Professional Documents
Culture Documents
Zhao 2004
Zhao 2004
nanostructure and functionalization methods on cata- temperature of 35 °C on the helium flow, the carbon
lyst properties are discussed. A comparison between monoxide pulses are injected into the quartz reactor and
these new catalysts and conventional catalysts such as the net volume of CO is recorded by TCD. The chemi-
Pd/R-Al2O3 and Pd/γ-Al2O3 is presented. sorption stoichiometry of one CO molecule per surface
palladium atom is assumed.
Experimental Section Catalyst Activity Test for DMO Synthesis. Cata-
lytic studies are performed in a quartz reactor where
Catalytic Synthesis of Carbon Nanofibers. Car- the catalyst sample (1.80 g) is packed at a fixed bed.
bon nanofibers are synthesized on 20 wt % Ni/γ-Al2O3 The gas reactant, composed of 10% (v/v) methyl nitrite
and 20 wt % NiFe (1:1)/γ-Al2O3. A deposition-precipita- and 90% nitrogen, is generated through a reaction
tion method is used to prepare finely dispersed metal between sodium nitrite and methanol as shown in eq
catalysts. Details of the preparation procedures were 5. As the mixture of sodium nitrite and methanol is
described by Geus21 and Hoogenraad.22 After injection
of the diluted solution containing the active component 2NaNO2 + 2CH3OH + H2SO4 f
precursor, the subsequent precipitate is filtered, washed, 2CH3ONO + Na2SO4 + 2H2O (5)
dried overnight at 120 °C, and then calcined in station-
ary air for 5 h at 600 °C. vigorously stirred in a vessel, concentrated H2SO4 is
Typically 1.0 g of catalyst is loaded in a quartz reactor slowly dropped in to carry out the reaction, and a CH3-
mounted in a horizontal tubular furnace. The sample ONO-containing effluent is collected. The ratio of both
is initially reduced in a 25% hydrogen-argon stream gases (CH3ONO and N2) is controlled by the partial
at 600 °C for 3 h. Following this step, a desired mixture pressure of CH3ONO, allowing a constant composition
of carbon monoxide/hydrogen (4:1) is led into the reactor of desired reactant. The catalyst sample is reduced in a
and the reaction is allowed to proceed for a period of 16 hydrogen and nitrogen mixture (25% H2) at 350 °C for
h. All the gas flow rates are measured by mass flow several hours, and the temperature is then lowered to
controllers. Gas effluents are analyzed by GC. Each run 120 °C for synthesis of DMO. The gas reactant consist-
produces more than 10 g of carbon nanofibers. ing of methyl nitrite (mixed with nitrogen) and carbon
Oxidation of carbon nanofibers has been conducted monoxide is led into the reactor at a CH3ONO:CO:N2
to introduce surface functional groups. To study the molar ratio of 1:1.25:9. The flow rate of carbon monoxide
effects of functionaliztion of the catalyst properties, both remains at 15 mL/min, and the reaction lasts for 120
air and HNO3 have been used as oxidants. Part of the min. The concentration of methyl nitrite is determined
as-grown carbon nanofibers can be subsequently oxi- by GC (HP5890) equipped with a TDX-1-packed column.
dized overnight in the air at 400 °C. The rest of the as- After reaction, the collected solid is dissolved in metha-
grown carbon nanofibers are subsequently soaked in a nol; the solution is also analyzed using a GC instrument
4 M nitride acid aqueous solution for 6 days, then (HP5890) equipped with an OV-101 column to deter-
filtered, thoroughly washed in deionized water, and mine the DMO purity.
dried overnight at 120 °C.
