In Situ Synthesis of Multiwalled Carbon Nanotubes over LaNiO3 as

Jan 12, 2012 - ... as Support of Cobalt Nanoclusters Catalyst for Catalytic Applications ... made available by participants in Crossref's Cited-by Lin...
7 downloads 0 Views 4MB Size
Article pubs.acs.org/JPCC

In Situ Synthesis of Multiwalled Carbon Nanotubes over LaNiO3 as Support of Cobalt Nanoclusters Catalyst for Catalytic Applications Dongyan Xu,*,† Peng Lu,† Ping Dai,‡ Haizhen Wang,† and Shengfu Ji§ †

College of Chemical Engineering, Qingdao University of Science and Technology, Qingdao 266042, People’s Republic of China College of Electromechanical Engineering, Qingdao University of Science and Technology, Qingdao 266061, People’s Republic of China § State Key Laboratory of Chemical Resource Engineering, Beijing University of Chemical Technology, Beijing 100029, People’s Republic of China ‡

ABSTRACT: In situ synthesis of multiwalled carbon nanotubes (MWCNTs) was performed by chemical vapor deposition of methane on a LaNiO3 perovskite-type growth-promoter prepared by the sol−gel method. The as-produced MWCNTs were purified via a two-step procedure consisting of an oxidation in air and subsequent HNO3 treatment. Nearly monodisperse cobalt nanoparticles with an average size of 11.3 nm have been successfully decorated on MWCNTs by an impregnation−reduction method. The properties of the as-produced carbon nanotubes (CNTs) and Co/CNT catalysts were analyzed by X-ray diffraction (XRD), field-emission scanning electron microscopy (FESEM), transmission electron microscopy (TEM), X-ray photoelectron spectroscopy (XPS), and thermogravimetric analysis (TGA). It was found that the reduction of LaNiO3 and formation of CNTs occur almost simultaneously under methane atmosphere and that the in situ controlled synthesis of MWCNTs with relatively high purity and uniform diameters could be realized by the chemical vapor deposition method in a tubular reactor. The catalytic activity of the Co/CNT catalyst was tested for the hydrolysis reaction of sodium borohydride (NaBH4) in basic medium to produce hydrogen. The prepared Co/CNT catalyst calcined at 300 °C shows high catalytic activity for hydrogen generation from hydrolysis of NaBH4.

1. INTRODUCTION Carbon nanotubes (CNTs) present a unique atomic structure, very high aspect ratio, and extraordinary mechanical properties, which make them suitable for applications in the fields of heterogeneous catalysis.1−3 CNT-based catalysts often show a higher activity and/or selectivity than the catalysts made of the same active phase but deposited on other carbon materials, for example, activated carbon, carbon nanofibers, and carbon black.4−8 The reasons for this unusual behavior are their particular electronic properties, their high thermal conductivity, the high accessibility of the active phase, and the absence of any microporosity, thus eliminating diffusion and intraparticle mass transfer limitations. Compared to single-walled carbon nanotubes (SWNTs), multiwalled carbon nanotubes (MWCNTs) have attracted much attention for their potential applications in the fields of catalysis because they are much easier to produce in large quantities at a reasonable price.9−11 The most promising CNT growth technique, for both bulk production and localized growth on a surface, is catalytic chemical vapor deposition (CCVD). With this method, a variety of transition metals such as nickel, cobalt, molybdenum, and iron, namely, the growth-promoter, plays a crucial role in the control of size and quality of the prepared CNTs.12−14 The overall CCVD growth process can be viewed as a two-step process: (1) forming catalytically active nanoparticles during a © 2012 American Chemical Society

pretreatment step and (2) introduction of carbon feedstock and CNTs growth. It is widely recognized that the formation of discrete metal nanoparticles is vital for the deterministic control of CNTs growth and that in many cases the carbon nanotube diameter is largely governed by the size of catalytically active particles.15−18 Nonuniform particles lead mainly to a mixture of MWCNTs with different inner and outer diameters. Therefore, controlling the dispersion and size of the active particles is undoubtedly essential for the growth of desired MWCNTs. For that purpose, perovskite−type metal oxides with a well-defined structure such as LaNiO 3 , 19,20 La 2 Ni 2 O 5 , 21 and LaFexMoyMnzO322 were used as a growth-promoter precursor for the preparation of CNTs by CCVD. Prereduction of these perovskite precursors leads to the formation of uniform metallic particles deposited on lanthanum oxide and consequently enables tailored CNTs in terms of their diameter and number of walls. However, the controlled CNTs growth through a onestep procedure, which is performed by direct CCVD on the growth-promoter precursor without experiencing a separate prereduction procedure (step 1), has not been fully investigated. Received: November 15, 2011 Revised: January 8, 2012 Published: January 12, 2012 3405

