(498e) A Calorimetric Study of Ammonia Adsorption and Desorption Over a Cu Beta Zeolite Catalyst for Urea SCR Applications

Wilken, N., Chalmers University of Technology
Olsson, L., Competence Centre for Catalysis, Chalmers
Kamasamudram, K., Cummins Inc.
Li, J., Cummins Inc.
Yezerets, A., Cummins Inc.


Diesel engine exhausts contains excess oxygen and a standard 3-way catalyst cannot reduce NOefficiently, due to oxygen poisoning of the noble metal sites. Selective catalytic reduction (SCR) of NOx by urea is one possible solution to reduce NOx in diesel exhaust gases. The ammonia (NH3), formed from urea, reacts selectively with NOx to produce N2 and H2O over a catalyst. Among several potential catalysts, for SCR of NOx, copper and iron exchanged zeolites are at the forefront, especially for diesel exhaust cleaning. Over zeolite catalysts, NOx reacts with stored NH3. NH3 storage capacity inturn depends on the exhaust gas temperature and composition. Due to the transient nature of diesel exhaust temperature and composition, a thorough understanding of the adsorption and desorption of ammonia on the catalyst is critical to effectively conduct kinetic modelling of the SCR system. The objective of this study is to experimentally measure energetics of NH3 adsorption/desorption processes over a Cu-Beta zeolite catalyst and develop NH3 storage and NOx reduction kinetic models.


The catalyst powder was prepared by aqueous ion-exchange (IE) of a beta zeolite with a silica to alumina ratio of 38, from Zeolyst International. To introduce controlled amount of copper, the powder zeolite was ion-exchanged in two stages, IE with NaNO3 followed by  IE with Cu(CH3COO)2. Calorimetric measurements were done on 100 mg of powdered catalysts placed in a sintered quartz cell of the differential scanning calorimeter (DSC, Sensys from Setaram). Reactant or inert gas flow of 100ml/min was maintained through the reference and sample cells for catalyst surface coverages and heats of NH3 adsorption and desorption measurements critical for kinetic model development.

Results and discussion

We have investigated the adsorption and desorption of ammonia over a Cu-Beta zeolite under different catalyst temperature and gas composition conditions. Figure 1 shows the resulting thermogram from one experiment. The catalyst was first pre-treated for 15 min with 8vol% O2 in Ar at 500°C. The catalyst was then cooled to 150°C in Ar followed by exposure to 2000ppm of NH3 for 60min. On NH3 introduction, an exotherm of approximately 13.7mW was observed due to the NH3 adsorption. At about 5500sec the adsorption heat signal decreased to 0mW, indicating a saturation of the catalyst surface with NH3. A negative heat signal, due to the endothermic desorption of loosely bound NH3 is observed at 8000sec because of the change of NH3 containing reactant gas to inert argon flow.  The NH3 saturated catalyst is heated, starting at 11500sec, to 600°C in Ar only. During the temperature ramping an endotherm is observed, due to ammonia desorption. These calorimetric results together with gas phase measurements using FTIR resulted in an average heat of adsorption of ammonia of about 100 kJ/mol.

Figure 1. Thermogram of an experiment exposing a Cu-Beta catalyst to 2000 ppm NH3 for 60 min, followed by flushing with Ar only and heating the catalyst with 20 °C/min to 600°C. The sample size was 100 mg and the flow 100 ml/min.


Ammonia SCR is an important technology for reducing NOx from diesel and lean burn gasoline engines. Kinetic modeling is an important tool for predicting the catalyst performance and also to gain a deep understanding of the catalytic mechanisms. Since emissions from vehicles are very transient it is important to understand the adsorption of ammonia on the catalyst. One of the objectives of this work was to measure the heat of ammonia adsorption over Cu-Beta to apply these results to kinetic models. The measurments resulted in an average heat of adsorption of about 100 kJ/mol.  


This work has been performed within the Competence Centre for Catalysis and Cummins Inc. The authors would like to thank Cummins Inc. for the financial support. One author would also like to acknowledge the Swedish Research Council (Contract: 621-2003-4149 and 621-2006-3706) for additional support. The financial support for the micro calorimeter from the Swedish Research Council (Contract: 621-2003-4149 and 621-2006-3706) and for the FTIR from Knut and Alice Wallenberg Foundation, Dnr KAW 2005.0055, is gratefully acknowledged.