HCN + NO2 over Na- and Ba-Y,FAU
J. Phys. Chem. B, Vol. 109, No. 4, 2005 1489
+ O2 water is formed and in the absence of NO2 only the 2175
cm-1 feature is seen. In the HCN + NO2 reaction (with no water
added) only the 2165 cm-1 band is seen. Addition of H2O to
the HCN + NO2 system after elevated temperature reaction
results in the development of both of these IR features.
HCN or adsorbed cyanide species can be important inter-
mediates, or whether they are only byproducts of some side
reactions. We believe that HCN/CN is an important intermediate
in the overall reaction mechanism, and possibly the key
intermediate in the formation of N2. Our results strongly support
a mechanism in which N2 is formed by the direct reaction
between CN and NCO species and NOx. The alternative reaction
path that involves the formation of NH3 by hydrolysis, and the
subsequent NH3-SCR process is not supported by our findings.
We did not, under any circumstances, observe the formation of
4. Relevance to NOx Reduction. The formation of HCN in
NOx SCR-related studies has been reported for a number of
transition metal ion exchanged zeolite materials.7,8,26,27 The
amount of HCN formed varied substantially depending on the
catalyst material used. For example, Cant and co-workers21
reported that the amount of HCN produced in the reaction of
nitromethane with an NO2/O2 gas mixture was significantly
higher over Na-ZSM5 then over either H- or Co-ZSM5,
while the nitrogen yield was very similar over Na- and Co-
ZSM5. Furthermore, the temperature regime where significant
amount of HCN formation was seen varied significantly
depending on the cationic form of the zeolite.22,26 While the
maximum amount of HCN over Co-ZSM5 was observed at
∼520 K, on the Fe-ZSM5 it was seen at about 620 K, and at
460 K over Cu-ZSM5 when nitromethane was reacted with a
NO/O2 gas mixture. In the nonthermal plasma assisted catalytic
NOx reduction, HCN was also reported on both Na-Y zeolite
and Al2O3-based catalysts when the tests were carried out in a
synthetic exhaust gas mixture containing NO + O2 + propylene
+ H2O.23,24 The possibility of HCN or surface adsorbed cyanide
being an important intermediate in the NOx reduction by CH4
over Co-ZSM-5 has been suggested previously.8 These authors
have shown that CN species were reactive intermediates, and
N2 and CO2 were formed in the reaction of CN with NO2. The
possibility of HCN being an important intermediate in the SCR
of NO over Cu-ZSM-5 was considered by Cant and Liu.27 They
propose two possible reaction paths for the formation of N2 from
HCN. One is the reaction between adsorbed CN and NO2 to
form N2 and CO2 (reaction 2 above). They argue that this
reaction is feasible due to the presence of NO2 at the reaction
temperature of their study. The second reaction path to N2
formation is a two-step process, in which HCN is first
hydrolyzed to NH3, which in turn is converted to N2 by NH3-
SCR. Hydrolysis can proceed on acidic OH groups, or via
hydration to formamide followed by a standard acid-catalyzed
reaction on Bro¨nsted acidic sites. (These authors also report on
the formation of substantial amounts of C2N2 both in HCN +
NO + O2 and HCN + O2 reactions.)
+
NH3 or NH4 that are crucial in the NH3-SCR reaction (not
even in experiments when we added NH3+NO2+H2O onto the
BaY catalyst). When we exposed the BaY catalyst to water
following HCN + NO2 reaction, we did not observe the fast
hydrolysis of adsorbed NCO species or the appearance of IR
features that would support the formation of species containing
NHx (NH3, NH4+, etc.) entities that is mentioned in some
mechanistic studies.19 There was no increase in the IR signature
of CO2, which would be the other product of the proposed
hydrolysis, as well. We only observe a shift in the νCN frequency
of the adsorb CN- species from 2167 to 2174 cm-1 as these
species interact with zeolitic OH groups rather than NO+ species
in the absence of water (Figure 9). The results of these
experiments, however, cannot rule out the possibility that under
practical catalytic conditions when the concentration of water
is much higher than that in this study the formation of NHx-
containing intermediates (e.g., NH4NO2) becomes significant.
Our experimental set up, however, did not facilitate investiga-
tions under these very high water concentrations.
Conclusion
Both Na- and Ba-Y zeolites adsorb HCN molecularly, and
the strength of HCN adsorption is much higher on Ba-Y than
on Na-Y. Up to 473 K, no reaction between the catalysts and
HCN was observed. At 300 K, there is no reaction between
HCN and NO2 on Na-Y, while this reaction proceeds on Ba-Y
even at this low temperature. The IR absorption features seen
on the two catalyst samples in the HCN + NO2 reaction at 473
K are very similar, although the intensities of these features
are much lower on Na-Y than on Ba-Y, due to the much
higher catalytic activity of the latter. In the gas-phase N2, N2O,
NO, CO, and CO2 was observed after the reaction at 473 K,
while adsorbed cyanide and cyanate species were detected on
both catalyst surfaces. Postreaction mass spectrometric analysis
also revealed the formation of a significant amount of C2N2
formed by the dimerization of adsorbed cyanide species. The
results of this investigation strongly suggest that HCN/CN can
be an important intermediate in the overall NOx reduction on
these zeolite catalysts. Adsorbed CN- and NCO- ionic species
formed in the HCN + NO2 reaction are believed to be the
important surface intermediates. Their reaction with ionic NOx
species (in particular with NO+) can lead to the formation of
the N-N bond.
Also, as we have mentioned previously, the reaction between
CN and NO2 was observed to be much faster than that between
NCO and NO2. On the other hand, surface NCO was proposed
as the key intermediate in the formation of N2 over Ag-TiO2-
ZrO2 catalysts.25 The reaction between -NCO and NO to form
N2 was proposed to be much faster than that between -CN
and NO. In our study we find that both -NCO and -CN can
react with NOx and N2 can form in both reactions; however,
the rate of N2 formation in these two reactions cannot be
compared from the results of our study.
The series of reactions (1-6) we propose can explain both
the formation of N2, CO2, CO, N2O, and C2N2 in the overall
reaction, and the observed changes in -CN/ -NCO ratio as
the HCN/NO2 ratio is changed. At low NO2 levels, we see
mostly the formation of surface -CN species, while at higher
NO2 concentrations, surface -NCO is formed in the expense
of surface -CN, in accordance with reaction 6. The high
concentration of surface cyanide species observed throughout
this study also explains the relatively high levels of C2N2
measured by mass spectrometry. All of these results help us
answering our key question that we set out to address: whether
Acknowledgment. The authors gratefully acknowledge the
U.S. Department of Energy (DOE), Office of Energy Efficiency
and Renewable Energy, FreedomCAR, and Vehicle Technolo-
gies for support of this program. The work was performed as
part of a CRADA with the USCAR Low Emissions Technolo-
gies Research and Development Partnership (LEP), Pacific
Northwest National Laboratory (PNNL), and DOE/OFCVT. The
research described in this paper was performed at the Environ-
mental Molecular Sciences Laboratory, a national scientific user
facility sponsored by the DOE Office of Biological and