dx.doi.org/10.1021/jp211009g | J. Phys. Chem. C 2012, 116, 3405−3413

The Journal of Physical Chemistry C

Article

temperature under a nitrogen flow. The purification of raw CNTs was conducted in two steps. First, the crude CNTs product was calcined in air at 400 °C for 1 h to remove the amorphous carbon.29 Then, the sample was further mixed with nitric acid (HNO3, 65 wt %) and refluxed for 2 h at 120 °C to dissolve the residual growth-promoter.30 This procedure can also effectively functionalize the CNTs and introduce oxygenated functional groups onto the surface of the nanotubes, allowing for deposition of finely dispersed metal nanoparticles.8,31 Co/CNTs catalyst was prepared by impregnation of the purified and functionalized CNTs carrier with an aqueous solution of cobalt nitrate. The mass of all the precursors was calculated in order to obtain 10 wt % Co/CNT catalysts. The wet solid was dried overnight at 110 °C, and an excess amount of NaBH4 solution was dropped into the catalyst precursor to completely reduce Co2+.32 The sample was then dried and calcined under nitrogen atmosphere at 300, 400, and 500 °C, respectively. 2.3. Characterization. X-ray powder diffraction (XRD) patterns of the samples were recorded in a D/MAX-2500/PC X-ray diffractometer using Cu Kα radiation, operated at 40 kV and 30 mA. The XPS experiments were carried out with an ESCALAB 250 spectrometer. The spectra were recorded using monochromatic Al Kα radiation (hv = 1486.6 eV) as the excitation source. Photoelectrons were selected in energy with a hemispheric electron analyzer. Low-resolution survey spectra were recorded in the range 0−1200 eV to determine the elements present in the sample. High-resolution spectra were recorded for Co 2p and B 1s levels to determine the chemical state of these elements. The binding energy (BE) values were referenced to the C 1s peak of carbon at 284.6 eV. The CNTs and Co/CNT samples were characterized by transmission electron microscopy (TEM, JEOL JEM-2000EX) and high resolution transmission electron microscopy (HRTEM, JEM2010) with a point resolution of 0.194 nm. Characterization of the surface morphology of the as-prepared CNTs was examined using a JEOL JSM 5200 scanning electron microscope. TG/ DTA experiments were performed in a thermogravimetric analyzer (NETZSCH STA 409PC). About 6.6 mg of the asprepared LaNiO3 sample was loaded into a 5 mL alumina crucible. The system was heated from room temperature to 850 °C at a rate of 10 °C min−1 under 20 v% CH4−N2 (60 mL min−1). 2.4. Hydrolysis of Sodium Borohydride. NaBH 4 (Shanghai Aibi Chemistry Preparation Co., Ltd., 98 wt %) was used for the reaction. The kinetic studies with catalysts were carried out in batch operation and a 50 mL three-necked round-bottom flask was used as the reactor, which was kept at the preset temperature by using thermostatic circulator with a temperature variation within ±0.1 °C. One of the flask openings was equipped with a thermometer inserted into the solution to monitor the temperature. In a typical H2 generation experiment, 10 mL of 5 wt % NaBH4−0.5 wt % NaOH was initially put into the flask. The onset of the hydrogen generation reaction started from the moment when a fixed amount, ca. 200 mg, of Co/CNT catalyst was added to the NaBH4 solution. A wet gas meter was adopted to measure the cumulative volume of the generated hydrogen over time.

Proton exchange membrane fuel cells (PEMFCs) that utilize hydrogen as fuel are being developed as clean power generation devices for various uses including portable power, residential uses, and transportation applications. Chemical hydrides such as NaBH4 have attracted worldwide interest as a source to supply pure hydrogen to PEMFCs at room temperature, because it is stable in alkaline solution, nonflammable, nontoxic in nature, and with a hydrogen storage capability of 10.8 wt %.23 A critical area of activity in hydrogen production from NaBH4 involves the development of suitable catalysts. Cobalt catalysts supported on a variety of carriers, such as Ni foam,24 Al2O3,25 nanostructured poly(p-xylylene) films,26 activated carbon,27 and IR-120 resin beads,28 were found to be very effective to catalyze the hydrolysis of NaBH4. However, little attention has been paid to the deposition of cobalt nanoclusters on CNTs and its performance for catalyzing hydrolysis of NaBH4. In this study, we present an in situ synthesis of MWCNTs by direct CCVD on the growth-promoter precursor (one-step process). The one-step synthesis route is more convenient to operate and is expected to facilitate the formation and maintenance of uniform metallic particles due to the avoidance of possible agglomeration of these particles during the prereduction procedure (step 1). Subsequently, CNTs supported cobalt nanoclusters catalyst was prepared by an impregnation−chemical reduction method. Hydrogen generation from hydrolysis of NaBH4 was selected as a probe reaction to investigate the catalytic performance of the asprepared Co/CNT catalyst.

