CHEMCATCHEM
FULL PAPERS
haust temperature window. Process optimization was performed
by testing different configurations for the catalytic materials, as de-
scribed below.
VWTi owing to excellent oxidation and NO adsorption abilities,
which results in a higher NO to NO2 oxidation rate at LT.
In the CC-B configuration, MnCeTi maintains its redox prop-
erties (Figure 2) and oxidation activity (Figure 7). The small de-
creases observed are due to the reduced amount of MnCeTi in
the CC-B configuration. Finally, it can be concluded that the LT
activity of the CC configuration is mainly determined by the
active sites over MnCeTi, regardless of whether it is located in
the fore part (CC-B) or the rear part (CC-A) of the configura-
tion.
Examples of the configuration adopted for the CC process are pre-
sented in Figure 1 and summarized as follows:
1) CC-A: MnCeTi was set as the fore part (contacting the exhaust
gases earlier); VWTi was separately set as the rear part (contact-
ing the exhaust gases later). Each part represented half of the
total volume.
2) CC-B: VWTi was set as the fore part; MnCeTi was separately set
as the rear part. Each part represented half of the total volume.
3) CC-C: The two catalysts were mixed together in the same total
volume as used for CC-A or CC-B.
The mixing of MnCeTi and VWTi into a whole system (CC-C
configuration) resulted in a decreased efficiency (LT efficiency:
MnCeTi>CC-CꢂVWTi; HT efficiency: VWTi>CC-CꢂMnCeTi;
Figure 4). The H2-TPR results (Figure 2) revealed that the main
Mnn+ reduction process is significantly delayed (from 360 to
4728C), which leads to a redox activity located between the
two single catalysts. Such a delayed reducibility in CC-C can ex-
plain the limited NO oxidation activity measured (Figure 7),
which results in poor LT SCR performance for this configura-
tion. In addition, this configuration cannot avoid significant
consumption of NH3 by O2 oxidation at HT, which finally results
in decreased HT SCR activity (Figures 7 and 4).
4) CC-D: VWTi was set as the fore part; MnCeTi was set as the
rear part. The former represented one third of the total
volume; the latter represented two thirds of the total volume.
5) CC-E: VWTi was set as the fore part; MnCeTi was set as the rear
part. The former represented one fourth of the total volume;
the latter represented three fourths of the total volume.
Characterization
XRD was performed with a X-ray diffractometer (D8 FOCUS, Bruker)
using CuKa radiation (l=0.15406 nm). Diffractograms were collect-
ed in the 2q range between 20 and 808 (step size: 0.058; step
time: 5 s). Phase identification was performed by indexation using
the JCPDS database. Surface areas were measured by N2 physisorp-
tion at 77 K on a TriStar 3000 instrument from Micromeritics by
using a 10-point BET method. The elementary chemical composi-
tion was determined with an Optima 3000 XL spectrometer
(Perkin–Elmer). Before the analysis, materials were dissolved in an
acid solution under microwave heating.
Conclusions
Herein, the catalyst combining method is found to be a con-
venient way to design a highly efficient and selective catalytic
reduction of NO with NH3 process. Fundamental evidences are
given for the rational design of the process and its efficiency
maximization (assembly order and catalytic bed volume). An
adequate configuration using MnCeTi and VWTi active materi-
als ensures an excellent SCR performance (NO conversion
>85%, N2 yield >70%, and decreased N2O production) over
a broad operating temperature window (from 150 to 4008C).
The surface chemical valence of Mn in the MnCeTi catalyst was in-
vestigated with a Thermo Fisher Scientific ESCALAB 250Xi X-ray
photoelectron spectrometer using AlKa radiation under ultrahigh
vacuum. Binding energy was calibrated with respect to the C1s
value of the contaminated carbon at 285.0 eV.
Experimental Section
Materials
Before the H2-TPR experiment, each material (30 mg) was pretreat-
ed at 5008C in 20% O2/He flow (flow rate: 20 mLminꢁ1) for 30 min.
After cooling the material to 308C, first 5% H2/He flow (flow rate:
20 mLminꢁ1) was stabilized and then the temperature of the reac-
tor was increased from 30 to 8008C (ramp rate: 58Cminꢁ1). H2 con-
sumption along with temperature was recorded online with a ther-
mal conductivity detector (TCD). A water trap was used to elimi-
nate H2O generated from material reduction .
Materials [MnOx(10 wt%)–AOy(10 wt%)/TiO2 and V2O5(4 wt%)–
AOy(10 wt%)/TiO2, in which AOy =CeO2, WO3, ZrO2] were prepared
by using the wet impregnation method. Commercial TiO2 (anatase,
Sigma–Aldrich) with a surface area of 50 m2 gꢁ1 was used as a sup-
port. The desired amounts of Mn(NO3)2, metallic precursor of A
[Ce(NO3)3, (NH4)10(H2W12O42)·4H2O, or Zr(NO3)4], and TiO2 were
mixed as a suspension solution, with continuous stirring at 408C
for 24 h. After impregnation, the solvent was removed with
a rotary evaporator at 808C. The samples were dried at 1208C for
12 h and then calcined in air at 5008C for 6 h. The as-synthesized
samples were labeled as MnOx–AOy/TiO2. V2O5–AOy/TiO2 were pre-
pared by using a similar method, except Mn(NO3)2 was replaced
with NH4VO3.
NH3-TPD was performed on AutoChem II 2920 equipped with
a TCD. The sample (ꢂ50 mg) was first pretreated in He at 5008C
for 1 h and then saturated with NH3 (4% in He; flow rate:
20 mLminꢁ1) for 1 h. After a purge with pure He at RT, NH3 desorp-
tion was performed in the range of 50–5008C (heating rate: 108C
minꢁ1) and in He flow (flow rate: 20 mLminꢁ1).
As a prerequisite of this study, MnCeTi and VWTi confirmed their
respective high activity for the LT and HT reactions in comparison
to other credible compositions (see the section on Material selec-
tion in the Supporting Information). On the basis of the idea of
combining high SCR activity, high N2 selectivity, and broad operat-
ing temperature window, we developed a CC process that was
found to be highly efficient all over the classical automobile ex-
NO/O2-TPD experiments were performed with a homemade tem-
perature-programmed characterization system equipped with
a fixed-bed continuous-flow reactor and a quadrupole mass spec-
trometer (OmniStar MS200). The sample (50 mg) was pretreated
under 1000 ppm NO + 10% O2 balanced by Ar at 3008C for 1 h
(total flow rate: 100 mLminꢁ1). Subsequently, the reactor was
cooled to 308C in the same atmosphere and purged with Ar
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 2263 – 2269 2268