Research Stories

by Nicholas Gerbis

People tend to think about glass as a human invention. Not even close. In its natural form, glass has existed on Earth almost since the planet began to cool billions of years ago.

Volcanoes produce glass as obsidian. The dark, translucent substance forms when lava rich in silica cools. Glassy rocks called tektites also form when large meteorites smash into the planet's surface. Sometimes that melted material is hurled hundreds of miles from the impact site. It cools into bottle-green moldavite.

Humans have been using glass almost as long as we have been using tools. During the Stone Age, we smashed obsidian with harder rocks, knapping it like flint into arrowheads and cutting tools of remarkable sharpness. The material is stilled used in scalpels today, thanks to its ability to form edges of almost molecular thinness. Our ancestors also polished obsidian to make crude early mirrors. It was even used it to decorate the eyes of the statues of Rapa Nui on Easter Island.

C. Austen Angell studies glass. In fact, he has studied glass-forming substances for almost 50 years. Angell is a Regents Professor of Chemistry at Arizona State University. He says that one of the central mysteries of glass research–the riddle within the enigma–is resolving exactly what occurs when glass is formed, not from silica or other typical glassmakers, but from pure water. A better understanding of "glassy" water is key to unraveling the mysteries of glass formation in general.

Glass is much more than the stuff of windows and mirrors, polished stones and melted sand. The scientific term "glass" refers generally to a highly viscous, non-crystalline sub-state of matter. Glass can be made from a variety of substances.

Glass remains a bit of an enigma, one that physical chemists have struggled for decades to crack. The theory of the nature of glass and of the transition to and from the glassy state is a challenging problem. In 1995, Nobel laureate Philip Anderson called it the "deepest and most interesting unsolved problem in solid-state theory."

"Glassy water is a sort of Rosetta stone," Angell says. "It sits smack in the middle, between typical molecular glasses and high-temperature, inorganic-network glasses, such as ceramics and volcanic glasses. They all link together in some way. But finding the link between them is difficult."

This "missing link" of glass research, as controversial as it is full of explanatory promise, has remained elusive–until now. Angell has developed a novel approach to the problem. His idea allows him to address several of glassy water's central mysteries. The results of his work appeared in the February 1, 2008, issue of the journal Science.

Water has many odd and amazing properties thanks to its molecular structure and powerful hydrogen bonds. Angell had to overcome these properties before he could solve the problem.

These same properties are the reason why water is the only natural substance commonly found in all three states (solid, liquid and gas) within the naturally occurring range of Earth temperatures. They also explain why ice floats on water. Most substances are denser in their solid form than they are as liquids. These properties are vital for life as we know it, because they give water its very high surface tension. This helps water move through roots and capillaries. It also allows it to absorb a great deal of heat without getting hot, which contributes to better thermoregulation.

Water has a high heat capacity. The capacity is measured by the amount of heat needed to increase the temperature of a volume of water by one degree Kelvin. Scientists use the Kelvin scale as a measure of absolute temperature. It begins at absolute zero, which is equal to -273.1 degrees Celsius or 459 degrees below zero on the Fahrenheit scale.

Water has a tendency to behave differently at different temperatures, Angell says. This plays a crucial role in its behavior at the glass transition point. The ASU scientist had to find a way to account for these tendencies in order to solve the problem.

It may seem strange to think of water forming glass. Glassy water is, however, the most common form of water in the universe. It glazes interstellar dust grains and accounts for most of the water in comets. Understanding why glassy water, not ice, is so common requires a better understanding of glass in general.

For scientists, the term "glass" refers to a subset of amorphous solids. These are substances that cool and become rigid without taking on a crystalline structure. There are many ways to create amorphous solids. Glass typically forms when a substance cools so rapidly that its constituent molecules do not have enough time to organize into an ordered latticework. For slowly crystallizing substances like obsidian and glycerol, "too rapidly" might mean one Kelvin per week. For others, including water, it means more than one million Kelvin per second.

"When crystals are avoided, the liquid solidifies continuously over a wide range of temperatures," Angell says. "This is compared with the sudden solidification at a single temperature for crystals. We call it ‘vitrification' as opposed to ‘freezing,' which happens at a single temperature."

Under normal Earth temperatures and pressures, water resists forming a glass. Pure water's small molecules have a dense network of hydrogen bonds. They are so adept at rapidly rearranging themselves that they easily form crystalline ice instead of amorphous glass. However, when water does form glass, it behaves very oddly during its transition to and from a glassy state. Consequently, glassy water–water in an amorphous solid state, not ice–is perhaps the least understood example of one of the least understood materials on earth.

The majority of glassformers undergo a rapid change in heat capacity as they change from liquid to glass or vice-versa. This change is called the "glass transition." For example, substances solidifying into glass lose their ability to reconfigure from higher to lower energy arrangements. They essentially "get stuck" in their configuration. This limits their ability to absorb heat. The converse occurs when solid glass "melts" into a viscous liquid.

The temperature ranges over which glass transitions take place in most substances can be a narrow band of a few degrees or a broad range of many. But in most materials the heat capacity jump itself is quite dramatic. Even in solutions containing water as the chief component, there is usually a 100 percent change in heat capacity at the glass transition.

