Molecular Beams and Nanoreactors

New methodology is being developed in our group to measure the kinetics of catalytic reactions under well-defined conditions.  The main idea behind this work is to perform dynamic measurements on single-crystal surfaces and under a control environment but mimicking the high coverage and multi-component conditions typical of atmospheric catalytic reactions.  In the past kinetic parameters have been determined for high probability reactions such as CO oxidation or NO reduction by using an effusive molecular beam source in an arrangement derivative from that originally developed by King and Wells but where the emphasis is placed on measuring reaction rates at high gas fluxes and surface coverages.  Our initial setup was first tested for the uptake of CO on Pt(111) as a function of CO beam flux, surface-to-doser distance, and surface temperature [J. Liu, M. Xu, T. Nordmeyer and F. Zaera, J. Phys. Chem. 99 (1995) 6167].  A homogeneous flux across the surface was obtained by placing the ~10 mm-diameter sample somewhere between 5 and 25 mm away from the beam source, an 12 mm-diameter capillary array.  Only moderately low fluxes were used, equivalent to local pressures at the surface on the order of 1E-6 to 1E-4 Torr, but the fact that the measured fraction of the beam intercepted by the sample reproduces reasonably well estimates from calculations at many doser-to-surface distances attests to the collimated nature of the beam.

Schematic representation of the effusive molecular beam setup used for our dynamic kinetic measurements.   This design is based on that originally developed by King and Wells, and consists of a collimated beam directed towards the surface while the scattered molecules are detected as a function of time by mass spectrometry.  A flag is used to intercept the beam at will in order to control the experimental sequence.

In order to be able to design instruments capable of reproducing atmospheric-pressure environments, some effort has been placed on trying to understand the factors that control the collimated nature of the beam as its flux is increased.  Our goal is to direct high-flux beams to the front face of a single-crystal sitting in a UHV environment in order to achieve molecular impinging frequencies comparable to those encountered in typical catalytic systems.  An instrument was built to measure the beam flux and profile of the gas beams emanating from capillary arrays by mounting that gas source on a x-y-z stage in front of a skimmer connected to a differentially-pumped detection chamber [J. M. Guevremont, S. Sheldon and F. Zaera, Rev. Sci. Instr. 71 (2000) 3869].  Initial measurements with such an apparatus have indicated that molecular flow, a requisite for maintaining beam directionality, can be attained with backing pressures of up to 30 Torr (see figure below).  This leads to an impinging frequency roughly equivalent to a pressure on the surface of about 30 mTorr, sufficient to perform steady-state experiments on most catalytic systems.

Conductance and peaking factor of beams produced by an array of 10 µm diameter capillaries as a function of backing pressure.  The conductance of this source remains constant in the molecular flow regime until the drop seen at pressures between 5 and 10 Torr.  It does increase again once it reaches a viscous flow regime, and the peaking factor, a measure of collimation, also improves somewhat after reaching a minimum about 50 Torr.  Inset: Spatial distribution of the gas beam generated by the 10 µm-diameter capillary array, highlighting again the peaked nature of the resulting beam.

Beam profile measurements were performed as well.  An example of the results obtained from those studies is shown in the figure below.  There, z represents the distance between the surface and the beam, and x the relative transverse distance between the centers of the beam and the skimmer.  It is seen that even though the absolute intensity of the center of the beam decays rapidly as it is moved away from the skimmer, the width deteriorates at a much slower rate.  Also encouraging is the fact that similar profiles have so far been obtained with backing pressures of up to 10 Torr, which represent a total flux equivalent of about 10 mTorr on the surface.

Three-dimensional spatial profile of an Ar molecular beam generated by using the capillary array described in the text.  Notice that even though the total flux decays rapidly with distance, the spread does not deteriorate in any significant way.  This collimated nature of the beam is critical for the design of high-flux catalytic experiments.

An alternative approach also being pursued in our laboratory is the use of a retractable nanoliter-sized reactor to be placed right on the front surface of a single-crystal solid sample.  This instrumentation aims to address three kinetic issues not previously resolved: (1) to allow for the measurement of low-probability reactions such as most hydrocarbon conversions; (2) to identify the changes in catalytic activity and selectivity induced by variations in pressure over the whole range from UHV to atmospheric; and (3) to discriminate between the reactivity of the crystalline surface of interest and that of the associated polycrystalline edges and wires used to mount the sample.  In our design, the reactants are introduced via a small micro-capillary and trapped within a volume of a few nanoliters directly above the surface of the catalyst by an outer tube that also serves as the conduit for gas pumping and sampling.  The products are then detected in a separate differentially-pumped chamber equipped with a mass spectrometer.  A picture of the prototypical instrument we have already developed for this is provided below.

Picture of the prototypical nanoliter-sized reactor built in our laboratory for kinetic studies of catalytic reactions.  A single crystal is attached to the UHV manipulator seen in the upper part of the picture via two electrical feedthroughs, which are also used for cooling and heating of the sample.  The retractable probe that approaches the sample from the bottom left consists of two concentric tubes, an inner capillary used as the gas inlet, and an outer cone to define the volume of the reactor.  Gas flow and gas detection are achieved by differential pumping of the trapped volume into a second chamber equipped with mass spectrometry.

The gas flow behavior of this system has been characterized as a function of the separation of this nano-reactor from the sample, and the gas detection during catalysis has been tested for the oxidation of carbon monoxide on Pd(111). Preliminary data from the latter measurements are shown below.

Data from a preliminary kinetic run for the oxidation of carbon monoxide on Pd(111) at 490 K using the nanoliter-sized reactor described above.  The uncalibrated mass spectrometer signals for CO, O2 and CO2 are shown here as a function of time as the gas mixture is introduced into the reactor.  A CO:O2 1:2 mixture was used, and a final total flux of approximately 7E19 molec/cm^2 was attained, equivalent to a pressure of 200 mTorr on the surface.  The small size of signal for the CO2 production may be explained by a combination of a lower mass spectrometer sensitivity and/or a viscous flow pattern in the reactor.