Compositional effects in Ru, Pd, Pt, and Rh-doped mesoporous tantalum oxide catalysts for ammonia synthesis

Article abstract of DOI:10.1021/ic062385d

A series of early metal-promoted Ru-, Pd-, Pt-, and Rh-doped mesoporous tantalum oxide catalysts were synthesized using a variety of dopant ratios and dopant precursors, and the effects of these parameters on the catalytic activity of NH₃ synthesis from H₂ and N₂ were explored. Previous studies on this system supported an unprecedented mechanism in which N-N cleavage occurred at the Ta sites rather than on Ru. The results of the present study showed, for all systems, that Ba is a better promoter than Cs or La and that the nitrate is a superior precursor for Ba than the isopropoxide or the hydroxide. ¹⁵N-labeling studies showed that residual nitrate functions as the major ammonia source in the first hour but that it does not account for the ammonia produced after the nitrate is completely consumed. Ru₃(CO)₁₂ proved to be a better Ru precursor than RuCl₃-3H₂O, and an almost linear increase in activity with increasing Ru loading level was observed at 350°C (623 K). However, at 175°C (448 K), the increase in Ru had no effect on the reaction rate. Pd functioned with comparable rates to Ru, while Pt and Rh functioned far less efficiently. The surprising activities for the Pd-doped catalysts, coupled with XPS evidence for low-valent Ta in this catalyst system, support a mechanism in which cleavage of the N-N triple bond occurs on Ta rather than the precious metal because the E_a value for N-N cleavage on Pd is 2.5 times greater than that for Ru, and the 9.3 kJ mol^-1 E_a value measured previously for the Ru system suggests that N-N cleavage cannot occur at the Ru surface.

Full text of DOI:10.1021/ic062385d

Inorg. Chem. 2007, 46, 5084−5092
Compositional Effects in Ru, Pd, Pt, and Rh-Doped Mesoporous
Tantalum Oxide Catalysts for Ammonia Synthesis
Chaoyang Yue,† Longhui Qiu,† Michel Trudeau,‡ and David Antonelli*,†
Department of Chemistry and Biochemistry, UniVersity of Windsor, 401 Sunset AVenue,
Windsor, Ontario N9B 3P4, Canada, and M. L. Trudeau, Emerging Technologies, Hydro-Qu e´ bec
Research Institute, 1800 BouleVarde Lionel-Boulet, Varennes, Quebec J3X 1S1, Canada
Received December 13, 2006
A series of early metal-promoted Ru-, Pd-, Pt-, and Rh-doped mesoporous tantalum oxide catalysts were synthesized
using a variety of dopant ratios and dopant precursors, and the effects of these parameters on the catalytic activity
of NH
3
2 2
synthesis from H and N were explored. Previous studies on this system supported an unprecedented
mechanism in which N−N cleavage occurred at the Ta sites rather than on Ru. The results of the present study
showed, for all systems, that Ba is a better promoter than Cs or La and that the nitrate is a superior precursor for
Ba than the isopropoxide or the hydroxide. 15N-labeling studies showed that residual nitrate functions as the major
ammonia source in the first hour but that it does not account for the ammonia produced after the nitrate is completely
consumed. Ru
3
(CO)12 proved to be a better Ru precursor than RuCl
3
‚
3H
2
O, and an almost linear increase in
C (448 K), the increase
activity with increasing Ru loading level was observed at 350
°C (623 K). However, at 175
°
in Ru had no effect on the reaction rate. Pd functioned with comparable rates to Ru, while Pt and Rh functioned
far less efficiently. The surprising activities for the Pd-doped catalysts, coupled with XPS evidence for low-valent
Ta in this catalyst system, support a mechanism in which cleavage of the N
−
N triple bond occurs on Ta rather
N cleavage on Pd is 2.5 times greater than that for Ru, and
a
E value measured previously for the Ru system suggests that N−N cleavage cannot occur at the
than the precious metal because the E
a
value for N
−
the 9.3 kJ mol-
1
Ru surface.
Introduction
The activation of dinitrogen has been one of the holy grails
the use of dinitrogen as a feedstock for the synthesis of fine
organics under mild conditions has still remained elusive.
Thus, the discovery of new and milder ways of selectively
cleaving and functionalizing the N-N triple bond is one of
the great challenges of modern inorganic chemistry. Over
the years, various supported transition metal-based catalytic
systems have been studied for ammonia synthesis. For this
particular process, Ru is generally accepted to be superior
to Fe or other transition metals; therefore, most current efforts
have focused on the investigation of the effects of different
1,2
of inorganic chemistry for several decades, and recent
advances in organometallic3 and materials chemistry
-7
8-10
only confirm the importance of this reaction. While the
synthesis of ammonia using transition metal catalysts is one
of the most important catalytic processes in all of chemistry,
*
To whom correspondence should be addressed. E-mail: danton@
uwindsor.ca. Fax: +1(519) 973-7098.
†
University of Windsor.
Hydro-Qu e´ bec Research Institute.
9-17
‡
supports and promoters.
Common supports include main
(
(
(
(
1) Schl o¨ gl, R. Angew. Chem., Int. Ed. 2003, 42, 2004.
group oxides such as magnesia, alumina, and silica, while
2) Fryzuk, M. D.; Johnson, S. A. Coord. Chem. ReV. 2000, 379, 200.
3) Yandulov, D. V.; Schrock, R. R. Science 2003, 301, 76.
4) Fryzuk, M. D.; Love, J. B.; Rettig, S. J.; Young, V. G. Science 1997,
(11) Aika, K.; Hori, H.; Ozaki, A. J. Catal. 1972, 27, 424.
(12) Aika, K.; Shimazaki, K.; Hattori, Y.; Ohya, A.; Ohshima, S.; Shirota,
K.; Ozaki, A. J. Catal. 1985, 92, 296.
(13) Murata, S.; Aika, K. J. Catal. 1992, 136, 110.
(14) Aika, K.; Takano, T.; Murata, S. J. Catal. 1992, 136, 126.
(15) Cisneros, M. D.; Lunsford, J. H. J. Catal. 1993, 141, 191.
(16) Murata, S.; Aika, K. J. Catal. 1992, 136, 118.
(17) Aika, K.; Kawahara, T.; Murata, S.; Onishi, T. Bull. Chem. Soc. Jpn.
1990, 63, 1221.
275, 1445.
(
(
(
(
(
5) Laplaza, C. E.; Cummins, C. C. Science 1995, 268, 861.
6) Pool, J. A.; Lobkovsky, E.; Chirlik, P. J. Nature 2004, 427, 527.
7) Kim, J.; Rees, D. C. Nature 1992, 360, 553.
8) Hoshino, K. Chem.sEur. J. 2001, 7, 2727.
