Research

                                             Electrochemical Energy Storage

 

 

 

 

 

 

 

 

 

Improving electrochemical energy storage technologies such as batteries, supercapacitors, and fuel cells requires tuning the chemical and physical properties of electrode/electrolyte interfaces under applied voltages.  We develop and apply sophisticated first-principles simulation techniques to understand basic physical properties of electrochemical interfaces that are relevant to energy storage devices.  Utilizing GPU-accelerated computer codes, novel Poisson solver algorithms, and ab initio force fields, we study complex physical processes such as interfacial surface structuring, inner-layer capacitance and dielectric saturation, and voltage-induced phase transitions.

 

                       Redox Chemistry at Electrochemical Interfaces

 

The interface between electrodes and electrolytes hosts a fascinating array of chemical processes that are dictated by voltage-modulated redox chemistry.  Numerous redox-active processes may be highly coupled and depend sensitively on the microscopic properties of the electrochemical interface--because of this complexity, many such processes that occur in batteries and related devices are often not well understood.  Furthermore, depending on the particular device operating conditions, chemical processes may be either thermodynamically or kinetically controlled.  We are developing new multi-scale modeling approaches to predict electron transfer, redox chemistry, and chemical reactivity at electrochemical interfaces.  Our computational approaches integrate quantum chemical methods, machine learning, and molecular simulation in novel ways, enabling first-principles property prediction through the use of ab initio force fields.  We seek to connect with experimental electrochemical, spectroscopic, and scattering measurements to enable comprehensive understanding of these systems.  

 

               Chemical Reactions Modulated by Strong Electric Fields

             

 

Solvent effects on chemical reactions are ubiquitous in organic chemistry, yet have proven historically difficult to predict from first-principles quantum chemistry.  The challenge is that high-level electronic structure methods must be utilized for describing bond-breaking/bond-forming processes with chemical accuracy, yet long-range energetic and entropic interactions with the solvent significantly alter transition state barriers.  Recently developed machine-learning approaches offer a fresh and exciting direction for research progress.  We are developing a novel combination of physics-based and machine-learning methods to capture the essential quantum mechanical interactions and complex collective variables that dictate chemical reactions in solution.  We utilize these approaches to study chemical reactions in ionic media, in which the strong electric fields from the surrounding ions rival the intrinsic strength of chemical bonds.

 

                      Statistical Mechanics of Heterogeneous Ionic Media

 

Statistical mechanical theories in combination with increasingly powerful computer simulations have provided detailed knowledge and understanding of liquids, soft matter, and other condensed-phase systems.  However, many of the organic electrolytes that are utilized in modern energy storage devices are not well-described by the basic statistical mechanical theories, and these theories may require significant extension/modification.  We are fundamentally interested in the statistical mechanics of room-temperature ionic liquids and organic electrolytes comprised of low-dielectric solvents, for which strong ion-correlation renders mean-field theories entirely inappropriate.  We are particularly interested in how these electrolytes behave in confinement, at interfaces, and under strong applied electric fields for which linear response theories may or may not be valid.