Carbonaceous Deposit as a Preferential Reductant in the Reduction of NO with C2H4 in Excess Oxygen over Alumina

Article abstract of DOI:10.1246/cl.2003.794

Carbonaceous deposit (denoted as C(d)) which was formed on alumina by C₂H₄ pyrolysis at 550°C reacted selectively and preferentially with a mixture of NO and O₂ in the presence of C ₂H₄, suggesting th

Full text of DOI:10.1246/cl.2003.794

7
94
Chemistry Letters Vol.32, No.9 (2003)
Carbonaceous Deposit as a Preferential Reductant in the Reduction of
NO with C H in Excess Oxygen over Alumina
2
4
Ã
Ryuta Fujii, Sayaka Kitagawa, Kou Takahashi, Noriyasu Okazaki, and Akio Tada
Department of Applied and Environmental Chemistry, Kitami Institute of Technology, Koencho, Kitami 090-8507
(Received May 26, 2003; CL-030462)
Carbonaceous deposit (denoted as C(d)) which was formed
20 min. Such excess N2-yield is apparently caused by the
C(d), because the integrated excess N2-yield is almost equal
to the integrated N2-yield in the C(d)-SCR. Figure 1 also shows
the concentration ratios (denoted as r) of N2/(CO + CO2) in the
outlet gases. Interestingly, the r for the C(d)-C2H4-SCR was
higher than the r for the C2H4-SCR at the early stage, while
the r for the C(d)-SCR was very low.
ꢀ
on alumina by C2H4 pyrolysis at 550 C reacted selectively and
preferentially with a mixture of NO and O2 in the presence of
C2H4, suggesting that the C(d) acts as an effective intermediate
during the selective reduction of NO with C2H4 too.
The selective catalytic reduction of nitrogen oxides (NOx)
by hydrocarbons (HC) in excess oxygen (denoted as HC-
SCR) has been regarded as a promising process for removing
NOx in exhaust gases, and many catalysts were proposed such
In order to investigate in detail what happens in an excess
N2-yield period, a great deal of C(d) needs be accumulated,
and so the ethene pyrolysis was carried out at 800 C. In the
ꢀ
1
,2
as zeolites, metal oxides, and supported noble metals. On alu-
mina, one of such oxides, carbonaceous deposit (denoted as
C(d)) was formed as an intermediate during C3H6-SCR, and
500
0.5
0.4
0.3
0.2
0.1
0.0
Excess
N yield period
2
4
00
300
00
3
,4
found to act as an effective reductant of NO. On the other
hand, there has been reported another intermediate such as iso-
5,6
cyanate (–NCO) for HC-SCR over alumina or doped alumina.
Thus, there are at least two reaction pathways for HC-SCR; one
is via C(d) and the other is via another intermediate. Which
pathway is more predominant? There has, however, been no an-
swer so far, because there is very few report on HC-SCR over
2
100
0
,
,
,
C2H4-SCR
C(d)-C H -SCR
C(d)-SCR
3
,4
alumina covered with a considerable amount of C(d). There-
fore, the present paper deals with C2H4-SCR, C(d)-SCR, and
C(d)-C2H4-SCR over alumina catalyst, and discusses the reac-
tivity of the C(d) by comparing with that of C2H4 gas.
2 4
0
10
20
30
40 110
120
Time on stream / min
ꢀ
Alumina was produced by heating its precursor at 600 C
for 24 h, which was prepared by the controlled hydrolysis of
Figure 1. Conversion (open points) of NO to N2 during
C(d)-SCR, C2H4-SCR, and C(d)-C2H4-SCR over alumina
and concentration ratio (filled points) of N2/(CO + CO2)
aluminum triisopropoxide. It was then pressed into a disc under
À2
2
00 kgÁm , followed by crushing, sieving (355–500 mm) and
ꢀ
in the outlet gases. Ethene pyrolysis temp. = 550 C, reac.
ꢀ
ꢀ
again heating at 800 C for 4 h in air. Every reaction including
C2H4-SCR was performed using a fixed-bed flow tubular reac-
temp. = 550 C.
À3
tor under conditions as follows: W/F was 0.18 gÁsÁcm (cata-
500
00
300
00
0.5
À1
lyst, 0.40 g; total flow rate, 130 mLÁmin ) and the composition
Excess
N yield period
,
,
,
C H -SCR
2 4
of the typical reactant gas was NO (1000 or 0 ppm), O2 (2%),
C2H4 (500 ppm), and helium as a balance. The accumulation
2
C(d)-C H -SCR
2 4
4
0.4
0.3
0.2
0.1
C(d)-SCR
of C(d) on alumina was carried out by the pyrolysis of C2H4
ꢀ
(
1000 ppm) at 550 or 800 C for 3 h in the above reactor. The
outflow gas was analyzed by gas chromatography with a Molec-
ular Sieve 5A (separation of H2, N2, O2, CO, CH4, and NO) and
Porapak Q (separation of CO2, C2H4, N2O, and H2O). The
2
length of Molecular Sieve 5A and Porapak Q columns was
À1
100
0
3
m. Helium was used as the carrier gas (30 mLÁmin ).
ꢀ
The ethene pyrolysis was carried out at 550 C, followed by
ꢀ
0
.0
C(d)-SCR or C(d)-C2H4-SCR at 550 C at which NO conver-
sion was known to reach a maximum in C2H4-SCR over alumi-
na. For comparison, C2H4-SCR was also conducted. As can be
0
20
40
60
80
100 120
7
Time on stream / min
seen in Figure 1, the NO conversion in the C2H4-SCR reached
the plateau immediately after the start, while the NO conversion
in the C(d)-C2H4-SCR was higher from the start, and then de-
creased to the plateau; the latter surpassed the plateau for
Figure 2. Conversion (open points) of NO to N2 during C(d)-
SCR, C2H4-SCR, and C(d)-C2H4-SCR over alumina and con-
centration ratio (filled points) of N /(CO + CO ) in the outlet
2
2
ꢀ
ꢀ
gases. Ethene pyrolysis temp. = 800 C, reac. temp. = 550 C.
Copyright ꢀ 2003 The Chemical Society of Japan

