Reactivity of Rh+(CO)2 during the NO-CO and CO-O2 Reactions over Rh/Al2O3

Article abstract of DOI:10.1006/jcat.1999.2702

Exposure of Rh⁺(CO)₂ on Rh/Al₂O₃ to NO causes CO desorption and adsorption of NO as Rh-NO⁺; exposure of Rh⁺(CO)₂ to NO/H₂ produced N₂O at 573 K. The presenc

Full text of DOI:10.1006/jcat.1999.2702

Journal of Catalysis 189, 247–252 (2000)
Article ID jcat.1999.2702, available online at http://www.idealibrary.com on
RESEARCH NOTE
Reactivity of Rh+(CO)2 during the NO–CO and CO–O2
Reactions over Rh/Al2O3
Khalid A. Almusaiteer, Steven S. C. Chuang,1 and Cher-Dip Tan
Department of Chemical Engineering, The University of Akron, Akron, Ohio 44325-3906
Received July 20, 1999; revised September 27, 1999; accepted September 28, 1999
tive adsorbates for the NO–CO reaction on Pd/Al2O3 (33).
Exposure of Rh+(CO)2 on Rh/Al2O3 to NO causes CO desorption
and adsorption of NO as Rh–NO+; exposure of Rh+(CO)2 to NO/H2
produced N2O at 573 K. The presence of both reductant (i.e., CO)
and oxidant (i.e., NO or O2) in the reactant pulse is needed to initiate
and sustain the NO–CO redox reaction cycle for CO2 formation.
The use of this technique on Rh/Al2O3 did not lead to useful
information to determine the Rh+(CO)2 role in the NO–CO
reaction due to extensive modification of the Rh catalyst
surface state during the addition of O2 as a poison (35).
In the present study, we have examined the Rh+(CO)2
reactivity and its role in CO2 formation by exposing
Rh+(CO)2 to NO, 13CO/NO, NO/H2, air, and 13CO/air. Di-
rect IR spectroscopic evidence is obtained to show involve-
ment of Rh+(CO)2 in CO2 formation during the 13CO/NO
pulse reaction over Rh/Al2O3.
c
2000 Academic Press
Key Words: gem-dicarbonyl; NO–CO reaction; Rh catalyst; ad-
sorbed CO; transient studies; reactivity; redox cycle; adsorbed NO;
infrared spectroscopy; mass spectrometry; oxidative disruption; re-
ductive agglomeration; reaction mechanism.
EXPERIMENTAL
INTRODUCTION
The 2 wt% Rh/Al2O3 and 0.2 wt% Rh/Al2O3 catalysts
were prepared by incipient wetness impregnation of RhCl3
2H2O (Alfa Chemicals) onto a -alumina support (Alfa
Chemicals, 100 m2/g). The catalysts were dried overnight in
air at room temperature and calcined by flowing air at 723 K
for 6 h and then reduced by flowing H2 at 723 K for 6 h. The
dispersion of metal crystallite on Al2O3 was determined to
be 9% for 2 wt% Rh/Al2O3 and 94% for 0.2 wt% Rh/Al2O3
by H2 pulse chemisorption at 300 K.
Details of the experimental apparatus have been de-
scribed elsewhere (34) and will be briefly discussed here.
The experimental system consists of (i) the gas flow sys-
tem with a GC six-port sampling valve and mass flow con-
trollers for the NO, 12CO, 13CO, air, H2, and He flow, (ii) the
in situ IR reactor cell with self-supporting catalyst disk,
and (iii) the analysis section with a Nicolet Magna 550 IR
spectrometer for recording IR spectra of adsorbed species
and Balzers QMG 112 quadruple mass spectrometer (MS)
for the analysis of reactants and products. The gaseous re-
sponses for m/e ratios corresponding to m/e = 28 (12CO and
N2), m/e = 29 (13CO), m/e = 30 (NO), m/e = 44 (12CO2 and
N2O), and m/e = 45 (13CO2) were monitored by the MS.
The CO2 and N2O (m/e = 44) responses were separated by
using the response ratio of the CO+2 (m/e = 44) primary
ionization and the CO+2 + (m/e = 22) double ionization.
The nature of sites associated with Rh+(CO)2, rhodium
gem-dicarbonyl, and the mechanism of its formation have
been subjects of extensive studies (1–20) because of funda-
mental interests in understanding the role of Rh surface
states and their adsorbates in automobile catalysis (21–
25). Rh+(CO)2 can be produced from CO adsorption on
the oxide-supported Rh crystallites, oxidized Rh, RhCl3,
Rh(NO3)3, and partially decomposed Rh carbonyls (2, 9,
16, 19, 26–30). Rh+(CO)2 has been found to be a major CO
adsorbate on the Al2O3- and SiO2-supported Rh catalysts
during the NO–CO and CO–O2 reactions (5, 7, 9, 10, 17).
We have attempted to address the role and reactivity
of Rh+(CO)2 in the NO–CO reaction by in situ infrared
(IR) coupled with various transient techniques (16, 17, 28–
32). In situ IR coupled with steady-state isotopic transient
kinetic analysis (SSITKA) revealed a rapid exchange be-
tween gaseous 13CO and Rh+(12CO)2, but failed to pro-
vide unambiguous information to determine the role of
Rh+(CO)2 in the reaction (31). The selective enhance-
ment/poisoning technique, which we recently developed,
has unambiguously identified Pd0–NO and Pd0–CO as ac-
1 To whom correspondence should be addressed. E-mail: schuang@
uakron.edu.
247
0021-9517/00 $35.00
c
Copyright
2000 by Academic Press
All rights of reproduction in any form reserved.