2. EXPERIMENTAL SECTION 2.1. Preparation of LaNiO3. LaNiO3 perovskite precursor was prepared by a citric acid sol−gel combustion method. Typically, a stoichiometric amount of lanthanum nitrite (La(NO3)3·6H2O, 99.0 wt %, Shanpu, Shanghai) and nickel nitrite (Ni(NO3)2·6H2O, 99.0 wt %, Aibi, Shanghai) was dissolved in distilled water to form a homogeneous solution. An aqueous solution of citric acid (C6H8O7·H2O, 99.5 wt %), as a complexing agent, was added to the mixed solutions of nitrates in a molar ratio of 1 for citric acid to total ions (Ni2+ and La3+) and then heated at 110 °C for dehydration. During the dehydration process, a polycondensation reaction happened between citric acid and nitrates, leading to the formation of a transparent gel. The gel was subsequently transferred to an oven at 300 °C. After a few minutes, the gel burned in a selfpropagating combustion manner until the gel was completely burnt out to form a loose ash. Finally, the ash was calcined in air at 700 °C for 2 h and used as the promoter precursor to grow CNTs. 2.2. Synthesis of CNTs and Preparation of Co/CNT. CNTs were synthesized by the CCVD method using LaNiO3 as a growth-promoter precursor and methane as a carbon source in a tubular reactor. Unlike the more widely used two-step CCVD procedure consisting of a catalyst prereduction step and subsequent CNTs growth step in the literature, our one-step procedure is more convenient to operate. The catalyst precursor (0.15 g) was placed in a quartz boat and inserted into a horizontal tubular reactor. The reactor temperature was raised from room temperature to 700 °C at a rate of 10 °C min−1 in a nitrogen gas flow (50 mL min−1). Once the temperature reached 700 °C, CH4 (99.99 wt %) gas was introduced into the reactor at a rate of 25 mL min−1 and held for 1.5 h. After the reaction, the oven was cooled down to room

3. RESULTS AND DISCUSSION 3.1. Characterizations of LaNiO3 and CNTs. It was reported that the phase of LaNiO3 could form when the 3406

dx.doi.org/10.1021/jp211009g | J. Phys. Chem. C 2012, 116, 3405−3413

The Journal of Physical Chemistry C

Article

samples. The average Ni crystallite size calculated with Scherrer’s equation is about 10.3 nm. The representative morphologies of the as-synthesized and purified CNTs are shown in Figure 3. The raw CNTs, with lengths and diameters in the range of tens of micrometers and nanometers, respectively, can be observed in Figure 3a,c at different magnifications. It is evident that the CNTs are randomly entangled. The produced CNTs are very numerous and contain an abundance of impurities (bright areas as marked with arrows) including growth-promoter particles (metal and/ or metal oxide) and/or nontubular carbonaceous impurities. These impurities are mainly attached on the bundles of CNTs making the surface of these CNTs unsmooth. Oxidation of raw CNTs has been widely used to remove various impurities.35,36 In comparison, SEM images of the purified CNTs show clearly that the purification consisting of air oxidation and liquid-phase oxidation via acid treatment led to an efficient elimination of the impurities (Figure 3b,d). In should be mentioned that, to use CNTs as heterogeneous catalyst supports, the entire surface of CNTs also needs to be oxidized for functionalization and provision of enough anchor sites so that highly dispersed catalysts could be achieved.9,10 TEM image of the as-prepared CNTs is presented in Figure 4a. It can be seen that the raw CNTs connect with a large number of growth-promoter particles, which was confirmed to be La2O3 according to XRD measurements (Figure 2a). Figure 4b shows distinctly that the attached growth-promoter particles and other possible impurities disappeared on the purified CNTs. Both closed and open ends of the CNTs can be found. Additionally, some Ni metal nanoparticles confirmed by XRD measurement locate at the closed ends of the CNTs. Both TEM and XRD measurements demonstrate that Ni particles play catalytic action in the growth of CNTs, and La2O3 just acts in the role of dispersing the active Ni particles. After controlled synthesis of CNTs, La2O3 can be easily eliminated by acid treatment. The open end is probably induced by Ni releasing from the carbon tube end in certain cases, for example, mechanical disturbing37 and acid purification.38 Figure 4c shows the characteristic tubular structure of CNT with a hollow core and multiple walls. The typical inner and outer diameters of CNTs are 12.5 and 30 nm, respectively. An outer diameter distribution histogram, measured by counting 128 CNTs on the TEM images, is presented in Figure 4d. The solid line of histogram corresponds to the Gaussian fit of the diameter distribution. It is easily noted that the purified CNTs display a narrow outer diameter distribution with values centered at 20− 30 nm. It is expected that the inner diameters of CNT are determined by the particle sizes of nickel originated from the reduction of LaNiO3 under methane atmosphere. In fact, the inner diameter of CNTs obtained from TEM images is nearly consistent with the size of Ni particles detected with XRD measurements. The HRTEM image confirms the presence of capped ends (Figure 5a). From a tubular segment (Figure 5b), it can be noticed that the graphene layers are almost parallel to the tube axial direction and that the average interlayer spacing in the walls is about 0.34 nm, which agrees well with the (002) plane lattice parameter of CNTs. Figure 5c depicts a close view of a ∼30 nm nickel nanoparticle located at the tip of a multiwalled CNT and encapsulated within the nanotube, which suggests a growth by the well-known tip mechanism.39 Figure 6 shows the TG-DTG curves of LaNiO3 growthpromoter under methane flow. Two peaks are present on the

precursor ash containing La2O2CO3 and NiO was calcined at 600 °C.33 As shown in Figure 1, the diffraction peaks at 2θ

