{"id":12,"date":"2020-04-24T17:52:50","date_gmt":"2020-04-24T17:52:50","guid":{"rendered":"http:\/\/elements.chem.umass.edu\/zlinqcgroup\/?page_id=12"},"modified":"2026-07-29T21:51:42","modified_gmt":"2026-07-30T02:51:42","slug":"research","status":"publish","type":"page","link":"https:\/\/elements.chem.umass.edu\/zlinqcgroup\/research\/","title":{"rendered":"RESEARCH"},"content":{"rendered":"\n

Quantum Chemistry and Artificial Intelligence for Complex Chemical Systems<\/h2>\n\n\n\n
\"\"<\/figure>\n\n\n\n

The Lin Group develops hybrid methods in first-principles quantum mechanics (QM) and physics-informed artificial intelligence (AI) to understand complex chemical systems. We connect calculated electronic structures and molecular dynamics to experimental observables, including molecular spectra, reaction mechanisms, and material properties. Our studies aim to establish computational models that are accurate enough to reveal atom-level molecular and material mechanisms and efficient enough to guide experiments and discovery.<\/p>\n\n\n\n

Artificial Intelligence and Quantum Mechanics for Electronic Structures and Molecular Spectroscopy<\/h3>\n\n\n\n

We develop hybrid QM\/AI methods that make accurate electronic structure calculations more scalable and turn complex spectroscopic measurements into molecular information. Our work spans both forward structure-to-property prediction which calculates experimental observables, such as energies, forces, mechanisms, and signals, from molecular structures, and inverse property-to-structure inference, which converts experimental observables back to candidate molecular structures.<\/p>\n\n\n\n

For forward structure-to-property prediction<\/strong>, we develop machine learned density functionals, including ML-\u03c9PBE<\/strong> [1<\/a>,2<\/a>] and GNN-\u03c9PBE [3<\/a>], to predict electronic structures and optical properties of organic semiconducting molecules and radicals. We also develop FB-GNN-MBE<\/strong> [4<\/a>,5<\/a>] and FB-GNN-QE<\/strong>, two hybrid methods that integrate fragment-based quantum chemistry, such as many-body expansion (MBE) and quantum embedding (QE) theory, with fragment-based graph neural networks (FB-GNNs), to learn fragment-fragment and fragment-environment interactions in large complex systems. <\/p>\n\n\n\n

For inverse property-to-structure inference<\/strong>, we convert spectral signals, such as infrared (IR), Raman, nuclear magnetic resonance (NMR), and mass spectrometry (MS), into two-dimensional (2D) topology and three-dimensional (3D) structures, using foundation models of generative AI, such as transformers and diffusion models. We also elucidate reaction mechanisms and dynamics in different chemical contexts, such as heterogeneous catalysis, astrochemistry, and enzymatics, based on real-time inference of spectral signals.<\/p>\n\n\n\n

Heterogeneous Catalysis, Gas Sensors, and Inorganic Materials<\/h3>\n\n\n\n

We use QM and QM\/AI modeling to understand how local electronic and chemical environment, such as surface structure, solvation effect, adsorbate vibration, intermediate transport, and nuclear quantum effects, control the behavior of different inorganic materials. Our work span electrocatalysis, thermal catalysis, photocatalysis, gas sensors, and high-entropy ceramics, which support reactions through coupled electron, proton, solvent, adsorbate, and surface dynamics.<\/p>\n\n\n\n

For electrocatalysis<\/strong>, we decode the atom-level reaction mechanisms that remove harzadous molecules or synthesize value-added feedstock using electroreduction, including oxygen reduction reaction<\/strong> (ORR) to synthesize hydrogen peroxide (H2<\/sub>O2<\/sub>) [6<\/a>], nitrate reduction reaction<\/strong> (NO3<\/sub>RR) to synthesize ammonia (NH3<\/sub>) [7<\/a>], and carbon dioxide reduction reaction<\/strong> (CO2<\/sub>RR) into C2+<\/sub> products, and provide design principles of various electrocatalysts, including doped graphene, single crystal copper (Cu), and 2D metal-organic frameworks (MOFs). <\/p>\n\n\n\n

For gas sensors<\/strong>, we focus on the sensing capacity of pure and doped indium oxide (X:In2<\/sub>O3<\/sub>) surfaces toward nitrogen dioxide (NO2<\/sub>) [8<\/a>], correlate it to the local atomic arrangement and electronic density of the adsorption site, and propose design principles for high-capacity sensors. <\/p>\n\n\n\n

Physical Organic Chemistry and Organic Materials<\/h3>\n\n\n\n

We apply QM modeling to determine how molecule-level mechanisms, such as molecular structures, molecular arrangement, electronic configurations, vibrational configurations, energy levels, and local chemical environments regulate behaviors of molecular assemblies and functional materials, such as reactivity and function. Through close collaboration with experimental groups, we investigate how molecular design can control molecular and material behaviors.<\/p>\n\n\n\n

Explore Our Work<\/h3>\n\n\n\n

If you are interested in our research, you are welcome to read our publications<\/a>, download our recruitment poster<\/a> (also below), explore our GitHub repository<\/a> for open-source codes and models, meet the group members<\/a>, and view research openings<\/a>. Our work has been supported by the National Science Foundation (NSF), Research Corporation for Science Advancement (RCSA), American Chemical Society Petroleum Research Fund (ACS PRF), Bezos Earth Fund, and UMass Amherst.<\/p>\n\n\n\n

\n
\"\"<\/figure>\n<\/figure>\n\n\n\n

<\/p>\n","protected":false},"excerpt":{"rendered":"

Quantum Chemistry and Artificial Intelligence for Complex Chemical Systems The Lin Group develops hybrid methods in first-principles quantum mechanics (QM) and physics-informed artificial intelligence (AI) to understand complex chemical systems. We connect calculated electronic structures and molecular dynamics to experimental …<\/p>\n

RESEARCH<\/span> Read More »<\/a><\/p>\n","protected":false},"author":88,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"class_list":["post-12","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/elements.chem.umass.edu\/zlinqcgroup\/wp-json\/wp\/v2\/pages\/12","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/elements.chem.umass.edu\/zlinqcgroup\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/elements.chem.umass.edu\/zlinqcgroup\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/elements.chem.umass.edu\/zlinqcgroup\/wp-json\/wp\/v2\/users\/88"}],"replies":[{"embeddable":true,"href":"https:\/\/elements.chem.umass.edu\/zlinqcgroup\/wp-json\/wp\/v2\/comments?post=12"}],"version-history":[{"count":16,"href":"https:\/\/elements.chem.umass.edu\/zlinqcgroup\/wp-json\/wp\/v2\/pages\/12\/revisions"}],"predecessor-version":[{"id":1020,"href":"https:\/\/elements.chem.umass.edu\/zlinqcgroup\/wp-json\/wp\/v2\/pages\/12\/revisions\/1020"}],"wp:attachment":[{"href":"https:\/\/elements.chem.umass.edu\/zlinqcgroup\/wp-json\/wp\/v2\/media?parent=12"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}