APPLIED

CATALYSIS
A: GENERAL
ELSEVIER Applied Catalysis A: General 162 (1997) 121-131

TPR and TPD studies of C u O / C e O 2 catalysts for

low temperature CO oxidation
a*
Meng-Fei Luo ' , Yi-Jun Zhong b, Xian-Xin Yuan a, Xiao-Ming Zheng a
~lnstitute of Catalysis, Hangzhou University, Hangzhou 310028, China
bDepartment of Chemistry, Zhejiang Normal Universi~,, Jinhua 321004, China

Received 27 November 1996; received in revised form 18 February 1997: accepted 26 February 1997

Abstract

Copper oxide supported on cerium dioxide (CuO/CeO2) catalysts were prepared and used for carbon monoxide oxidation in
stoichiometric carbon monoxide and oxygen. The catalysts were characterized by means of XRD, H2-TPR and CO-TPD
studies. The CuO/CeO2 catalysts exhibit high catalytic activity in CO oxidation, showing markedly enhanced catalytic
activities due to the combined effect of copper oxide and cerium dioxide. The activity of the CuO/CeO2 (15%) catalyst
prepared by impregnation is higher than that prepared by co-precipitation. CeO2 promotes the hydrogen reduction activity of
copper, so that CuO/CeO2 catalysts show a different behavior with respect to pure CuO. Two reducible copper species were
observed in all CuO/CeO2 catalysts. CO-TPD experiments revealed that CuO/CeO2 catalysts can adsorb CO, while pure CuO
and CeO2 cannot. Combining the results of TPR, TPD study, and the catalytic activity measurements, it is proposed that the
well dispersed CuO which can adsorb CO and which is reducible at low-temperature is responsible for low-temperature CO
oxidation. The bulk CuO which cannot adsorb CO and which is reducible at high-temperature contributes little to the
oxidation activity. © 1997 Elsevier Science B.V.

Keywords: Carbon monoxide oxidation; Copper oxide; Cerium dioxide

1. Introduction tion [2-5]. Due to the limited availability of precious

metals, attention has been given to base metals as
CO oxidation catalysts with high activity at or near catalysts, especially copper oxide [6,7]. During CO
room temperature are useful for several applications oxidation, copper oxide can exhibit activities per unit
including closed-cycle CO2 lasers and air purification surface area similar to those of noble-metal catalysts
[1]. Many different catalysts have been prepared and such as platinum [8]. Cerium oxide has been widely
tested for low-temperature CO oxidation activity. For used in the automotive three-way catalytic converter
example, noble metal catalysts such as Pd/SnO2, Pt/ as an oxygen storage medium and thermal stabilizer.
SnO2, Au/MnOx, Au/c~-Fe203 and A u / C o 3 0 4 have The interaction of ceria with precious metals (Pd, Pt,
been demonstrated to be very effective in CO oxida- Rh) and its effect on catalytic activity have been
intensively studied [9]. Lin and co-workers [ 10,11 ]
*Corresponding author. Fax: +86 571 8070107. found that C u - C e - O composite catalyst is active for

0926-860X/97/$17.00 ~© 1997 Elsevier Science B.V. All rights reserved.

PII S0926-860X(97)00089-6
122 M.-F. Luo et al./Applied Catalysis A: General 162 (1997) 121-13l

