Impact of Mass Spectrometer Capillary Probe on the
Measured Concentration in a Monolith Reactor
Hoang Nguyen, Dan Luss*, and Michael P. Harold*
Detailed spatiotemporal information is crucial to elucidating the behavioral features of heterogeneous catalytic reactions. The technique called capillary inlet mass spectrometry (Spaci-MS), first developed by Partridge and coworkers at Oak Ridge National Laboratory (ORNL), enables spatially-resolved concentration measurements using a fused-silica capillary positioned in a monolith channel during a catalytic reaction . Suction gas samples are analyzed with a mass spectrometer to provide time-resolved species concentrations while translation capability of the probe enables their spatial dependencies. Currently, there is a debate on the veracity of the Spaci-MS technique because the probe may influence the conversion in the measured channel due to blockage of the flow by the probe. Research carried out by Deutschmann and coworkers  used computational fluid dynamic modeling with spatially-resolved mass spectrometer measurements to show that during catalytic methane partial oxidation the blockage effect was significant even when the probe occupied only 3.5% of the channel area. It is of interest to the catalysis and reaction engineering community to further elucidate the utility of Spaci-MS method.
The objective of this work is to systematically study the impact of the capillary probe during in situ measurements in a monolith reactor. To this end, we carried out propylene oxidation on Pt/Al2O3 washcoated monoliths with a range of channel densities (100, 200, 400, and 600 cpi) with concentration sampling by probes of two different sizes (OD = 170 & 363 µm). We positioned another stationary optical fiber probe (OD = 125 µm) in an adjacent channel to monitor the intra-channel temperature using the recently-developed coherent Optical Frequency Domain Reflectrometry (OFDR) technique, applied for the first time to catalytic reactions by Nguyen et al. . Our approach is to compare the Spaci-MS measured propylene concentration profile and the OFDR measured temperature profile. A spatial shift in the location of the concentration (temperature) decrease (increase) would indicate a disturbance by the probe on the extent of reaction.
Steady state profiles were obtained for lean propylene oxidation (0.2 vol.% C3H6, 10 vol.% O2) for different combinations of the channel size and probe diameter, and the feed gas space velocity and temperature. Our findings reveal a complex interaction of several factors, including, in addition to the probe diameter and channel size, the axial position of the probe, as well as the feed flow rate and temperature.
Typical results in Figure 1 show the spatial dependence of propylene concentration and temperature when the gas mixture having a space velocity of 17400 hr-1 and a temperature of 200°C was fed to a 600 cpi monolith (Pt loading = 23.8 g/.ft3). In this experiment, the 363 µm OD probe occupied about 10% of the channel area. A comparison of the two profiles indicates a negligible shift between the location of the maximum temperature and the propylene depletion point (Dz ≈ 4 mm). In this case the probe had a small impact. In contrast, Fig. 2 shows the results of an experiment in which the gas was fed with a space velocity of 12630 hr-1 and a temperature of 180˚C to a 400 cpi monolith (Pt loading = 96 g/.ft3). A comparison of the temperature and concentration profiles revealed a major probe impact (Dz ≈ 12 mm). Furthermore, the same shift was obtained for two different capillary probe sizes (170 µm and 363 µm OD). One might have expected that the larger probe would block more flow and lead to a shift in the concentration profile upstream because of the lower flow rate and higher residence time, following the findings of Deutschmann et al. . However, we suspect that the larger probe increases both the blockage and the flow suction, thereby compensating for the blockage effect. To confirm this conjecture, we carried out the same experiment but with a higher total flow rate (space velocity = 17400hr-1) while keeping other operating conditions constant. Figure 3 shows that the concentration profile measured by the bigger probe was shifted towards the upstream of the reactor compared to that of the smaller probe. This result indicates the stronger impact of the blockage effect compare to the suction effect.
These and other experiments to be described indicate that suction effect, channel density, and reaction rate (catalytic activity) are important and should be taken into account in the analysis of Spaci-MS data. Additional experiments will be described for different combinations of the channel and probe size. We will discuss the important factors that contribute to the observed trends, and conditions are determined for which the impact of the probe is negligible.
Figure 1: Simultaneously measurements of temperature and concentration inside the 600 cpi channel monolith reactors.
Figure 2: Intra-channel concentration measurements by two capillary probes (170 µm and 363 µm) and temperature measurement by the optical fiber (OD = 125 µm) in a 400 cpi monolith at 12630 hr-1 space velocity respectively.
Figure 3: Intra-channel concentration measurements by two capillary probes (170 µm and 363 µm) and temperature measurement by the optical fiber (OD = 125 µm) at 17400 hr-1 space velocity in a 400 cpi monolith respectively.
 J.S. Choi, W.P. Partridge, C.S. Daw, Spatially resolved in situ measurements of transient species breakthrough during cyclic, low-temperature regeneration of a monolithic Pt/K/Al2O3NOx storage-reduction catalyst, Appl Catal a-Gen 293 (2005) 24-40.
 H. Nguyen, M.P. Harold, D. Luss, Optical frequency domain reflectometry measurements of spatio-temporal temperature inside catalytic reactors: Applied to study wrong-way behavior, Chem Eng J 234 (2013) 312-317.