Physicists Bring Topological Band Theory to the Heart of Chemical Reactions

Some of the most consequential ideas in modern physics arrive as unexpected imports. Topological band theory, the framework that explains why certain insulators conduct electricity only along their edges and why electrons can behave as if they were massless, has spent two decades reshaping condensed-matter physics. Now a team of theorists at The Pennsylvania State University has carried that machinery into a domain where few thought it could survive: the world of chemical reactions that are governed by the conservation of orbital symmetry. In a study published in Nature Physics, Ziren Xie, Amir Mirzanejad and Lukas Muechler introduce a many-body formalism, built on Green’s functions, that extends the topological classification of reaction pathways to molecules in which electrons interact so strongly that the standard theory breaks down entirely.

The reactions in question are among the most celebrated in all of chemistry. In 1965, R. B. Woodward and Roald Hoffmann formulated their rules for the conservation of orbital symmetry, explaining why some pericyclic reactions proceed effortlessly while others, apparently similar, are essentially impossible under thermal conditions. The rules classify reaction pathways as symmetry allowed or symmetry forbidden depending on whether the energies of the relevant molecular orbitals cross as the molecule contorts along its reaction coordinate. A crossing of orbitals of different symmetry signals a forbidden pathway; an avoided crossing between orbitals of matching symmetry signals an allowed one. These selection rules earned Hoffmann a share of the Nobel Prize in Chemistry and have steered synthetic organic chemistry ever since.

Yet the Woodward-Hoffmann framework rests on a quiet assumption inherited from molecular orbital theory: that each electron can, to a first approximation, be treated as moving in the average field of all the others. Chemists have long known that this single-particle picture fails near certain geometries. As a bond forms or breaks, electrons can become strongly correlated, their motions entangled in ways that no independent-particle description can capture. In such regimes the very concept of a molecular orbital energy band, and with it the topological invariants that classify band structures, loses its footing. The topological band theory that works so beautifully for crystals assumes well-defined quasiparticles and a noninteracting or weakly interacting reference system, and neither survives the strong static correlation that plagues, for example, the automerization of cyclobutadiene or the ring opening of butadiene.

The Penn State team’s solution is to abandon orbitals as the fundamental objects and work instead with the single-particle Green’s function, a quantity that encodes all the information about how an electron propagates through the interacting many-electron system. In the language of Green’s functions, the familiar features of band theory reappear as poles: sharp peaks where the function blows up, corresponding to well-defined quasiparticle excitations. But Green’s functions possess a second, less familiar kind of feature: zeros, points where the function vanishes entirely. In strongly correlated systems, zeros are not mathematical curiosities. They can carry topology in their own right, a fact that condensed-matter physicists have exploited in recent years to understand Mott insulators and other systems where interactions dominate. The new work shows that this duality is exactly what chemistry needs.

The central result is strikingly clean. When the researchers computed the Green’s functions of reacting molecules along symmetry-forbidden pathways, they found that the obstruction to the reaction does not appear as a crossing of poles, as molecular orbital theory would predict, but as a crossing of zeros. In other words, the forbidden nature of the pathway is encoded in the vanishing of the electron propagation amplitude rather than in the divergence of an orbital energy. For symmetry-allowed pathways, by contrast, the poles behave in the familiar way, avoiding each other as the orbitals of matching symmetry mix. The distinction between poles and zeros thus becomes a many-body diagnostic that separates dynamic correlation, which perturbs quasiparticle energies, from static correlation, which destroys them and replaces pole crossings with zero crossings.

To make this diagnostic quantitative, the authors constructed symmetry-resolved topological invariants: integer quantities computed from the Green’s function within each irreducible representation of the molecular symmetry group. These invariants play the role that Chern numbers and symmetry indicators play in topological insulators, but they are defined for finite molecules along a reaction coordinate rather than for crystals in momentum space. Crucially, they are end-point diagnostics: one computes the invariant at the reactant geometry and at the product geometry, and if the values differ, the pathway is topologically forbidden. No continuous, symmetry-preserving deformation of the electronic structure can connect the two ends, because the invariant can only change when the Green’s function develops a singularity, which is precisely the zero crossing that characterizes the forbidden route.

