Multifunctional Catalysts for Singlewall Carbon Nanotube ...nanofast.ucdavis.edu/Publications/Ch7.pdf137 7 Multifunctional Catalysts for Singlewall Carbon Nanotube Synthesis T. Guo*

137

7 Multifunctional Catalysts for Singlewall Carbon Nanotube Synthesis

T. Guo*

7.1. INTRODUCTION

Single-wall carbon nanotubes (SWNTs) are usually produced with the help of metal catalysts that first break down various carbon species such as graphite, carbon clusters, amorphous carbon, or hydrocarbons and then release them in the form of SWNTs. These catalysts therefore must perform several functions in order to produce high quality, high yield, long SWNTs, and the finesse of these catalysts is controlled by the balance between those functions. Such a production is critical to future applications such as telecommunication, molecular electronics, sensing, and mechanical devices. In general, multicomponent materials such as bimetallic nanoparticles (bi-MNPs) are much better catalysts than those made of single component materials because the two metals in the bi-MNPs can enhance certain functions by playing complementary catalytic roles. Furthermore, some functions can only be offered by bi-MNPs because the two metals are structurally arranged in such specific manners that the combined properties would otherwise be impossible to access by monometallic nanoparticles (mono-MNPs).

Bimetallic catalysts have been proven superior with improved activity, selectivity, stability, and resistance to poisoning than their monometallic counterparts.1,2 For example, bi-MNPs are preferred heterogeneous catalysts in petroleum reforming processes.3 Bi-MNPs of large diameters (>50 nm) have also been used to catalyze the growth of carbon filaments.4 Those catalysts have also been used in ammonia synthesis * Ting Guo, Chemistry Department, University of California, Davis, 95616

MULTIFUNCTIONAL CATALYSTS FOR SINGLEWALL NANOTUBE SYNTHESIS 138

by barium promoted Ru catalysts supported on boron nitride.5 More recently, bi-MNPs (e.g., Ni/Co) have been used to produce SWNTs, and it has been shown that the yield increases by many folds in comparison to that using their monometallic counterparts.6-9 In addition, bi-MNPs have also been used to produce SWNTs of different diameters, cleanness, and lengths.

Previous studies on bi-MNPs have revealed that their superior catalytic abilities can generally be attributed to the unique geometric and electronic structures they possess. The structural information obtained on those catalysts has enabled researchers to understand the structure-function relationship associated with these nanoparticles. For instance, the catalytic mechanisms of mono-MNPs and bi-MNPs used in both petroleum reforming processes and in multi-walled nanotube production have been satisfactorily explained only after the structures of these catalysts were determined.3,7 J. H. Sinfelt and colleagues have employed the extended x-ray absorption fine structure (EXAFS) technique to thoroughly investigate bi-MNPs in petroleum reforming processes. In a Ru/Cu system, they have found that Cu atoms concentrate in the surface of small Ru/Cu bi-MNPs (1 nm diameter).3 Such surface segregation helps regulate the rates of hydrogenolysis, which requires large patches of Ru atoms to function. In contrast, hydrogenolysis dominates in pure Ru nanoparticles, leading to undesirable results. Another catalyst system, Pd-Pt bi-MNPs, which is best known for being more resistive to sulfur poisoning due to the special surface arrangement, has also been investigated by many research groups.1,10,11 In the example mentioned above, barium has been used to assist Ru in catalyzing ammonia synthesis. In this case, barium acts as a promoter, meaning that it alone possesses no activity.5 Reacting with oxygen, it forms an overlayer to enhance the catalytic ability of Ru MNPs. All these studies have clearly demonstrated that it is the special structure of those bi-MNPs that leads to the novel physical and chemical properties, resulting in enhanced activities such as catalytic functions. Similarly, the structural factor may also be the key to understanding the catalytic growth mechanisms of SWNTs.

Many investigations from both theoretical and experimental perspectives have been directed towards understanding the growth mechanisms of SWNTs.12-25 Transmission electron microscopy (TEM), x-ray photoelectron spectroscopy (XPS), and other methods are routinely employed in these experimental investigations of both SWNTs and metal catalysts. As a result of these efforts, a great wealth of morphological data on the dimensions of nanoparticles present in the samples after the nanotube growth is now available, including that of the overall composition, size distribution, and in some cases, crystallinity.9,26 It has been shown experimentally that the overall compositions of MNPs can alter the yield and diameter of SWNTs. It has also been discovered that bi-MNPs are generally much better catalysts for SWNT production. For example, the presence of Ni/Co bi-MNPs can produce ca. 100% SWNTs.8

Theoretical studies have been performed on the growth of SWNTs as well.23,27,28 It has also been suggested that metal atoms may be responsible for the growth.20 Most other models, however, consider that the growth is the result of the interaction of metal nanoparticles and the carbon feedstock.21

Despite those advancements, detailed information such as the surface composition of multicomponent nanoparticle catalysts is still lacking. As a result, the exact mechanisms through which MNPs catalyze the growth of SWNTs are still largely unknown.29 It is


now obvious that more structural information is needed before one can understand the growth mechanisms of SWNTs.

