Materials Horizons Emerging Investigator Series: Prashun Gorai, Colorado School of Mines, USA

Prashun Gorai is a research assistant professor in the Department of Metallurgical and Materials Engineering at the Colorado School of Mines (CSM) and is also affiliated with the Materials Science Center at the National Renewable Energy Laboratory (NREL). He obtained his BTech in Chemical Engineering from the Indian Institute of Technology Madras in 2008, and his PhD (also in Chemical Engineering) from the University of Illinois at Urbana-Champaign (UIUC) in 2014. Subsequently, he worked as a postdoctoral fellow with Vladan Stevanović at CSM and NREL, from 2014–2017. He received a graduate student fellow award from the American Vacuum Society in 2009. In 2012 and 2013, he was a recipient of the Dow Chemical graduate fellowship. The Royal Society of Chemistry has recognized him as an outstanding peer reviewer for Journal of Materials Chemistry A in 2018 and 2019. At CSM and NREL, his team utilizes first-principles computations and data informatics to accelerate the discovery of novel functional materials for thermoelectrics, photovoltaics, solid-state energy storage, and power electronics. In the quest for novel functional materials, his team aims to map unexplored/under-explored chemical spaces with targeted computational searches. Beyond discovery, his team is also interested in predicting and modeling defect properties of materials, with the eventual goal of achieving “doping by design”.

Read Prashun Gorai's Emerging Investigator Series article “Computational discovery of promising new n-type dopable ABX Zintl thermoelectric materials” and read more about him in the interview below:

MH: Your recent Materials Horizons Communication focuses on your new CRISP (chemical replacements in structure prototype) method for computational materials discovery. How has your research evolved from your first article to this most recent article and where do you see your research going in the future?

PG: To provide some context, I started my graduate research (with Edmund Seebauer, UIUC) as an experimentalist studying the diffusion of ion-implanted dopants in Si and native oxygen defects in oxides. In my initial foray into computation, I developed drift-diffusion continuum models to extract parameters by fitting to experimental diffusion data. It was in the fourth year of my PhD, during a prolonged characterization instrument downtime, when I delved into first-principles computations to explain the atomistic mechanisms that gave rise to the unique defect diffusion behavior in my experiments. Shout out to my co-adviser, Elif Ertekin (UIUC), for welcoming a computational noob into her research group! My graduate studies culminated in a series of papers that successfully combined experiments and computations. While the experience of combining theory and experiments was quite satisfying, there was something missing.

I had so far utilized computations to retrospectively explain experimental observations. My postdoctoral work with Vladan Stevanović and Eric Toberer (CSM) was a turning point – the focus shifted from “explaining” to “predicting”. I used high-throughput calculations, along with tractable descriptors, to computationally search and discover novel materials for thermoelectrics. Some of the candidate materials demonstrated relatively high performance in experiments, which endorsed these predictions. I extended these efforts to search for defect-tolerant semiconductors, and wide bandgap materials for power electronics. In our materials discovery endeavors, we realized a common limitation that prevented experimental realization of the predicted superior performance – doping, which is governed by the defect properties of materials. Using first-principles defect modeling, we are now able to predict whether a material can be suitably doped. Functional materials discovery enabled by high-throughput searches combined with doping predictions has been rewarding. However, there are still aspects of materials discovery that are missing.

Doping remains a bottleneck. The current approach is to computationally survey the pool of known/reported materials, predict their properties, identify candidates, and hope they can be doped. This approach has been successful to a certain extent, but requires a bit of good fortune to find dopable materials. What if we could “design” new (unreported) materials that have the desired superior functional and doping properties? In our paper, we explore the CRISP approach as a route to realize “doping by design”. The goal is not to simply discover new materials, but new materials with designed properties. The hypothesis behind CRISP is simple and intuitive. We utilized this method to discover new n-type dopable Zintl phases, which are rare among Zintl compounds. In the future, I see opportunities to extend this method beyond thermoelectrics, to applications where conventional searches within known materials have had limited success.

MH: What aspect of your work are you most excited about at the moment?

PG: There are two aspects that we are most excited about: (1) in the phase space of plausible materials, the known phases represent only the tip of the iceberg. I am optimistic that there are opportunities in unexplored/under-explored chemical spaces to unearth high-performing functional materials as well as unconventional phenomena. As humans, we are innately curious to explore. (2) Finding computational routes for materials discovery that achieve “doping by design”. CRISP is one such route but there may be other more robust approaches.

MH: In your opinion, what are the most important questions to be asked/answered in this field of research?

PG: Almost a decade ago, the Materials Genome Initiative sparked a widespread interest in computation-aided accelerated materials discovery. The most enticing aspect was the promise of “inverse design”. While we, the computational materials community, have made rapid progress in the discovery aspect, it can be debated whether we have come close to “designing” materials. There have been notable efforts in recent years but inverse design remains a formidable challenge. So, how can we truly realize inverse design in the next phase of materials discovery?

To realize the predicted high functional performance, materials still need to be optimized in experiments. Depending on the properties, such optimization may involve fine tuning growth conditions, regulating carrier densities, alloying, etc. High-throughput combinatorial experimental techniques have accelerated this otherwise laborious optimization step. Within this paradigm, computations can play a complementary role in facilitating speedy convergence to the optimal synthesis routes, alloy compositions, etc. Mainstream computational approaches are not well suited to enable this complementary role. In what ways can computations facilitate experimental optimization?

MH: What do you find most challenging about your research?

PG: As theorists, we predict properties and computationally discover novel materials. We work closely with our experimental collaborators, who try to realize our predictions in the lab. I have recognized that, as theorists, we are often not cognizant of the experimental limitations. For instance, there are synthesis challenges involving air sensitive, volatile, and toxic elements, incongruent melting, etc. Also, it is challenging when theoretically predicted properties cannot be directly measured in experiments.

MH: At which upcoming conferences or events may our readers meet you?

PG: I will be at the Materials Research Society Fall meeting 2020 in Boston, if the conference is not cancelled. I will also be attending the European Materials Research Society Spring meeting 2021 in Strasbourg, and possibly the International Conference on Thermoelectrics 2021 in Krakow. I can be reached via email at pgorai@mines.edu. For those who are social media savvy, contact me on Twitter @prashungorai.

MH: How do you spend your spare time?

PG: I need my creative releases. So, I spend my spare time cooking different types of cuisines from around the world, doing a bit of photography, and reading about graphic design and fine arts. When my schedule allows, I travel to places that are off the beaten path.

MH: Can you share one piece of career-related advice or wisdom with other early career scientists?

PG: I will take the liberty of sharing two pieces of advice. Early in my PhD, I was stuck in my research because I could not analytically solve a complex set of partial differential equations. After a few months of frustration, I decided to contact a professor in the Mathematics department, without even expecting a response. The professor not only responded, but he helped me solve the PDEs. The analytical solution was the subject of one of my first papers and also a key element of my PhD research. So, do not be afraid to reach out to your peers or senior academicians, whether you are looking for collaborations, to learn something new, or for mentorship. Secondly, if you are an early-career computational researcher, make allies in the experimental community and forge collaborations, which will help you close the theory–experiment loop.

---