(689e) Understanding the Distributed Intra-Catalyst Effect of Sulfation On the Water Gas Shift Reaction in a Lean NOx Trap Catalyst

Authors: 
Choi, J., Oak Ridge National Laboratory
Pihl, J. A., Oak Ridge National Laboratory
Toops, T. J., Fuels, Engines and Emissions Research Center, Oak Ridge National Laboratory
Yezerets, A., Cummins Inc.
Parks, J. E., Fuels, Engines and Emissions Research Center, Oak Ridge National Laboratory
Ottinger, N. A., Oak Ridge National Laboratory


Introduction

The Lean NOx Trap (LNT) catalyst is an aftertreatment technology for abatement of nitrogen-oxide emissions from lean-burn vehicle engines. The water-gas-shift (WGS) reaction occurs during LNT regeneration; WGS can contribute to regeneration reactions and be used in catalyst-system control strategies. Exhaust sulfur can limit the overall LNT performance, but its impact on specific LNT reactions varies, and in previous work we have developed a conceptual model of LNT sulfation based on measurements of distributed-intra-catalyst sulfur impact on NOx storage/reduction (NSR) and oxygen storage capacity (OSC) functions. In this work we study the impact of LNT sulfation on the transient WGS distribution within the operating catalyst.

Materials and Methods

A model Al2O3/Pt/Ba LNT catalyst (47 cells per cm2) was evaluated in a bench reactor system described previously. The experiments were performed in 60/5-s lean/rich cycling mode at 325C and a space velocity of 30000h-1. Three feed-gas mixtures were investigated to characterize uninhibited WGS performance (no O2 nor NO in lean phase), and with competition from OSC (10% O2 in lean phase) and NSR (300 ppm NO, 10% O2 in lean phase). Three sulfation levels were evaluated; 0.00 (baseline), and 0.9 & 1.7 g S/Lcat. Spatiotemporal profiles of gas composition were measured by SpaciMS (Spatially resolved capillary-inlet Mass Spectrometer) at 0, ¼, ½, ¾, and 1 relative LNT locations along a monolith channel.

Results and Discussion

Whereas in the baseline fresh condition, NOx was stored primarily in the front LNT half, sulfation progressively poisoned the catalyst in a plug-like manner, shifting the NSR region further down the catalyst axis. The first and second sulfations poisoned the first quarter and half of the catalyst, respectively, with respect to the NSR function. OSC measurements allowed measurement of the Pt active surface area distribution, which was relatively uniform along the catalyst axis. Sulfation had a minor impact on OSC (i.e. oxygen adsorption and reduction on Pt surface). In neutral conditions, WGS performance generally mirrored that of the Pt distribution. WGS was very sensitive to sulfation, and its sulfation front preceded that of the NSR function by about a quarter catalyst length. OSC has a minor mitigating impact on WGS sulfur degradation; recovering 5-10% of the ca. 95% performance loss associated with sulfation. This is apparently due to oxygen readily displacing S from Pt sites in OSC conditions. Sulfation appears to primarily impact the nature of the active sites relevant to WGS rather than the number of Pt sites. Notably, the WGS sulfation impact is broader and more sensitive than that of NOx storage, and WGS is highly sensitive to initial sulfation. Apparently in sulfated conditions, a zone where surface S levels sufficient to degrade WGS but not NOx storage, exists immediately downstream of the NSR poisoned zone at the catalyst front, and causing broadening of the WGS-zone affected by sulfation. We will discuss these results and suggest conceptual models of how S impacts the various LNT functions. This research has broad relevance to fundamental reaction understanding as well as applied uses including on-board diagnostics (OBD) or catalyst-systems control.