carbon monoxide oxidation. Complete oxidation of placed in a TPR cell at 20°C, into which H2-N2
carbon monoxide occurs at 100°C in air. Cu-Ce-O (5 : 95) mixed gas was introduced. The water pro-
composite catalyst also showed excellent resistance duced by the reduction was trapped on a 5A molecular
to water poisoning. By means of XPS and XRD sieve. The temperature of the sample was programmed
L i n e t al. [10,11] suggested that a strong inter- to rise at a constant rate of 20°C/min and the amount of
action exists between the transition metal and cerium H2 uptake during the reduction was measured by a
oxide. thermal conductivity detector (TCD).
The present work is concerned with TPR and
TPD studies of CuO/CeO2 catalysts and with
2.3. CO-TPD
oxidation activity for carbon monoxide. Our aim is
to gain some direct evidence of the synergism between
TPD experiments were performed after CO adsorp-
CuO and CeO2. Such information may aid the devel-
tion on samples. A 100 mg sample was loaded in a
opment of CuO/CeO2 catalysts which can replace
quartz microreactor and heated in He at 650°C for
noble metal catalysts for the low-temperature CO
0.5 h, then cooled to 20°C in a flow He (40 ml min 1).
oxidation.
Adsorption of CO (>99.5%) was performed at 20°C by
injecting several consecutive pulses of 1 ml until the
amount of CO of effluent stream is constant (with
2. Experimental
TCD). Excess of adsorbate was removed by allowing
the sample to remain in He flow until no significant
2.1. Preparation of catalysts
amount of adsorbate could be detected (approximately
0.5 h). The sample was then ramped to 500°C at a
CeO2 was prepared by thermal decomposition of
linear heating rate of 20°C min -1 in flowing He
cerous nitrate, Ce(NO3)3, for 4 h at 650°C in air. The
(40 ml min 1). The analysis of the effluent gases
BET surface area of the obtained CeO 2 material is
(CO and CO2) was performed with a mass spectro-
55 m 2 g-1. The supported CuO/CeO2 catalysts were
meter (ZHD-OIS instrument).
prepared by the conventional wet impregnation
The amount of CO adsorbed on the catalysts was
method using an aqueous solution of Cu(NO3)2. Pre-
measured by a pulse method at 20°C. The pretreat-
pared samples were dried overnight in an oven at
ment was the same with CO-TPD. 0.5 ml of CO was
120°C and then heated in air for 4 h at 650°C. The
injected with a syringe repeatedly until no adsorption
loading of CuO is 0.25 ~ 15%, respectively. The CuO/
was observed, the amount of catalyst was chosen so as
CeO2 (CuO 15%) samples were also calcined at 400,
to allow more than 10% of the first injected CO to be
550, 800 and 950°C to examine the influence of
adsorbed. The amount of CO adsorbed was calculated
calcination temperature. The catalyst prepared by
from the difference between amounts of injected and
the impregnation method is denoted as CuOICeO2
eluted CO.
(X%).
For the comparison, catalysts were also prepared by
coprecipitating Cu(NO3)2 and Ce(NO3)3 solutions by 2.4. Activity measurements
adding NaOH. The precipitate was washed with
deionized water and then dried for about 5 h at 120°C. Catalytic activity measurements were carried out in
The dried sample was heated in air for 4 h at 650°C. a fixed bed reactor (0.6 cm i.d.) using 150 mg of
This catalystis denoted as CuO (15%)-CEO2. Nacontent catalyst of 20~60 mesh size. The total gas flow rate
of the catalyst is 96.25 ~tg/g by atomic adsorption was set at 80 SCCM. The gas consists of 2.4% CO and
spectrometry. 1.2% 02 in N2. The catalysts were directly exposed to
80 SCCM of reaction gas as the reactor temperature
2.2. He-TPR stabilized at reaction temperature, without any pre-
treatment. The analyses of the reactor effluent were
TPR measurements were made in a flow system. performed with an on-line SP-2304 model gas chro-
10 mg catalyst was pretreated in air at 500°C and matography with a TCD.
 M.-E Luo et al./Applied Catalysis A: General 162 (1997) 121-131 123

3. Results and discussion The reduction profile of pure CuO is characterized by

a single peak at 380°C. A two-step reduction profile is
3.1. Temperature-programmed reduction observed for all CuO/CeO2 catalysts, indicating the
existence of two CuO species. TPR peak temperature
The TPR profiles for these C u O / C e O 2 catalysts are of CuO/CeO2 catalyst is much lower than that of pure
shown in Fig. 1. The reduction of pure CeO2 began at CuO. The intensity of the low-temperature peak
400°C (not shown in Fig. 1), and the signal of TPR is increases with CuO loading from 0.25% to 3.0%.
much weaker than that of CuO/CeO2 (0.25 %) catalyst. However as the CuO loading further increased to

&

Ii
11

;uO

0
O..
,(A)
E 15%
~.== t0%
0
L) 5%
q)
3%
0
"0 2%
>,,
I 1%
0.5%

I ! I 1
100 200 300 400 500
Temperature(°C)
Fig. 1. TPR profiles of CuO/CeO2 catalysts pretreated at 500°C. (A) CuO (15%)-CEO2; (B) HNO3 treated with CuO/CeO2 (15%) dried at
650°C; (C) HNO3 treated with CuO/CeO2 (15%) dried at 200°C.
124 M.-F. Luo et al./Applied Catalysis A." General 162 (1997) 121-131