As mentioned earlier, a special case that may help shed light on the growth mechanisms is the significant enhancement of SWNT yield when bi-MNPs are used. In order to address the question of why bi-MNPs are better catalysts, many research groups have begun to investigate the structures of these nanoparticles. In this chapter, we will present recent results obtained by our group and other research groups. Although the results on several catalytic systems will be presented, we will focus on the Ni/Co system. The characterization method used here is EXAFS, which has been employed to investigate other bi-MNP or multicomponent systems. In addition to studying bi-MNPs that are used in space or free of support, we will also consider MNPs on support as a special case of multifunctional catalysts, to demonstrate the limits of using EXAFS to investigate an assembly of nanoparticles and nanotubes.

In the following sections, we will first describe the synthetic methods to make nanoparticles and nanotubes in Sec. 7.2.1, and the details of EXAFS technique in Sec. 7.2.2. We will then present the results of the catalytic systems that we have studied, including mono-MNPs (Sec. 7.3.1) and bi-MNPs (Sec. 7.3.2). In the end, we will briefly discuss other multicomponent catalyst systems (Sec. 7.3.3).

7.2. EXPERIMENTAL

7.2.1. Sample preparation

We will describe the syntheses of mono-MNP Co, and bi-MNP Co/Ni and Co/Mo nanoparticles in this section. References to the methods for making other nanoparticles discussed in Section 7. 3.2.3 and 7.3.3 are listed in the text therein.

7.2.1.1. Co Nanoparticles. The Co mono-MNPs were made using the high temperature decomposition method developed by Alivisatos and colleagues.30 Using this method, bi-MNPs can be made as well.31 Different sizes of Co mono-MNPs were made, and these particles were used to synthesize carbon nanotubes. Three ratios of three surfactants, trioctylphosphine oxide (TOPO), oleic acid (OA) and trioctyl amine (TOA), were used in the nanoparticle synthesis to produce distinctive sizes of Co MNPs at 3, 5, and 12 nm, respectively. These nanoparticles were deposited on clean Si substrates, and the assembled samples were placed in a high temperature furnace. Methane, hydrogen, and argon gases were fed into the reactor at high temperatures (>1000 ºC). Nanotube growth using these Co MNPs was conducted in the furnace at 1100 ºC. A detailed description of the production of nanotubes and nanoparticles will be published in another paper.32

7.2.1.2. Co/Ni Nanoparticles. The Ni/Co bi-MNPs were obtained from [email protected] They contained the product of materials produced in a laser vaporization chamber by vaporizing metal-graphite composite targets with 0.6 at. % of Co and 0.6 at. % of Ni. In the EXAFS measurements, micrograms of the nanotube-bi-MNP samples were deposited onto Scotch tapes, which were then placed in the synchrotron x-ray beam. We chose to examine the product from laser vaporization because it was in this case that the highest yield of SWNTs was obtained (~ 100%). It indicated that these nanoparticles possessed certain properties that differed from that of a simple mixture of the two metals.


7.2.1.3. Co/Mo and Other Nanoparticle Catalysts. A brief description of Co/Mo(W) on SiO2 system will be given below. The preparation methods used to produce other catalysts are not provided here. Readers who are interested in these methods should consult with the references listed in the relevant sections (7.3.1 and 7.3.2).

The Co–Mo catalysts were prepared by impregnating a silica gel support with aqueous solutions of cobalt nitrate and/or ammonium heptamolybdate. The total metal content was kept at 6%. Calcined samples at 500 ºC were used for EXAFS/XANES measurements.34

7.2.2. EXAFS

The main characterization method used here is extended x-ray absorption fine structure (EXAFS) spectroscopy. We will briefly describe the experimental setup, and then discuss the data analysis method we have employed.