Chemistry Letters Vol.32, No.9 (2003)
795
C(d)-C2H4-SCR, the maximum NO conversion appeared after
an induction period, and the integrated excess N2-yield became
larger (Figure 2). Unexpectedly, the r for the C(d)-C2H4-SCR
was markedly lower than the r for the C2H4-SCR, indicating
that the C(d) reduced NO less selectively than C2H4 gas. The
r for the C(d)-SCR increased gradually, indicating that the reac-
tivity of the C(d) during the C(d)-SCR was improved by inter-
acting with a mixture of NO and O2. Interestingly, the r for the
C(d)-SCR nearly synchronized with that for the C(d)-C2H4-
SCR at the early stage.
5
4
3
00
00
00
200
C(d)-C2H4-SCR
C H -SCR
1
00
0
,
,
2 4
C H +O
2
4
2
0
20
40
60
80
100 120
ꢀ
When the ethene pyrolysis was carried out at 550 C, the
Time on stream / min
catalyst bed became gray, indicating that it was partly covered
with C(d). The r for the C(d)-SCR was lower than the r for the
C(d)-C2H4-SCR. This will be explained as follows: The C(d)
interacts with C2H4 molecules to form a new reactive C(d),
and it reacts with NOx over alumina to result in C(d)-NCO-like
species. The r for the C(d)-SCR in which the C(d) was formed
Figure 3. Outlet C2H4 concentrations during various reactions
in the presence of alumina (filled points) and in the absence of
ꢀ
alumina (open points). Ethene pyrolysis temp. = 800 C, reac.
ꢀ
temp. = 550 C.
8
6
00
00
ꢀ
at 550 C was higher than the r for the C(d)-SCR in which the
ꢀ
Reductant for C(d)-C H -SCR
2
4
C(d) was formed at 800 C. This is probably due to difference in
C(d)
C(d)+C H
C H
2 4
2
4
reactivity which reflects difference in crystallinity of carbon in
the C(d); the higher the temperature of the ethene pyrolysis, the
higher the content of graphite in the C(d). In this connection, ac-
tive charcoal and graphite were used instead of the C(d) in C(d)-
SCR and it was found that graphite was inactive, while active
400
2
00
0
3
charcoal was somewhat active but its reactivity (1:3 Â 10 ppm
ꢀ
N2/g-carbon) was much lower than C(d) formed at 800 C
4
(
1:7 Â 10 ppm N2/g-carbon) and the r for active-charcoal-
0
20
40
60
80
100 120
SCR under the same reaction conditions as the C(d)-SCR shown
in Figure 2 was 0.01 (not shown). In addition, a mechanical
mixture of active charcoal (0.014 g, equal to the amount of
Time on stream / min
Figure 4. Water concentration during C2H4-SCR(n), C(d)-
C H -SCR(Ã), C(d)-SCR(ꢁ), and C H + O (l) over alumi-
ꢀ
C(d) formed at 800 C) and ꢀ-alumina (0.40 g) was used instead
2
4
2
4
2
ꢀ
ꢀ
na. Ethene pyrolysis temp. = 800 C, reac. temp. = 550 C.
of the C(d) in the C(d)-SCR in Figure 2, resulting in 70 ppm of
N2 with the r of 0.03 (not shown), while the N2 concentration in
the C(d)-SCR was 170 ppm (Figure 2).
1
70 ppm, almost correspond to the respective H2O concentra-
tions which were estimated from the outlet C2H4-shortages.
The H2O concentrations for the C2H4-SCR (at NO conversion
ꢀ
When the ethene pyrolysis was carried out at 800 C, the