15%, the intensity of the low-temperature peak does the untreated ones. The copper content of treated
not change. The low-temperature peak position is catalyst is about 0.9% by chemical analysis. This
unchanged with increased CuO loading around indicates that a large number of Cu atoms in catalyst
160°C. By contrast, the position of the high-tempera- was removed by this process. At the same calcination
ture peak shifts to higher temperature with increasing temperature, the same TPR peak is also observed with
CuO loading; the intensity also increases with the different immersed time (4~40 h). The temperature of
loading. Lin et al. suggested that bulk CuO can be TPR peak of the catalyst (by immersed) at 650°C
easily dissolved in nitric acid and bulk CuO particles calcination is much lower than that at 200°C, similar
in catalyst were removed by immersing the sample in to CuO/CeO2 (1%) catalyst. No lower temperature
nitric acid. In this work, the CuO/CeO2 (15%) catalyst peak (about 160°C) is observed in the catalyst (acid
was immersed in concentrated nitric acid (12ml treated) dried at 200°C. This indicates that the higher
HNO3/g catalyst) for 40 h and then filtered, washed temperature calcination made copper oxide disperse
with deionized water and dried at 200 or 650°C for 1 h. well. Therefore, we believe that the finely dispersed
The TPR profiles of the catalyst after these treatments CuO can also be dissolved in nitric acid. Therefore, we
are also shown in Fig. 1. The peak area of the catalyst propose that there are two reducible copper species in
treated with HNO3 was significantly less than that of CuO/CeO2 catalysts. One is represented by small CuO

0
5 A
0

0
0
0

,
0
m

13.
0

~41' 41,41,41,~ ~II,

I I !
1 O0 200 300
Temperature(°C)
Fig. 2. TPD spectra of CO2 after CO adsorption over CuO/CeO 2 catalyst. • 15%; O 5%; • 3%;/~ 2%; • 1%; [] 0.5%; ~ 0.25%; × HNO 3
treated with CuO/CeO2 dried at 650°C; • CuO, CeO2 or HNO3 treated with CuO/CeO2 dried at 200°C.
 M.-F. Luo et al./Applied Catalysis A: General 162 (1997) 121-131 125

particles which are finely dispersed, and interacting observed during CO-TPD. TPD spectra of CO2 after
with the CeO2 particles they reduce at lower tempera- CO adsorption at 20°C are given in Fig. 2. From
ture. Another is represented by larger particle bulk Fig. 2, we can see that there is only one CO2 desorp-
CuO, they reduce at higher temperature. From Fig. 1, tion peak on CuO]CeO2 catalyst at about I I0°C,
overlapping peak for CuO (15%)-CEO2 catalyst can with no peak on pure CuO or CeO2. This indicates
be seen. The temperature of the high-temperature peak that almost no CO is adsorbed on pure CuO and CeO2
of CuO (15%)-CEO2 is similar to CuO/CeO2 (15%). surfaces. The area of CO2 desorbed peak increases
But the temperature of the low-temperature peak of with increased CuO loading from 0.25% to 3.0%.
the former is higher than that of the latter. We suppose However at CuO loading higher than 3.0%, the area
that a small amount of Na remainder (96.25 ktg/g) and of CO2 desorbed peak no longer increased. The CO
prepare method cause the higher shift of the low- adsorption amount of the CuO/CeO2 (15%), CuO/
temperature peak. CeO2 (5%), CuO/CeO2 (3%), CuO/CeO2 (1%),
CuO/CeO2 (0.25%), pure CuO and CeO2 was 5.0,
3.2. T P D o f C O 4.9, 4.8, 3.3, 1.2, 0.05 and 0.03 mug, respectively. The
adsorption amount of pure CuO and CeO2 is very
For CO-TPD studies several doses of CO pules were small. The CO adsorption amount of the catalysts is
injected until excess CO was detected in He stream consistent with the CO2 desorption amount. Fig. 3
and the adsorbed species were then desorbed by shows the deconvoluted peak area of low-temperature
raising the temperature. Almost no CO (m/e=28) is peak in Fig. 1 as a function of the desorbed amount of

