Zoe Wang
Carleton College
Faculty Advisor: Prof. Troy Van Voorhis
Research Supervisor: Shaun Weatherly
Department: Chemistry
Biography
Zoe Wang is a rising sophomore at Carleton College studying mathematics and religion.
This summer, she worked in the Chemistry Department in the Van Voorhis group, which aims
to model electron motion in molecules. This problem’s complexity grows exponentially
because one electron’s motion is correlated with every other electron in a given system. Zoe
spent the summer learning about quantum entanglement, information structures, and electron
correlation to develop new, efficient methods to reduce this complexity. Her experience has
motivated her to continue learning about scientific computing and applied math. During the
school year, Zoe is also the Programming Director for Carleton’s student-led radio station,
co-captain of the women’s ice hockey team, and a Perlman Learning and Teaching Center
Student Fellow.
Exploring the Role of Information Structures in Developing Quantum
Chemistry Algorithms
Zoe Wang1, Shaun Weatherly2, and Troy Van Voorhis2
1Department of Mathematics and Department of Religion, Carleton College
2Department of Chemistry, Massachusetts Institute of Technology
Understanding the fundamental quantum systems that make up molecules is crucial to
predicting their physical and chemical properties. These systems are electronic orbitals:
wavefunctions that contain all possible degrees of freedom for the system. In practice, finding
an orbital’s exact wavefunction is impossible because electron motion is correlated: every
electron’s behavior is tied to the motion of all neighboring electrons, which affect one another
as well. Electron correlation makes finding the wavefunction scale exponentially in
complexity, and therefore impossible to compute. In this work, we aim to study the
fundamental structure of electronic correlation. We look at how electronic correlation can be
quantified as orbital mutual information—how much information different orbitals have about
one another—to help explain the limitations of existing quantum chemistry algorithms and
guide the development of new, `correlation-aware’ theories and methods. For example, we
show that mutual information structures can explain why the density matrix renormalization
group (DMRG) fails for non-linear systems. Accurately and efficiently taking advantage of
entanglement structures will make these algorithms more optimal, allowing scientists to
describe fully complex chemical reactions from first principles.
