Welcome

We are engaged in theory and modeling of materials using a spectrum of computational methods, including electronic structure, molecular dynamics, machine learning, as well as our own algorithms. Our recent work has four primary directions:

  1. Innovation of monolayer and few layer materials (i.e. graphene, MoS2) for structural phase change electronic memories, electromechanical properties for NEMS, and van der Waals engineering.
  2. Machine learning to predict materials properties, including for batteries and other energy technologies.
  3. Development of fast algorithms and data-driven computational models for complex chemistry and structural phase transitions, including pyrolysis and shock compression of energetic and other organic and inorganic materials.
  4. Experimental collaborations to realize our predictions in the lab.

See the Impact page for some of the exciting results that we have discovered.

 

Group News Feed

Group member Austin Sendek to present at the 2017 Stanford Global Climate and Energy Project (GCEP) Symposium, October 17-18

Austin received the Distinguished Student Lecturer Award for his talk on machine learning-guided design of solid Li-ion electrolytes in the GCEP Student Energy Lectures series this summer and will present his talk at the 2017 GCEP Symposium at Stanford's Frances C. Arrillaga Alumni Center from October 17-18. More details will be shared when they are available.

Prof. Reed participates in a MIT Faculty Forum online panel on graphene and other two-dimensional materials. You can watch the video on YouTube here!

Congratulations Graduates!

Dr. Yao Li and Dr. Yuan Shen received their PhDs and were hooded by Evan at the Physics and Applied Physics graduation ceremony held on June 18, 2017. Congratulations Yuan and Yao! More pics here.

We identify 1173 two-dimensional layered materials and 487 materials that consist of weakly bonded one-dimensional molecular chains.

Moreover, we discover 98 weakly bonded heterostructures of two-dimensional and one-dimensional subcomponents within bulk materials, opening up new possibilities for the much-studied assembly of van der Waals heterostructures. These materials were found by using a novel data-mining algorithm that we developed to screen the Materials Project database of over 50,000 inorganic crystals. Click HERE to download the complete lists.

Clean energy scientists launch the Alt-E Fund and Professor Reed joins as a technical advisor

Scientists from the University of Colorado, Boulder and the National Renewable Energy Laboratory have launched the Alt-E Fund, the world's first foundational energy research fund that seeks to change the funding landscape by harnessing the passion of millions of citizens through grassroots funding efforts. The Alt-E Fund aims to collect and disburse funds for foundational clean energy research projects around the country at universities, national labs, and other organizations. Professor Reed brings his expertise in theory and modeling of materials for energy applications to the technical advisory board.

A defined Lifshitz model provides fast van der Waals computations for layered materials

We discover that a defined Lifshitz-based model can provide van der Waals (vdW) potentials to within 8-20% of advanced electronic structure calculations (ACFDT-RPA) while being orders of magnitude faster. Using this fast model, we study the vdW binding properties of 210 three-layered heterostructures and discover the potential for repulsive three-body vdW effects.

MathWorks blogs about Sendek et al.'s recent work using machine learning to build a better battery. Join the Facebook discussion here and here!

Austin Sendek's submission to the Global Energy Forum Video Competition is selected as one of this year's winners. Congratulations, Austin! The submitted video can be viewed here. The full director's cut is available here.

Yuan Shen successfully defends his PhD thesis.

Machine learning enables large-scale screening of candidate solid-state electrolytes for lithium-ion batteries

We screen 12,000 candidate materials comprising all known lithium-containing inorganic crystalline solids and provide a list of 21 promising structures.

Yao Li successfully defends her PhD thesis.

Imagine a "cool" data-storage technology that's just a few atoms thick

A news article on the Stanford Engineering website about recent group work on monolayer materials.

Lenson Pellouchoud successfully defends his PhD thesis.

Electrostatic gating drives structural phase transitions in monolayer materials

Two-dimensional transition metal dichalcogenides undergo structural semiconductor-to-semimetal phase transition under electrostatic gating of several volts.

L1 regularization enables efficient model reduction of large-scale chemical reaction networks

We use L1 regularization to reduce a large chemical reaction network observed from molecular dynamics simulations of shock compressed liquid methane, and find that CH4 decomposition can be modeled with less than 9% relative error using 11% of reactions.

Congratulations to Qian Yang for receiving a speaking award for her Fall MRS meeting talk.