Preparation of Carbon Nanofiber Supported Pd
Results and Discussion
Catalysts. Supported catalysts containing 1 wt %
palladium are prepared by an incipient wetness im- Characterizations of Carbon Nanofibers. From
pregnation method with an aqueous solution containing the TEM images (Figure 1a,c) of the as-grown carbon
PdCl2 and HCl. The wet solid is dried at 80 °C overnight nanofibers, metal particles and alumina support can
and then calcined in stationary air at a temperature of also be observed, while the metal particles are removed
350 °C for 5 h. The resulting dry solid is reduced under by treating the carbon nanofibers in a HNO3 solution
flowing pure hydrogen at 350 °C for 5 h. The samples as shown in Figure 1b,d. By examining these images,
are cooled in the presence of argon to room temperature one may appreciate that the diameter distribution of
and are then stored in a sealed vessel. For Pd/γ-Al2O3 the carbon nanofibers is relatively narrow and that the
and Pd/R-Al2O3 catalyst preparation, similar procedures diameter of the formed carbon nanofibers is almost less
are followed except for carrying out the calcination and than 50 nm. This is possibly due to a fine dispersion of
the reduction for 16 h.23 metal particles on the support, since it is well-known
Catalyst Characterization. Surface area and tex- that the deposition-precipitation method can yield
tual properties of the supports are measured by nitrogen small particles with a narrow distribution. However,
adsorption at 77.4 K with a Micromeritics ASAP 2010 when the metal powder is used as the catalyst, the
apparatus. The samples are degassed for 5 h at 190 °C, diameter of the produced carbon nanofiber is larger and
prior to the measurements. the distribution is wider. This phenomenon is in good
The microstructure and the morphology of various agreement with the observation in the literature.24
supports and supported palladium catalysts are studied During further HRTEM studies some very interesting
using a JEOL JEM 2010 transmission electron micro- features have been observed concerning the influence
scope with an accelerating voltage of 200 kV and a of the catalyst composition on the nanostructure of
lattice resolution of 0.18 nm. Suitable transmission carbon nanofibers generated from the CO/H2 (4:1)
specimens are prepared by ultrasonic dispersion of the mixture at 600 °C. In Figure 2a, the orientation of the
respective samples in ethanol, followed by dropping the graphite platelets is aligned in a direction parallel to
suspension onto a holey carbon-coated copper grid. the fiber axis and some defects are also observed,
Metal dispersions are measured using a CO pulse defined as CNF-00 (the angle between graphite platelet
chemisorption procedure in a Micromeritics AutoChem e and the axis of the carbon nanofiber is 0). These
(II) apparatus. The weighted catalysts are directly defects appear as a bamboo-like structure, indicating
reduced in the mixture of 10% H2/Ar at a temperature some breakage of the tubular carbon nanofiber growth.
of 150 °C for 30 min and then degassed by helium for This phenomenon can possibly be linked to reconstruc-
20 min at 150 °C. After the catalysts are cooled to the tion of the metal particles during the catalytic growth
Ind. Eng. Chem. Res., Vol. 43, No. 16, 2004 4597
Figure 1. (a, top left) TEM image of the fresh carbon nanofibers produced by disproportionation of CO on 20% Ni/Al2O3 (CNF-00) at the
conditions CO:H2 ) 4:1, Ptot ) 1 bar, and T ) 600 °C. (b, top right) TEM image of the carbon nanofibers after treatment in HNO3 solution
produced by disproportionation of CO on 20% Ni/Al2O3 (CNF-00-HNO3) at the conditions CO:H2 ) 4:1, Ptot ) 1 bar, and T ) 600 °C. (c,
bottom left) TEM image of the fresh carbon nanofibers produced by disproportionation of CO on 20% NiFe (1:1)/Al2O3 (CNF-05) at the
conditions CO:H2 ) 4:1, Ptot ) 1 bar, and T ) 600 °C. (d, bottom right) TEM image of the carbon nanofibers after treatment in HNO3
solution produced by disproportionation of CO on 20% NiFe (1:1)/Al2O3 (CNF-05-HNO3) at the conditions CO:H2 ) 4:1, Ptot ) 1 bar, and
T ) 600 °C.
process. It has been revealed that carbon nanofiber concentration, the crystal will instantly release carbon
formation involves surface reactions leading to surface onto the back of the Ni particle.