Figure 1. XRD pattern of LaNiO3 powder.

values of 23.1, 32.8, 40.5, 47.2, 53.4, 58.4, 68.8, 78.6° are characteristic of the LaNiO3 rhombohedral phase (JCPDS Card Files, No. 33-0711), indicating a perovskite structure obtained after calcination of the precursor ash at 700 °C. The XRD profile of the as-synthesized CNTs is shown in Figure 2a. Compared to Figure 1, it can be found that the

Figure 2. XRD patterns of (a) crude and (b) purified CNTs.

characteristic diffraction peaks of LaNiO3 disappeared after the CCVD reaction. Additionally, the peaks ascribed to La2O3 (JCPDS Card Files, No. 54-0213) and cubic Ni (JCPDS Card Files, No. 04-0850) appeared on the XRD profile, indicating that the structure of perovskite was destroyed during the process of CCVD owing to the reduction of LaNiO3. It is known that the XRD pattern of the CNTs is similar to that of graphite because the hexagonal ring structure of graphene sheets remains unchanged in the carbon nanotubes. Thus, the peaks at 25.8 and 42.8° correspond to the (002) and (100) planes of the CNTs. The shift of (002) reflection from 2θ value of 26.4° for graphite to 25.8° for the CNT also reveals an increase in the interlayer distance from 0.335 nm for graphite to 0.344 nm for the carbon nanotubes.34 As shown in Figure 2b, only the peaks ascribed to CNTs and Ni0 remained, whereas the peaks indexed to La2O3 disappeared in the purified CNT 3407

dx.doi.org/10.1021/jp211009g | J. Phys. Chem. C 2012, 116, 3405−3413

The Journal of Physical Chemistry C

Article

Figure 3. SEM images of (a) crude and (b) purified CNTs. Panels (c) and (d) are high-magnification images of (a) and (b), respectively.

methane and the formation of CNTs. These results imply that the formation of CNTs occurred immediately once the active Ni0 site was formed. Thermogravimetric analysis can provide information regarding the carbonaceous impurity content, residual catalyst content, and the crystalline defect density of the nanotubes because of the differences in the reactivity of the various forms of carbon causing preferential burn off at different temperatures and the defects in nanotube walls leading to a lowertemperature oxidation and gasification of the carbon.40,41 Thermogravimetric analysis was performed to study the oxidation behavior of CNTs with a heating rate of 15 °C min−1 up to 800 °C in 50% O2−N2 (Figure 7). An initial weight loss from ambient temperature to 200 °C for both the crude and purified samples is less than 0.5 wt %, which is ascribed to the removal of physisorbed water. A second weight loss from 200 to 480 °C can be attributed to the burning of amorphous carbon on the outer walls of CNTs and some surface structure defects.42 No obvious difference of weight loss between the crude and purified CNTs in this temperature range was found, suggesting that there was almost no amorphous carbon formed accompanying with the CNTs growth. The third weight loss corresponds to the complete oxidation of CNTs. The main weight loss starts at 480 °C, and the maximum loss appears at 620 °C for the crude sample, while these temperatures for the purified sample are 500 and 640 °C, respectively. It is worth mentioning that no weight gain appears during the initial stage of the thermal treatment, arising from the oxidation of exposed metal particles. This implies that the nickel particles are exclusively encapsulated within the CNTs, which was also confirmed by the HRTEM measurement. The residues of the crude and purified samples are found to be 20 and 9.1 wt % of the original samples, respectively. XRD results show that the

Figure 4. TEM images of (a) crude, (b, c) purified CNTs, and (d) histograms of outer diameter distributions.

profile of DTG. Corresponding to the DTG peaks, a weight loss (inside Figure 6) and a marked weight increase can be observed. The weight loss at 630 °C is attributed to the reduction of LaNiO3 completed by the reaction between LaNiO3 and CH4. When the temperature increases to 640 °C, a substantial weight increase resulted from the cracking of 3408

dx.doi.org/10.1021/jp211009g | J. Phys. Chem. C 2012, 116, 3405−3413

The Journal of Physical Chemistry C

Article

Figure 5. HRTEM images of (a) capped end, (b) segment of a CNT, and (c) a Ni nanoparticle encapsulated within a CNT.

Figure 6. TG-DTG curves of LaNiO3 catalyst at methane ambience. Figure 8. XRD patterns of Co/CNT uncalcined and calcined at different temperatures. (a) Uncalcined, (b) 300 °C, (c) 400 °C, and (d) 500 °C.