9) Bielawa, H.; Hinrichsen, O.; Birkner, A.; Muhler, M. Angew. Chem.,
Int. Ed. 2001, 40, 1061.
(
10) Fishel, C. T.; Davis, R. J.; Garces, J. M. J. Catal. 1996, 163, 148.
5084 Inorganic Chemistry, Vol. 46, No. 12, 2007
10.1021/ic062385d CCC: $37.00
© 2007 American Chemical Society
Published on Web 05/12/2007

Ru, Pd, Pt, and Rh-Doped Mesoporous Tantalum Oxide Catalysts
the promoters are usually alkali metals, alkali metal oxides/
hydroxides, and alkali earth metal oxides because of their
strong electron-donating abilities, which are believed to
fixed bed. H
in the TGA experiments. NH
2
and N
2
(99.999%) were used in all catalytic runs and
produced from the reaction was
3
absorbed by a 0.01 N HCl aqueous solution and was quantitatively
determined by the “indophenol blue method”.24 XRD patterns (Cu
KR) were recorded on a Siemens D500 θ-2θ diffractometer. XPS
data were obtained on a Physical Electronics PHI-5500 using charge
neutralization, and all peaks were referenced to the carbon C-(C,
H) peak at 284.8 eV. TEM and EDS studies were conducted using
a H9000 HR-TEM spectrometer operated at 300 kV. Nitrogen
adsorption and desorption data were collected on a Micromeritics
13
modify the reactivity of the active Ru center. Recently, we
developed a new series of Ru-doped mesoporous Ta oxide
catalysts which convert dinitrogen into ammonia over the
18
temperature range of 22-350 °C (295-623 K). This system
was an extension of work we conducted on bis(toluene) Ti
and bis(toluene) Nb reduced mesoporous Ti, Ta, and Nb
oxides, which demonstrated that metallic phases on the
surface of the mesostructure are capable of stoichiometrically
ASAP 2010. Hydrogen differential thermal analysis (H -TDA)
measurements were performed on a TGA/SDTA851 (Mettler
2
cleaving the N-N triple bond of dinitrogen at room
Toledo) over the temperature range from room temperature to 450
temperature19-21
-1
°
4
C at a heating rate of 2 °C min , in a stream of 96% argon and
% hydrogen with a flow rate of 30 cm min . Mass spectrometry
and subsequently producing ammonia, as
3
-1
long as ambient moisture was present. Since bis(arene)
complexes constitute an exotic and expensive source of
electrons in any catalytic process, we developed a new Ru-
experiments were conducted at the University of Alberta using an
Applied Biosystems Mariner TOF mass spectrometer equipped with
an electrospray source. Aqueous solutions exposed to the synthesis
gas for 1 h were infused into the spectrometer at a rate of 25 µL
doped mesoporous Ta oxide system which uses H
a source of electrons and protons in the reduction of N
NH . The activities of this new system at 350 °C are
2
as both
2
to
-
1
sec
.
3
Synthesis. (a) Mesoporous Ta Oxide. Mesoporous tantalum
comparable to those for standard Ru-doped Haber catalysts
but drop off substantially after the first hour. Arrhenius plots
in this study provided surprising activation energies of only
oxides were synthesized following the ligand-assisted templating
method developed by Antonelli and Ying.25 The template was
removed by stirring of the mixture for 24 h with 1.5 equiv p-toluene
sulfonic acid, with respect to the original amount of amine template
used, in 1:1 methanol/diethylether. This was followed by a second
washing in the same solvent mixture for 24 h and three additional
washings in pure methanol, followed by oven-drying at 120 °C for
-
1 18
9.3 kJ mol , roughly 10% of that reported for the tradi-
tional Ru-based Haber systems and many times lower than
the threshold Ea for N
eV or 58-96 kJ mol ).
2
cleavage on a Ru surface (0.6-1.0
-
1
22,23
XPS studies on the material
-
3
8
h and heating at 300 °C at 10 Torr for 24 h. This procedure
was necessary to remove the last traces of the template (BP ) 180
C) and was verified by elemental analysis and the complete absence
during several stages of the process show strong evidence
for the involvement of reduced Ta species in the support
mesostructure. These data suggest a new mechanism, in
which the Ru acts as an interface to transfer electron density
from hydrogen to neighboring Ta sites on the oxide support,
which in the reduced form are then able to cleave dinitrogen.
The electrochemical involvement of the Ta oxide support in
°
-
1
of C-C and C-N vibrations from 1000 to 1500 cm in the IR.
b) Ru-Doped Mesoporous Ta Oxide. In a general procedure,
wt % ruthenium-doped tantalum oxide catalyst was prepared by
impregnating mesoporous Ta oxide with Ru (CO) (99%) in
(
5
3
12
tetrahydrofuran (THF). After it was stirred overnight, the mixture
was evaporated in a rotary evaporator and dried in situ at 70 °C
for 4 h. The resulting yellow powder was then placed in a reaction
tube and evacuated at 300 °C for 3 h. The temperature ramping
2
the N -cleavage process is not possible with traditional
supports, such as magnesia and alumina, which do not
possess variable oxidation states. This is particularly intrigu-
ing given that Fryzuk has found that low-valent Ta com-
time was 60 min. When ruthenium chloride hydrate (RuCl
99.98%) was used as the ruthenium precursor, template-free
mesoporous Ta oxide was impregnated with RuCl ‚xH O in a
3 2
‚xH O,
4
plexes readily activate dinitrogen under mild conditions. To
3
2
better understand this new system, the dependence of the
activities on Ru precursor, Ru loading levels, precious metal
dopant other than Ru (i.e., Pd, Pt, Rh), promoter metal,
promoter precursor, promoter loading levels, and hydrogen
activation temperature have been determined.
methanol suspension. After the mixture was stirred overnight, the
organic solvent was removed in a solvent storage flask on a Schlenk
line.
(c) Mesostructured Pt-Doped (or Rh-, Pd-) Ta Oxides. These
doped mesostructured oxides were synthesized by the ligand-
assisted templating method25 modified to introduce Pt (or Rh, Pd)
during the initial hydrolysis/condensation step. In an Ar glovebox,
Experimental Section
1
2 5 5
5.0 g of Ta(OC H ) was added to 2.1 g of liquid dodecylamine
Materials and Equipment. All chemicals were obtained from
Aldrich unless otherwise stated. Catalytic experiments were con-
ducted in a U-shaped Pyrex reactor with a sintered glass frit as the
in a 150 mL conical flask, and the mixture was stirred for 10 min.
Then, 1.3 g of platinum(IV) chloride powder was added, and it
was stirred for 1 day at 50 °C to form a uniform orange mixture.
