific software:

• NWChem / NWChemEx - Scalable electronic structure codes for exascale computing

• MADNESS - Multiresolution adaptive numerical environment for scientific simulation

• COLUMBUS - Multi-reference configuration interaction for excited states

• FHI-aims / ELSI - All-electron electronic structure and eigenvalue solver infrastructure

• PySCF (GPU) - Vendor-portable GPU acceleration for Python-based quantum chemistry

• GAtor / Genarris - Genetic algorithms for molecular crystal structure prediction

• miniGAP - AI/ML proxy app for chemistry and materials

• GPUAM - GPU-accelerated molecular property calculations