5%
5 - 3 % ~x~t
/ OO, o
~ 4

a o

0 I 1 I I I
0 1 2 3 4 5
Desorbed amount of CO2(a.u.)
Fig. 3. Deconvolutedpeak area of low-temperaturepeak in Fig. 1 as a function of desorbed amount of CO2 in Fig. 2.
126 M.-E Luo et al./Applied Catalysis A: General 162 (1997) 121-131

CO2 in Fig. 2. The plot in Fig. 3 show perfect straight (thermal decomposition of C u ( N O 3 ) 2 at 650°C for 4 h)
line between the area of low-temperature TPR peak and CeO2 is quite low, and the activity of the CuO/
with the area of CO2 desorbed peak. Almost no CO CeO2 catalysts is much higher than that of the pure
was adsorbed on the catalyst treated by nitric acid CuO and CeO2. The light-off temperature of CuO
dried at 200°C, but this catalyst dried at 650°C can (15%)-CEO2 catalyst prepared by coprecipitation is
adsorb CO; the area of CO2 adsorption peak is similar about 85°C higher than that of C u O / C e O 2 (15%)
to CuO/CeO2 (1%) catalyst (Fig. 2). It is evident these prepared by impregnation. This indicates that activity
CO2 peaks are due to bidentate carbonates species of the supported catalyst is higher than that of copre-
produced by interaction of CO with lattice oxygen in cipitation catalyst. The activity of CuO/CeO2 catalysts
the finely dispersed CuO. Therefore, the CO adsorp- increases with CuO loading from 0.25% to 3%, then
tion site of CuO/CeO2 catalyst is the finely dispersed remains constant for a higher loading of CuO (from
CuO species. 3% to 15%). The order of CO oxidation activity of
CuO/CeO2 is consistent with intensity of the low-
3.3. Catalytic activity for CO oxidation temperature peak (Fig. 1) and CO2 adsorption peak
(Fig. 3) of catalysts. This also indicated that CO
Fig. 4 shows the light-off curves over CuO/CeO2 oxidation activity is related to reduction temperature
catalysts prepared by impregnation. None of these of finely dispersed CuO species. The easier the reduc-
catalysts were pretreated before running the reaction. tion of the finely dispersed CuO species is, the higher
This is particularly important in practical applications the oxidation activity of catalyst. The light-off tem-
where pretreatment is not possible or when the cata- perature of C u O / C e O 2 (>3%) is about 60°C lower
lysts must be stored in or exposed to air before use. At than that of the Cu-Ce-O composite catalysts reported
each reaction temperature, the reaction was continued in the literature [10]. Since pure CuO and CeO2
for about 1 h to achieve steady-state activity. From support has a very low activity, there is synergistic
Fig. 4, it can be seen that the activity of pure CuO interaction between the CuO and CeO2 support that is

1 O0 ,L~.~ll~-~r~. I-~ + ×/
x~×

g 80 /~
0 1O0- ,.~>---*S-..~'o
oa0
.-
/ o/ /
/ /
.o 50 - .

e- ×
O° 20 / 0 ~."~/
200 300
I
400
+ x
0 "" ~ / I ,4 t I I I I
80 80 I00 120 140 160 180
T e m pe ratu re( ~C)
Fig. 4. Effect of copper loading in CuO/CeO2 catalyst on CO oxidation activity. © 15%; • 10%; [] 5%; • 3%; /~ 2%; • 1%; ~) 0.5%; x
0.25%; + CuO (15%)~3eO2; ~ CuO; • CeO2.
 M.-F. Luo et [Link] Catalysis A: General 162 (1997) 121-131 127

J~ ~. ,. ~ _ 3°/o_~

L__..JLS~L_.~ _._/~ :_LJ_~ _._- ~ I S'/o

20 40 60 80 90
2O
Fig. 5. XRD pattern of the CuO/Ce02 catalyst (Cu Kc0.