The X-ray photoelectron spectroscopy (XPS) technique is a valuable characterization tool based on the electron binding energy which provides information about chemical composition of the surface. The survey scan of XPS of the Co/CNT catalyst calcined at 300 °C is given in Figure 9a. The observed peak at 284.6 eV in Figure 9a represents sp2 carbon−carbon bond in the CNTs. In order to ascertain the oxidation state of the cobalt species on the sample surface, a curve-fitting procedure of the Co 2p photoelectron peaks was performed (Figure 9b). The Co 2p core level spectrum was characterized by two components due to spin−orbital splitting Co 2p3/2 and Co 2p1/2, along with two shakeup satellites. The Co 2p3/2 and Co 2p1/2 peaks of cobalt species, located at binding energies of 781.1 and 796.5 eV, respectively, are the characteristic of surface Co2+ species.46,47 These peaks may be attributed to the presence of CoO because it has a strong shakeup satellite peak 5 eV higher than its main peak and has a spin−orbit coupling of around 15.5 eV.48 The absence of metallic cobalt in the XPS analysis indicates clearly that the surface of the particles is fully oxidized and the oxide layer is thicker than ca. 3 nm, the penetration depth of the photoelectron analysis. The exposure of cobalt nanoparticles to ambient air during the XPS sample preparation may cause the oxidation of the cobalt metal since the cobalt(0) nanoclusters are sensitive to aerobic atmosphere.49The quantitative analysis from the XPS measurement reveals that the cobalt content is 8.7 wt % which is lower than the theoretical value of 10 wt %. For bulk cobalt boride synthesized by reduction of cobalt salt with NaBH4 in a solution, the peaks corresponding to both elemental Co 2p and B 1s levels can be detected.43,50 However,

Figure 7. TG curves of (a) crude and (b) purified CNTs.

residue of the crude sample includes LaNiO3 and NiO, while that of the purified sample is only NiO (not shown). This suggests that the purified CNT contains about 7.1 wt % Ni which is encapsulated by the CNTs and cannot be removed by mild acid treatment adopted in this study. 3.2. Co/CNT Characterization. Figure 8 presents the XRD patterns of the Co/CNT catalysts uncalcined and calcined at different temperatures. For both uncalcined and calcined samples, the characteristic diffraction peaks of CNTs at 2θ = 26.0 and 43.0° were observed. As to the catalyst samples calcined below 400 °C, no peaks ascribed to any cobalt species were found due to the amorphous structure.43−45 When the sample was calcined at 500 °C in flowing N2 for 2 h, the weak diffraction peaks of nanocrystalline Co phase (JCPDS card 150806) became detectable. In addition to metallic Co, CoO phase (JCPDS card 48-1719) was also found in the catalyst calcined at 500 °C (Figure 8d), which is most possibly induced by the oxidation of Co during catalyst storage. 3409

dx.doi.org/10.1021/jp211009g | J. Phys. Chem. C 2012, 116, 3405−3413

The Journal of Physical Chemistry C

Article

Figure 9. XPS spectra of (a) survey scan and (b) Co 2p levels for Co/CNT catalyst.

the particles located on the outer surface of the tubes is about 11.3 nm. It was reported that the metal nanoparticles are mainly deposited outside CNTs when their inner diameter is below 20−30 nm.53 Therefore, the Co nanoparticles were mainly decorated on the outer surface of the CNTs if no special filling procedures were taken during the impregnation process even though the CNTs were partially opened by purification. The high dispersion of Co metals is ascribed to the functional groups of CNTs derived from HNO3 treatment. It is suggested that deprotonation of the functional groups occurs whenever the HNO3-treated CNTs are in contact with water, thus leaving a negatively charged carbon surface, which improves significantly the interaction between the CNT and metal cations of the precursors used during the subsequent impregnation.11 The better dispersed Co nanoparticles are expected to provide active sites with larger surface areas, enhancing Co utilization during the hydrolysis. 3.3. Catalytic Performance. To understand the effect of calcination on the catalytic activity, kinetic studies on hydrolysis of NaBH4 were carried out with the catalysts uncalcined and calcined at various temperatures. Figure 12 presents the hydrogen volume as a function of time at the reaction temperature of 30 °C. It is obvious that the calcination treatment exerts remarkable influence on the catalytic performance of Co/CNT. Compared to uncalcined Co/CNT, the catalyst calcined at 300 °C displays rather high activity. However, a further increase of the calcination temperature from 300 to 500 °C caused an evident decrease of hydrogen generation rate. Thus, an appropriate calcination temperature benefits the catalytic activity, but a higher calcination temperature will induce partial crystallization of Co and decrease its catalytic activity. Numerous studies have demonstrated that the amorphous Co-based catalysts exhibit high activity for the hydrolysis of NaBH4 even though it is not clear whether these catalysts are CoxB, Co−B, Co0, or Co oxide because powder XRD is not efficient in identifying an amorphous or poorly crystallized compound.54 It was suggested that the increase in calcination temperature may enhance the crystallization degree and increased the grain size of cobalt and thus resulting in a lowered catalytic activity of the catalyst.55 Lee et al. also drew a similar conclusion that calcination at temperatures above 300 °C led to decomposition and sintering of the Co−B catalyst, reducing the catalytic activity to the hydrogen generation reaction.44 Hydrogen generation kinetics was measured at a solution temperature ranging from 303 to 318 K by using a Co/CNT catalyst and 10 mL of 5 wt % NaBH4−0.5 wt % NaOH (Figure