Angell says the heat capacity change at the glass transition of pure water is as little as 2 percent. That is nearly impossible to detect. In fact, water's glass transition zone is characterized by a kind of "no man's land." Researchers have difficulty observing the transition in the range of temperatures in which it occurs. It is one aspect of glassy water that has baffled researchers for years.

When heat is applied to pure water in its glassy state, its heat capacity remains nearly flat. That changes when its temperature reaches 136 K (-214.9 F). At that point it begins to ramp up slowly. When the temperature reaches 150 K (-189.7 F) things change quickly. Instead of turning into a highly viscous liquid, the glassy water abruptly crystallizes into ice. Scientists are left to wonder how the tiny heat capacity of the viscous liquid might join up with the high heat capacity of normal liquid water on the other side of the data gap.

Angell says that approaching the transition zone from above produces a similarly enigmatic result. When laboratory water is supercooled, its heat capacity remains constant until around 250 K (-9.4 F). At this point it arcs rapidly upwards until it becomes immeasurable due to crystallization at around 236 K (-34.6 F). Scientists wondered what could be happening in the gap between 150 K and 226 K?

The researchers could apply thermodynamic principles to data above and below the gap to limit how water would behave if it did not crystallize. Still, they could not be sure that the process would be continuous.

Angell and others, in fact, had proposed other theoretical possibilities. These included an exotic "strong-liquid-to-fragile-liquid" first-order transition. The dilemma was that they had no way to measure directly what was actually transpiring on micron-sized droplets undergoing rapid cooling at one million Kelvin per second. Scientists call this "quenching."

The problem seemed intractable. Then Angell noticed work done by one of his post-doctoral fellows. Working at the Tokyo Institute of Technology in Japan, Masaharu Oguni had hit upon a way of stopping water from crystallizing by absorbing it into nanoporous silica glass. He successfully measured its heat capacity right through the "no man's land." Remarkably, his results showed a heat capacity spike consistent with both thermodynamic analysis and direct measurements on supercooled water before it crystallized.

"Water's heat capacity suddenly goes crazy on deep supercooling. And before we can see what is happening, it crystallizes," Angell says. "One trick for finding out what is going on in there is to put the water in confinement–to make it so nanoscopic so that it forgets how to crystallize."

Oguni had squeezed water into 18-angstrom pores, containers roughly five times the scale of atoms and chemical bonds. The tiny space deprived it of the elbow room required for crystallization. It appeared to be the breakthrough Angell needed. Was this finally a window on exactly what was transpiring in the 150 K to 226 K gap? Unfortunately, it proved too good to be true.

Later work by Oguni revealed the problem. His nanoconfined water results were highly sensitive to the confinement media used. More troubling, work by others showed that even traditional glassformers had their usually sharp and distinct glass transitions "smeared out" when they were nanoconfined. It all pointed to the same conclusion, Angell says. Nanoconfined liquids are actually a "different breed" and not representative of glass formation in bulk liquids, including water.

It was a disappointing setback, but Angell still had a few ideas.

"Before the nanoconfined water data came along, we had the supercooled water anomaly and the knowledge that water can be quenched to get this glass. We knew that water exists in some form in ‘no man's land,'" Angell says. "Now it seems we just can't use the porous glass data to say definitively how it behaves, so we are back on our wits."

He had thermodynamic data on supercooled water. He also had data from the glassy states and structures of analogous substances. Combined, the data gave Angell sufficient information to bracket water's possible behavior in "no-man's land." Together with the unreliable nanodata, it also inspired him to formulate a new model of water's behavior near the glass transition.

Angell realized that water's unusual network of hydrogen bonds could cause it to behave like a class of crystalline materials that undergo an "order-disorder transition" during heating from their low temperature ordered states. All of this happens between 140 K and 240 K.

This pseudo-crystalline state has a higher degree of order compared to "normal" liquid water. It also has lower heat capacity. This would explain why there is only a miniscule heat capacity jump at the glass transition, Angell says. It also accounts for the decades of confusion on whether or not water actually has a glass transition.

"It supports the idea that we've got a different sort of thermodynamics in water than we do in other molecular glass-forming liquids," Angell says.

According to the ASU scientist, there are provocative similarities between water's behavior in "no man's land" and the classical tetrahedral-network glasses, like silicon dioxide, as they are heated far above their glass transitions in computer simulations. Making water behave like other molecular liquids requires that you add something to break up the tetrahedral network, he says.

Angell might well have helped solve a central puzzle of solid-state physics. However, his findings must first be converted from experimental observations into a theory that meshes with the overall theoretical framework describing glass.

"I think in its present form, the theory needs some modifications before it will provide the entire picture," Angell says. "Nobody's going to buy it for a while yet. But, the more I see, the more I think it's going to turn out to be the best story we can have. We'll see as time goes on–this may be my legacy."

For more information about the research, contact C. Austen Angell, Ph.D., Department of Chemistry and Biochemistry, 480.965.7217. Send email to caangell@asu.edu