100

60 °C
o',t 80 m

o,I
0
0
o 6O
t-
O

40
(-
o
0 2O _ 55°C
) o o ~ o o o o o o (:z:~oooooooo

0 I I I I I I
0 200 400 600 800 1000 1200
Time(min)
Fig. 6. The percentage yield of CO2 over CuO/CeO 2 (15%) catalyst as a function of reaction time.
128 M.-F. Luo et al./Applied Catalysis A." General 162 (1997) 121-131

responsible for the high activity of carbon monoxide the catalytic activity, TPD, and TPR results, we con-
oxidation at low temperature. Fig. 5 shows the XRD clude that the finely dispersed CuO species that can
patterns of these catalysts. At low CuO loading (<5%) absorb CO mainly contribute to the catalytic activity.
no visible CuO crystal phases can be observed. As the Only a small amount of copper oxide (around 3%) is
CuO loading increased, the crystal phase of CuO needed to promote the CO oxidation activity of CeO2,
becomes apparent. On the basis of the activity results, and excess copper forms bulk CuO that cannot adsorb
we propose that only a small amount of copper (3%) is CO and contributes little to the activity.
needed to form active sites for the oxidation reaction, Deo et al. [12] reported that bridge-bonded vana-
and that excess copper forms bulk CuO particles dium oxide, such as Z r - O - V is more reducible than
contributing little to the activity. Lin and co-workers that of terminal bond oxygen (V=O). Brogan et al. [ 13]
[10,11] suggested that the activity of C u - C e - O cata- pointed out that when Pt is supported on CeO> the
lyst derives primarily from the combination of finely surface oxygen can be easily removed at low tem-
dispersed copper-cerium oxide system, while the CuO perature owing to Pt-O-Ce linkage. More recently,
particles have negligible contribution. The data shown Dow et al. [14] pointed out that the shared oxygen ion
in Fig. 4 are consistent with this assertion. The light- Z r - O - C u can be reduced more easily than unshared
off temperature of the catalyst treated nitric acid at capping oxygen ion. Therefore, we suggest that the
650°C calcination was about 70°C, which is much low-temperature peak in Fig. 1 is similar to the shared
lower than that at 200°C calcination. On the basis of oxygen ion C e - O - C u the shared oxygen ion can be

100

,-,t 80
¢w

0 00
0
o 60
e-
0
50
m 40
e.-

0
0 2O 0
1 00 200 300

0 I I I I
50 60 70 80 90 1 00
Temperature(°C)
Fig. 7. Effect of calcination temperatureof the CuO/CeO2 (15%) catalyston light-offtemperatureof CO oxidation. A 400°C; • 550°C; ©
650°C; • 800°C; ~>950°C.
 M.-F. Luo et al./Applied Catalysis A: General 162 (1997) 121-131 129

reduced more easily than unshared capping oxygen ently does not decay with reaction time. The activity
ion. The shared capping oxygen ion is also adsorption of this catalyst is close to Au/MnOx catalyst, and
sites of CO [15]. The shared oxygen ions reacts with clearly superior to the Pt/SnOx catalyst reported in
CO to produce CO2 with rose desorption temperature. the literature [5].
Jernigan et al. [16] pointed out that CO oxidation over
CuO catalyst belongs to a redox mechanism, the rate- 3.4. Effect of calcination temperature
limiting is the reduction of CuO by CO. Based on the
above discussion, we attribute the enhanced catalytic Fig. 7 shows that the effect of calcination tempera-
activity of CuO/CeO2 to the shared oxygen ion C e - O - ture of the CuO/CeO2 (15%) catalyst on light-off
Cu, which can provide active oxygen species and CO temperature of CO oxidation. The light-off tempera-
adsorption sites for CO oxidation. ture reaches a minimum at 650°C calcination. The
The CO oxidation activity of CuO/CeO2 (15%) activity of the catalyst calcined 650°C is the highest.
catalyst as function of reaction time is shown in After 950°C calcination, the light-off temperature
Fig. 6. The activity of the CuO/CeO2 catalyst appar- markedly increases. Fig. 8 shows the TPR profiles

5
v

,
O
i

o_
E j °°c
t-
O
o 800 °C
t-

650°C
£ P------" I
550 °C
T
L____#
400°C

100 200 300 400 500

Temperature(°C)
Fig. 8. TPR profiles of CuO/CeO2 (15%) catalyst calcined at different temperatures.
130 M.-F. Luo et al./Applied Catalysis A: General 162 (1997) 121-131

d~
cq
0
(0
0 JZ,\
O.)
//7 'g\k

I I
100 200 300
Temperature(°C)
Fig. 9. TPD spectra of CO 2 after CO adsorption over CuO/CeO2 (15%) catalyst calcined at different temperatures. O 400°C; • 550°C; A
650°C; • 800°C; <) 950°C.

