Scott A. Roberts, Ph.D.

Distinguished Research & Development Chemical Engineer

Author profile picture

Distinguished Research & Development Chemical Engineer

Google Scholar



ResearchGate Profile

(505) 844-7957

Sandia National Laboratories, New Mexico
P.O. Box 5800
Albuquerque, NM 87185-0836


Dr. Scott A. Roberts is a Distinguished Research and Development Chemical Engineer in the Thermal/Fluid Component Sciences Department, Engineering Sciences Center at Sandia National Laboratories in Albuquerque, New Mexico. He has a B.S. in Chemical Engineering from the University of Kansas and a Ph.D. in Chemical Engineering from the University of Minnesota, under the supervision of Prof. Satish Kumar. He has worked at the Sandia National Laboratories since 2010 on a variety of projects, please see the research page for more information.


  • Lithium dendrite growth published in Cell Reports Physical Science

    In collaboration with Purdue University, Ph.D. student Julia Meyer developed a model for the impact of external pressure and surface roughness on separator deformation and lithium plating in lithium metal batteries. This work was recently published in Cell Reports Physical Science and is available as open access at DOI: 10.1016/j.xcrp.2023.101364.

  • Kintsugi imaging paper published in JES

    Imaging as-manufactured battery electrodes is important for determining their performance, yet it is often difficult to segment images from scanning electron microscopy due to the electrons’ ability to see through pores. In a paper recently published in the Journal of the Electrochemical Society (doi: 10.1149/1945-7111/ac7a68), Scott, along with collaborators at Imperial College and ThermoFisher Scientific, developed a new "Kintsugi" method of unambiguously resolving all three phases in battery electrode imaging.

  • Postdoctoral openings in hypersonics TPS M&S

    I am actively recruiting for multiple post-doctoral positions in modeling and simulation of thermal protection system (TPS) materials for hypersonic vehicles. Requires a Ph.D in a science/engineering discipline with prior modeling and simulation experience and U.S. citizenship. If you’re interested, apply at the link below and reach out to me with questions.

    Job posting on Sandia’s website

  • Effect of battery electrode manufacturing heterogeneity on transport properties

    Ph.D. student Chance Norris published his first first-author paper in ACS Applied Materials & Interfaces! In this work, we study the heterogeneity of commercially manufactured graphite electrodes for lithium-ion batteries using x-ray computed tomography and computational simulation. This heterogeneity spans multiple length scales, from the particle shape/morphology to that spanning multiple images samples. We show that heterogeneity at all scales influences the eventual transport properties of the electrode. DOI: 10.1021/acsami.1c19694

  • Cryo SEM identifies relationship between pressure and short circuits

    In a collaborative paper recently published in iScience, we use novel cryogenic electron microscopy to show the relationship between eternally applied pressure, lithium dendrite growth, and short circuits in lithium batteries. While pressure improves Li inventory retention at high currents, higher pressures promotes sh ort circuits. Thicker or multiple separators only masks this effect. Available open source at DOI: 10.1016/j.isci.2021.103394

  • Method of calculating diffusion coefficient from GITT published in ACS AEM

    What is “the diffusion coefficient”? A harder question than we first thought. In a new paper in ACS Applied Energy Materials, we show a more objective and accurate way to calculate the diffusion coefficient from GITT data specific to its end use. We realized that the diffusion coefficient used in a non-ideal (non-Fickian) diffusion model is different than the standard Fickian one. The value you calculate depends on its end use. Also interesting was that the non-ideal diffusion model fit discharge data across discharge rates much better than the Fickian model, even when each used their best fit diffusion coefficients. DOI: 10.1021/acsaem.1c02218