Research on Remote: USU Chemistry Grad Students Awarded Prestigious Internships

An unprecedented three Utah State University doctoral scholars were selected last spring for highly competitive internships, for summer 2020, offered by Los Alamos National Laboratory. Nikita Fedik, Maksim Kulichenko and Nikolay Tkachenko, all students of Professor Alex Boldyrev in USU’s Department of Chemistry and Biochemistry, immediately began preparations for moving, temporarily, to the storied, U.S. Department of Energy facility’s community, situated in northern New Mexico.

“The chances of being selected for this internship are slim and I was so excited to pursue this amazing opportunity,” Tkachenko says. “We all began searching for summer housing and organizing travel arrangements.”

Alas, just weeks after receiving the exciting news, the COVID-19 pandemic progressed and the students learned they wouldn’t be able to travel to Los Alamos, after all. But that didn’t mean to the internships were cancelled.

“Yes, it was disappointing news, as we all looked forward to being at the historic lab and exploring New Mexico,” Kulichenko says. “But, on the plus side, we would save on rent and travel.”

Making the best of a less-than-ideal situation, the Aggies unbooked summer Los Alamos accommodations and dove into their internships – remotely – from Logan.

“Our LANL mentors contacted us and welcomed us to the team,” Fedik says. “Even though we weren’t on-site, we met via WebEx or phone calls every day.”

Because the students are theoretical chemists, they weren’t hampered by the need to be at the lab.

“We were granted access to the lab’s supercomputers, which enabled us to work successfully from Utah,” says Tkachenko, who is pursues quantum chemistry.

Quantum computing, he says, is fundamentally a different type of calculation and the solution of an electronic structure problem is thought to be the nearest future implementation of small, noisy quantum computers.

“While the complexity of the exact solution of electronic structure problems on classical computers grows exponentially with the number of electrons, we can use the advantage of quantum calculations to make the problem complexity polynomial with respect to the number of simulated particles,” Tkachenko says. “During my LANL internship, I was developing new techniques for reducing complexity of quantum chemistry calculations on quantum computers.”

As a result of his internship efforts, Tkachenko is lead author of a peer-reviewed paper.

“This is an extraordinary accomplishment for a doctoral student,” Boldyrev says.

Fedik and Kulichenko’s efforts will also result in published papers.

Fedik’s internship focused on improving accuracy of theoretical methods for transition metals.

“Transition metals are ubiquitous – iron and titanium, for example – and very important,” he says. “Besides industrial applications, they play an essential role in biological processes. For instance, hemoglobin, responsible for oxygen transport in our blood, is a biopolymer containing iron.”

Fedik and his LANL mentors used machine learning to replace current theoretical methods for transition metals, which are based on computationally demanding quantum mechanical calculations.

“We believe that our models will replace or augment current quantum mechanical calculations, allowing for fast, large-scale simulations involving hundreds of thousands of atoms,” he says.

Kulichenko also delved into machine learning, aimed at replacing computationally demanding chemistry calculations.

“This is usually done by pre-trained artificial neural networks,” he says. “However, neural networks require a proper dataset to be trained on. I worked toward developing a technique of constructing a dataset, which would cover as much chemical space as possible, while keeping a manageable dataset size.”

Kulichenko, Fedik and Tkachenko join USU alum Ivan Popov PhD’17, who was also a LANL intern during his Utah State years. Popov was subsequently named an Oppenheimer Distinguished Postdoctoral Fellow at LANL, where he currently serves as a computational chemist.

Boldyrev reports his current students’ LANL mentors were so impressed with the interns’ performance, they’ve invited the Aggies to continue their research with LANL into the next year. Beyond the welcome opportunity and funding support, he says the remote research experience has provided a critical learning experience and preparation for the future.

“The end of the pandemic won’t signal the end of remote research,” he says. “This is the wave of the future and opens doors for increased collaboration across distances and borders. And collaborating remotely brings advantages, including enabling researchers to share resources, such as access to supercomputers.”