carbon formation on the gas-metal interface. The However, the defected nanostructure of carbon nanofi-
formed carbon segregates, diffuses, and precipitates on bers produced on the NiFe (1:1)/Al2O3 catalyst is not
the back of the metallic particles.25 The carbon concen- clearly observed in a typical image (Figure 2b). The
tration in the particles is determined by a balance graphite platelets are not completely parallel to the axis
between the surface reaction and diffusion of carbon of carbon nanofibers, and there is a small angle apart
through the particle. When the surface reaction rate is from the axis, about 5-10° (CNF-05). Anyhow, the angle
faster than the diffusion rate, the carbon site coverage is so small that the structure is very close to that of a
and the carbon concentration as well will increase with carbon nanotube. The difference in nanostructure of
time. As a consequence, the Ni lattice is expanded and different catalysts can possibly be explained by the
the Ni-Ni bond length increases when a relatively high crystallographic orientation on the active metal particle
amount of carbon is inserted into the lattice, since the faces.
nickel atom has a covalent radius of 1.15 Å compared The measured surface area and pore volumes of
to carbon having a covalent radius of 0.77 Å. It implies different samples are shown in Table 1. The surface
that accumulation of carbon atoms in the Ni crystal areas of carbon nanofibers are high and close to that of
destabilizes the lattice, and creates some driving force γ-Al2O3. The measured surface area of R-Al2O3 is very
for reconstruction of Ni crystals. At a certain carbon low as expected. It is found that the surface area of the
4598 Ind. Eng. Chem. Res., Vol. 43, No. 16, 2004
Figure 2. (a, left) HRTEM image of carbon nanofibers produced by disproportionation of CO on 20% Ni/Al2O3 (CNF-00) at the conditions
CO:H2 ) 4:1, Ptot ) 1 bar, and T ) 600 °C. (b, right) HRTEM image of carbon nanofibers produced by disproportionation of CO on 20%
NiFe (1:1)/Al2O3 (CNF-05) at the conditions CO:H2 ) 4:1, Ptot ) 1 bar, and T ) 600 °C.
carbon nanofiber produced on 20 wt % NiFe/Al2O3 is and the specific features of the surface of the carbon
larger than that on 20% Ni/Al2O3. Formation of the nanofibers.26,27 The surface ratio of basal and edge
larger surface area can be explained by the fact that planes can be deliberately controlled by selecting the
there must be more graphite edge exposed to the surface growth condition; therefore, carbon nanofiber based
of CNF-05 compared to that of CNF-00. Although catalysts can be prepared intentionally to achieve some
oxidation in air yields slightly higher surface area and desired properties. Particularly, steps or terraces on the
larger pore volume than in HNO3, no significant effects surface of a small-diameter carbon nanofiber might
on surface area using different oxidants have been benefit anchoring of the active component precursor,
observed. The difference is only about 8 m2/g. and thus help to disperse the palladium particles very
Characterization of Supported Palladium Cata- finely on the support.
lysts. The size and distribution of palladium particles On the other hand, the surface of as-grown carbon
on different supports have been investigated by HR- nanofibers is hydrophobic due to the existence of basal
TEM. Only palladium supported on carbon nanofibers planes of graphite platelets. This extremely valuable
with functionalization in air are selected for TEM advantage for a common inert substrate consequently
studies, and the results are also compared to those for has a weak point in that the active component precursor
palladium supported on R-alumina (see Figure 3a-c). is difficult to anchor on the surface. Functionalization
The TEM images (Figure 3) show that palladium of the surface to introduce oxygen-containing groups by
particles on R-alumina (Figure 3a) are larger than 10 the oxidant is an essential step in application of carbon
nm, while the palladium particles on carbon nanofiber nanofibers. Oxygen-containing groups are considered as
supports (Figure 3b,c) present a much smaller size: anchoring sites for favoring a higher dispersion of the
most palladium particles are smaller than 5 nm, and metal. Mojet and Hoogerraad28 studied the preparation
the particle size distribution is relatively narrow. Simi- of carbon nanofiber supported palladium catalysts using
lar particle sizes are observed on CNF-00 (Figure 3b) an ion-exchange technique and found that the palladium
and on CNF-05 (Figure 3c), probably due to their similar ions interacted directly with the basal plane. Moreover,
carbon nanostructures as discussed above. Furthermore, Simonov29 has done extensive research on H2PdCl4
metal dispersions and the average diameter of pal- produced in aqueous solution adsorbed on the surface
ladium particles measured by the CO pulse chemisorp- of a graphite-like carbon material and also found that
tion method are in agreement with the examined TEM there was weak interaction between the basal plane and
results as shown in Table 1. The high dispersion of the the active phase, but no direct interaction of the Pd(II)
metal particles could be attributed to high surface area complex and oxygen-containing groups. According to
Ind. Eng. Chem. Res., Vol. 43, No. 16, 2004 4599
Figure 3. (a, top left) TEM images of R-alumina supported palladium particles: 1 wt % Pd/R-Al2O3. (b, top right) TEM images of CNF-
00-HNO3 supported palladium particles: 1 wt % Pd/CNF-00-air. (c, bottom) TEM images of CNF-05-HNO3 supported palladium
particles: 1 wt % Pd/CNF-05-air.
