Application of CuOx-CeO2 catalysts as selective sensor substrates for detection of CO in H2 fuel

Article abstract of DOI:10.1039/b807241h

The applicability and mechanism of CuO_x-CeO₂ as a catalytic microsensor substrate enabling 100% selective detection of low concentration CO in gas mixtures with H₂ is described. The Royal Society of Chemistry.

Full text of DOI:10.1039/b807241h

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Application of CuOx–CeO2 catalysts as selective sensor substrates for
detection of CO in H2 fuelw
Christopher S. Polster and Chelsey D. Baertsch*
Received (in Berkeley, CA, USA) 29th April 2008, Accepted 9th June 2008
First published as an Advance Article on the web 4th August 2008
DOI: 10.1039/b807241h
The applicability and mechanism of CuOx–CeO2 as a catalytic
microsensor substrate enabling 100% selective detection of low
concentration CO in gas mixtures with H2 is described.
CO oxidation selectivity in process streams containing
B10 000 ppm CO or more in H2 excess.4 The presence of
common reformate gases such as H2O and CO2 has been
reported to decrease CO oxidation rates with little deleterious
effect on the selectivity of CO oxidation.5 This work describes
the potential of CuOx–CeO2 catalysts for the selective detec-
tion of CO at much lower concentrations in H2.
Proton exchange membrane fuel cells (PEMFCs) using H2 fuel
are a promising source of portable power for electronic devices
and transportation.1 However, as CO is a common poison for
PEMFC electrodes and also a common contaminant in H2, it
is critical to be able to detect and quantify CO contamination
at low levels in concentrated H2 fuel. A selective catalyst for
low level CO detection in H2 is presented in this work. Sensors
for on-line detection of CO in portable devices must be small
in size, cost, and power consumption, and possess high
sensitivity. These requirements have been met with microelec-
tromechanical system (MEMS) signal transducers.2 However,
all current microsensor technologies use materials that are
inherently unselective to CO, give rise to false positive re-
sponses in the presence of H2, and require arrays of sensors for
gas identification.3
CuOx–CeO2 catalysts (4.5 at% Cu as measured by atomic
absorption) were synthesized by the urea gelation technique5a
from Cu(NO3)2Á3H2O and (NH4)2Ce(NO3)6 precursors fol-
lowed by calcination at 923 K. The BET surface area of the
prepared catalyst is 115.8 m2 À1. Insertion of Cu into the
g
CeO2 lattice reportedly leads to the formation of an active and
selective mixed metal oxide redox catalyst.6 Neither Cu, CuO,
nor Cu2O are observed by XRD, supporting favorable incor-
poration of Cu into the ceria structure, however a Ce : Cu
atomic surface ratio of B7 : 1 (from XPS) suggests that some
of the Cu is segregated to the catalyst surface. Reactions were
carried out in a packed-bed reactor with high purity gases
introduced and mixed upstream using mass flow controllers.
Selectivity is reported as the CO2 production rate divided by
the sum of CO2 and H2O production rates. Predicted sensor
responses (RCO and RH2) are calculated as a product of the
measured oxidation rate with the molar heat of reaction. An
Agilent microGC with Plot Q and molecular sieve columns
A new sensor paradigm and capability is presented for using
intrinsically selective catalysts coupled with MEMS thermal
transduction. As shown in Scheme 1, a catalytic substrate
selectively and exothermically oxidizes CO, providing a mea-
surable temperature rise that is directly proportional to CO
gas concentration and measured on chip using a resistive
temperature device or similar signal transducer. The technical
challenge giving rise to this high chemical specificity is the
development of a catalyst that oxidizes CO, but not H2, at
process conditions; this is the focus of this communication.
CuOx–CeO2 catalysts have been used in preferential oxida-
tion (PROX) operations with very high (up to 100%)
Scheme 1 Selective catalytic sensor operation.
School of Chemical Engineering, Purdue University, 480 Stadium
Mall Drive, West Lafayette, Indiana, 47907, USA.
E-mail: baertsch@purdue.edu; Tel: +1 765 496 7826
w Electronic supplementary information (ESI) available: Experimental
details as well as X-ray diffraction pattern for CuOx–CeO2 (Fig. S1).
See DOI: 10.1039/b807241h
Fig. 1 Steady state CO conversion and CO2 selectivity over CuOx–
CeO2 as a function of temperature. Reaction conditions: 50% H2, 1%
CO, 0.5% O2, balance He, total flow = 100 sccm (100 cm3 minÀ1),
Wcat E 100 mg.
ꢀc
This journal is The Royal Society of Chemistry 2008
4046 | Chem. Commun., 2008, 4046–4048