This mixture was then removed from the glovebox, and 100 mL
of deionized water was added. The resulting precipitate was aged
in solution as described above for mesoporous Ta oxide and isolated
by filtration. The as-synthesized materials were then dried under
vacuum for 10 h at 100 °C and for another 10 h at 300 °C (this
step sublimes out ∼80% of the amine template as measured by
(
(
(
(
(
(
18) Yue, C.; Trudeau, M.; Antonelli, D. M. Chem. Commun. 2006, 18,
1918.
19) Vettraino, M.; He, X.; Trudeau, M.; Schurko, R.; Antonelli, D. M. J.
Am. Chem. Soc. 2002, 124, 9567.
20) Vettraino, M.; He, X.; Trudeau, M.; Drake, J. E.; Antonelli, D. M.
AdV. Funct. Mater. 2002, 3, 174.
21) Lezau, A.; Skadtchenko, B.; Trudeau, M.; Antonelli, D. M. J. Chem.
Soc., Dalton Trans. 2003, 4115.
2
IR), followed by a heat treatment in a flow of H for 5 h at 350 °C
22) Dahl, S.; Taylor, P. A.; T o¨ rnqvist, E.; Chorkendorff, I. J. Catal. 2000,
192, 381.
(24) Yue, C.; Trudeau, M.; Antonelli, D. M. Can. J. Chem. 2005, 83, 308.
(25) Antonelli, D. M.; Ying, J. Y. Angew. Chem., Int. Ed. Engl. 1996, 35,
426.
23) Egeberg, R. C.; Larsen, J. H.; Chorkendorff, I. Phys. Chem. Chem.
Phys. 2001, 3, 2007.
Inorganic Chemistry, Vol. 46, No. 12, 2007 5085

Yue et al.
Figure 1. TEM of 5% Ru-doped mesoporous Ta oxide.
to reduce the Pt and encourage metal cluster formation. Washing
of the product before reduction of the precious metal salt led to
significant leaching of the dopant. The materials were then subjected
to acid treatment, multiple washings, and heat treatment under
vacuum, exactly as described for pure mesoporous Ta oxide.
Removal of the template was monitored by IR and elemental
analysis to ensure complete extraction of any residual organic
material. An analogous procedure was used to synthesize meso-
structured Rh- (RhCl
d) Ba(NO -, CsNO
Typically 0.5 g of M-Ta oxide (M ) Ru, Pt, Pd, Rh) was stirred
in a solution of the nitrate [Ba(NO , Alfa Aesar, 99.95%, metal
basis; CsNO , 99%; La(NO ‚xH O, 99.9%] and solvent (H O/
EtOH ) 50:50) for 4 h. The solid was collected and subsequently
dried in an oven at 120 °C overnight. The powders were then moved
to a U-shaped Pyrex reactor with a sintered glass frit as the fixed
bed and were further dried at 100 °C in vacuo overnight.
3
‚xH
2
O) or Pd-doped (PdCl
2
) Ta oxides.
(
3
)
2
3
-, and La(NO -Promoted Catalysts.
3 3
)
3 2
)
3
3
)
3
2
2
Figure 2. (A) EDS spectrum of 5% Ru-doped mesoporous Ta oxide. (B)
XPS spectra of the Ru 3p 1/2 region for (a) parent Ba-Ru/Ta oxide catalyst,
(
b) after catalytic run at 175 °C, 3 h, and (c) catalyst after H2 regeneration
at 350 °C, 2 h.
any reflections for Ru metal, indicating that the metal
particles were below the 6 nm threshold for XRD detection.
However, energy dispersive X-ray spectroscopy (EDS)
(e) Ba(OH)
2
- and Barium Isopropoxide-Promoted Catalysts.
Barium hydroxide (Ba(OH)
2
, tech. ∼95%) was impregnated into
template-free mesoporous M/Ta (M ) Ru, Pt, Pd, Rh) oxide in
methanol in an argon glove box (to avoid carbonate formation).
After the mixture was stirred for 4 h, the solvent was removed in
an air-free solvent storage flask on a Schlenk line at 100 °C. The
(Figure 2A) provided firm evidence for the presence of Ru
in the materials. Quantitative values were obtained by
comparison of the peak intensities to a sample of known Ru
concentration. A broad low-intensity emission at 485.2 eV
3 2 2
barium isopropoxide (Ba[OCH(CH ) ] , Alfa Aesar, 20% w/v in
isopropanol) analog was prepared by impregnation in isopropanol.
The solvent was removed, and the solid was dried as described
above for the hydroxide promoters.
in the X-ray photoelectron spectroscopy (XPS) (Figure 2B,
1
trace a) assigned as the Ru 3p /
2
emission establishes that
the Ru was slightly oxidized during the synthesis of the
parent catalyst. After hydrogen activation and catalysis (traces
b and c), the Ru is reduced to the metallic state (483.5 eV).
These materials also survive the Ba doping and subsequent
heat treatment stages of catalyst preparation. Figure 3 shows
the nitrogen adsorption-desorption isotherms of the meso-
porous Ta oxide starting material, the Ba-Ru/Ta catalyst
Results and Discussion
Characterization. Since heterogeneous catalysts for am-
monia synthesis typically exhibit complicated multicompo-
nent behavior in which the support, metal dopant, and
promoter all play an important role, we undertook a series
of compositional studies to establish important trends which
may shed more light on the reaction mechanism. In these
studies, XRD, nitrogen adsorption, and TEM (where ap-
propriate) were used to confirm the retention of the meso-
structure in the catalysts after doping. The TEM image in
Figure 1 clearly shows the disordered ∼20 Å wormhole pore
structure and the 50-100 nm size of the individual meso-
porous particles in the 5% Ru catalyst. The retention of the
mesostructure after Ru doping was also confirmed by the
retention of the reflection at d ) 32 Å in the XRD of all
samples studied. Extensive TEM studies at higher magnifica-
tion did not reveal individual Ru metal grains located within
the pores. This was expected because XRD did not show
2 2
before H reduction, and this same catalyst after H reduction.
BET surface areas are summarized in Table 1. Hydrogen
heat treatment at 350 °C for 4 h results in a decrease in
2
-1
specific BET surface area from 541 to 242 m g and a
corresponding decrease in the cumulative Horvath-Kawazoe
3
-1
(HK) volume from 0.282 to 0.092 cm g , with the pore
size remaining at 23.3 Å. The nitrogen adsorption-desorp-
tion isotherms show characteristics of both type I (for
microporous) and type IV (for mesoporous) because of a
broader pore-size distribution and the pore size on the cusp
between the microporous (d < 20 Å) and mesoporous (20
26
< d < 500 Å) regimes. This behavior is typical of the C12
-
templated mesoporous Nb, Ta, and Ti oxides studied in our
5086 Inorganic Chemistry, Vol. 46, No. 12, 2007

Ru, Pd, Pt, and Rh-Doped Mesoporous Tantalum Oxide Catalysts
Figure 4. H2-DTA of Ru/Ta oxide catalysts promoted with Ba(NO3)2:
a) Ba/Ru ) 1 and (b) Ba/Ru ) 3.