this point, the functionlization of the carbon nanofiber It was previously reported in the literature that the
surface makes CNFs more hydrophilic, and they thus conversion of methyl nitrite on Pd/γ-Al2O3 was lower
benefit from impregnation and stabilization of the active than on Pd/R-Al2O3.4 Uchumi4 concluded that R-Al2O3
phase.30 and activated carbon were more preferable supports
Catalyst Activity Test. According to thermodynamic than γ-Al2O3 for the gas-phase synthesis of DMO,
analysis, the reaction shown as eq 2 can be regarded as although the function of the support had not been well
an irreversible one.7 The reaction of a CO and CH3ONO understood. A redox reaction mechanism has been
(1.25:1) mixture is performed on the various supported generally accepted to explain the reaction. As mentioned
palladium catalysts at 120 °C. Analysis of the CH3ONO above, the crucial step in this catalytic process is the
concentration is done once every 30 min. In most cases, redox step between Pd(0) and Pd(II) to form a catalytic
the reaction reaches a steady state after 10 min. The cycle. The needs for reoxidation or reduction depend on
conversion of CH3ONO is found to be strongly depend- the reaction nature. Experimental results show that Pd-
ent upon the nature of the support material as shown (0) catalyzes the formation of DMO (eq 2) (forming
in Figure 4. Inspection of these results indicates that reaction) while Pd(II) catalyzes the formation of DMC
Pd/CNF-05 after oxidation treatment in the air offers (eq 4).31 The key step in the redox mechanism can
the highest activity, resulting in a CH3ONO conversion simply be expressed as
at 85%. The activities of palladium catalysts on different
CH3ONO
supports can be arranged as follows: Pd(0) y\ z Pd(II) (6)
CO
and CNF-05 has been synthesized on NiFe (1:1)/γ-Al2O3. (15) Chambers, A.; Nemes, T.; Rodriguez, M. N.; Baker, R. T.
Palladium particles are homogeneously deposited on the K. Catalytic Behaviour of Graphite Nanofiber Supported Nickel
outer edges of the carbon nanofibers by the standard Particles. 1. Comparison with Other Support Media. J. Phys.
Chem. B 1998, 102, 2251.
incipient wetness impregnation method, and also on (16) Park, C.; Baker, R. T. K. Catalytic Behaviour of Graphite
γ-alumina and R-alumina supports. The diameter of the Nanofiber Supported Nickel Particles. 2. The Influence of the
palladium particles of the carbon nanofiber supported Nanofiber Structure. J. Phys. Chem. B 1998, 102, 5168.
catalyst is mainly less than 5 nm, which is much smaller (17) Chambers, A.; Baker, R. T. K. Catalytic Behaviour of
than those on R-alumina. Finely dispersed palladium Graphite Nanofiber Supported Nickel Particles. 3. The Effect of
particles on carbon nanofibers are explained by the high Chemical Blocking on the Performance of the System. J. Phys.
surface area and oxygen-containing groups on the Chem. B 1999, 103, 2454.
surface. (18) Gao, R.; Tan, C. D.; Baker, R. T. K. Ethylene Hydroformy-
lation on Graphite Nanofiber Supported Rhodium Catalysts. Catal.
