Soon after I met Tarun Chitra, he shattered all of my illusions about glass.
It was a Friday evening in early September, the air was still summery, and we were hanging out at a friend's house before going out dancing.
Chitra works with one of my roommates at a biochemistry research company called D.E. Shaw Research based in New York City. He is trying to figure out how glass molecules behave, but doing so on a level beyond the typical models and simulations physicists use.
Rather than using simplified models that essentially treat molecules like balls in a ball pit, Chitra uses a custom-built supercomputer to run intricate simulations of glass molecules that represent actual atoms bonded together with springs.
In addition to furthering physicists' understanding of how glass works, Chitra's research could inform the production of better glass technologies in everything from tough gorilla glass to better optical lenses to strong, ultra-thin glasses that can be used in fiber optics and nanotechnology.
Glass isn't a solid, but it's not quite liquid either
In my experience, asking someone to talk about their research can either be immensely rewarding or immensely tedious, but I was on the prowl for good science stories, so I figured I'd give it a shot. Plus, I liked Chitra's style: scruff and disheveled hair that I decided meant he prioritized thinking over grooming, a subtly nasal voice that I felt would pair well with explaining physics, and socks featuring colors not made by Crayola, which I coveted for myself.
Right off the bat, Chitra threw down a couple facts that surprised me. First, glass isn't solid (which I knew)—but it's not quite liquid either (which is what I thought it was). Secondly, glass doesn't just refer to the silica-based material that we use to make bottles and windows. In fact, pretty much any liquid, including water, can be made into a glass if it is cooled past its melting point fast enough.
Then Chitra started talking about movies and contra-dancing. And that's when I knew for sure that I was in for a good science lesson. By the end of it, it was clear to me that glass is freakishly weird, and scientists know relatively little about it.
We might mistake glass for a solid because it's hard and retains its shape when we pick it up, Chitra said. But if we were to zoom in on a glass, down to a microscopic level, we'd see something different. "Glass is surprising because it's so rigid, but when you measure its molecular properties, it looks exactly like a liquid," he told me. The molecules in a glass are scattered all over the place rather than packed into the neat lattice structure that gives solids their stiffness.
Part of the confusion about where glass stands in relation to solids and liquids has to do with timescale, according to Chitra. Think of it in terms of frame rates for a movie, he suggested. "A solid is a single frame that never changes. A liquid is a movie at normal speed. And a glass is a movie that moves so slowly that it takes your whole lifetime to move one second."
So even though a single frame of glass molecules looks the same as a single frame of liquid molecules, the frame rates for the two are very different. The frame rate of glass is so slow that it appears to act like the static single frame of a solid. "It's as if I gave you a single picture and said, 'Tell me whether this was taken by a still camera or a video camera,'" said Chitra. "Glass is like watching a movie so slowly that it basically looks like the frame doesn't change."
But it's not like glass is just a really slow-moving liquid either. In fact, glass departs from liquid is some key ways.
Quenching molten basalt in water produces a glass.
To understand this, it helps to know how glass is formed. Making glass involves dropping the temperature of a liquid really quickly, a process called quenching. Cooling a liquid past a rough, experimentally determined point called the glass transition temperature (this temperature differs for each material, and also depends on the rate of cooling) produces a glass. Not quenching enough produces a supercooled liquid, a slushy material that flows like honey but acts similarly to glass on a molecular level.
"Think of a supercooled liquid as something that's on the way to being a glass," said Chitra.
Supercooled liquids are highly viscous, like honey.
Strange things happen when you quench a liquid. Once the liquid drops below its melting point, it wants to form a solid really badly. But because you are cooling the liquid so rapidly, the molecules in the liquid don't have enough time to pack themselves neatly into sheets. Which means the end product isn't exactly a solid.
Usually when matter changes from one phase to another, it gives off energy called latent heat. Water releases latent heat when it freezes into ice. But when liquids transition into glass, they release no such heat.
So there is clearly something going on with glass that isn't the happening with solids and liquids. But exactly how glass behaves at a molecular level is still a mystery. That's what Chitra is trying to figure out. It seems glass is governed by a different set of rules entirely, he said. To explain this, he offered three dancing analogies.
"Solid molecules are like a bunch of 60-year-olds in a nursing home, on a slow dance floor," he told me. Each person finds a partner and they vibrate around a little bit. They're not moving far from where they start, and all their neighbors stay the same. If you zoom out, it looks like they're not moving at all.
"A liquid is like a singles bar," continued Chitra. "People are dancing, going crazy, forgetting who their partners are, forgetting people's names, kind of moving around crazily. It's very high paced."
That leaves glass. "Glass is kind of like an Eastern European folk dance or a contra dance," Chitra said. "It looks like it's very structured for a while, nothing's changing, and then all of a sudden you kind of have random changes."
Glass molecules behave like dancers in a contra dance.
In a Hungarian or Romanian folk dance, for instance, there might be 100 people divided into ten groups of ten people, and each of those ten dancers hold hands in a circle. Within their groups, the folk dancers kind of behave like the dancers at the liquid singles bar. They keep switching partners until they dance with everyone. But unlike the people at a singles bar, the folk dancers stay localized to their circles. For a glass, explained Chitra, the folk dancers remain in the same circles forever. For a supercooled liquid, they stay in their circle for a length of time and then suddenly switch.