(
Figure 3. Nitrogen adsorption-desorption isotherms: (a) pure mesoporous
Ta oxide; (b) catalyst Ba-Ru/Ta (Ru ) 5 wt %, molar ratio Ba/Ru ) 1)
before H2 activation; and (c) catalyst from trace b after H2 activation at
ratio moves the exothermal peak toward a higher reduction
temperature with a stronger peak. No exothermal peak was
observed below 450 °C in the absence of Ru, which provides
evidence for the role of Ru in lowering the activation barrier
for nitrate reduction in this system.
3
50 °C for 4 h, followed by catalytic run at 175 °C.
Table 1. Surface Areas of Catalyst and after Subsequent Steps in the
Catalytic Process
BET surface
In previous studies, much of the initial high activity of
2
-1
no.
catalyst processing conditions
area (m g
)
the catalysts occurred only if this H
was not conducted and the entire process was performed in
the presence of both N and H . Thus, if the catalyst is first
reduced in H for 1-4 h and then a catalytic run is started,
2
-only activation step
1
2
mesoporous Ta oxide starting material
5% Ru3(CO)12-doped Ta oxide after
541
470
decomposition in vacuo at 300 °C, 3 h
2
2
3
4
5
6
7
parent catalyst Ba-Ru/Ta, Ru ) 5 wt %, Ba/Ru ) 1
parent catalyst after H2 reduction at 350 °C, 4 h
material from 4 after vacuum treatment at 300 °C, 2 h
material from 5 after catalytic run at 175 °C, 3 h
material from 6 after H2 regeneration at 350 °C, 2 h
370
242
237
229
220
2
the activities are generally 50-100 times lower. This
suggests that nitrate may be involved in the generation of
ammonia; however, studies have shown that nitrate reduction
proceeds through an N
pathways indistinguishable as long as N
To further investigate the possible role of nitrate during this
2
intermediate, making these two
is present in excess.
group. The HK pore-size distribution shows a major peak at
3.3 Å, demonstrating that the doping process has not
compromised the narrow pore-size distribution of the parent
mesostructure.
2
2
high-activity phase of the process, a labeling study was
15
conducted using N-labeled Ba(NO
3
)
2
, with unlabeled N
2
Effect of Reduction Temperature. Before catalytic runs
as a feedstock. Mass spectrometry (MS) analysis of the
ammonia produced over the first hour confirmed that 92%
with N
with H
2
and H
2
, the catalysts were first reduced and activated
2
at elevated temperature. This step ensures that all
15
is generated from the N-labeled nitrate source. If ammonia
nitrate present in the promoter decomposes to the oxide, the
active form of the promoter. Catalysts can be used without
this step; however, this leads to some ambiguity of the results
over the first hour, when nitrate decomposition may occur
parallel to ammonia formation (see below). Thus, the rate
of ammonia synthesis was measured over Ru/Ta oxide
2
production from nitrate went through an N intermediate as
14
previous work suggests, then more N ammonia would be
detected during this first hour because the amount of nitrogen
-1
passing through the system (1 mL s ) during this time
greatly exceeds that contained within the surface nitrate. After
nitrate is fully decomposed, the MS showed that all ammonia
formed originates from N , ruling out ammonia formation
2
from any residual nitrate during this second, lower activity
phase. Further kinetic studies demonstrated that after nitrate
reduction the catalyst only produces new ammonia when
catalysts promoted with Ba(NO
of the reduction temperature was studied. The Ru precursor
used for these studies was Ru (CO)12 because this precursor
3 2 3
) and CsNO , and the effect
3
proved to be the most effective in our previous studies and
has also been used successfully in mesoporous silica-based
exposed to both N
drops off significantly at H
functions at nearly the same rate at H
2
and H
2
; however, the catalyst activity
/N ratios of less than 1:1 but
/N ratios as high as
2
Haber systems. Figure 4 shows the H -TDA results of
2
2
Ba(NO -promoted Ru catalysts. There is a strong exother-
3 2
)
2
2
mic peak between 190 and 250 °C that can be attributed to
the reduction of promoter precursors and liberation of
gaseous nitrogen species from surface nitrate. The reduction
of nitrate in the presence of Ru has been shown to occur
between 200 and 250 °C.1 A higher promoter/Ru molar
5
:1. For example, at a 4:1 H
2
/N
2
ratio at 623 K and 2% Ru
-1
loading, the rate was 2.1 µmol h g, but at an H
2
/N
2
ratio
-1
of 1:4 for the same catalyst, the rate was 0.8 µmol h g.
For these studies, the flow rates were held constant as
compared to the total flow rate of the 1:1 control experiment.
2,27
(
26) Kresge, C. T.; Leonowicz, M. E.; Roth, W. J.; Vartuli, J. C.; Beck, J.
(27) Liang, C.; Wei, Z.; Luo, M.; Ying, P.; Xin, Q.; Li, C. Stud. Surf. Sci.
Catal. 2001, 138, 283.
S. Nature 1992, 359, 710.
Inorganic Chemistry, Vol. 46, No. 12, 2007 5087

Yue et al.
vaporize more easily than the latter metal oxide at high
temperature. Although catalysts with lower H
2
reduction
temperature at 300 °C have somewhat higher surface areas
2
-1
at the range of 240-260 m g , the lower activities are more
likely related to the lower decomposition percentage of the
promoter precursors.
Effect of Different Promoters. The effect of alkali and
alkaline earth metal promoters on Ru-based Haber systems
has been an area of much study. Electronegativities of these
promoters have been used as a rough criterion to evaluate
their mechanism of promotion, which often correlates well
with the rate of ammonia production; however, this is not
Figure 5. Activity of ammonia synthesis for 5 wt % Ru-Cs/Ta and
Ru-Ba/Ta oxides (promoter/Ru ) 1 mol/mol) as a function of the reduction
temperature. The rate was measured at 623 K.
9
,10,16
always the case.
Alkali metals are effective electron
donors; however, these metals are quite sensitive to oxygen-
containing compounds such as water and dioxygen and,
consequently, are converted into oxides or hydroxides.