View Article Online
Isothermal oxidation reactions with variable conversion
were conducted to separate the convoluting effects of
temperature and conversion on selectivity. Such kinetic mea-
surements have not previously been reported for this system,
yet they provide substantial insight into catalyst function,
performance, and mechanism. Fig. 2 shows CO2 selectivity
over CuOx–CeO2 as a function of CO conversion and tem-
perature during reaction with 50% H2 in either 1% CO (a) or
100 ppm CO (b). With a 1% CO feed, 100% selectivity is
achieved at low conversion (o10%) from 333–403 K and at
333 K, 100% selectivity is maintained from 0–90% conver-
sion. Selectivity, however, decreases with both increasing
temperature and increasing CO conversion at temperatures
above 333 K. With a 100 ppm CO feed, 100% CO oxidation
selectivity is not observed at any measured temperature or
conversion; selectivity still decreases with increasing tempera-
ture, but is independent of CO conversion.
H2 and CO oxidation rates are reported in Table 1 as a
function of temperature and feed composition (single reactants
vs. CO–H2 mixtures). For both independent oxidation reac-
tions (just CO or H2 independently) and PROX reactions, CO
oxidation rates are higher than H2 oxidation rates at all
temperatures. PROX selectivities are much higher (particu-
larly at lower temperatures) than ideal selectivities calculated
from independent CO and H2 oxidation rates. Thus, factors in
addition to the kinetic differences between CO and H2 oxida-
tion must contribute to this catalyst’s ability to oxidize CO in
H2 with 100% selectivity.
From the results presented in Fig. 2 and Table 1 it is clear
that the required 100% CO oxidation selectivity is achieved by
operating (1) at lower temperatures where CO oxidation is
appreciably more favorable than H2 oxidation and (2) at high
CO gas concentrations and/or low CO conversion (high CO
conversion leads to a reduction in the effective CO pressure
from inlet conditions). Both conditions are critical for achiev-
ing high CO oxidation selectivity and minimizing H2 oxidation
to H2O. The high CO oxidation selectivity over CuOx–CeO2
appears to result from a combination of kinetic and adsorp-
tion factors. Conditions which should lead to an increase in
CO coverage (lower temperature, higher CO gas concentra-
tion) provide high CO oxidation selectivities.
Fig. 2 Steady state CO2 selectivity as a function of CO conversion
and temperature over CuOx–CeO2. Reaction conditions: (a) 50% H2,
1% CO, 0.5% O2, balance He. (b) 50% H2, 100 ppm CO, 50 ppm O2,
balance He, total flow = 50–400 sccm, Wcat E 2–20 mg.
was used for product gas analysis with a lower limit on H2O of
B2 ppm.
Catalytic activity during oxidation of 1% CO in the presence
of 50% H2 (typical relative concentrations of a steam reformed
H2 source7) is shown in Fig. 1 as a function of temperature. As
previously reported at these gas concentrations,4,5a,8 the cata-
lyst maintains 100% selectivity up to nearly 363 K.
For sensors, it is critical to maximize selectivity to the target
gas in order to avoid false positive responses. 100% selectivity
(or an infinite CO/H2 response ratio, RCO/RH2) is desirable.
Table 2 shows the selectivity and predicted sensor response as
the percentage that would be attributed to reaction with H2.
Table 1 Comparison of rates and selectivities from pure component oxidation reactions and PROX reactions
CO Ox. ratee (mol mÀ2
CO onlya
s
À1) Â 109
H2 Ox. ratee (mol mÀ2
H2 onlyb
s
À1) Â 109
Selectivityd (%)
H2 + COc
H2 + COc
Ideal
PROX
Temp./K
333
353
363
373
a
0.6
1.8
5.5
11.1
1.5
2.0
3.8
4.0
0.2
0.6
2.0
2.9
0.0
0.4
0.9
1.4
75
75
73
79
100
83
81
74
b
c
CO oxidation conditions: 200 ppm CO, 100 ppm O2, balance He. H2 oxidation conditions: 25% H2, 100 ppm O2, balance He. PROX reaction
d
conditions: 25% H2, 200 ppm CO, 100 ppm O2, balance He. Calculated by the CO2 production rate divided by the sum of CO2 and H2O
e
production rates. Differential rates reported (CO or O2 conversion o10%).
ꢀc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 4046–4048 | 4047

