Hopkins Geochemist Develops Promising Computer Model
The development, devised by Dimitri Sverjensky, may usher in a wide range of benefits, in areas as diverse as pollution control, chemical production and geological research. He has recently developed a theory to describe how protons attach to negatively charged ions on the surfaces of solids in water. The new theory enables the model to be used for a large variety of general applications involving the reaction of minerals, and chemicals called crystalline solids, with water.
Dr. Sverjensky discussed the new theory May 24 during a meeting of the American Geophysical Union in Baltimore. The Hopkins professor of geochemistry, in the Morton K. Blaustein Department of Earth and Planetary Sciences, will lead the symposium, entitled Geochemistry of Mineral-Water Interfaces, which is sponsored by the Geochemical Society.
The computer-model technique could represent a significant improvement over conventional, and sometimes prohibitively expensive, methods of finding mineral deposits and analyzing other reactions deep underground.
"We need to be able to predict how minerals react with water, not only under surface conditions but also at very high temperatures and pressures," Dr. Sverjensky said. "We need theoretical geochemistry because it's very expensive to drill holes in the ground, extremely expensive."
The method eliminates the need to conduct separate laboratory experiments for each mineral, since the calculations can be done with computers. And, because the Earth contains about 3,000 minerals, all existing in various combinations with each other, the computerized approach is an exceedingly practical way to predict how minerals in nature will react with water. It could take many years to measure the multitude of reaction rates for the plethora of mineral combinations and different "water types" by using traditional laboratory methods, Dr. Sverjensky said.
Water exists in many different chemical compositions that can vary drastically, with a wide range of acidity and numerous kinds of chemical constituents. The fact that the Earth contains so many varieties of water is another reason that theoretical models are needed, since the different compositions affect how minerals react with water.
The technique could prove especially helpful for industry, where thousands of crystalline solids are essential for a multitude of chemical processes. By using the computer-model technique, the reactions of various minerals and crystalline solids, in any combinations, could be predicted, reducing laboratory costs. Dr. Sverjensky's research has been supported for three years by grants from Du Pont.
The model has other economic implications, as well, because understanding how water dissolves minerals locked up in rock is essential to tracking the formation of ore deposits. Learning how fast various minerals dissolve also is essential to analyzing a key environmental process called weathering, in which soil is ultimately produced when acidic water reacts with minerals. Since it might prove to be the first method to accurately predict the rates of various chemical reactions between minerals and water, the method could be useful in learning how long soil formation takes and how fast pollutants migrate underground.
The fundamental factor that controls how fast many minerals dissolve in water is how many protons are contained in the water. The more protons, the more acidic the water is. The surfaces of the crystal structure of minerals called oxides and silicates contain negatively charged oxygen ions. These attract protons that exist in the water, in a chemical reaction called surface protonation, which is the process of protons in water attaching to the surface of a solid. Earth's crust and mantle are made up almost entirely of oxides and silicates.
Other scientists have demonstrated that the rates at which minerals dissolve are directly related to the number of protons that attach to a surface. And research shows that the amount of proton absorption can be predicted simply by knowing the crystal structure of a given mineral.
"So if you can predict the number of protons that are stuck on the surface, then you've got the rates," Dr. Sverjensky said. "It's actually a terribly simple calculation. I do it all on the Macintosh."
He describes his use of the new theory of surface protonation in a scientific paper published this month in Geochimica et Cosmochimica Acta, an international geochemistry journal. A paper on the application of the new theory for predicting reaction rates is being submitted for publication in another journal.
"Up until now, although people have realized the connection between rates and surface protonation, they have always had to have experimental measurements of the surface protonation to use for the rates," Dr. Sverjensky said. "I'm saying that we now have a theory for predicting surface protonation, instead of individual experimental studies of each mineral one at a time."
Go to Headlines@HopkinsHome Page