Because of this concern, oxide promoters might be more
practical than the metal analogs although the promotion
The reaction thus depends more on H
on N concentration, which is quite different from standard
Ru-based catalysts, which have a negative order in H
because of blocking of the Ru sites by H binding. This
2
concentration than
2
2
1
7
2
effects are somewhat lower. Although the most widely
accepted mechanism for ammonia synthesis over Ru-based
catalysts is one in which electron density from the promoter
is first donated to the Ru metal grains and then transferred
to the dinitrogen antibonding orbital to weaken the robust
N-N triple bond, this is still a matter of contention because
the detailed role of the promoter has never fully elucidated.10
Some disagreements have arisen from conflicting experi-
mental observations; however these may be attributed to
some extent to the different support materials and promoters
used in the various catalytic systems in question.1
Previous studies by a number of groups have shown that
alkali metal Cs and alkali earth metal Ba promoters are the
two best choices for Ru-based Haber systems. The supports
employed in these studies include active carbon (AC), oxides
(MgO, Al O ), zeolite (X, Y), and MCM41 mesoporous
suggests that reduction of the active Ta sites is the rate
determining step, and this is consistent with the low activities
and the observation from bis(toluene) Ti reduced materials
2
that reduction of N on a reduced Ta oxide surface is
instantaneous. These and other kinetic effects will be the
subject of a more detailed study later, but they underscore
the unique mechanism of our system. The low activity at
higher temperatures where standard Ru-based Haber catalysts
operate further supports this mechanism where electron
density is transferred from the Ru grains to neighboring Ta
sites, effectively crippling the ability of the Ru grains to
3,14,17
2
mediate the cleavage of N and shifting this role to the
reduced Ta sites. This is not observed in Ru catalysts
supported on other metals, such as Mg or Si, because these
metals do not have variable oxidation states like Ta.
2
3
Figure 5 shows that the activity of Ru-doped mesoporous
Ta oxide catalysts is steady for reduction temperatures
between 250 (523 K) and 350 °C(623 K) but that it decreases
substantially for both promoters when the reduction tem-
perature is increased to 400 °C (673 K). While these rates
are clearly lower than those for standard Haber catalysts,
the much greater molar mass of the Ta oxide support rela-
tive to silica and magnesia,18 makes these values more
favorable. The BET surface area of Cs- or Ba-promoted
silica. One anomaly is that rare earth metal La promoters
are the most effective for Al O systems. This has fully been
2
3
explained by a SMSI (strong-metal support interaction)
effect, which is related to coverage by the promoter of the
precious metal surface when this surface is reduced to a lower
2
8
valency. SMSI was used to explain the strong suppression
of the ability of a surface to chemisorb H (and CO) and
2
has been observed for catalytic metals (e.g., group VIII noble
metals) on titania with H activation at ordinary temperature.
2
2
-1
catalysts remain about 200-220 m g after 4 h of hydrogen
reduction at 350 °C (Table 1). In contrast, the surface area
Other documented SMSI materials include oxides of nio-
bium, manganese, and lanthanum. For the sake of complete-
ness, all three of these promoters were selected and tested
in our Ru-doped mesoporous Ta oxide catalyst system. The
activities of these catalysts are compared in Table 2. These
results clearly show that Ba is the best promoter for our
Ru/Ta oxide system. La(NO ) -promoted catalysts did not
2
-1
drops to 120-150 m g if hydrogen treatment is performed
at 400 °C. While loss of surface area may explain the drop
in activity of the catalyst after heat treatment, it may also be
caused by sintering of Ru metal particles because of the
structural loss of the support and increased diffusion of Ru
particles at higher temperatures, although we could not
confirm this by XRD. Conversely, the loss of activity may
be related to a migration of promoter oxides from the Ru
metal surface or an evaporation of promoter oxides from the
3
3
show any activity in our system below 250 °C but are active
above this threshold. This could be attributed to the SMSI
effect of La, as observed in the alumina systems. The
rationalization of why the Ba promoter should work best in
a system that clearly may involve a different mechanism than
the standard Haber catalysts (insofar as the variable oxidation
17
catalyst, as has been observed previously. The decrease in
activity at higher temperature is more extensive for the
CsNO
3 3 2
-promoted catalyst than for the Ba(NO ) -promoted
analogue, possibly, because the former metal oxide tends to
(28) Tauster, S. T. Acc. Chem. Res. 1987, 20, 389.
5088 Inorganic Chemistry, Vol. 46, No. 12, 2007

Ru, Pd, Pt, and Rh-Doped Mesoporous Tantalum Oxide Catalysts
Table 2. Promoter Effect on the Rate of Ammonia Synthesis over 5 wt
%
Ru/Ta Oxide
activity
promoter
element
M)
M/Ru
BET
(µmol g-1) over 3 h
a
promoter (mole surface area
precursor ratio)
2
-1
(
(m g
)
298 K 448 K 623 K
none
Cs
Ba
319
213
220
208
0.22
1.44
1.78
0
0.28
1.72
4.35
0
0.55
6.37
11.56
8.19
CsNO3
Ba(NO3)2
La(NO3)3
1
1
1
La
a
BET surface areas were measured after catalysts were activated by H2
at 623 K for 4 h and a catalytic run at 623 K for 3 h.
states of the Ta oxide support may be involved) requires
some knowledge of the role of the support in the traditional
systems. The presence of a K promoter in Fe-based Haber
systems not only greatly enhances the reactivity of previously
inactive Fe surfaces (e.g., Fe(110), Fe(210)), but also lowers
Figure 6. Activity of ammonia synthesis on 5 wt % Ba-Ru/Ta as a
function of Ba/Ru mol ratio: (2) rate was measured at 448 K on the catalyst
reduced at 623 K and (9) rate was measured at 623 K on the catalyst reduced
at 623 K.
2
9
electron transfer occurs to oxygen rather than ruthenium.34
This is supported by the observation that the addition of
promoters to the catalytic systems in question generally does
not cause a significant change in the activation energies.
However, when Ba(NO ) is used as promoter for Ru/MgO
the energy of adsorption of ammonia on the Fe metal.
Therefore, as the concentration of NH on the surface
decreases, more active sites remain available for dissociative
adsorption of N . Studies on thermal desorption spectroscopy
of NH by Benndorf and Madey suggested a similar
3
2
3
3
2
30
mechanism for catalysis on Na-doped Ru (001). Aika and
co-workers have proposed that rate enhancement by alkali
or alkali earth metal promoters in Ru-based systems result
from electronic promotion. Electron density is donated from
the promoter to the Ru, lowering its ionization potential,
which allows the electrons at the Ru surface to be readily
catalyst, a large decrease in apparent activation energy for
1
0
ammonia production is observed. This unusual change is
apparently not caused entirely by the effect of Ba because
this drop in activation energy does not occur in the Ba-Ru/
zeolite X system; the apparent activation energy of Ba-Ru/
zeolite X catalyst is actually similar to the values for alkali-
exchanged zeolite X. Therefore, the combined interaction
of promoter, the support, and the Ru metal is possibly
responsible for the change in activation energy. Also,
according to Aika’s mechanism, Ba should never be superior
to Cs as an electronic promoter. However, Ba functions as
a better promoter than Cs in almost all studied Ru catalyst
donated into the N
activation energy for dissociative adsorption of N
2
antibonding orbital, thus reducing the
3
1
2
. When
this observation is related to our system, it stands to reason
that the effectiveness of electron transfer from Ru to Ta
should rely in part on the potential at the surface of the Ru
particles. Thus, the decrease of the binding energy of
electrons at the Ru surface should lead to a more effective
transfer of electrons to the Ta centers in the generation of
low-valent sites, as long as a suitable electron pathway exists
between the Ru and the Ta. This would mean that promoters
which decrease the binding energy of Ru surface electrons
would be more effective in the generation of reduced Ta
sites than those promoters which increase the binding energy
of surface electrons and would lead to an increase in the
3
5
systems. With many competing factors potentially at play
and evidence for involvement of low-valent Ta in the process,
it is not yet possible to delineate the exact reason why Ba
functions as the best promoter in our system. The simplest
explanation is that it facilitates electron transfer between the
Ru and the Ta by modifying the Fermi level of the Ru, but
this neglects possible direct interactions between Ba and Ta,
Ba and O, or other factors mentioned above. Clearly more
experimental work needs to be done on both the mechanism
and structure of the active site before more definite conclu-
sions can be drawn.