no peaks belonging to boron were found on the CNT supported cobalt catalyst (Figure 9a). Some similar results have been reported.28,49,51 Our previous studies have also found that the activated carbon supported cobalt and/or nickel catalysts prepared with the same method as this study contain a very small amount of boron by inductively coupled plasma (ICP) analysis.52 Therefore, it is more prone to obtain metal instead of metal−metalloid alloy nanoparticles on the surface of supporter materials with the chemical reduction method using borohydride. It should be noted that the surface of the catalyst sample during XPS measurement is different from that of the in situ formed one during the reaction. However, the XPS measurement can help to confirm the presence of cobalt metal rather than CoxB or Co−B on the surface of CNTs. Microstructure and its selected area electron diffraction (SAED) pattern of the Co/CNT characterized by TEM observation are shown in Figure 10. The SAED pattern

Figure 10. TEM images of Co/CNT catalysts. (a) Uncalcined and (b) calcined at 500 °C.

shown in Figure 10a is characteristic of a carbon nanotube with a hexagonal graphite crystalline structure. The rings in the pattern correspond to (002), (004), and (006) planes of CNT. As to the sample calcined at 500 °C in N2, the SAED pattern (Figure 10b, inset) exhibits bright distinct dots, implying the highly crystalline nature of Co nanoparticles, which is consistent with the XRD pattern plotted in Figure 8d. The HRTEM images of Co/CNT sample calcined at 300 °C (Figure 11a,b at different magnifications) show that the Co particles disperse well on the outer walls of CNTs. A bar graph depicts the size distribution of the nanoparticles outside the tubes (Figure 11c). This graph shows that the average size of 3410

dx.doi.org/10.1021/jp211009g | J. Phys. Chem. C 2012, 116, 3405−3413

The Journal of Physical Chemistry C

Article

Figure 11. (a, b) HRTEM images of Co/CNT catalyst at different magnifications. (c) Cobalt oxide particle size distribution.

where k0 is the rate constant (mL min−1 g−1), Ea is the activation energy (kJ mol−1), R is the gas constant (8.3143 kJ mol−1 K−1), and T is the reaction temperature (K). An Arrhenius plot, in which ln k is plotted against the reciprocal of absolute temperature (1/T), was plotted in Figure 14. From the

Figure 12. Plots of the volume of hydrogen versus time for the hydrolysis of NaBH4 catalyzed by Co/CNT catalysts uncalcined and calcined at different temperatures.

Figure 14. Arrhenius plot for catalyzed hydrolysis of 5 wt % NaBH4− 0.5 wt % NaOH solution with Co/CNT catalyst.

slope of a straight line, the activation energy was calculated to be 33.8 kJ mol−1. This value of activation energy compares favorably with an activation energy of 66.7 kJ mol−1 for Co/IR120,28 44.3 kJ mol−1 for Co−Mo−B/Ni foam,55 and 57.8 kJ mol−1 for Co−B/C.52 The favorable activation energy values obtained in the present work are possibly attributed to the high dispersion of Co on the surface of CNTs, the high accessibility of the active phase, and the absence of any microporosity, thus eliminating diffusion and intraparticle mass transfer limitations. Further detailed investigations on the promotional mechanism of CNTs on the reaction are in progress. It should be noted that the rigorous comparison of activity between different catalysts reported in various literature reports is very difficult because the concentration and alkalinity of the reactant, reaction conditions, and catalyst types (bulk or supported) are different. It is known that Ni is also an active component for catalytic hydrolysis of NaBH4.56 However, the nickel particles are all encapsulated in the CNTs after purification, which cannot be contacted by the reactant. The hydrolysis reaction with purified CNT was also investigated in this study. No obvious activity of the purified CNT itself for the hydrolysis reaction was found. Therefore, these Ni particles contribute little to the activity of the Co/CNT catalyst.

Figure 13. Plots of the volume of hydrogen versus time for the hydrolysis of NaBH4 catalyzed by the Co/CNT catalyst at different reaction temperatures.

13). As expected, the hydrogen generation rate increases with temperature. The initial hydrogen generation rate k (mL min−1 g−1) was used to determine the activation energy by the following Arrhenius equation:

