COMMUNICATION
Selective catalytic reduction of NOx by hydrocarbons enhanced by
hydrogen peroxide over silver/alumina catalysts
Petr Sazamaa and Blanka Wichterlova´*b
Received (in Cambridge, UK) 27th May 2005, Accepted 20th July 2005
First published as an Advance Article on the web 17th August 2005
DOI: 10.1039/b507553j
could entail considerable problems in storage of hydrogen or its
preparation on board. Even though the application of hydrogen
peroxide in SCR-NOx would also bring some difficulties the
observed positive effect of hydrogen peroxide on NOx conversion
to nitrogen over Ag/alumina described here could be considered in
the reduction of NOx in the exhaust gases of lean-burn
combustion processes.
It is shown that hydrogen peroxide enhances substantially
selective reduction of NOx to nitrogen with hydrocarbons over
Ag/alumina catalysts.
Ag/alumina is one of the promising catalysts for selective catalytic
reduction of nitrogen oxides to nitrogen (SCR-NOx) by hydro-
carbons, providing high and stable activity at temperatures
.350 uC when using long-chain paraffins (the main components
of diesel fuel) as reductants. However, at temperatures ,350 uC
the activity is low. A great improvement has been achieved by the
discovery that hydrogen added to a hydrocarbon feed dramatically
increases the conversion of NOx, especially at low temperatures
and high gas space velocities.1,2 Shibata et al.3 suggested that the
hydrogen effect stems from the enhanced rate of oxidation of
propane to acetates, which are more reactive. UV-Vis and EXAFS
measurements of the state of the silver has led to the conclusion
that the presence of hydrogen causes the formation of small
metallic charged Ag clusters,4 which are the active sites responsible
for the enhanced rate of the SCR-NOx reaction.
A 2 wt.% Ag/alumina catalyst was prepared by the procedure
described in ref. 8. Catalytic tests, performed in a quartz tubular
micro-reactor, employed a reactant mixture modelling the exhaust
gas composition, and consisting of 1000 ppm NO, 6.0% O2, 12.0%
H2O, 600 ppm n-C10H22, 0 or 2000 ppm H2O2 and 0 or 2000 ppm
H2 and the rest helium. The gas components in the feed were
controlled by mass-flow controllers, and n-decane was fed via
saturators maintained at the desired temperature and using helium
as a carrier gas. Water and hydrogen peroxide vapour were added
by using a linear dosing device and aqueous solution of 3.2%
H2O2. The flow of the gas mixture and the weight of catalyst
corresponded to GHSV of 60 000 h21. The reaction products were
analyzed under reaction steady-state conditions using an on-line
connected NO/NOx chemiluminescence analyzer (Horiba CLA-
355K) and gas chromatograph (Hewlett Packard 6090).
Experimental details are given in ref. 5.
However, time-resolved monitoring of the effect of the addition
of hydrogen to and its removal from the reactant stream on the
NOx conversion and Ag cluster formation has not yielded
evidence for their direct relationship.5 Moreover, the presence of
metallic Ag clusters on alumina was also detected in the SCR-NOx
reaction performed in the absence of hydrogen.5,6 These findings
led us to the conclusion that hydrogen itself probably takes part
in the SCR-NOx reaction.5 This conclusion was supported by
the observation that the presence of both CO and H2 induced the
formation of Ag clusters, but only hydrogen increased the
conversion of NOx to nitrogen.7
Conversion of NO in the reaction with n-decane at 470–520 K
without co-fed hydrogen peroxide or hydrogen was low (Fig. 1a).
The NO conversion was increased considerably by addition of
hydrogen peroxide into the reactant stream. This was accompanied
by higher conversion of n-decane and higher selectivity to CO2
(Fig. 1b). Both molecular nitrogen and nitric dioxide were formed.
The greatest enhancement of NO conversion was achieved at low
temperatures.
Based on these results, but without providing any direct experi-
mental evidence, we suggested in ref. 5 that the mechanism by which
hydrogen functions could be explained by the formation of Ag
hydride with subsequent formation of hydroperoxy and hydroxy
radicals. These radicals then enhance the individual reaction steps of
the SCR-NOx process as well as the observed NO–NO2 oxidation.
In the present study we used hydrogen peroxide as a source of
radicals that would enhance the rate of SCR-NOx, if the radical
mechanism is operative in the hydrogen-co-assisted reaction.
When using SCR-NOx, particularly for mobile diesel engines, a
method employing hydrogen to increase the rate of SCR-NOx
Table 1 shows a comparison of decane-SCR-NO co-assisted by
hydrogen peroxide over Al2O3 and Ag/Al2O3 at 523 K. The
activity of alumina itself in the absence of hydrogen peroxide in the
feed was negligible and the effect of the added hydrogen peroxide
was remarkable. The NO conversion to nitrogen was significantly
higher over Ag/alumina as well as in the reaction co-assisted by
hydrogen peroxide. The positive effect of hydrogen on CH-SCR-
NO has been observed only with silver on alumina or Ag-zeolite
catalysts.9 Analogously to hydrogen, with Cu- and Fe-ZSM-5
catalysts the increase in NOx conversion in the presence of
hydrogen peroxide was not observed.
aJ. Heyrovsky´ Institute of Physical Chemistry, Academy of Sciences of
the Czech Republic, Prague 8, 182 23, Czech Republic.
E-mail: petr.sazama@jh-inst.cas.cz; Fax: +420 286582307;
Tel: +420 26605 3325
When hydrogen was employed as an additive over Ag/alumina,
the conversion of NO to NO2 was lower and the yield of N2 was
higher compared to the reaction using hydrogen peroxide (Table 2).
Thus, in comparison with hydrogen, in addition to improving the
reduction of NO to N2, hydrogen peroxide enhances NO oxidation
to NO2 much more. This effect is more pronounced at low
bJ. Heyrovsky´ Institute of Physical Chemistry, Academy of Sciences of
the Czech Republic, Prague 8, 182 23, Czech Republic.
E-mail: wichterl@jh-inst.cas.cz; Fax: +420 286582307;
Tel: +420 26605 3595
4810 | Chem. Commun., 2005, 4810–4811
This journal is ß The Royal Society of Chemistry 2005