248
ALMUSAITEER, CHUANG, AND TAN
RESULTS AND DISCUSSION
Formation of Rh+(CO)2
In situ IR studies have shown that CO adsorbs as gem-
dicarbonyl [Rh+(CO)2], linear CO [Rh0–CO], and bridged
CO [Rh02–CO]on supported Rh catalysts(1–20, 29–31). The
concentration of each species on the Rh surface is gov-
erned by the oxidation state of the Rh surface which de-
pends on the Rh crystallite size, redox environment (i.e.,
NO/CO ratio), and reaction temperature (1–20, 28–32).
Small Rh crystallites, high NO/CO ratio, and low tem-
perature (i.e., below the light-off temperature) favor the
Rh+(CO)2 formation and vice versa for large Rh crystallites
(7).
Attempts to study the reactivity of preadsorbed Rh0–
CO and (Rh)2–CO toward gaseous NO at temperatures
above 473 K have failed due to its rapid desorption (16).
It should be noted that desorption of Rh0–CO at 573 K in
the present study was due to its low thermal stability rather
than its reaction with oxygen impurities in He flow. Poten-
tial oxygen impurities in He flow was verified by flowing
He over a reduced Rh/Al2O3 bed/containing adsorbed CO
before entering the IR reactor cell. Rh0–CO and (Rh)2–CO
in the IR reactor cell was flushed by He and no CO2 was
observed.
Rh+(CO)2 produced from the NO–CO reaction persisted
in He flow at temperatures as high as 633 K on 0.2 wt%
Rh/Al2O3 (32). To facilitate Rh+ formation, the prereduced
2 wt% Rh/Al2O3 and 0.2 wt% Rh/Al2O3 were oxidized
by flowing NO which has been shown to be effective in
facilitating the oxidative disruption of Rh0 crystallites to
isolated Rh+ sites(19). FlowingCO over the NO-pretreated
catalyst at 573 K produced Rh+(CO)2.
Reaction of Rh+(CO)2 with NO
Figure 1a illustrates the experimental approach involv-
ing a 1-cm3 NO pulse over Rh+(CO)2 on 2 wt% Rh/Al2O3.
Figures 1b and 1c show the IR spectra of adsorbates and
MS analysis of the IR reactor cell effluent during a 1-cm3
NO pulse over Rh+(CO)2 on 2 wt% Rh/Al2O3 at 573
K. Pulsing NO over Rh+(CO)2 produced cationic NO
[Rh–NO+] at 1910 cm 1 and dinitrosyl [Rh+(CO)2] at 1839
and 1761 cm 1 as well as an increase in the MS intensity of
N2O and m/e = 28. CO2 was not produced during the pulse,
indicating the absence of the reaction between Rh+(CO)2
and gaseous NO. The absence of the reaction was further
verified (not shown here) by the exposure of Rh+(CO)2 to
various NO pulse sizes ranging from 0.5 to 10 cm3 on 2 wt%
FIG. 1. (a) Experimental approach: pulsing 1 cm3 of NO over
Rh+(CO)2, (b) IR spectra collected during a 1-cm3 NO pulse over 2 wt%
Rh/Al2O3 at 573 K, and (c) MS analysis of the reactor effluent.
Reaction of Rh+(CO)2 with 13CO/NO
The presence of CO along with NO could promote the
Rh/Al2O3 and 0.2 wt% Rh/Al2O3 catalysts. The lack of re- reduction of Rh+ to Rh0, facilitating NO dissociation and
activity of Rh+(CO)2 toward gaseous NO could be due to CO2 formation. The presence of both NO and CO reac-
the absence of Rh0 sites for NO dissociation and the insuf- tants may be needed to activate Rh+(CO)2 for the reaction
ficient concentration of Rh–NO during the NO pulse over with NO. Pulsing 10 cm3 of 13CO/NO at a molar ratio of
Rh+(CO)2.
1 : 1 was carried out over 2 wt% Rh/Al2O3 and 0.2 wt%