7.2.2.1. Setup. EXAFS spectroscopy can reveal the immediate chemical environment around an absorbing element in composite materials and is ideal to study multicomponent systems such as bi-MNPs.35 EXAFS is also complementary to many other ways to probe MNPs such as electron microscopy because it can detect the whole sample rather than only those visible to the electron microscope.2 Furthermore, because EXAFS can be used at high temperatures and x-rays are highly penetrating, it is applicable to probe metal species through the carbon background during the nanotube growth.

EXAFS was performed on the K edges of Ni and Co in Ni/Co bi-MNPs, standards, and Co in Co mono-MNPs. The standards were Ni and Co thin foils (3-µm thickness).

FIGURE 7.1. Schematic of the EXAFS setup. The fluorescent signal, IF, is used with the ionization chamber (IC) #1 to obtain the EXAFS raw data on the sample. A standard is measured at the same time using detector #2 and #3. Harmonics are suppressed with the second crystal monochromator.


The bi-MNPs were the raw soot of a few micrograms containing both webs of SWNTs and Ni/Co bi-MNPs.

The measurements were performed at the beamline 4-1 at the Stanford Synchrotron Radiation Laboratory (SSRL) with a double silicon crystal (220) monochromator, as shown in Figure 7.1. Harmonics were suppressed by detuning a second crystal spectrometer. No efforts were made to control the temperature of the samples. EXAFS data was obtained by detecting the fluorescence of K lines from Ni and Co in the bi-MNP samples or that from Co in Co MNPs using a Lytle detector. Simultaneously, EXAFS signals of the standards were detected in the transmission mode with ion chambers. When detecting fluorescence, thin metal foil filters placed in front of the Lytle detector were used to reduce the scattering of the incident x-rays by the sample and sample holder. Mn and Fe filters were used when detecting Co and Ni fluorescence, respectively. The experimental schematic is shown in Figure 7.1.

In order to reduce the amount of interference from Co fluorescence on the Ni EXAFS data in the near edge region of the Co/Ni bi-MPNs, a LN2 cooled, 13-element, high-energy resolution (300-eV FWHM) Ge detector was used to replace the Lytle detector. The parameters of each element of the detector were optimized separately.

In grazing angle EXAFS measurements, a similar Ge detector was used for collection of a large solid angle of fluorescent photons with high detection efficiency. The incoming x-rays irradiated a sample at 0.2 degrees grazing angle. The sample was a flat Si substrate on which Co nanoparticles were deposited. That angle was smaller than the ~ 0.24 degree critical reflection angle for Si at the photon energy of the incident x-rays (up to 11 keV).

7.2.2.2. EXAFS Data Analysis. All EXAFS data as a function of the energy of the incoming x-ray photons was collected using the XAS program at SSRL. The raw data was analyzed using ATOMS, EXAFSPAK and FEFF7.36 Subtractions of polynomial-fit pre-edge, spline-fit atomic absorption curves, and k3 factor multiplication were performed using the built-in programs in the EXAFSPAK package. Using k= )(2 0EEm − , the EXAFS data in energy space was converted to that in k reciprocal space. Only 3.5 – 11 Å-1 k range was used to define the EXAFS data. This range was chosen so that the distortion to the Ni EXAFS data caused by the Co fluorescence on the Ni EXAFS measurements could be minimized because the K edges of Ni and Co are only 500 eV apart. The radial distribution function (RDF) in real space, which is determined by the number of neighboring atoms, their average disordering, and the type of atoms, was then obtained via Fourier transformation of the EXAFS data in k space. A 1 – 6 Å range was used for displaying the RDFs. Theoretical phases and amplitudes for each element were calculated by the FEFF7 program. Fitted structural parameters (average neighbors, N, average disorder σ2, and bond length R) were then obtained using the calculated amplitudes and phases for each coordination shell. Because the structural parameters for the first coordination shell can be analyzed with high accuracy, we employed Fourier filtering by Fourier transformation of the first major peak of the RDF to k space (3.5 – 11 Å-1) to obtain the filtered EXAFS data. The filtering window was set to be 1 to 3 Å, and a 0.2-Å slope was used to reduce ripple effect caused by Fourier filtering. An optimization program in the EXAFSPAK package was then used to fit this data and to obtain the fitted parameters.


