Research Overview and Lab Tour
We use the tools of surface science and chemical reaction dynamics to understand how reactions occur on surfaces. The mechanistic insights we gain uncover new ways of controlling surface chemistry. Our work is relevant to both heterogeneous catalysis and chemical vapor deposition processes. It also provides detailed insight into the energy flow dynamics of complex chemical systems with high internal state densities. 

We study the energetics of activation for important surface reactions, including methane activation on transition metals. Use of supersonic molecular beams and high resolution infrared laser excitation allows us to prepare an ensemble of gas-phase reagents in a single excited quantum state with well defined rotational, vibrational, and translational energies. Varying the beam conditions, or the laser excitation wavelength allows us to systematically change how energy is partitioned among these degrees of freedom and identify which forms of energy best activate reaction.

In our experiments, a supersonic expansion creates a directed beam of molecules with a narrow, yet tunable, translational energy distribution. Light from an infrared laser intersects this molecular beam and excites many of the molecules to a single vibrational state. These molecules, with their well-defined energies, then pass into an ultrahigh vacuum chamber and interact with the surface of interest. By observing how their reaction probability varies as a function of translational energy and vibrational state, we can infer which nuclear motions most effectively promote reactivity and construct a detailed molecular picture of the reaction mechanism.

Two beam-surface scattering apparati are now operational in our laboratory. The figure  highlights  key features of these machines. The reagents expand continuously through an orifice to form a supersonic molecular beam. Infrared light intersects the molecular beam and excites a fraction of the reagents to the desired rovibrational quantum states. A computer-controlled servo system (not shown) actively stabilizes laser frequency to better than 1 MHz and locks the laser output to the absorption feature of interest. A pyroelectric bolometer, which can be placed into the beam path, provides a direct measure of laser excitation of molecules in the beam. A quadrupole mass spectrometer (QMS) located on the beam axis permits time-of-flight analysis of the molecular beam. Auger electron spectroscopy (AES) verifies surface cleanliness and quantifies reaction products. The QMS also quantifies the gas-phase products of temperature programmed desorption (TPD) experiments. (See Publications, McCabe et al., Rev. Sci. Instrum., 2000; and Juurlink et al., J. Phys. Chem. B, 2000)

The heterogeneously-catalyzed steam reforming of methane is the chief industrial preparation of H2. In this reaction, methane reacts with water over a nickel catalyst to produce carbon monoxide and hydrogen gas. Kinetic measurements have revealed that the dissociation of methane onto the nickel catalyst limits the overall rate of this multi-step reaction. Molecular beam measurements and computational studies suggest that there is a significant energy barrier (ca. 70 kJ/mol) to dissociative chemisorption. Our experiments quantify and compare the reactivity of methane molecules containing different kinds of vibrational energy in order to identify the molecular origin of this activation energy.

Our ability to quantify the reactivity of reagents in select quantum states removes the internal state averaging inherent to conventional beam-surface and bulb studies of methane activation, and it permits much more detailed insight into the mechanistic basis for activation and the rates for energy redistribution processes that compete with reaction.

We have found that energy in a C-H stretching mode and translational energy directed along the surface normal are equally effective at promoting reaction on Ni(100) (Juurlink et al., Phys. Rev. Lett., 1999), but that the C-H stretch is more effective than translation in promoting reactivity on Ni(111) (Smith et al., Science, 2004).

We have also explored how energy deposited into different vibrational coordinates promotes reactivity. If energy redistribution prior to reation were fast so that statistical models apply, we would expect expect that reactivity depends only on the total vibrational energy of the molecule. In contrast to this statistical prediction, we find that vibrational energy in a pure bending vibration is even less effective at promoting reactivity than is an equivalent amount of energy in a C-H stretching mode. (Juurlink et al., Phys. Rev. Lett., 2005). The differing reactivity of these two vibrational modes is clear evidence for vibrational mode specific reactivity - i.e. reactivity depends not only on the energy of the reagent, but also on the nature of the excited vibration. These observations suggest that the crucial portion of the reaction coordinate may involve both C-H stretching and deformation character.

We have exploited our enhanced understanding of gas-surface chemistry to identify new approaches to controlling chemical reactivity at surfaces. Homogeneous and inhomogeneous broadening mechanisms that affect laser excitation in the beam further collimate the incident flux of vibrationally excited reagent molecules and permit spatially resolved deposition. (Juurlink et al., J. Phys. Chem. B, 2000). More recently, we have demonstrated an ability to use gas-phase laser exctation of a gas-phase reagent to exert chemical control of the product yield of a heterogeneously catalyzed reaction (Killelea et al., Science, 2008, and Killelea et al., submitted, 2008).

This work has been supported by the National Science Foundation.


A Tour of the Lab

The laser table is located near the center of our lab. A Krypton ion laser is located in the left center of the photo. The red dot is light scattered from the 676-nm beam emitted by this laser. Two mirrors steer the light into the color-center laser located on the front corner of the table. This red light excites a lithium-doped RbCl crystal that is housed inside the shiny metal dewar at the far end of the color-center laser. Liquid nitrogen inside the dewar keeps the color center crystal cold and suppresses the recombination and annihilation of the optically active color center sites in the laser gain medium. Other items visible on the laser table include a wavemeter for absolute wavelength determination, two Fabry-Perot interferometers with free spectral ranges of 300MHz and 1.5GHz, and detectors for laser power measurements. (1GHz corresponds to about 0.03 cm-1, and 1MHz corresponds to about 0.00003 cm-1 in the mid-infrared.) The oscilloscope on the right displays the signal from one of the scanning Fabry-Perot interferometers and indicates that the color center laser is lasing on a single longitudinal mode. The molecular beam machine is located behind the laser table. The end of the ultrahigh vacuum main chamber and the light blue stand on which it rests are visible near the large LN2 dewar.