ln k = ln k 0 − (Ea /RT ) 3411

dx.doi.org/10.1021/jp211009g | J. Phys. Chem. C 2012, 116, 3405−3413

The Journal of Physical Chemistry C

Article

(17) Rümmeli, M. H.; Schäffel, F.; Kramberger, C.; Gemming, T.; Bachmatiuk, A.; Kalenczuk, R. J.; Rellinghaus, B.; Büchner, B.; Pichler, T. J. Am. Chem. Soc. 2007, 129, 15772−15773. (18) Chiang, W. H.; Sankaran, R. M. J. Phys. Chem. C 2008, 112, 17920−17925. (19) Germán, S. G.; Joöl, B.; Catherine, B. D. Catal. Today 2010, 149, 365−371. (20) Maneerung, T.; Hidajat, K.; Kawi, S. Catal. Today 2011, 171, 24−35. (21) Kuras, M.; Petit, P.; Petit, C. Carbon 2011, 49, 1453−1461. (22) Moura, F. C. C.; Tristão, J. C.; Lago, R. M.; Martel, R. Catal. Today 2008, 133−135, 846−854. (23) Liu, B. H.; Li, Z. P. J. Power Sources 2009, 187, 527−534. (24) Liang, Y.; Wang, P.; Dai, H. B. J. Alloys Compd. 2010, 491, 359− 365. (25) Ye, W.; Zhang, H. M.; Xu, D. Y.; Ma, L.; Yi, B. L. J. Power Sources 2007, 164, 544−548. (26) Malvadkar, N.; Park, S.; Urquidi-MacDonald, M.; Wang, H.; Demirel, M. C. J. Power Sources 2008, 182, 323−328. (27) Xu, D. Y.; Dai, P.; Guo, Q. J.; Yue, X. H. Int. J. Hydrogen Energy 2008, 33, 7371−7377. (28) Liu, C. H.; Chen, B. H.; Hsueh, C. L.; Ku, J. R.; Tsau, F.; Hwang, K. J. Appl. Catal., B 2009, 91, 368−379. (29) Chiang, I. W.; Brinson, B. E.; Smalley, R. E.; Margrave, J. L.; Hauge, R. H. J. Phys. Chem. B 2001, 105, 1157−1161. (30) Duan, X. Z.; Zhou, J. H.; Qian, G.; Li, P.; Zhou, X. G.; Chen, D. Chin. J. Catal. 2010, 31, 979−986. (31) Azadi, P.; Farnood, R.; Meier, E. J. Phys. Chem. A 2010, 114, 3962−3968. (32) Patel, N.; Fernandes, R.; Miotello, A. J. Catal. 2010, 271, 315− 324. (33) Li, Y. Y.; Yao, S.; Wen, W.; Xue, L. H.; Yan, Y. W. J. Alloys Compd. 2010, 491, 560−564. (34) Chen, P.; Wu, X.; Sun, X.; Lin, J.; Ji, W.; Tan, K. L. Phys. Rev. Lett. 1999, 82, 2548−2551. (35) Colomer, J.-F.; Piedigrosso, P.; Willems, I.; Journet, C.; Bernier, P.; Tendeloo, G. V.; Fonseca, A.; Nagy, J. B. J. Chem. Soc., Faraday Trans. 1998, 94, 3753−3758. (36) Hou, P. X.; Liu, C.; Cheng, H. M. Carbon 2008, 46, 2003−2025. (37) Guevara, J. C.; Wang, J. A.; Chen, L. F.; Valenzuela, M. A.; Salas, P.; García-Ruiz, A.; Toledo, J. A.; Cortes-Jácome, M. A.; AngelesChavez, C.; Novaro, O. Int. J. Hydrogen Energy 2010, 35, 3509−3521. (38) Tagliaferro, A. Phys. E: Low-Dimens. Syst. Nanostruct. 2007, 37, 58−61. (39) Snoeck, J. W.; Froment, G. F.; Fowles, M. J. Catal. 1997, 169, 240−249. (40) Chen, C.; Dai, Y.; Huang, J.; Jehng, J. Carbon 2006, 44, 1808− 1820. (41) McKee, G. S. B.; Flowers, J. S.; Vecchio, K. S. J. Phys. Chem. C 2008, 112, 10108−10113. (42) Das, N.; Dalai, A.; Mohammadzadeh, J. S. S.; Adjaye, J. Carbon 2006, 44, 2236−2245. (43) Fernandes, R.; Patel, N.; Miotello, A. Appl. Catal., B 2009, 92, 68−74. (44) Lee, J.; Kong, K. Y.; Jung, C. R.; Cho, E.; Yoon, S. P.; Han, J.; Lee, T. G.; Nam, S. W. Catal. Today 2007, 120, 305−310. (45) Zhao, J. Z.; Ma, H.; Chen, J. Int. J. Hydrogen Energy 2007, 32, 4711−4716. (46) Sciortino, L.; Giannici, F.; Martorana, A; Ruggirello, A. M.; Liveri, V. T.; Portale, G; Casaletto, M. P.; Longo, A. J. Phys. Chem. C 2011, 115, 6360−6366. (47) Bechara, R.; Balloy, D.; Dauphin, j. Y.; Grimblot, J. Chem. Mater. 1999, 11, 1703−1711. (48) Ji, L.; J. Lin, J.; Zeng, H. C. J. Phys. Chem. B 2000, 104, 1783− 1790. (49) Metin, Ö .; Ö zkar, S. Energy Fuels 2009, 23, 3517−3526. (50) Ma, H.; Ji, W. Q.; Zhao, J. Z.; Liang, J.; Chen, J. J. Alloys Compd. 2009, 474, 584−589. (51) Rakap, M.; Ö zkar, S. Appl. Catal., B 2009, 91, 21−29.