Carbon nanofibers as supports result in a significant Today 2001, 65, 19.
improvement in the gas-phase synthesis of dimethyl (19) Ros, T. G. Rhodium Complexes and Particles on Carbon
oxalate compared to the alumina supported catalyst. Nanofibres. Ph.D. Dissertation, Utrecht University, Utrecht, The
Experimental results show that palladium supported on Netherlands, 2002.
CNF-05 funtionalized by air oxidation offers the highest (20) Zhang, Y.; Zhang, H.; Lin, G.; Chen, P.; Yuan, Y.; Tsai, K.
activity for DMO synthesis. A small metal particle size R. Preparation, Characterization and Catalyic Hydroformylation
and relatively low metal-support interaction can en- properties of Carbon Nanotubes-supported Rh-phosphine Catalyst.
Appl. Catal., A 1999, 187, 213.
hance the redox properties. Enhanced carbon monoxide
(21) Boellaard, E.; van der Karaan, A. M.; Geus, J. W. Prepara-
adsorption and possible Fe promotion effects are factors tion of Supported Mono- and Bimetallic Catalysts by Deposition-
that have been proposed to explain the high activity and precipitation of Metal Cyanide Complexes. Stud. Surf. Sci. Catal.
selectivity over carbon nanofiber based catalysts. 1995, 91, 931.
(22) Hoogenraad, M. S.; Onwezen, M. F.; van Dillen, A. J.; Geus,
Acknowledgment J. W. Stud. Surf. Sci. Catal. 1994, 101, 1331.
(23) Geus, J. W.; van Veen, J. A. R. Preparation of Supported
This work was done under the support of the National Catalysts. Stud. Surf. Sci. Catal. 1993, 79, 459.
Science Foundation of China and the Development (24) Park, C.; Rodriguez, N. M.; Baker, R. T. K. Carbon
Project of Shanghai Priority Academic Discipline. Deposition on Iron-Nickel during Interaction with Carbon Mon-
oxide-Hydrogen Mixtures. J. Catal. 1997, 169, 212.
(25) Hoogenraad, M. S. Growth and Ultilization of Carbon
Literature Cited Fibrils. Ph.D. Dissertation, Utrecht University, Utrecht, The
(1) Nakamura, A.; Matsuzaki, T. A New Oxidation System Netherlands, 1995.
Using Nitrite Oxidatants. Res. Chem. Intermed. 1998, 26, 213. (26) Toebes, M. L.; van Dillen, J. A.; de Jong, K. P. Synthesis
(2) Wang, J.; Ma, X.; Han, S.; Xu, G. Preparation of Diethyl of Supported Palladium Catalysts. J. Mol. Catal. A 2001, 173, 75.
Oxalate by CO Coupling Reaction in a Gas-solid Fixed Bed (27) Likholobov, V. A. Catalysis by Novel Carbon-based Materi-
Reactor. J. Nat. Gas Chem. 2000, 9, 319. als. In Catalysis by Unique Metal Ion Structures in Solid Matrics;
(3) Jiang, X.; Su, Y.; Lee, B.; Chien, S. A Study on the Synthesis Centi, G., et al., Eds.; Kluwer Academic Publishers: Dordrecht,
of Diethyl Oxalate over Pd/R-Al2O3 Catalysts. Appl. Catal., A 2001, The Netherlands, 2001.
211, 47. (28) Mojet, B. L.; Hoogenraad, M. S.; van Dillen, J. A.; Geus,
(4) Uchiumi, S.; Ataka, K.; Matsuzaki, T. Oxidative Reactions J. W.; Koningsberger, D. C. J. Chem. Soc., Faraday Trans. 1997,
by a Palladium-Alkyl Nitrite System. J. Organomet. Chem. 1999, 93 (24), 4371.