REACTIVITY OF Rh+(CO)2
249
FIG. 2. (a) IR spectra collected during a 10-cm3 13CO/NO pulse at a molar ratio of 1 : 1 over 2 wt% Rh/Al2O3 at 573 K and (b) MS analysis of the
reactor effluent. (c) IR spectra collected during a 10 cm3 13CO/NO pulse at a molar ratio of 1 : 1 over 0.2 wt% Rh/Al2O3 at 573 K and (d) MS analysis
of the reactor effluent.
Rh/Al2O3 at 573 K to determine Rh+(CO)2 reactivity in
The dramatic enhancement of Rh+(CO)2 reactivity in the
the presence of gaseous NO and CO. Investigation of two presence of both NO and CO could be due to not only CO-
different Rh loadings allowed the determination of the Rh induced reductive agglomeration of Rh+ to Rh0 crystallites
particle size effects on the Rh+(CO)2 reactivity and verifi- but also high CO coverage. The former is evidenced by a
cation of the IR and MS observations. Figure 2a shows that significantly higher intensity of the band at 1997 cm 1 than
1
pulsing 13CO/NO over Rh+(12CO)2 at 2098 and 2017 cm
that at 2048 cm 1 (shown in Fig. 2a), indicating that the Rh0–
caused the disappearance of Rh+(12CO)2 and the forma- 13CO band produced from reductive agglomeration is over-
1
tion of Rh+(13CO)2 at 2048 and 1977 cm and Rh0–NO
lapped with the asymmetriccomponent ofRh+(13CO)2. The
at 1630 cm 1. Figure 2b shows that the pulse produced formation of Rh0 is further evidenced by the presence of
13CO2 and 12CO2 and achieved nearly complete NO con- Rh0–NO at 1630 cm
.
1
version. Although 12CO2 can be formed from the reaction
The intensity of Rh+(12CO)2 is approximately the same
of gaseous 12CO desorbed from Rh+(12CO)2, the close re- (i.e., an equal amount of Rh+(12CO)2 species) for both
lationship between the decrease in Rh+(12CO)2 intensity 2 wt% Rh/Al2O3 and 0.2 wt% Rh/Al2O3 catalysts as shown
and the increase in 12CO2 formation suggests that 12CO2 in Figs. 2a and 2c; however, Figs. 2b and 2d show that the
formation is a result of the reaction between Rh+(12CO)2 2 wt% Rh/Al2O3 produced more 12CO2 and 13CO2 as well
and adsorbed oxygen from dissociated NO.
as higher NO conversion than 0.2 wt% Rh/Al2O3. This high

250
ALMUSAITEER, CHUANG, AND TAN
activityofthe 2 wt% Rh/Al2O3 catalyst maybe attributed to band at 2020 cm 1, and (iii) N2O formation. A decrease
its large, reduced Rh0 crystallites. The hydrogen chemisorp- in Rh+(CO)2 would cause an intensity reduction in both
tion study shows that the 2 wt% Rh/Al2O3 catalyst has bands at 2091 and 2020 cm 1 while an increase in Rh0–CO
1
˚
an average Rh crystallite size of 52 A whereas 0.2 wt% would increase the band intensity in the 2010- to 2060-cm
Rh/Al2O3 is highly dispersed with a Rh crystallite size of region. Variation of the bands at 2091 and 2002 cm 1 can be
˚
11 A. The relationship between the NO–CO reaction ac- further highlighted by the difference spectra presented in
1
tivity and the Rh crystallite size has also been established the inset in Fig. 3a which shows a negative band at 2091 cm
for Rh/Al2O3 and shows that the catalyst activity for the as well as positive bands at 2002 and 3735 cm 1. The nega-
reaction increased with increasing Rh crystallite size (36).
tive band at 2091 cm 1 reflects a decrease in the Rh+(CO)2
1
concentration while positive bands at 2020 and 3735 cm
Reaction of Rh+(CO)2 with NO/H2
correspond to an increase in the concentration of Rh0–CO
and an isolated OH group, respectively. The variation of
these species in the presence of H2 has been attributed to
the following reaction (8),
To further elucidate the role of reducing agents in CO2
formation, Rh+(CO)2 was exposed to an NO/H2 pulse. Puls-
ing 0.5 cm3 of NO/H2 at a molar ratio of 1 : 1 over Rh+(CO)2
on 0.2 wt% Rh/Al2O3 at 573 K as shown in Figs. 3a and
3b caused (i) a decrease in Al–NCO at 2237 cm 1, (ii) a
0
1 H2(g) + Rh+CO2
Rh –CO + OH + CO(g),
→
2
1
decrease in the band at 2091 cm and an increase in the
FIG. 3. (a) IR spectra collected during a 0.5-cm3 13CO/H2 pulse at a molar ratio of 1 : 1 over 0.2 wt% Rh/Al2O3 at 573 K and (b) MS analysis of the
reactor effluent. (c) IR spectra collected during a 0.5 cm3 13CO/air pulse at a molar ratio of 1 : 1 over 0.2 wt% Rh/Al2O3 at 573 K and (d) MS analysis
of the reactor effluent.

