During a chemical reaction, a molecule can go through some changes; for example, its atoms can change position or the molecule can break up or become part of other molecules. To see exactly what happens, one must precisely track the positions of all the atoms in a molecule as the reaction unfolds.
Such a technique does not yet exist, but Los Alamos theorists Suxing Hu (now at the Laboratory for Laser Energetics, University of Rochester) and Lee Collins, along with Barry Schneider of the National Science Foundation, recently used supercomputer simulations to explore a technique that could one day do the trick.
The technique’s potential was proved experimentally in 2005 by researchers at the University of Arizona. The main result of that experiment was a plot of the distance, as a function of time, between two iodine atoms drifting apart after a laser pulse split an iodine molecule in two. This plot is now referred to in the scientific literature as a “movie.”
To make the movie, the Arizona researchers first hit an iodine molecule sporting an extra electron with a 100-femtosecond pulse of infrared laser light. (A femtosecond is a million-billionth of a second.) The pulse broke the molecule into an ordinary atom and an atom with an extra electron (an ion). Then the researchers hit the atom/ion pair with another 100-femtosecond pulse, this time of ultraviolet laser light, to eject the extra electron.
The movie was possible because a subatomic particle such as an electron can behave like a wave and because any wave can be diffracted (scattered) by two closely spaced objects to produce multiple waves. When these waves contact (interfere with) each other, they create a distinctive interference pattern of regularly spaced bright spots that can be analyzed to reveal the distance between the two objects that caused the scattering.
In the Arizona iodine experiment, the electron wave was diffracted by the nuclei of the two iodine atoms produced when each iodine molecule was split apart. The experiment involved huge numbers of split iodine molecules, resulting in huge numbers of ejected electrons that, after being diffracted, impinged on a flat-plate detector some distance from the scene of the molecular crime to form an interference pattern. Analysis of the pattern revealed the distance between the two iodine nuclei.
By running the experiment many times, while systematically increasing the time between the infrared and ultraviolet pulses, the researchers ended up with a series of interference patterns that showed—one pattern (one “movie” frame) at a time—the increasing distance between the two atoms as they drifted apart.
This arcane cinematic adventure probably wouldn’t win an Academy Award, but the experiment proved the technique’s very-real potential for precisely capturing the motion of individual atoms. However, some features of the experiment could be altered to make it more suitable for making movies of chemical reactions of more general interest. The theorists took that fact into account when designing their supercomputer simulations.
For one thing, the spatial resolution of the experiment, the smallest distance (between the two iodine atoms) that could be captured in one frame, was about 1 nanometer (one-billionth of a meter), acceptable for the iodine experiments but too big for chemical movies involving molecules that contain hydrogen, such as hydrocarbon molecules or biomolecules. The distance between adjacent hydrogen atoms is about 0.1 nanometer.
The spatial resolution can be sharpened by increasing the energy of the ejected electron, whose maximum energy is the energy of the light particle (photon) in the pulse that ejects it, at least for the most common case, in which a molecule absorbs only one photon.
To make the spatial resolution about 0.1 nanometer, the photon energy would need to be a few hundred electronvolts (eV).
Moreover, although the 100-femtosecond ultraviolet pulse in the Arizona experiment was short enough to stop (as a camera “stops”) the motion of the relatively ponderous iodine atoms, hydrogen atoms are distinctly “nonponderous.” As the lightest atoms, hydrogen atoms move much more quickly than iodine atoms do. To stop their motion, the pulse would need to be a minimum of a few femtoseconds.
Taking all these considerations into account, the theorists decided to simulate the response of a positively charged hydrogen molecule to an intense, ultrashort pulse of x-ray laser light. A positively charged hydrogen molecule is the simplest molecule imaginable: one electron and two protons, the protons being the molecule’s two atomic nuclei. Even so, each simulation required 15 hours and 480 processors on Los Alamos’ Coyote supercomputer. The theorists have since run simulations on 1,000 processors on Los Alamos’ Lobo supercomputer and 5,000 processors on the National Science Foundation’s TeraGrid computer at Oak Ridge National Laboratory. The method seems to scale well with increasing numbers of processors.
In the simulations, the x-ray pulse ejected the electron from the molecule, and the molecule’s two protons diffracted the electron’s wave function.
The theorists found that when the photon energy of the x-ray pulse was set high enough (greater than 350 eV), a clear interference pattern was produced that resolved the distance between the molecule’s two protons—0.133 nanometer. This is about the shortest distance between two atoms in any molecule.
Such exquisite resolution is possible because the ejected electron (with an energy of 350 eV) behaves like a wave with a wavelength about equal to the distance between the two protons.
Moreover, the duration of the simulated x-ray pulse was so short (one-quarter or one-half a femtosecond) that it could easily freeze the relative positions of the protons in time, producing a very clear interference pattern.
In addition, the x-ray pulses in the simulations were of very high intensity. That fact made it easier to image the protons’ positions because the electric field of each pulse was so high that it quickly pulled the electron from its molecular orbit and, as the field reversed direction several times during the pulse’s duration, drove the electron back and forth through the molecular interior before it was ejected. During its intramolecular trip, the electron passed near the protons many times, so in effect, the protons diffracted the electron’s wave many times, producing a highly intense interference pattern especially easy to measure.
The electron ejected from a positively charged hydrogen molecule (H2+) by an x-ray photon is diffracted by the molecule’s two protons to form an interference pattern. The ejected electron’s angular distribution gives the distance between the two protons. The blue arrow shows the direction of x-ray-photon propagation. The yellow arrow shows how the photon’s plane of polarization is oriented.
A few years from now, it’s likely that several pulsed x-ray sources now being developed will have the photon energy, pulse duration, and beam intensity required to perform experiments like the simulated ones.
However, making useful chemical movies will require an important addition. In the simulations, only one laser pulse was used—to produce a single interference pattern. To make a multiframe movie, two pulses will be required. Similar to the iodine experiment, researchers will use the first pulse to initiate the chemical reaction and the second to liberate an electron to form a diffraction pattern. They’ll also need to increase the time delay between the two pulses in successive repetitions of the experiment, as in the iodine experiment.
In this issue...