of CuO/CeO2 (15%) catalyst calcined at different the basis of the above results, we attributed the
temperature. Fig. 9 shows the TPD spectra of CO2 enhanced catalytic activity to synergistic interaction
after CO adsorption on these catalysts. The CO of CuO and CeO2, which results in the easier reduced
adsorption amount of CuO/CeO2 (15%) catalyst cal- CuO species and CO adsorption on CuO/CeO2 cata-
cined at 800°C and 950°C was 1.8 and 0.20 ml/g, lyst.
respectively. After 950°C calcination the double TPR
peak disappeared and TPR profile is characterized by
single peak centered at about 380°C (Fig. 8), and CO2 4. Conclusions
desorption peak also disappeared (Fig. 9). Compari-
son of Figs. 8 and 9 indicates that the light-off tem- CuO/CeO2 catalysts of different CuO loading have
perature of CO oxidation is consistent with the been prepared and tested for low-temperature carbon
temperature of low-temperature TPR peak and the monoxide in stoichiometric mixtures of carbon mon-
area of CO2 desorbed peak. This also indicates that oxide and oxygen. The CuO/CeO2 catalyst prepared
easier reduction CuO species on catalyst is a possible via impregnation show higher CO oxidation activity
active site of CO oxidation, which can adsorb CO. On than that of the Cu-Ce-O composite catalyst, and do
 M.-F. Luo et al./Applied Catalysis A: General 162 (1997) 121-131 131

not exhibit deactivation of catalytic activity. The finely [2] T. Engel, G. Ertl, Adv. Catal. 28 (1979) 1.
[3] M. Harute, S. Tsubota, T. Kobayshi, H. Kageyama, M.J.
dispersed CuO on CeO2 is responsible for low-tem-
Genet, B. Delmon, J. Catal. 144 (1993) 175.
perature CO oxidation, and the bulk CuO contributes [4] D.R, Schryer, B.T. Upchurch, J.D. Van Norman, K.G. Bromn,
little to the activity. The TPR investigation of CuO/ J. Schryer, J. Catal. 122 (1990) 193.
CeO2 catalyst showed that CeO2 promotes the reduc- [5] S.D. Gardner, G.B. Hoflund, B.T. Upchurch, D.R. Schryer,
tion of CuO species. The observation of two reduction E.J. Kielin, J. Schryer, J. Catal. 129 (1991) 114.
[6] T.-J. Huang, T.-C. Yu, S.-H. Chang, Appl. Catal. 52 (1989)
peaks in H2-TPR suggests the presence of two differ-
157.
ent CuO species. The low-temperature TPR peak of [7] T.-J. Huang, T.-C. Yu, Appl. Catal. 72 (1991) 275.
CuO/CeO2 is attributed to the finely dispersed CuO [8] J.K. Hummer, Prog. Energy Combust. Sci. 6 (1980) 177.
species that can adsorb CO, and the high-temperature [9] H.C. Yao, Y.F.Y. Yao, J. Catal. 86 (1984) 254.
peak is attributed to the bulk CuO species which [10] W. Lin, M. Flytzani-Stephanopoulos, J. Catal. 153 (1995)
304.
cannot adsorb CO.
[11] W. Lin, M. Flytzani-Stephanopoulos, J. Catal. 153 (1995)
317.
[12] G. Deo, I.E. Wachs, J. Catal. 146 (1994) 323.
Acknowledgements [13] M.S. Brogan, T.J. Dines, J.A. Cairns, J. Chem. Soc. Faraday
Trans. 90 (1994) 1461.
This work was supported by Zhejiang Provincial [14] W.-P. Dow, Y.-P. Wang, T.-J. Huang, J. Catal. 160 (1996)
Natural Science Foundation of China. 155.
[15] W.-P. Dow, T.-J. Huang, J. Catal. 160 (1996) 171.
[16] G.G. Jernigan, G.A. Somorjai, J. Catal. 147 (1994) 567.
References

[1] D.R. Schryer, B.T. Upchurch, B.D. Sidney, K.G. Bromn, G.B.
Hoflund, P.K. Herz, J. Catal. 130 (1991) 314.