Taking a couple of steps to the left the lasers are in full view.

The long black Coherent laser (left) is the krypton ion laser that pumps the Burleigh F-center laser (right). The molecular beam machine is visible in the background

Walking around the right hand side of laser table, we see the beam machine.

The main chamber is in the foreground with the interconnected second and first differential pumping stages and the molecular beam source chamber extending to the left.

The differential pumping stages further collimate the molecular beam and reduce the effusive gas load on the main chamber. Thus, nearly all molecules interacting with the nickel surface in the main chamber originate from the supersonic expansion, and not from background gases that may be present in the molecular beam source.

Surface chemistry is highly sensitive to the presence of contaminants on the surface of interest. In order to ensure that our surface is clean and free of unknown contamination, the nickel surface we study is housed in an ultrahigh vacuum chamber (the main chamber), whose pressure is maintained at about 1x10-10 Torr. This level of vacuum minimizes the number of gas-phase species that may interact with, stick to, and contaminate the surface under study. The (brown) manipulator on top of the main chamber and the cold finger extending down from it support the nickel sample. We use a 1-cm disk cut from a single crystal of metallic nickel and polished to expose the (111) (hexagonal-close-packed) or (100) (square surface mesh) crystalline face. The manipulator allows the nickel crystal to be rotated about a vertical axis and translated in three dimensions within the UHV environment. The RF box in front of the main chamber provides RF and DC potentials to the quadrupole mass spectrometer (QMS) in the main chamber. The QMS is located on the machine's beam axis. It is mounted on the end of the main chamber by the small flange visible just above the RF box.

All four compartments of the machine are independently pumped by diffusion pumps (DPs). The DPs are backed by direct drive mechanical pumps that are located in a separate pump room behind the white wall. The main chamber and second stage are protected from backstreaming DP oil by liquid nitrogen-cooled (LN2) traps. The white trap under the main chamber is clearly visible in the picture. The first stage is protected by a water baffle. The source chamber has no trap to keep the pumping speed high.

The first and second stages and the main chamber can be isolated from their DPs by gate valves.The main chamber and second stage have pneumatic gate valves that are individually interlocked to the DP's cooling water flow and DP temperature, the foreline pressure, and the cryogen level in the corresponding LN2 trap.

Walking toward the left along the machine, we end up at the source chamber end. The following picture shows the machine with the source in the foreground.

The source chamber houses the supersonic molecular beam source, which consists of a metal nozzle with a 25um diameter orfice and a 1mm diameter skimmer. The plexiglass box on top of the source chamber contains a step-down tranformer that provides the high current needed to heat the nozzle resistively.

The first differential pumping stage, located in the middle section of the machine, contains a multi-pass cell where infrared light from the color-conter laser intersects the molecular beam and a mechanical shutter for precise control of molecular beam exposure times.

The second differential pumping stage lies between the end of the chamber's middle section and a vertical wall welded inside the right-hand part of the chamber. This wall separates the second stage from the main chamber. The second differential stage contains a slotted chopper wheel that can rotate at speeds up to 400Hz. This wheel modulates the molecular beam and, in conjunction with a quadrupole mass spectrometer located on the beam axis in the main chamber, permits time-of-flight measurements that characterize the translational energy of molecules in the supersonic molecular beam. The second differential stage also contains a bolometric detector for direct measurement of infrared absorption by molecules in the molecular beam. A sliding valve in the second differential chamber can completely isolate the ultrahigh vacuum main chamber from the source and differential pumping stages.

From this point, turning to the left, we can view some of the machine's control electronics.

All the controls at this end of the machine are home built. The chopper motor driver (top right and top left panels), the automatic LN2-trap controller (red rectangles), and the foreline pressure gauges (middle right) are not labeled in the picture.

Looking along the back side of the machine, one sees the remaining controls next to the main chamber. The machine itself is now on your immediate right.

The left rack (gray) contains all the controls for the electron gun and hemispherical electron energy analyzer that comprise the auger electron spectrometer, as well as the ion gauge controllers for the 2nd differential pumping stage and main chamber. The right rack (blue) contains the QMS controller. Variable transformers for the main chamber bake-out heaters and an ion pump controller are also located in these racks. On the right hand side of the picture one can see the closest of the pneumatic gate valves and a part of the hemispherical analyzer that is wrapped up in aluminum foil for bake-out.

Turning to the right and looking into the main chamber through a large window, one sees the nickel crysal at the end of the cold finger.


The nickel crystal is suspened from two tungsten rods extending from the bottom of the cold finger. It is surrounded by a copper cold shield that minimizes radiative heating of the crystal by the chamber walls. The cold shield is in thermal contact with a LN2 reservoir. A thoriated tungsten filament located behind the crystal permits electron bombardment heating of the crystal. Liquid nitrogen cooling coupled with electron bombardment heating permits the crystal temperature to be varied from 78K to over 1000K. The entire crystal support assembly can be rotated around a vertical axis and translated in 3 dimensions. Combinations of crystal rotation and translation position the crystal for dosing (the molecular beam enters the main chamber from the right in this picture), argon ion sputter cleaning, or for auger electron spectroscopy, electron energy loss spectroscopy, or temperature programmed desorption measurements.