|
Mazziotti explains his method in the Nov. 19 issue
of Physical Review Letters. Further details will follow in early
December in the Journal of Chemical Physics.
The key to understanding whether or not a
particular chemical reaction will occur depends on a detailed
statistical description of the electrons' positions in the molecules
involved. Until now, scientists have found it necessary to attempt to
represent the motion of all the electrons in the molecule of
interest-a daunting task requiring vast quantities of computer power.
"Just a single water molecule has 10 electrons," Mazziotti said.
But in the 1950s, researchers theorized that it
should be possible to accurately and more efficiently calculate the
electronic properties of a molecule using only a pair of electrons
representing many-even hundreds-of electrons in a molecular system.
Mazziotti compares the feat to assembling a set of architectural
blueprints, which represent in two dimensions a structure that can be
built in three dimensions.
An architect follows certain rules to ensure that a
builder can translate a two-dimensional sketch into a
three-dimensional structure. "In the same way atoms and molecules
consist of many electrons, but there is a way to represent all of the
electrons rigorously with only two electrons. Certain rules have to be
followed to ensure the two-electron 'sketch' of the molecule
accurately represents all the electrons in the atom or molecule,"
Mazziotti explained.
Mazziotti's Physical Review Letters paper realizes
a dream that scientists have pursued for 50 years by introducing a set
of instructions for accurately and efficiently computing with a pair
of electrons that represent the many electrons of the molecule. These
instructions dramatically reduce the amount of computer time and
memory required to compute the electronic properties of a molecule.
Now Mazziotti can do some of the same calculations on his desktop
computer that previously required Japan's Earth Simulator, the world's
largest supercomputer.
"David has really made a huge contribution in
turning the dreams of 50 years ago into useful tools," said Bob Erdahl,
a professor of mathematics at Queens University in Kingston, Ontario.
Erdahl said Mazziotti's Physical Review Letters paper has applications
to his own research in computing how behavior at the subatomic level
brings about macroscopic changes in materials, such as the transition
to the superconducting state.
"I'm certainly going to look very closely and try
to incorporate David's latest innovation into my work. I think we will
very quickly be able to beat other approaches in this area of solids
and compute things that were out of reach before," Erdahl said.
Erdahl is especially interested in determining why
superconductivity manifests itself only in two-dimensional layers
rather than in three-dimensional solids. "The computations are of
course very difficult to do. These methods that David is developing
and that we're developing are very helpful in attacking that problem."
While Erdahl works in mathematical physics,
understanding the electronic energies that atoms and molecules possess
also affects almost every area of chemistry. One such area, Mazziotti
said, has broad applications includes the chemistry of free
radicals-highly reactive unpaired electrons.
In atmospheric chemistry, free radicals are
instrumental in reactions leading to ozone depletion and the creation
of greenhouse gases. Another area is the combustion of hydrocarbon
fuels, which creates a variety of carbon-based radicals.
"A lot of people want to know which radicals are
present in a given combustion process and what reactions those were
undergoing because that's going to affect fuel efficiency," Mazziotti
said.
A third area is medicine, because radical-type
reactions are common in the human body. Mazziotti noted that hydroxy
urea therapy combats sickle-cell anemia by forming a radical that
triggers a cascade of additional reactions. "There is a 40 percent
reduction in mortality for patients who receive hydroxy urea treatment
for sickle cell," he said.
Despite the advances that Mazziotti and others have
contributed to the representation of electrons in atoms and molecules,
further advances could be in the offing. He said the field is
experiencing a new wave of research. "We're not done by any means." |
