Samuel Blau

Harvard University

For Samuel Blau, the motive for his research into photosynthetic molecules is simple: “Our generation has to make some big strides – socially, politically, technologically – if we want this only planet that we’ve found that can support life to still be habitable in a hundred years.”

Blau also believes his Harvard University education has equipped him well to help solve the problem. “I’m trying to figure out where I can have an impact,” the Department of Energy Computational Science Graduate Fellowship (DOE CSGF) recipient says. “I feel like that’s my responsibility as someone with a unique skill set.”

For now, Blau focuses his abilities on how energy moves in photosynthesis, the process green plants and some bacteria use to convert light into food. He’s particularly interested in several species of deep-water-dwelling algae that turn sparse photons into chemical energy with nearly 100 percent efficiency. Working under Alan Aspuru-Guzik, Blau builds computer simulations of the algae’s chromophores – their light-harvesting machinery – to understand how they do it. If he succeeds, we someday may make artificial photosynthetic molecules that have equal efficiency.

But “simulating these systems ends up being incredibly challenging,” Blau adds. Calculations must track the complex movements of electrons and excitons, “balls of energy” produced when photons boost electrons into higher energy states. They require computational methods for time-dependent quantum mechanics – the system’s evolution under the strange physics in which an electron can be both a particle and a wave.

Blau: Crystal

To make the problem tractable, standard solutions make some simplifications, like calculating the atomic system at zero temperature. But “temperature ends up being incredibly important here,” Blau says. Heat-induced vibrations are crucial to energy transport and the simulation must include them. Standard methods also often leave out the environment surrounding the chromophore and how it influences photosynthesis.

It’s understandable that scientists would simplify these calculations, Blau says. “Things get really messy really quickly” when details are included. So he’s devoted his graduate research to leveraging high-performance computing systems to develop more accurate simulations that eliminate some of these approximations.

“I was getting concerned, because I just wasn’t sure if I believed some of the most cited simulations in the field,” Blau adds. His focus on accuracy means “I’ve progressed more slowly, but I have more confidence in my results.” It’s also provided skills “to push the boundary on what simulations are even possible.”

Accuracy is especially important because Blau is investigating two proteins from slightly different algae genuses. Although the complexes have similar structure, their properties differ significantly, he says. If he and Aspuru-Guzik had stuck with standard methods, “it wasn’t clear to us that we would be able to get to the level of detail we need to reproduce these very different properties observed in these very similar systems.”

Blau’s 2014 Argonne National Laboratory practicum focused on an alternative, highly accurate but computationally demanding quantum chemistry method: coupled cluster. Working with DOE CSGF alumnus Jeff Hammond, Blau contributed to the development of Aquarius, a framework for tensor computations, which are vital to coupled cluster calculations. Devin Matthews created Aquarius based on his work with fellow DOE CSGF alumnus Edgar Solomonik and former Argonne intern Martin Schatz. During his practicum, Blau implemented the excited-state equation of motion method in Aquarius to extend its capability. Now he’s developing a collaboration that would use Aquarius to compute properties of photovoltaic materials.

As for his doctoral research, Blau aims to make his simulations good enough to match experimental data. That would let computational scientists model millions of photosynthetic compounds to find the most promising ones for lab testing.

“The dream is to be able to screen and predict” chemical structures for light-harvesting systems that could provide renewable energy, he adds. “But we’re still very far away from that.”

Graduation, however, is near. Blau aims to get his degree in 2017. He worked at Lawrence Berkeley National Laboratory while still in high school, so Blau feels at home in the DOE system. He hopes to return to his native Bay Area and work there again, focusing on research to help stave off climate change.

Image caption: Crystal structure of the biliprotein PC645, a light-harvesting complex found in Chroomonas cryptophyte algae. Image courtesy of Samuel Blau.