The framework was put to the test on textbook systems. In benzene and cyclobutadiene, the pole-versus-zero analysis cleanly distinguished dynamic from static correlation, with cyclobutadiene’s notorious diradical character manifesting as Green’s function zeros. Along the electrocyclic ring closure of butadiene to cyclobutene, the conrotatory and disrotatory pathways, one allowed and one forbidden under the Woodward-Hoffmann rules, displayed exactly the expected behavior: poles avoided crossing on the allowed route, while zeros crossed on the forbidden one. The team also examined a 4π photoswitch, a molecular system relevant to the design of light-driven molecular machines, and found that the zero crossings persist even when the molecular symmetry is weakly broken, suggesting that the topological signature is robust enough to matter in realistic, slightly asymmetric chemical environments.

The work forges a bridge between two research communities that have largely talked past each other. On one side stands the topological quantum chemistry program, which since 2017 has catalogued how band representations and symmetry indicators classify crystalline materials, and its earlier molecular antecedents, including topological Chern indices in molecular spectra and Muechler’s own work on perplectic structures in molecules. On the other side stands the physics of strong interactions, where Green’s function zeros have been shown to underlie the breakdown of Luttinger’s theorem, to generate emergent quasiparticles at Luttinger surfaces, and to produce topological phase transitions in Mott insulators without any gap closing of single-particle states. The molecular setting turns out to be an ideal laboratory in which these ideas meet, because reaction coordinates provide a natural one-dimensional parameter, and molecular symmetry groups supply the classification machinery.

There are practical implications as well. Selection rules are the currency of chemical intuition, and a many-body generalization of them promises to extend that intuition to reactions where single-reference methods fail. The authors suggest that tracking Green’s function zeros directly could become a computational tool for predicting whether a proposed pathway is viable, particularly for photochemistry and for the design of molecular switches, where static correlation is endemic. The distinction between dynamic and static correlation, currently a source of endless trouble for quantum chemistry methods, gains a sharp spectral definition: dynamic correlation moves poles, static correlation creates zeros. That kind of crisp criterion could guide the choice of electronic structure methods before expensive calculations are attempted.

The study also resonates with a broader theme in contemporary physical chemistry: the growing recognition that geometric and topological phases shape molecular dynamics. Berry phases around conical intersections, geometric-phase effects in nonadiabatic dynamics, and even direct experimental observations of geometric-phase interference in molecules have all matured over the past decade. The new formalism adds a complementary piece, showing that topology does not merely decorate molecular motion but can fundamentally determine which reactions are possible. What began as a set of empirical rules drawn on blackboards in the 1960s now stands revealed as a special case of a deep many-body principle, one in which the fate of a chemical reaction is written not in the energies of orbitals but in the zeros of a Green’s function. For a field that has spent sixty years reasoning with orbital pictures, that is a genuinely new way of seeing an old problem.

Subject of Research: Topological classification of orbital-symmetry-controlled chemical reactions using many-body Green's functions

Article Title: Topological transitions in orbital-symmetry-controlled chemical reactions

Article References: Xie, Z., Mirzanejad, A., & Muechler, L. (2026). Topological transitions in orbital-symmetry-controlled chemical reactions. Nature Physics. https://doi.org/10.1038/s41567-026-03455-5

Image Credits: AI Generated

DOI: 10.1038/s41567-026-03455-5

Keywords: topological band theory, Green's functions, Woodward-Hoffmann rules, orbital symmetry, strong correlation, chemical reactions, molecular orbital theory, topological invariants, pericyclic reactions, Mott insulators, quantum chemistry, Nature Physics