As the x-ray photon energy scans from (say) 20 eV below to 20 eV above the photoionization threshold for 1s electrons in molecules comprising atoms with low atomic numbers and p bonding, a progression of typical spectral features appears. The sequence begins at the lower energy end with the familiar p* (Feshbach) resonances that are so useful in molecular fingerprinting and in calibrating monochromators, progresses through a number of Rydberg series just below threshold, and ends above threshold with overlapping features due to doubly excited (two electrons in excited states) molecules and shape resonances.

A s* shape resonance above the ionization potential results from a potential (centrifugal) barrier that temporarily prevents electrons excited in the resonance energy range from leaving the molecule.

The shape resonance results from a potential barrier that temporarily prevents electrons excited in the resonance energy range from leaving the molecule. Their apparent simplicity, broad energy widths, and well-defined symmetries have made shape resonances a popular object of study in gas-phase molecular science and a common tool for probing orientations and bond lengths in molecules adsorbed on solid substrates. Unfortunately, picking out shape resonances from other spectral features is not so straightforward, thereby raising questions about the validity of parameters extracted from their analysis. For example, peaks due to double excitation can be experimentally indistinguishable from shape resonances.

Fragmented Molecules Reveal Their Secrets

Compared to complex solids or large biological molecules, small molecules with only a few atomic nuclei surrounded by electrons are simple structures. But, belying their relative simplicity, small molecules display a bewildering forest of spectral signatures when beams of x rays illuminate them over a range of wavelengths. Since deciphering spectra is the key to understanding molecular behavior, any new tool that can selectively cut through the forest and classify specific spectral features is more than welcome.

Enter an international team working at the ALS, which has hit upon a way to conclusively identify hitherto elusive shape resonances, relatively broad and featureless spectral peaks that researchers widely believe to be present in molecular x-ray spectra but nonetheless find hard to pin down. The key is the fragmentation into both positively and negatively charged entities (cations and anions, respectively) that often occurs after x-ray absorption. In experiments on carbon monoxide, where a shape resonance is known to exist, it turns out that the resonance does not show up when measuring the yield of oxygen anions but it does for cations. Measuring both species thereby unmasks the shape resonance.

Against this background, carbon monoxide presents itself as a valuable test bed. It displays a broad peak that is known to be due to a shape resonance and is readily distinguishable from sharper features at somewhat lower energy from doubly excited molecules. Led by researchers from the University of Nevada, Las Vegas, the international team extended the well-known technique of ion-yield spectroscopy to examine negatively charged ion species, as well as the commonly measured positively charged ions, that are created when the molecule breaks into fragments during the decay process following x-ray excitation. The high spectral resolution and signal intensity achieved at ALS Beamline 8.0.1 made anion spectroscopy feasible.

Partial oxygen-ion-yield spectrum at the carbon K edge of carbon monoxide shows the typical collection of spectral features, but the shape resonance is conspicuously absent.

Several pathways exist for creation of ions from simple photoemission, which leaves a CO²⁺ ion, to photofragmentation, which can result in several cation species and O^- as the primary anion. Measuring the O^- production as a function of photon energy near the carbon K edge yielded a spectrum containing the various features enumerated above but with no apparent contribution from the known shape resonance. Detailed comparison of cation and anion spectra above the photoionization threshold conclusively demonstrated the absence of the resonance clearly visible in the cation spectra. Anion measurements at the oxygen K edge were only somewhat less conclusive, owing to more overlap with the region of doubly excited states.

Detailed comparison of partial cation- and anion-yield spectra above the carbon K edge graphically demonstrates the absence in the anion spectrum of the broad shape resonance seen in the cation spectra.

The main message is that shape resonances are completely absent in anion yields. Moreover, there is nothing special about carbon monoxide, so the researchers argue this new approach should apply to many small molecules, thereby providing a new tool to examine core-level resonant processes in general and sort out shape resonances in particular.

Research conducted by W.C. Stolte (University of Nevada, Las Vegas, and ALS); D.L. Hansen (Jet Propulsion Laboratory); M.N. Piancastelli (University of Rome, "Tor Vergata," and ALS); I. Dominguez Lopez (Centro Nacional de Metrologia, Mexico); A. Rizvi and A.S. Schlachter (ALS); O. Hemmers and D.W. Lindle (University of Nevada, Las Vegas); H. Wang (Lund University, Sweden); and M.S. Lubell (City College of New York).

Research Funding: National Science Foundation; U.S. Department of Energy (DOE), EPSCoR; National Research Council. Operation of the ALS is supported by the U.S. DOE, Office of Basic Energy Sciences.

Publication about this research:W.C. Stolte, D.L. Hansen, M.N. Piancastelli, I. Dominguez Lopez, A. Rizvi, O. Hemmers, H. Wang, A.S. Schlachter, M.S. Lubell, and D.W. Lindle, "Anionic Photofragmentation of CO: A Selective Probe of Core-Level Resonances," Phys. Rev. Lett. 86, 4503 (2001).

ALSNews Vol. 187, October 31, 2001