In the following, several sources of error are estimated. First, since the sample size was small, the absorption of Ni fluorescent photons by surrounding Co in the samples was negligible. Second, since the x-ray energy at which Ni EXAFS was measured was higher than that Co absorption edge, the detected photons for Ni could include fluorescence from Co. This distortion produced by the Co K line fluorescence on the low energy portion of the Ni EXAFS spectrum was estimated to be less than 3% because of the lack of modulation of Co fluorescence (EXAFS) in this region. When the Ge detector was used, the amount of distortion was reduced by over one order of magnitude, making this error negligible. After the fitting, the amplitudes of the Fourier transform were higher than the original ones due to the reduction in the range of data in real space. However, since the parameter fitting process was actually performed in k space, no distortion was introduced by this procedure. After fitting, the three parameters defining the first coordination shell, N (average coordination number), σ2 (disorder or the Debye-Waller factor) and R (bond distances), were obtained.

Error analysis was carried out to estimate the standard deviation for each fitted parameter. The error (the F value) given in the EXAFSPAK package was used. The standard deviation of a specific fitting parameter was obtained when the F value was doubled while scanning the parameter.37 This method is compatible to the χ2 method, in which the experimental errors are included as noise.38 We found that F ≤100 generally corresponded to a good fitting as indicated by the χ2 method, after considering the experimental noise in the EXAFS data.

7.2.3. TEM

High resolution TEM (JEOL JEM-200 CX operated at 200 keV) was used to perform selected area diffraction (SAD) and dark field imaging measurements on these nanoparticles at the National Center for Electron Microscopy (NCEM) at Berkeley. Micro-micro diffraction was also taken on individual nanoparticles. The composition of these particles on an individual nanoparticle basis was measured using energy dispersive x-ray spectroscopy (EDX) with the TEM. A regular TEM (Phillips CF-12) was used to inspect Co nanoparticles. A scanning electron microscope (SEM, Phillips XL30SFEG) was used for inspecting nanoparticles and nanotubes on Si wafers.

7.3. RESULTS

In general, tip growth and root growth models are used to explain the growth of SWNTs.39,40 The former requires no interaction of the MNPs and the substrate while the latter has to rely on such interaction to anchor the MNPs on the substrate. In this section, both mono-MNPs and multi-MNPs are studied. Our presentation will focus on the results of EXAFS of the Ni/Co binary system that functions in space without support (tip growth). In addition to other binary systems that are used either in space or with a weak support, we will also show preliminary results on mono-MNPs so as to demonstrate that they are much less efficient, especially when they are used as pre-made catalyst nanoparticles on support.


7.3.1. Monometallic Catalysts on Support

Mono-MNPs made during the nanotube synthesis were originally used as catalysts. For instance, Co, Ni, Fe, and other first row transition metals were used to grow SWNTs. Other investigations using metals such as Gd, Pt, Mo, Y, La, Sc, W, and other metals as catalysts were reported as well.22,41 Since the majority of them are made in situ and in space, the support-catalyst interaction can be neglected. However, more recent results employing pre-made catalysts deposited on substrates in general yield fewer nanotubes. Thus it is important to obtain structural information on those MNPs to understand the causes for such change.

In this section, we will show that mono-MNPs on support behave differently from those in space due to the influence of support-catalyst interaction on the growth. Although direct measurements are lacking, we will show that it is possible to obtain the desired structural information on those systems, even during the catalytic process. On the other hand, nanotube synthesis using these nanoparticles is similar to those in gas phase if these nanoparticles are lifted off from the substrate by nanotubes.

7.3.1.1. Results on Fe/SiO2 System. As mentioned earlier, morphologically identical MNPs behave differently in catalyzing the growth of SWNTs. This is best manifested in the experiments performed by Dai et al.42

In their experiments, Fe nanoparticles made from ferritin proteins were used as catalysts. Si substrates and the CVD method were used for nanotube synthesis.42 Their results clearly indicated that SWNTs can be grown from these isolated Fe nanoparticles. However, some of the morphologically identical nanoparticles failed to initiate the growth of SWNTs. It is unknown to the authors as to the causes for such a disparity. It is suspected that the substrate or contaminates may be the reasons for the inactivity.

7.3.1.2. Results on Other Mono Systems. Other groups have used the CVD method to grow SWNTs and only few nanotubes were found among numerous nanoparticles. For example, Lieber et al. recently reported on the synthesis of SWNTs from pre-made mono-MNPs.43 No structural information was available on those mono-MNPs. They found that only a small number of the nanoparticles participated in the catalytic process. This is in contrast to gas phase synthesis of SWNTs such as HiPCO method which can convert carbon to SWNTs at very high yields.14

SWNTs can also be produced by much larger metal nanoparticle catalysts, following the so-called sea-urchin growth model. No direct, in situ structural data is available to provide the information on the morphology of the nanoparticles during the catalytic growth of these short, straight SWNTs.44

7.3.1.3. Results on Co System. In this section, we will show that EXAFS can be performed on pre-made MNPs deposited on substrates to reveal structural information.