4. CONCLUSIONS In this work, we demonstrate that in situ multiwall CNTs synthesis can be realized by direct chemical vapor deposition of methane on a perovskite-type LaNiO3 growth-promoter precursor. The results show that the reduction of LaNiO3 and formation of CNTs occur almost simultaneously under methane atmosphere and that CNTs with high purity and uniform diameters can be synthesized by this one-step chemical vapor deposition method in a tubular reactor. With the purified CNTs as support, cobalt nanoparticles with an average size of 11.3 nm can be decorated on the surface of CNTs. The Co/ CNT catalyst calcined at 300 °C displays high catalytic activity for hydrogen generation from hydrolysis of sodium borohydride, but higher calcination temperature will induce part crystallization of Co particles and decrease its catalytic activity. The favorable activation energy value for Co/CNT catalyst is possibly attributed to the high dispersion of Co and the easy mass transfer of reactants and products on the surface of CNTs.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected].



ACKNOWLEDGMENTS The work was supported by Key Project of Chinese Ministry of Education (208076), The Foundation for Outstanding Young Scientist in Shandong Province (BS2009CL004). Professor Jian Zhang and Associate Professor Rui Wang of the Institute of Metal Research, Chinese Academy of Sciences are acknowledged for HRTEM measurements and fruitful discussions.



REFERENCES

(1) Yoon, B.; Wai, C. M. J. Am. Chem. Soc. 2005, 127, 17174−17175. (2) Zhang, J.; Liu, X.; Blume, R.; Zhang, A.; Schlögl, R.; Su, D. S. Science 2008, 322, 73−77. (3) Abbaslou, R. M. M.; Soltan, J.; Dalai, A. K. Appl. Catal., A 2010, 379, 129−134. (4) Serp, P.; Corrias, M.; Kalck, P. Appl. Catal., A 2003, 253, 337− 358. (5) Tessonnier, J.-P.; Pesant, L.; Ehret, G.; Ledoux, M. J.; PhamHuu, C. Appl. Catal., A 2005, 288, 203−210. (6) Alberto, A. V.; Wang, D.; Dimitratos, N.; Su, D. S.; Trevisan, V.; Prati, L. Catal. Today 2010, 150, 8−15. (7) Zhang, W.; Chen, J.; Swiegers, G. F.; Ma, Z.-F.; Wallace, G. G. Nanoscale 2010, 2, 282−286. (8) Zhang, W.; Sherrell, P.; Minett, A. I.; Razal, J. M.; Chen, J. Energy Environ. Sci. 2010, 3, 1286−1293. (9) Ovejero, G.; Sotelo, J. L.; Romero, M. D.; Rodríguez, A.; Ocaña, M. A.; Rodríguez, G.; García, J. Ind. Eng. Chem. Res. 2006, 45, 2206− 2212. (10) Hull, R. V.; Li, L.; Xing, Y.; Chusuei, C. C. Chem. Mater. 2006, 18, 1780−1788. (11) Tessonnier, J.-P.; Ersen, O.; Weinberg, G.; Pham-Huu, C.; Su, D. S.; Schlögl, R. ACS Nano 2009, 3, 2081−2089. (12) Nessim, G. D. Nanoscale 2010, 2, 1306−1323. (13) MacKenzie, K. J.; Dunens, O. M.; Harris, A. T. Ind. Eng. Chem. Res. 2010, 49, 5323−5338. (14) Kappen, P.; Halstead, B.; Rider, A.; Pigram, P. J.; Brack, N. J. Phys. Chem. C 2009, 113, 4307−4314. (15) Takagi, D.; Homma, Y.; Hibino, H.; Suzuki, S.; Kobayashi, Y. Nano Lett. 2006, 6, 2642−2645. (16) Cheung, C. L.; Kurtz, A.; Park, H.; Lieber, C. M. J. Phys. Chem. B 2002, 106, 2429−2433. 3412

dx.doi.org/10.1021/jp211009g | J. Phys. Chem. C 2012, 116, 3405−3413

The Journal of Physical Chemistry C

Article

(52) Xu, D. Y.; Wang, H. Z.; Guo, Q. J.; Ji, S. F. Fuel Process. Technol. 2011, 92, 1606−1610. (53) Ersen, O.; Werckmann, J.; Houlle, M.; Ledoux, M. J.; PhamHuu, C. Nano Lett. 2007, 7, 1898−1907. (54) Demirci, U. B.; Miele, P. Phys. Chem. Chem. Phys. 2010, 12, 14651−14665. (55) Dai, H. B.; Gao, L. L.; Liang, Y.; Kang, X. D.; Wang, P. J. Power Sources 2010, 195, 307−312. (56) Dong, H.; Yang, A. X.; Ai, X. P.; Cha, C. S. Int. J. Hydrogen Energy 2003, 28, 1095−1100.

3413

dx.doi.org/10.1021/jp211009g | J. Phys. Chem. C 2012, 116, 3405−3413