576, 279. (29) Simonov, P. A.; Romanenko, A. V.; Prosvirn, I. P.; Moroz,
(5) Matsuzaki, T.; Nakamura, A. Dimethyl Carbonate Synthesis E. M.; Bornin, A. I.; Chuvilin, A. L.; Likholobov, V. A. On the
and Other Oxidative Reaction Using Alkyl Nitrites. Catal. Surv. Nature of the Interaction of H2PdCl4 with the Surface of Graphite-
Jpn. 1997, 1, 77. like Carbon Materials. Carbon 1997, 35, 73.
(6) Huo, Z.; Li, Z.; Wang, B.; Ma, X.; Xu, G. Influence of Support (30) Schoegl, R. Carbons. In Preparation of Solid Catalysts;
Properties on CO Vapor-Phase Catalytic Coupling Reactions. Ertl, G., Knoezinger, H., Weitkamp, J., Eds.; Wiley-VCH: Wein-
Chem. React. Eng. Technol. 2002, 18, 31 (in Chinese). heim, Germany, 1999.
(7) Ma, X.; Xu, G.; Chen, J.; Chen, H. Kinetics Carbon Monoxide
(31) Pawlow, J. H.; Sadow, A. D.; Sen, A. Palladium(II)-
Catalytic Coupling to Ethyl Nitrite in Gaseous Phase. J. Chem.
Catalyzed Terpolymerization of Alkane-R,ω-Dinitrite Esters, Alk-
Ind. Eng. 1995, 46, 50 (in Chinese).
enes, and Carbon Monoxide to Polysuccinates. Organometallics
(8) Iijima, S. Helical Microtubules of Graphitic Carbon. Nature
1997, 16, 5659.
1991, 354, 56.
(9) Rodriguez, N. M. A Review of Catalytically Grown Carbon (32) Mann, D. J.; Halls, M. D.; Hase, W. L. Direct Dynamics
Nanofibers. J. Mater. Res. 1993, 8, 3233. Studies of CO-Assisted Carbon Nanotubes Growth. J. Phys. Chem.
(10) De Jong, K. P.; Geus, J. W. Carbon Nanofibers: Catalytic B 2002, 106, 12418.
Synthesis and Applications. Catal. Rev.sSci. Eng. 2000, 42, 481. (33) You, Q.; Xu, W.; Luo, Y. Studies of the Kinetics of
(11) Rodriguez, N. M.; Chambers, A.; Baker, R. T. K. Catalytic Synthesizing Diethyl Oxalate by Vapor Phase Catalytic Coupling
Engineering of Carbon Nanostructures. Langmuir 1995, 11, 3862. Carbon Monoxide and Ethyl Nitrite. J. Zhejiang Univ. [Nat. Sci.
(12) Hoogenraad, M. S.; van Leeuwarden, R. A. G. M.; van Ed.] 1991, 25, 512 (in Chinese).
Breda Vriesman, G. J. B.; Broersma, A.; van Dillen, A. J.; Geus, (34) Meng, F.; Xu, G.; Guo, Q. Kinetics of the Catalytic Coupling
J. W. Stud. Surf. Sci. Catal. 1995, 91, 263. Reaction of Carbon Monoxide to Diethyl Oxalate over Pd-Fe/R-
(13) Ledoux, M. J.; Vieira, R.; Pham-Huu, C.; Keller, N. New Al2O3 Catalyst. J. Mol. Catal. A 2003, 201, 283.
Catalytic Phenomena on Nanostructured (Fibers and Tubes)
Catalysts. J. Catal. 2003, 216, 333. Received for review September 17, 2003
(14) Planeix, J. M.; Coustel, N.; Coq, B.; Brotons, V.; Kumbhar, Revised manuscript received February 18, 2004
P. S.; Dutartre, R.; Geneste, P.; Bernier, P.; Ajayan, P. M. Accepted February 26, 2004
Application of Carbon Nanotubes as Supports in Heterogeneous
Catalysis. J. Am. Chem. Soc. 1994, 116, 7935. IE030728Z