REACTIVITY OF Rh+(CO)2
251
which indicates conversion of Rh+ to Rh0 sites and lation is supported by the results of a 13CO/NO pulse over
Rh+(CO)2 to Rh0–CO.
Rh+(CO)2 on which CO promotes reduction of Rh+ to Rh0
Although the presence of H2 reduces part of Rh+ to Rh0 sites for Rh0–NO adsorption and dissociation.
for NO dissociation, Rh+(CO)2 remains inactive toward
gaseous NO. Adsorbed oxygen from dissociated NO is un-
Reaction of Rh+(CO)2 with Air and 13CO/Air
able to access adsorbed CO (i.e., Rh+(CO)2 and Rh0–CO)
for CO2 formation. We speculate that the concentration of
Since CO in Rh+(CO)2 was not accessible for the reac-
Rh0 sites produced from H2-induced reductive agglomera- tion with adsorbed oxygen from dissociated NO during the
tion is too low and these sites are too far away from the CO NO/H2 pulse, the reactivity of Rh+(CO)2 toward adsorbed
sites to react with CO and produce CO2. Thus, we postulate oxygen was further investigated by pulsing air and 13CO/air
that a sufficiently high concentration of both Rh0 and Rh+ over Rh+(CO)2 on 0.2 wt% Rh/Al2O3. Pulsing 0.5 or 10 cm3
as well as both the adsorbed reductant and oxidant may be of air over Rh+(CO)2 (not shown here) did not cause any
needed to initiate and sustain the NO–CO redox reaction variation in the Rh+(CO)2 intensity and did not produce
cycle as shown in Fig. 4b. In this cycle, Rh0–NO dissociates CO2 as well. Lack of reactivity of Rh+(CO)2 toward gaseous
to produce N2 and Rh+ sites (step 3) for Rh+(CO)2 forma- O2 has also been reported on Rh/SiO2 (10).
tion (step 4); CO reduces Rh+ to Rh0 (step 5), leading to
Addition of 13CO to air for the pulse dramatically modi-
CO2 formation. CO-induced reductive agglomeration (i.e., fied the reactivity of Rh+(CO)2 toward oxygen. Figures 3c
step 5) has been well-established (3, 5, 7, 19). Our postu- and 3d show that pulsing 0.5 cm3 of 13CO/air at a molar ratio
FIG. 4. (a) Proposed mechanism of the NO–CO reaction and (b) steps involved in the NO–CO redox reaction cycle.