reaction rate if reduced Ta is indeed involved, which all E
a
18
data from our system imply. This is also consistent with a
high dependency of the rate on H concentration because
increased surface coverage of Ru by H would lead to a more
2
2
Effect of Promoter/Ru ratio. Since Ba(NO
to be the best promoter in our system, the amount of
Ba(NO was systematically varied, and the activity was
measured. The results are shown in Figure 6. In the case of
wt % Ru/Ta oxide catalyst, a Ba/Ru ratio of 1:1 gave the
3 2
) was found
highly reducing surface better able to transfer electrons to
the Ta.
3 2
)
The role of the Ba in our system may not be this simple;
however, various other groups have conducted studies that
suggest that electron density is not transferred from alkali
5
maximum activity at lower temperature, while 3:1 proved
to be the best at higher temperature. The surface of activated
32
promoters to the Ru metal and the electrostatic interaction
is limited to the vicinity of the alkali3 or that efficient
3
2
mesoporous Ta oxide, with the BET surface area of 242 m
-
1
-1
2+
g , can accommodate 7.13 mmol g of Ba ions or 7.07
(29) Somorjai, Gabor A. Introduction to Surface Chemistry and Catalysis;
Wiley-Interscience: New York, 1994.
30) Benndorf, C.; Madey, T. E. Chem. Phys. Lett. 1983, 101, 59.
31) Aika, K.; Ohya, A.; Ozaki, A.; Inoue, Y.; Yasmuri, I. J. Catal. 1985,
(
(
(33) Lang, N. D.; Holloway, S.; Nørskov, J. K. Surf. Sci. 1985, 150, 24.
(34) Over, H.; Bludau, H.; Skottke-Klein, M.; Moritz, W.; Ertl, G. Phys.
ReV. B 1992, 46, 360.
(35) Wachs, I. E.; Chen, Y.; Jehng, J.-M.; Briand, L. E.; Tanaka, T. Catal.
Today 2003, 78, 13.
9
2, 305.
32) Uner, D. O.; Pruski, M.; Gerstein, B. C.; King, T. S. J. Catal. 1994,
14, 530.
(
1
Inorganic Chemistry, Vol. 46, No. 12, 2007 5089

Yue et al.
mmol g- of Ru when Ba ions or Ru are closely packed
1
2+
2
5
2+
on the surface. The surface quantities of Ba and Ru with
promoter/Ru ) 1 and 5.0 wt % Ru loading are about 0.4
-
1
and 0.4 mmol g , respectively, and cover only about 5.6
and 5.7 % of the surface of Ta oxide, respectively. According
to Aika’s mechanism, in the case of an acidic support like
Ta oxide, a small amount of an alkaline promoter will be
consumed to neutralize the support and so a higher promoter/
Ru is preferred. Jacobsen and co-workers, by using atomic-
resolution in situ TEM, observed that Ba atoms distributed
5
as single atoms close to the crystal edges (B sites) and bound
Figure 7. Activity of ammonia synthesis at 448 (2) and 623 K (9) as a
function of Ru loading.
to oxygen are responsible for the electronic promotion of
the catalyst.36 Since very few such active sites are present
in the catalyst, the majority of Ba atoms from the relatively
Effect of Absolute Ru Loading. Ru/Ta oxides prepared
37-39
large, optimal Ba/Ru ratio (0.3-2.0)
reservoirs to maintain a constant coverage of mobile Ba
atoms in the vicinity of all B -type sites in the catalyst.
are used as
from Ru (CO) using different Ru loading were prepared
3
12
and doped with Ba(NO ) at a Ba/Ru ratio of 1. The rates of
3
2
5
ammonia synthesis from these catalysts at 175 (448 K) and
350 °C (623 K) are shown in Figure 7. The rate increases
linearly with Ru loading at 350 °C, but it increases only
slightly before leveling off at 175 °C. This pattern of greater
dependency on catalyst loading levels at higher temperature
is consistent with the trends relating activity and the Ba/Ru
ratio shown in Figure 6. Aika and co-workers have observed
a linear relationship up to 5 wt % of Ru loading over their
While Ru is the active agent for the standard Haber process
at high temperature, with all bond breaking and forming
processes occurring on the metal surface, Arrhenius plots
-
1
on our system provide an E
a
of 9.3 kJ mol , which is a
fraction of the minimum 58-96 kJ/mol required to break
22,23
an N-N triple bond on a pristine Ru surface.
XPS data
also provide evidence for low-valent Ta on the surface during
reaction conditions, and our previous work shows that
reduced mesoporous Ta oxide cleave N instantaneously at
2
room temperature. The activity of the catalyst at elevated
temperature shows a strong dependence on the Ru/Ba ratio
and a maximum activity at 3:1, similar to traditional Haber
catalysts. However, the plot at lower temperature shows far
less of a dramatic dependence on this ratio and a maximum
at 1:1. Two mechanisms may be at work in varying
proportions depending on temperature. At higher temperature,
1
4
Ru/MgO systems, which is not unexpected since with the
entire catalytic reactions take place on the Ru clusters in
these systems. However, in our Ru/Ta oxide catalysts, the
Ru could also function as an interface to funnel electrons
2
from H to the Ta reducible sites at lower temperature. The
fact that activity does not increase with Ru loading at 175
°C is in accord with a new mechanism, at least at this
temperature, in which the number of reduced Ta active sites
might be more important than Ru and the rate of reaction
depends more strongly on electron transfer between the Ru
and the Ta. This is in accord with the low activation energies
2
the majority of N cleavage occurs at the Ru surface, while
at lower temperature, where traditional Haber catalysts are
generally inactive, the second mechanism involving reduced
Ta predominates. However, for this to be true, an inflection
at higher T in the Arrhenius plot reflecting the greater
-
1
(9.3 kJ mol ) measured previously, which suggest a small
number of highly active sites. Since six electrons are required
2
per N , several Ta centers must be involved. This can occur
activation barrier of N
2
cleavage at Ru would be observed,
in several steps at Ta sites spread through the mesostructure
or in one step at an active site involving several low-valent
Ta sites. In the latter case, the binding of N might be
2
compared to binding to the FeS core in nitrogenase. Once
the electrons are transferred to the Ta, the reaction proceeds
rapidly, consistent with the almost instantaneous cleavage
but this was not the case because the plot remained linear.