Within the folk dance, it also seems like the groups themselves are behaving differently. When scientists study materials, they often measure a parameter called relaxation time. In general, molecules want to exist at the lowest energy level they can. Relaxation time refers to the amount of time it takes molecules to find a configuration relative to their neighbors that minimizes their energy.
In the framework of Chitra's dancing analogy, we can think of relaxation time as the amount of time dancers spend with a partner before moving onto a new one. With both solids and liquids, the relaxation times of molecules are pretty much the same regardless of which molecules you measure and how many you measure.
"So if I'm in a singles bar, and I take ten random people and ask them roughly how much time they spend dancing with each partner, the average answer would be pretty much the same as if I asked 100 people," said Chitra.
But in a glass or supercooled liquid, the answers vary a lot more, according to Chitra. "If I take ten people and they're all near each other, they'll have the same answer," he said. "But if I take ten people from a different part of the room, they'll have a very different answer. And if I ask 100 people, I'll see many different local environments, like different cliques."
The result is that glasses are made of molecular patches with very different viscosities, Chitra said. "One patch might have the viscosity of water, while another is basically solid and never changes. Or one patch is like honey, and another patch is like motor oil."
The exact mechanism behind this phenomenon, called dynamic heterogeneity, is one of many questions that elude glass researchers. It's clear that something weird is happening with glass, thermodynamically speaking.
Some scientists think glass is its own phase of nature that still operates under predictable thermodynamic rules, which tries to minimize energy in the system. Others think glass falls out of thermodynamic equilibrium, meaning glass molecules stay at their initial positions, even if those positions are not the most energetically favorable. Non-equilibrium would explain why glass molecules retain memory of their starting position in a way that liquid molecules do not.
Dynamic heterogeneity also sums up Chitra pretty well. Apart from researching glass, he regularly rock climbs, goes to all-night electronica concerts on weeknights, cycles 25-50 miles a week, and cultivates longstanding interests in coffee, the tech startup scene, computer science, and machine learning.
He traces his interest in science back to his chemist father and mathematician mother, along with watching many Nova episodes, particularly one about magnetic levitation trains that hooked him on superconductors for life. His early education consisted of Montessori school and he spent much of his childhood participating in math and science Olympiads. When he was 11 years old, he won second place on Kids Jeopardy (out of three contestants, he's quick to clarify).
At age 10, Chitra got into online pirating in a big way, first to watch rated R movies, but then as a point of pride. Along with a friend from his middle school, he ran an Internet Relay Chat (IRC) piracy chat room called warezpunks.
"I had this desire to be the owner of the biggest IRC chat room," said Chitra. That dream came true: on the week of his 12th birthday, warezpunks was the biggest piracy channel on IRC.
Around that time, Chitra developed a concurrent obsession with Alexander Shulgin, the biochemist who introduced MDMA to psychologists in the 1970s and is often dubbed "the godfather of psychedelics."
I pointed out that he seemed like quite the renegade 10-year-old. "In retrospect it sounds like I was pretty badass, but in actuality I'm sure I was by far the nerdiest person around," said Chitra.
At Cornell, where Chitra went to college, he took an average of eight classes a semester and graduated with degrees from the arts college and the engineering college—a double major in math and art history from arts and an applied physics degree from the engineering school. After trying research projects in biology, bioengineering, theoretical economics, and theoretical computer science, he found an academic home studying string theory.
For a while, Chitra thought he would go to grad school, but eventually decided there was no future in string theory. Instead he went to D.E. Shaw Research right after graduating in 2011, and has studied glass there ever since.
The research Chitra does at D.E. Shaw Research is unconventional within a company that is already somewhat unconventional. The company is funded by David Shaw, a computer scientist who made a fortune off gaming the financial market with high speed computer networks and owns a hedge fund called D.E. Shaw & Co, also based in New York City. With a healthy source of private funding, researchers at D.E. Shaw Research have a relatively long leash to pursue their intellectual interests in lone-ranger fashion. Chitra's research, though, is esoteric even among that of his coworkers, who are mostly conducting studies aimed at drug discovery.
Chitra thinks there are many goals to his research. One is to support experimental research on glass by determining what is happening on a molecular level. Another is to resolve long-standing disputes about glass formation in a highly contentious field of research. He also wants to explore dynamic heterogeneity with models based on actual atoms as opposed to toy models.
On another level, his work could help improve glass technologies, including making glasses that are just a few nanometers thick but extremely strong. The material could be useful for higher bandwidth fiber optics and nanotech.
But perhaps Chitra's biggest motivating factor is the opportunity to discover something completely new and unexpected. He told me that he is scheming up a weird model that doesn't totally fit within the parameters of existing theories about glass, but he wouldn't share any specifics. "It's more just something that explains on an intuitive level what the system kind of looks like" was the vague description I could get.
In adulthood Chitra seems to have retained many of the best qualities of his childhood. As he did in Montessori school, he approaches everything as a puzzle and is constantly on his feet, changing tasks because he can't sit still.
He is still the 10-year-old online piracy empire-builder with the need to do things his own way. Even when he was managing a team of 30 hackers at age 12, he wanted his chat room to retain a certain edge. While all the other piracy channels were serious, Chitra's had a joke room and free ASCII art memes. "After all, we were the warezpunks," he said.
Nowadays, the 25-year-old is putting his rascally intellect towards researching glass and learning to program supercomputers, among other projects. "I still want to be king, in a very childish way," he said.
This story is part of The Building Blocks of Everything, a series of science and technology stories on the theme of materials. Check out more here.
Lead photo: Marc Falardeau/Flickr