Influence of the type of reducing agent (H 2 , CO, C 3 H 6 and C 3 H 8 ) on the reduction of stored NO X in a Pt/BaO/Al 2 O 3 model catalyst Hussam Abdulhamid a,b , Erik Fridell a *, and Magnus Skoglundh a,b a Competence Center for Catalysis, Chalmers University of Technology, SE 412 96 Go ¨teborg, Sweden b Department of Materials Surface Chemistry-Applied Surface Chemistry, Chalmers University of Technology, SE 412 96 Go ¨teborg, Sweden In this investigation, a comparative study for a NO X storage catalytic system was performed focusing on the parameters that affect the reduction by using different reductants (H 2 , CO, C 3 H 6 and C 3 H 8 ) and different temperatures (350, 250 and 150 °C), for a Pt/BaO/Al 2 O 3 catalyst. Transient experiments show that H 2 and CO are highly efficient reductants compared to C 3 H 6 which is somewhat less efficient. H 2 shows a significant reduction effect at relatively low temperature (150 °C) but with a low storage capacity. We find that C 3 H 8 does not show any NO X reduction ability for NO X stored in Pt/BaO/Al 2 O 3 at any of the temperatures. The formation of ammonia and nitrous oxide is also discussed. KEY WORDS: NO X reduction; lean-burn; reducing agents; H 2 , CO, C 3 H 6 ,C 3 H 8 ; Pt/BaO/Al 2 O 3 catalyst. 1. Introduction Improvement of vehicle fuel economy could be achieved through the use of both diesel engines and the new generation of gasoline engines (lean-burn engines) that operate with a high air/fuel ratio. How- ever, these engines suffer from inefficient abatement of NO X emissions using the traditional three way catalytic systems. One promising technique to overcome the problem is to use the so-called NO X storage reduction catalysts in which NO X is stored under relatively long lean periods, and subsequently released and reduced under short rich periods [1–3]. Since lean-burn and diesel engines operate at relatively low temperatures (as a result of the high air/fuel ratio), it is of particular importance for the existing lean catalytic system to be able to store and reduce NO X at low temperatures. Basically, the NO X storage catalyst comprises noble metals (Pt, Pd and Rh) to promote oxidation and reduction, alkali and alkaline earth compounds (K 2 O, BaO, SrO, MgO) as storage media for NO X and support (Al 2 O 3 , Ceria) providing high surface area [3]. Several studies have focused on the storage step and the parameters affecting this step, e.g. type of storage medium [4–6], sulphur content [7,8], effect of noble metal [9,10], temperature [11,12], and gas composition like CO, CO 2 ,H 2 O and HC [13,14]. In this investigation, we will focus on the NO X reduction step and the different parameters that affect the regeneration step. The study was performed using a Pt/BaO/Al 2 O 3 model catalyst. Transient reactor studies were carried out at different temperatures (350, 250 and 150 °C) and by using either CO, H 2 ,C 3 H 6 or C 3 H 8 as reducing agent. 2. Experimental section 2.1. Catalyst preparation A monolith catalyst, containing c-Al 2 O 3 to provide a high surface area, BaO as storage medium and Pt to promote oxidation and reduction reactions, was pre- pared. The monolith sample of cordierite (2MgO Æ 2Al 2 O 3 Æ 5SiO 2 ) with a length of 15 mm and a cross- section consisting of 69 channels, with a cell density of 64 channels/cm 2 , was immersed in an alumina slurry (80 wt% c-Al 2 O 3 and 20 wt% boehmite as binder). Excess slurry was removed by blowing air through the channels, and the monolith was then dried in hot air at 95 °C for 5 min and calcined in air at 550 °C for 5 min. The procedure was repeated until the desired amount of alumina was deposited on each sample. The sample was finally calcined in air at 550 °C for 120 min. The same procedure was used to deposit the barium as BaCO 3 by using a Ba(NO 3 ) 2 solution followed by impregnation with NH 4 HCO 3 . Pt was then deposited on the sample by filling the monolith channels with a platinum nitrate [Pt(NO 3 ) 2 ] solution followed by drying at 95 °C for 8 h and finally by calcining at 550 °C for 120 min. 2.2. Catalyst characterization The specific surface area of the catalyst was determined by nitrogen adsorption according to the BET method using an ASAP 2010 instrument (Micrometrics) and was found to be 21.7 (m 2 /g). The platinum dispersion of the sample was determined by CO chemisorption experi- ments assuming one CO molecule adsorbed per surface Pt metal atom. The chemisorption experiments were carried out in a flow reactor after reducing the catalyst in 3 vol% H 2 in nitrogen for 30 min at 400 °C. After the reduction step, the reactor was flushed with nitrogen (600 mL/min) and the temperature was lowered to 0 °C. After 10 min the catalyst was instantly exposed to 50 ppm CO in N 2 * To whom correspondence should be addressed. E-mail: fridell@fy.chalmers.se 161 1022-5528/04/0700–0161/0 Ó 2004 Plenum Publishing Corporation Topics in Catalysis Vols. 30/31, July 2004 (Ó 2004) Nos. 1–4,