Another explanation for the different dependency of activity
on Ba/Ru ratio observed at different temperatures is that,
2 5
because of the much higher molecular weight of Ta O than
other supports, only 4.5 mmol Ta atoms present in 1 g of
Ta oxide. Excessive promoter could thus block the active
Ta sites, and for this reason, a somewhat lower ratio (Ba/Ru
2 3
of N and subsequent formation of NH in bis(toluene) Ti
reduced mesoporous Ta oxide. Thus, the limiting factor in
this mechanism would be the transfer of electrons to the Ta
and the low activities coupled with low activation barrier
can be explained by the presence of only limited effective
electron-transfer pathways in the material. Increasing the
temperature does not substantially improve electron transport
across the amorphous surface, but it allows the second
mechanism to become more important. This hypothesis also
suggests that improving electron transport across the surface
should lead to more reduced Ta sites and higher activities
and is supported by the fact that Ba is the best promoter and
is known to influence the energy of the electrons at the Ru
surface and thus their facility for mobility across the surface.
)
1) is preferable at lower (448 K) temperature. Catalysts
with an even higher ratio of 5:1 were also tested at 350 °C
and relatively low activity was found. In this case, XPS
studies showed that only about 85% Ba(NO
posed to BaO under our activation conditions.
3 2
) was decom-
(
(
(
(
36) Hansen, T. W.; Wagner, J. B.; Hansen, P. L.; Dahl, S.; Topsøe, H.;
Jacobsen, C. J. H. Science 2001, 294, 1508.
37) Kowalczyk, Z.; Jodzis, S.; Rar o´ g, W.; Zielinski, J.; Pielaszek, J.; Presz,
A. Appl. Catal. A. 1999, 184, 95.
38) Formi, L.; Molinari, D.; Rossetti, I.; Pernicone, N. Appl. Catal. A.
1
999, 185, 269.
39) Bielawa, H.; Hinrichsen, O.; Birkner, A.; Muhler, M. Angew. Chem.
001, 113, 1093.
2
5090 Inorganic Chemistry, Vol. 46, No. 12, 2007

Ru, Pd, Pt, and Rh-Doped Mesoporous Tantalum Oxide Catalysts
Table 3. Effect of Ru and Ba Precursor Compound
Table 4. Reaction Data for Pd/Ta Catalystsa
activity
activity
Ru precursor
compound
Ba/Ru
(mole surface area
ratio)
BET
BET
(µmol g h-1
)
-
1
(µmol g-1) over 3 h
a
Ba precursor
compound
promoter
catalyst precursor
synthesis
gas
surface area
2
2
-1
(5 wt %)
(m /g)
448 K 623 K
(m g
)
298 K 448 K 623 K
Ru3(CO)12
Ba(NO3)2
Ba(OH)2
Ba[OCH(CH3)2]2
1
1
1
1
1
1
219
252
232
225
249
241
1.38
0.42
0
0.34
0.16
0
5.52
0.75
0.13
1.87
0.51
0.08
10% Pd Ba(OH)2
10% Pd Ba(OH)2
10% Pd Ba(NO3)2 He
10% Pd Ba(NO3)2 1:1 N2 + H2
10% Rh Ba(NO3)2 1:1 N2 + H2
10% Pt Ba(NO3)2 1:1 N2 + H2
He
1:1 N2 + H2
279
243
271
222
238
229
0.00
0.302 0.73
0.00
0.67
0.02
0.01
0.00
0.00
1.61
0.00
2.21
0.06
0.18
0.00
1.10
0.04
0.08
RuCl3.xH2O Ba(NO3)2
Ba(OH)2
Ba[OCH(CH3)2]2
a
a
BET surface areas were measured after catalysts were activated by H2
Reaction data and BET surface areas were measured after catalysts
were activated by H2 at 623 K for 4 h and then heated under vacuum at
623 K for 3 h. All materials prepared using a Pd/Ba ratio of 1:1.
at 623 K for 4 (barium nitrate precursor) or 2 h (other barium precursors),
followed by a catalytic run at 623 K for 3 h.
RuCl
3
2
‚3H O could also be another poison and thus decrease
Effect of Ba and Ru precursor. Promoter precursor
12
the catalyst activity when used as a source of Ru.
effects of Ba(OH)
compared with that of Ba(NO )
2
and Ba(isopropoxide) were examined and
. The activities were
Pd, Pt, and Rh Catalysts. To more fully explore the
effects of composition on the catalytic behavior of this
system, we prepared a series of Pd, Pt, and Rh analogues to
the catalysts studied above. These metals are of special
3 2
measured in each case after hydrogen reduction was complete
and all ammonia formed during this initial period (arising
from nitrate or residual N
results of this study are shown in Table 3. All the Ba salts
are effective, with Ba(NO giving the best results at both
75 (448 K) and 350 °C (623 K). RuCl ‚3H O was also used
as an alternative of Ru precursor. These catalysts were
prepared by impregnating RuCl ‚3H O with mesoporous Ta
oxide in methanol and were further promoted by Ba(OH)
2
in the system) ruled out. The
interest because the very high activation energies of N
cleavage on the surface precludes the possibility of direct
cleavage by the precious metal under the conditions in
this study. For example, the E for N dissociation on a Pd
surface has been calculated as 2.6 eV (250 kJ mol ). This
compares to the somewhat lower E value for this process
on a Ru surface (0.6-1.0 eV or 58-96 kJ/mol), which is
still too high to allow direct N cleavage in our Ru-based
2
3 2
)
N
2
1
3
2
a
2
-1 23
3
2
a
2
and Ba(isopropoxide). The activities of thus prepared cata-
lysts were compared with that of the standard catalyst, and
the results are also shown in Table 3. The trend follows
2
-
1
system, given that the activation barrier is 9.3 kJ mol .
Skeptics may still argue that at high temperature we cannot
rule out some small amount of ammonia being formed on
Ru through this traditional mechanism because of a Boltzman
distribution of molecular energies in the system, especially
since some of our data on Ru and Ba loading could suggest
a second mechanism may be operative at higher temperature.
