There have have been several recent reports regarding the spatial resolution of gas species in monolith-supported automobile exhaust catalysts, with examples including characterization of DOCs (diesel oxidation catalysts) and LNTs (lean NOx traps). However, spatially resolving gas concentrations in NH3-SCR (selective catalytic reduction) applications is limited due to the difficulty in ppm-level NH3 detection in the presence of percent levels of water and N-2 using mass spectrometry. In this study, a gas-phase FT-IR analyzer was used for the first time to spatially resolve gas concentrations in a monolith-supported SCR catalyst. The reactions analyzed include standard SCR, fast SCR and SCR with NO2. The results show that the three SCR reactions proceed at significantly different rates, especially at temperatures below 300 degrees C, and can be correlated to the amount of catalyst used. For example, the catalyst lengths needed to achieve 80% NOx conversion at 300 degrees C are 2.4, 1.2 and 0.5 cm for conditions that meet the standard SCR, NO2-SCR and fast SCR reaction stoichiometries, respectively. For the standard SCR reaction, kinetic analysis, and spatially resolved NO oxidation and SCR results consistently indicate that the rate-determining step is NO oxidation. NH3 has an inhibition effect, as it suppresses NO oxidation by competitive adsorption on the active sites. At 300 degrees C, the outlet NOx conversion is not limited by the reaction kinetics, but by insufficient NH3 supply, since part of the NH3 is oxidized by O-2. Compared with 300 degrees C, higher NOx conversions are attained at 400 or 500 degrees C, which is due to significantly enhanced NO oxidation, and the resulting increase in NH3 reacting with NOx via SCR rather than O-2 via NH3 oxidation. For NO2-SCR, a considerable amount of N2O was formed at 250 degrees C but decreased with increasing temperature. The decreased N2O is due to improved selectivity in the NO2 -SCR to N-2, as well as N2O decomposition at the back part of the catalyst at high temperature. Finally, different SCR reaction patterns were identified when testing with NO:NO2 = 3:1 and 1:3. For NO:NO2 = 3:1, the SCR reactions proceed in series, namely through the fast reaction first, followed by standard SCR. The fast SCR and NO2-SCR reactions proceed in parallel for NO:NO2 = 1:3. The results indicate that if NO2 is the limiting reactant, fast SCR dominates, but if excess NO2 is available, the NO2 SCR reaction can proceed in parallel. (C) 2010 Elsevier B.V. All rights reserved.