Atomistic simulation of shock-induced silica crystallization

Using the Multi-Scale Shock Technique for molecular dynamics simulation of millions of atoms, we discover that SiO2, a prototypical good glass former, can be transformed to a very bad glass former upon the application of high pressure by the shock compression of meteor impact.

QBMSST is added to the public distribution of LAMMPS

The Quantum Thermal Bath Multi-Scale Shock Technique (QBMSST) has been incorporated into the public distribution of the LAMMPS molecular dynamics code by Yuan Shen with the key word fix_qbmsst. This method enables the simulation of shock compression of materials while incorporating some aspects of quantum nuclear effects in an approximate and computationally efficient fashion. These effects can play a significant role in the shock temperatures, kinetics, and potentially the chemical composition of the system at equilibrium.

Simulated coherent control of an isomerization reaction using THz electric field pulses

Using a tri-stable molecule (LiNC) and TDDFT-based Ehrenfest dynamics, our work shows that it may be possible to drive a prescribed reaction pathway with strong THz electric fields, without excessive heating, ionization, or electronic excitation of the target system.

Karel-Alexander Duerloo successfully defends his PhD thesis.

Yao Zhou is awarded a three year Stanford Graduate Fellowship.

Experimental Observations of Piezoelectricity in Two-Dimensional MoS2

The first experimental observations of piezoelectricity in single-layer 2D materials have been reported in two independent experiments in Nature and Nature Nanotechnology by researchers at Columbia University, Georgia Tech, UC Berkeley, Lawrence Berkeley National Laboratory, and the Chinese Academy of Sciences. Articles highlighting these experiments and our calculations have appeared in MRS Bulletin and a News and Views article in Nature Nanotechnology.

Strain Engineering in Monolayer Materials Using Patterned Adatom Adsorption

Our work shows that strains as large as 5% can be produced in monolayer materials using patterned adatom adsorption.

Deformations drive structural phase transitions in monolayer materials.

Two-dimensional transition metal dichalcogenides undergo structural metal-to-insulator phase transitions under tension.

Evan Reed gave two tutorials at the Lawrence Livermore National Laboratory Computational Chemistry and Materials Science program. These provide an introduction and summary of our recent work on "Emergent Electromechanical Properties of Nanoscale Materials," and "Atomistic Calculations of Dynamic Compression of Materials" including semiclassical quantum nuclear effects.

Graduate student Karel-Alexander Duerloo is awarded a Stanford Graduate Fellowship. Congratulations Alexander!

Electromechanical Bending in Boron Nitride Bilayers

Our work reveals a unique and manifestly nanoscale curvature-electric field coupling in boron nitride bilayers.

H and F coadsorption leads to piezoelectricity in graphene.

Motivated by a search for electromechanical coupling in monolayer materials, we have discovered that two types of piezoelectricity can be engineered into graphene when it is chemically modified with H and F.

Quantum corrections bring 40% lower pressure onset for methane dissociaton under shock compression.

We have developed a methodology for atomistic simulations of shock compressed materials that, for the first time, incorporates semi-classical quantum nuclear effects self-consistently. In our new method, the quantum nuclear effects are achieved with almost no additional computational expense.

Piezoelectricity in Two-Dimensional Materials

Our research has discovered that many of the widely studied two-dimensional monolayer crystals have excellent piezoelectric properties, making them ideally suited for applications in nanoscale technology.

Graduate student Karel-Alexander Duerloo has received the Best Instructor Award for his C programming course at the AHPCRC Summer Institute, held at Stanford.

Graduate student Lenson Pellouchoud is awarded a NASA Space Technology Research Fellowship. Congratulations Lenson!

A roadmap for engineering piezoelectricity in graphene

A news article on the NERSC website about recent group work on graphene.

Engineered Piezoelectricity in Graphene

Piezoelectric effects can be engineered into non-piezoelectric graphene through the selective surface adsorption of atoms. Published in ACS Nano.

> ACS Nano article highlight
> ACS Nano podcast: interview about article

 

 

 

 

Principal Investigator:
Evan Reed
evanreed _at_ stanford.edu
Tel: 650 723 2971
Fax: 650 725 4034
496 Lomita Mall
Stanford, CA 94305

Prospective Materials Science PhD students: Information about the admissions process can be found here.