So, to add further indirect verification of the involvement
exactly that using Ru
explanation for the observation that Ba(OH)
promoter than Ba(NO may be the result of the lower
basicity of BaO originating from the hydroxide rather than
Ba(NO . However, the fact that the hydroxide works at all
3
(CO)12 as the precursor. A possible
2
is a poorer
3 2
)
3 2
)
is further evidence that the ammonia formed in this system
cannot arise from residual nitrate. The reason that Ba
isopropoxide did not exert a promotion effect may be because
it did not decompose completely to BaO under our operating
of N
2
cleavage on Ta in our system, we exchanged Ru for
dissociation is energetic-
Pd, Pt, and Rh metals on which N
2
ally prohibitive and rarely, if ever, observed. Thus meso-
porous Ta oxide catalysts doped with metallic Pd, Pt, or Rh
were synthesized by solution impregnation of the metal
chloride followed by H reduction at elevated temperature
2
and characterized by XRD, nitrogen adsorption, TEM, EDS,
and XPS. Table 4 shows data for the 10% Pd catalysts using
conditions. Table 3 shows that RuCl
a precursor as Ru (CO)12. This is not surprising because of
results from standard Haber systems, and it has been related
3 2
‚3H O is not as effective
3
3+
to three major effects. Ru exists as Ru in the parent catalyst,
which has to be reduced before nitrate promoters decompose.
3 2 2
Ba(NO ) is not reduced with H below 450 °C in the absence
of Ru. On the other hand, because of the nature of the
precursor and the method of decomposition used, there is
2 3 2
either Ba(OH) or Ba(NO ) as a promoter at 350 (623 K)
and 175 °C (448 K). The catalyst activities for ammonia
production are roughly one-third those for the Ru analogues
and drop off further over several hours as shown in the table.
Blanks are also shown for the system using He as a feedstock
to rule out residual ammonia formed from some other
contaminant source. As with the Ru system, the catalyst also
already some Ru metal existing in parent Ru
before H treatment, so that Ru-catalyzed nitrate decomposi-
tion in this case should be more facile. Another major reason
that RuCl ‚3H O is a less effective precursor is that chloride
3
(CO)12 catalyst
2
3
2
ions are known to poison the catalyst. Aika and co-workers
found that much higher activities can be obtained after
removal of chloride at very higher temperatures, for example
does not function with either pure N
cleavage occurred on the Pd surface, on the basis of the E
for this process (250 kJ/mol, almost five times greater than
that for Ru), the rate of N
cleavage would be 1015 times
slower for a Pd/Ta catalyst. However, it is actually about
0% of the activity of the Ru catalyst. This high rate of
production relative to that predicted for any process involving
dissociation of N on Pd, firmly establishes that N cleavage
2 2 2
or pure H . If N
a
4
0
1
000 °C. Magnesia was also found to be one of the best
2
oxide supports for ammonia synthesis, especially when
14
3
Ru (CO)12 was used as precursor. The high activities might
3
be contributed from high Ru dispersion, the chloride-free
system, and the basicity of MgO. Finally, water vapor from
2
2
(40) Murata, S.; Aika, K. Appl. Catal. 1992, 82, 1.
must occur on the low-valent Ta sites. Figure 8 shows the
Inorganic Chemistry, Vol. 46, No. 12, 2007 5091

Yue et al.
metals generally provided ammonia yields over 3 h of less
-
1
-1
than 0.5 µmol h g . Yet the fact that any measurable
amount of ammonia is being formed using these metals is
further evidence for N
for cleavage on these metals is also prohibitive for direct
cleavage under the conditions of experiment.
While the activities in this new family of ammonia
2
cleavage on the Ta sites because the
E
a
N
2
synthesis catalysts are low at this early stage, improvement
of electron transport across the surface may lead to enhanced
rates. These observations aside, the new mechanism proposed
in this work supports much recent work in organometallic
2
chemistry revolving around N cleavage at early metal centers
and may lead to the development of multicomponent
organometallic systems employing early and late transition
metal centers. These studies also show that mesoporous Ta
oxides may be important in other areas of catalysis because
of their unique combination of controlled pore size, high
surface area, and variable oxidation states. Future studies
are ongoing to further clarify the mechanism and improve
the activities.
Conclusion
A study was conducted on the effect of composition on
activity for a series of metal-doped mesoporous Ta oxide
catalysts for conversion of N and H to ammonia. Labeling
2 2
studies established that the majority of ammonia formed in
the first hour originated from nitrate, but that once nitrate
was consumed, new ammonia from N
For the mesoporous Ta oxide catalytic system, Ru
and Ba(NO were found to be the most effective precursors
of Ru catalysts. La(NO was a poorer promoter at mild
conditions but a better promoter at higher temperature than
CsNO possibly because of the SMSI effect. Studies on Ru/
2 2
and H was formed.
3
(CO)12
Figure 8. XPS emissions for (A) N 1s, (B) Pd 3d, and (C) Ta 4f regions
in mesoporous Ta oxide doped with 10% Pd at a 1:1 Ba(OH)2 ratio after
heat treatment at 623 K for 4 h in pure H2, 4 h vacuum at 623 K, and 3 h
catalytic run with 1:1 H2 and N2.
3 2
)
3 3
)
3
XPS of the (a) N 1s, (b) Pd 3d, and (c) Ta 4f regions of the
catalyst after 1 h under operating conditions at 350 °C. The
N 1s region is obscured by the Ta 4p3/2 peak; however, there
is a small shoulder at 398 eV corresponding to surface bound
Ba ratio and absolute Ru loading show different behaviors
at 175 (448 K) and 350 °C (623K), suggesting that two
mechanisms may be at play, although further studies are
required to confirm this. Catalysts were also prepared with
Pd, Pt, and Rh in place of Ru. The Pt and Rh catalysts
showed poor activity, but the Pd catalyst was almost as active
2
0
ammonia. The Pd 3d region shows emissions consistent
with metallic Pd with small shoulders at higher binding
energies corresponding to a fraction of a second, possibly
oxidized species. The Ta 4f region shows a broad 5/2
emission with a maximum at 26.4 eV tailing off below 25.0
eV. For comparison, pure mesoporous Ta (V) oxide shows
this emission at 26.9 eV, and the Ru-doped catalyst shows
this emission at 26.2 eV at this stage of catalyst preparation.
The binding energy of this peak, the broadness of this
emission, and the tailing off to lower binding energies are
evidence of reduction to oxidation states lower than Ta(IV)
and confirm that low-valent Ta is indeed involved in the
process. The catalysts activities using Pt and Rh as precious
as the Ru catalyst. Since N dissociation on Pd is prohibitive,
2
this lends support to a mechanism involving cleavage at low-
valent Ta. Further kinetic investigations of this new catalytic
system are ongoing with an effort to better understand the
controlling factors and thus be able to develop strategies to
improve activity and longevity.
Acknowledgment. The authors would like to acknowl-
edge the Natural Sciences and Engineering Research Council
(NSERC) AGENO program for funding.
IC062385D
5092 Inorganic Chemistry, Vol. 46, No. 12, 2007