catalyst bed became black, indicating that it was completely
covered with C(d). In order to understand the time course of
the r for the C(d)-C2H4-SCR, concentrations of C2H4 and
H2O at the outlet of the reactor were studied (Figures 3 and
=
ca. 32%) and the C(d)-C2H4-SCR (at NO conversion = ca.
4
7%) are defined as [the inlet C2H4 conc. (500 ppm) À the out-
let C2H4 conc. (220 ppm)] Â 2 = H2O conc. (560 ppm) and [the
inlet C2H4 conc. (500 ppm) À the outlet C2H4 conc. (320 ppm)]
 2 = H2O conc. (360 ppm), respectively.
4). In the presence of alumina, the outlet C2H4 concentration
during the C2H4-SCR was 220 ppm, while in the absence of alu-
mina, the outlet C2H4 concentration was 500 ppm, indicating no
reaction occurred. These results clearly indicate that C(d) was
formed during the C2H4-SCR. Interestingly, the outlet C2H4
concentration (315 ppm) at the early stage during the C(d)-
C2H4-SCR was higher than that during the C2H4-SCR over alu-
mina, although the NO conversion during the C(d)-C2H4-SCR
was higher than that during the C2H4-SCR for a period of
In brief, the reactivity of C(d) depended on its preparative
temperature and its reaction atmosphere (existence of ethene);
ꢀ
during C(d)-C2H4-SCR, C(d) formed from C2H4 at 550 C re-
acted selectively and preferentially with a mixture of NO and
O2, while C(d) formed from C2H4 at 800 C reacted not selec-
ꢀ
tively but preferentially with the reactants, namely, its excess
N2-yield reflected not qualitative but quantitative effect of the
C(d).
30 min (Figure 2). This suggests that not C2H4 but the C(d) re-
duced NO, and this situation is likely to have took place for
References
4
2
0 min. Then the outlet C2H4 concentration decreased to
20 ppm, which is equal to that during the C2H4-SCR. In this
1
2
3
M. Iwamoto and H. Yahiro, Catal. Today, 22, 5 (1994).
H. Hamada, Catal. Today, 22, 21 (1994).
A. Obuchi, A. Ogata, K. Mizuno, A. Ohi, M. Nakamura, and H.
Obuchi, J. Chem. Soc., Chem. Commun., 1992, 247.
A. Obuchi, M. Nakamura, A. Ogata, K. Mizuno, A. Ohi, and H.
Obuchi, J. Chem. Soc., Chem. Commun., 1992, 1150.
G. R. Bamwenda, A. Ogata, A. Obuchi, J. Oi, K. Mizuno, and J.
Skrzypek, Appl. Catal., B, 6, 311 (1995).
period, the contribution of the C(d) as a NO reductant seems
to diminish gradually. The time course of the outlet C2H4 con-
centration during the C(d)-C2H4-SCR is consistent with that of
the corresponding H2O concentration; the H2O concentration
was initially zero, increased, and finally reached that during
the C2H4-SCR (370 ppm) through the plateau (170 ppm). No
formation of H2O is ascribed to the C(d) which completely cov-
ered alumina surface. The H2O concentrations, 370 and
4
5
6
7
K. Shimizu, H. Kawabata, H. Maeshima, A. Satsuma, and T. Hattori,
J. Phys. Chem. B, 104, 2885 (2000).
N. Okazaki, Y. Shiina, H. Itoh, A. Tada, and M. Iwamoto, Catal. Lett.,
49, 167 (1997).
Published on the web (Advance View) August 4, 2003; DOI 10.1246/cl.2003.794