by J. William Bell, NCSA Senior Science Writer

Champaign, IL — For all the transparency that its name implies, the greenhouse effect and its impact on global warming is a muddied topic. Many scientists believe that greenhouse gas emissions will cause average global temperatures to rise by almost 6 F degrees over the next 100 years. Other best guesses put the number at something closer to 3.5 F degrees. And contrarians maintain that there is no compelling reason to think that a rise in global temperature is caused by increases in greenhouse gas emissions in the first place.

Few, however, dispute that the levels of heat-trapping greenhouse gases such as carbon dioxide, methane, and chlorofluorocarbons have risen tremendously in the last 100 years. The atmospheric carbon dioxide level, for example, is up about 25 percent since the late 1800s, with most of this rise coming in the last 50 years alone. That level is higher than it has been in the last 160,000 years, and the burning of fossil fuels is the largest contributing factor.

“There’s no doubt that carbon dioxide levels have increased since the industrial revolution,” says James Kirkpatrick, a geology professor at the University of Illinois at Urbana-Champaign. “But is the carbon dioxide increase causing global warming through the greenhouse effect? And if it is, what can be done about it?”

Kirkpatrick and his colleague Andrey Kalinichev are currently working on the chemistry behind that second question. They create molecular dynamics models of carbon dioxide and other chemical species as they dissolve in water, as well as models of that water-carbon dioxide solution as it interacts with mineral surfaces. These simulations, which are being run on NCSA’s SGI Origin2000 supercomputer, will help researchers develop methods of “sequestering” carbon dioxide – injecting it deep into the ocean or a deep groundwater aquifers where it won’t interact with the atmosphere and won’t have the same negative environmental impact.

“We’re focused on the fundamentals here, but there’s a broad societal connection to the science,” says Kirkpatrick.

Carbon dioxide injection has been used for years in the oil and natural gas industry to enhance the amount of fuel extracted from underground deposits. Field testing of deep injection as a means of sequestering excess carbon dioxide created when burning fossil fuels, however, is a more recent undertaking, with many projects cropping up only in the last 10 years or so. Regardless of the intended goal, the process is easily understood. The gas is captured, highly pressurized, piped to a site, and then injected deep into the ground or ocean.

If injected into porous, isolated rock, the carbon dioxide interacts with whatever minerals are present and may not have a negative environmental impact. If that porous rock happens to be an oil reservoir, the carbon dioxide helps move the oil out the well. When injected into the water of an aquifer or the ocean, the carbon dioxide may simply dissolve into a solution with the water. If the pressure is high enough, the carbon dioxide will take on a “supercritical” liquidlike state, remain separate from the water, and not interact much at all.

“One of carbon dioxide’s main sinks is the ocean, anyway – most of it ends up dissolved there as a part of the natural carbon cycle. In a way, sequestration just speeds up that process,” says Kirkpatrick. And, according to a 1997 study by the Center for Energy and Environmental Studies at Princeton University, thousands of years of excess carbon dioxide produced by the burning of fossil fuels at the current rate could be managed using aquifer and ocean sequestration.

The overall impact is still uncertain, though. “Nobody knows these [products of sequestration] and their environments well enough to know which of these approaches might be best,” says Kalinichev, a visiting researcher at the U of I and head of the Physical Research Laboratory at the Institute of Experimental Mineralogy in Chernogolovka, Russia.

Kalinichev and Kirkpatrick’s research is the horse that has to go in front of the cart. Before undertakings like sequestration can be fully understood, the physical and chemical properties of water and carbon dioxide solutions and how they interact with their surroundings have to be brought into relief.

To do that, the team focuses closely on the hydrogen bonding between the molecules in their simulated solutions. When a hydrogen atom bonds to another atom that strongly attracts electrons, the resulting molecule is very polar, with one end strongly positive and one strongly negative. Hydrogen bonds form between the opposing ends of these polar molecules.

In the sorts of environments where carbon dioxide sequestration would be most common – under thousands of feet of earth or ocean – temperature and pressure vary widely, from near freezing to 400 F and with pressures of up to 1,000 times the atmospheric pressure at sea level. Accordingly, the team uses these two factors as their two most common thermodynamic variables.

“The dissolved species are dynamic objects. Hydrogen bonding is constantly changing. The models allow us to estimate lifetimes of different bonds under different conditions and states,” says Kalinichev.

Already they have discovered that hydrogen bonding is reduced at high temperatures, while pressure has little impact. The reduced bonding makes carbon dioxide – which does not readily bond, and thus dissolve, because it is not very polar – more soluble under what would be common sequestration conditions. And by understanding the hydrogen bonds, the team can also predict properties such as density, viscosity, diffusion rates, and heat capacity under changing conditions.

The models previously included only pure water, but they now consider carbon dioxide, carbonates like limestone that might make up a mineral wall that the water carbon dioxide mixture would interact with, and salts like sodium chloride, as well. Adding these compounds – and basing the models on first principle calculations – brings the simulations much closer to the real world. The amount of time required to complete these models is, however, and there’s still a great deal of work to be done. One picosecond of the molecular dynamics simulation requires about an hour on five Origin2000 processors, and a typical run tracks the behavior of only several thousand atoms over the course of several hundred picoseconds.

“We’re still simulating a relatively small number of molecules, therefore we are applying so called periodic boundary conditions to simulate bulk aqueous solutions and their interactions with mineral surfaces,” says Kalinichev. “But you have to begin with these mechanical descriptions of each molecule to extract the information that you want.”

Kirkpatrick adds, “Today’s science requires – absolutely requires – thinking on the molecular scale to understand what takes place on the macroscopic scale.”

This research is supported by the Center for Advanced Cement Based Materials, the National Science Foundation, and the Department of Energy Basic Energy Sciences Carbon Management Program, Geosciences Division.

============================================================