View Article Online
Table 2 CO2 Selectivity and corresponding predicted H2 response
however, they will adsorb irreversibly and poison the catalyst.
This is an important design consideration for these sensor
systems, as some hydrocarbon contaminants must either be
purified from the sample gas or degassed from the catalyst
surface intermittently during operation.
CO2 Selectivity (%) Predicted H2 responsea (%)
333 K
353 K
333 K
353 K
CO Conc. (ppm)
900
500
300
200
100
50
100b
100b
100b
100b
93
100b
97
83
80
81
71
0
0
0
0
6
11
0
2
15
18
16
26
For the first time, CuOx–CeO2 catalysts were demonstrated
as a promising catalytic substrate for low level detection of CO
in H2 fuel for sensors using reaction calorimetry. False positive
responses to H2 can be avoided during detection of CO at
concentrations as low as 200 ppm at 333 K. The catalyst’s high
selectivity results from a kinetic advantage of CO over H2
oxidation along with a competitive adsorption mechanism
with preferential adsorption of CO. Integration of this catalyst
with MEMS temperature sensors will provide new capabilities
in portable sensors for the selective quantitative detection of
CO in H2 while avoiding false positive responses.
87
a
Calculated as the heat evolved from H2 oxidation divided by the sum
of the heats evolved from both CO and H2 oxidation  100%. 25%
b
H2, stoichiometric O2 (wrt CO), balance He (o10% conversion). No
H2O detected. Minimum H2O detection limit is B2 ppm.
CuOx–CeO2 provides very high selectivities with no detectable
contribution from H2 (RCO/RH2 = N) at CO concentrations
down to o200 ppm and o500 ppm at operating temperatures
of 333 K and 353 K, respectively.
Notes and references
Even assuming that water is formed at its minimum detec-
tion limit, with a CO concentration of 300 ppm where no H2
oxidation is detected, the lower bound on RCO/RH2 is 30, and
in as low as 50 ppm CO at 333 K, only 11% of the catalytic
response is due to H2 background, corresponding to a sensor
response ratio (RCO/RH2) of 8. CO/H2 response ratios on the
order of 5–35, depending on substrate composition, gas com-
position (200–1000 ppm) and temperature, have been reported
for resistive type semiconductor gas sensors.9,10 Such results,
however, are reported as ratios of RCO and RH2 measured in
separate experiments containing either CO or H2 at equal
molar concentrations, while the studies presented here are
done under required detection conditions of CO mixed with
a large H2 excess (RCO/RH2 directly) and account for cross
sensitivity and chemical interactions. Cross sensitivity to oxi-
dizable gases other than H2 can also be a concern, but
preliminary studies show CuOx–CeO2 will not oxidize hydro-
carbons such as acetone or ethanol at these temperatures;
1 C. K. Dyer, J. Power Sources, 2002, 106, 31–34.
2 (a) I. Simon, N. Barsan, M. Bauer and U. Weimar, Sens. Actua-
tors, B, 2001, B73, 1–26; (b) H. L. Tuller and R. Mlcak,
J. Electroceram., 2000, 4, 415–425; (c) C. Hagleitner, D. Lange,
A. Hierlemann, O. Brand and H. Baltes, IEEE J. Solid-State
Circuits, 2002, 37, 1867–1878.
3 V. N. Mishra and R. P. Agarwal, Microelectron. J., 1998, 29,
861–874.
4 G. Sedmak, S. Hocevar and J. Levec, J. Catal., 2003, 213, 135–150.
5 (a) Y. Liu, Q. Fu and M. Flyzani-Stephanopoulos, Catal. Today,
2004, 93–95, 241–246; (b) G. Avgouropoulos, T. Ioannides, H. K.
Matralis, J. Batista and S. Hocevar, Catal. Lett., 2001, 73, 33–40.
6 X. Tang, B. Zhang, Y. Li, Q. Xin and W. Shen, Appl. Catal., A,
2005, 288, 116–125.
7 (a) C. D. Dudfield, R. Chen and P. L. Adcock, Int. J. Hydrogen
Energy, 2001, 26, 763–775; (b) G. Marban and A. B. Fuertes, Appl.
Catal., B, 2005, 57, 43–53.
8 G. Avgouropoulos and T. Ioannides, Appl. Catal., B, 2006, 67,
1–11.
9 U.-S. Choi, G. Sakai, K. Shimanoe and N. Yamazoe, Sens.
Actuators, B), 2005, 107, 397–401.
10 J. H. Yu and G. M. Choi, Sens. Actuators, B, 2001, 75, 56–71.
ꢀc
This journal is The Royal Society of Chemistry 2008
4048 | Chem. Commun., 2008, 4046–4048