252
ALMUSAITEER, CHUANG, AND TAN
of 1 : 1 over Rh+(CO)2 at 573 K caused the disappearance
of Rh+(12CO)2 at 2091 and 2020 cm 1, the appearance of
Rh+(13CO)2 at 2048 and 1987 cm 1, and 12CO2/13CO2 for-
mation. The significant broadening of the band in the 1950-
8. Basu, P., Panayotov, D., and Yates, J. T., Jr., J. Am. Chem. Soc. 110,
2074 (1988).
9. Dictor, R., J. Catal. 109, 89 (1988).
10. Li, Y. E., and Gonzalez, R. D., J. Phys. Chem. 92, 1589 (1988).
11. Buchanan, D. A., Hemandez, M. E., Solymosi, F., and Whit, J. M.,
J. Catal. 125, 456 (1990).
1
to 2070-cm region indicates the formation of Rh0–12CO
and Rh0–13CO species during the 13CO/air pulse. Depletion 12. Paul, D. K., and Yates, J. T., Jr., J. Phys. Chem. 95, 1699 (1991).
of Rh+(12CO)2 and formation of Rh0–12CO, Rh0–13CO, and
13. Logan, A. D., Sharoudi, K., and Datye, A. K., J. Phys. Chem. 95, 5568
(1991).
12CO2 clearly indicates that the presence of gaseous 13CO
14. Anderson, J. A., J. Chem. Soc., Faraday Trans. 87, 3907 (1991).
15. Chuang, S. S. C., and Pien, S. I., J. Catal. 135, 618 (1992).
16. Srinivas, G., Chuang, S. S. C., and Debnath, S., J. Catal. 148, 748 (1994).
promotesthe formation ofRh0 sites, facilitatingthe reaction
between Rh+(12CO)2 and adsorbed oxygen to form 12CO2.
17. Krishnamurthy, R., and Chuang, S. S. C., J. Phys. Chem. 99, 16727
(1995).
CONCLUSIONS
18. Berko, A., Menesi, G., and Solymosi, F., J. Phys. Chem. B 100, 17732
(1996).
Sufficiently high concentrations of both Rh0 and Rh+ as
well as both the adsorbed reductant and oxidant are needed
19. Novak, E., Sprinceana, D., and Solymosi, F., Appl. Catal. A 149, 89
(1997).
to initiate and sustain the redox cycle of the NO–CO and
CO–O2 reactions on the Rh catalyst. Results of this study
suggest that the reactivity of adsorbates can only be reli-
ably measured under conditions where all the reactants are
present on the catalyst. Studies involving exposure of the
first reactant adsorbate to the second or third reactant may
20. Conners, L., Hollis, T., Johnson, D. A., and Blyholder, G., J. Phys.
Chem. B 102, 10112 (1998).
21. Oh, S. H., Fisher, G. B., Carpenter, J. E., and Goodman, D. W., J. Catal.
100, 360 (1986).
22. Cho, B. K., Shanks, B. H., and Bailey, J. E., J. Catal. 109, 89 (1989).
23. Taylor, K. C., Catal. Rev.-Sci. Eng. 35, 457 (1993).
24. Shelef, M., Catal. Rev.-Sci. Eng. 36, 443 (1994).
not provide information directly relevant to heterogeneous 25. Simon Ng, K. Y., Belton, D. N., Schmieg, S. J., and Fisher, G. B.,
J. Catal. 146, 394 (1994).
26. Lamb, H. H., Gates, B. C., and Knozinger, H., Angew. Chem., Int. Ed.
Engl. 27, 1127 (1988).
catalysis of redox reactions.
ACKNOWLEDGMENT
27. Ichikawa, M., in “Advances in Catalysis,” Vol. 38, p. 283. Academic
Press, New York, 1992.
28. Chuang, S. S. C., and Debnath, S., J. Mol. Catal. 79, 323 (1993).
29. Srinivas, G., and Chuang, S. S. C., J. Phys. Chem. 98, 3031 (1994).
The authors gratefully acknowledge the support of this research by the
National Science Foundation under Grant CTS-942111996.
30. Chuang, S. S. C., Srinivas, G., and Mukherjee, A., J. Catal. 139, 490
REFERENCES
(1993).
31. Krishnamurthy, R., Chuang, S. S. C., and Balakos, M. W., J. Catal. 157,
512 (1995).
32. Chuang, S. S. C., and Tan, C. D., J. Catal. 173, 104 (1998).
33. Almusaiteer, A. K., and Chuang, S. S. C., J. Catal. 180, 161 (1998).
34. Chuang, S. S. C., Brundage, M. A., Balakos, M. W., and Srinivas, G.,
Appl. Spectrosc. 49, 1151 (1995).
35. Almusaiteer, K., Ph.D. Dissertation, UniversityofAkron, Akron, OH,
1999.
36. Oh, S. H., and Eickel, C. C., J. Catal. 128, 526 (1991).
1. Yang, A. C., and Garland, C. W., J. Phys. Chem. 61, 1504 (1957).
2. Yates, J. T., Jr., Duncan, T. M., Worley, S. D., and Vaughan, R. W.,
J. Chem. Phys. 70, 1219 (1979).
3. Solymosi, F., and Rasko, J., J. Catal. 63, 217 (1980).
4. Hecker, W. C., and Bell, A. T., J. Catal. 84, 200 (1983).
5. Solymosi, F., and Pasztor, M., J. Phys. Chem. 89, 4789 (1985).
6. Wong, C., and McCabe, R. W., J. Catal. 107, 535 (1987).
7. Solymosi, F., Bansagi, T., and Novak, E., J. Catal. 112, 183 (1988).