Using the pre-made Co mono-MNPs described above, we successfully produced carbon nanotubes. Figure 7.2 shows 12-nm Co nanoparticles synthesized recently. Normal EXAFS measurements were performed on these and two other sized nanoparticles, as shown in the right panel of Figure 7.2. The EXAFS data shows that large (12-nm) Co nanoparticles are partially oxidized while small ones (3 and 5 nm) are


more heavily oxidized. From Figure 7.2 it is also clear that the average numbers of first neighbors for Co nanoparticles (~4) is much less than that in bulk Co.

We also measured a sample of the 3-nm Co nanoparticles that were deposited on a Si wafer before the nanotube growth, as shown in Figure 7.3. Although it is possible to observe the first coordination shell, the signal-to-noise ratio has to be improved before a definite conclusion can be reached as to what the physical and chemical properties of these nanoparticles are. One improvement that can be readily made is to reduce the oxidized nanoparticles to their pure metallic forms before the measurement.

7.3.1.4 Discussion. It is obvious from the above studies that mono-MNPs on support behave very differently from each other when they are used in the root growth manner. Other results have also confirmed that monodisperse MNPs are catalytically unequal.43,45 It is unclear based on the current studies as to what is the key to the initiation of SWNT growth. Some catalysts are capable of producing SWNTs, while others are not. It will be extremely important to determine the causes for the initiation of SWNTs, and it is apparent that any analysis on these nanoparticles would have to be made on an individual catalyst basis, especially if the yield of producing SWNTs such as that prepared using the CVD method with these nanoparticles is low.

However, when the catalysts are lifted off the surface, such as those at high temperatures in the Co/Ni system, their catalytic ability is greatly enhanced. As will be shown later, that yield can reach 90%. In those cases, it is more valid to use bulk methods such as EXAFS.

0

0.5

1

1.5

2

2.5

3

3.5

1 2 3 4 5

12nm5nm3nmtotal

R (angstrom)

FGIURE 7.2. TEM image of 12–nm Co nanoparticles (left panel) and the EXAFS measurements on these particles (right panel). The total plot in the EXAFS is the theoretical model of ε-fcc of Co bulk. It is clear that the EXAFS signal of the Co particles, which is proportional to the number of nearest neighbors, is much smaller than that of the bulk. Also shown are the results of EXAFS of 3 – and 5-nm Co nanoparticles. They are clearly oxidized more extensively than that of the 12-nm particles.


Although the detailed structure is unknown, it has been shown experimentally that the inactive nanoparticles remain isolated during nanotube growth.42 It is thus reasonable to rule out the aggregation of those nanoparticles as the cause for such inactivity.

7.3.2. Binary Systems

In this section, we will focus on one kind of binary nanoparticle catalysts, i.e., the Ni-Co system, while reviewing the results of several other binary systems, e.g., Co-Mo/W and Ni-Y systems. EXAFS measurements on the first two systems are available.34,46 In the first system, which is an exemplary case where the growth is most likely completed in free space, we wish to demonstrate that these particles synthesized at the optimized temperature of 1200 °C possess unique structures after they are cooled down to room temperature. This finding may prove important to understanding the high-yield production processes of SWNTs. In the second system, Co clusters are found separated by Mo or W, which explains why bimetallic catalysts are better. In both systems, yields of SWNT production are much higher, making bulk structural determination methods such as EXAFS more valid. Other catalysts used for nanotube synthesis will also be discussed, although no structural information on those systems is yet available.

7.3.2.1. Results on Ni/Co System. The main purpose of this section is to show that special structural arrangements in binary systems may be the cause for the enhanced catalytic properties observed in those systems.46

0

0.2

0.4

0.6

0.8

1

1.2

0 1 2 3 4 5 6

Four

ier

Am

plit

ude

R (angstrom)

FIGURE 7.3. Grazing angle EXAFS measurement of 3-nm Co nanoparticles deposited on Si substrate. It is clear that the EXAFS signal of the Co particles, which is proportional to the number of nearest neighbors, is much smaller than that of the bulk that is shown in Figure 7.2.


Figure 7.4 shows the EXAFS result of Co in Ni/Co bi-MNPs after background correction. The EXAFS measurements on the Co and Ni standards and on Ni in bi-MNPs have yielded similar results. The insert in Figure 7.4 shows the near edge patterns of Co in bi-MNPs (solid line) and the bulk (dashed line). It is evident that the metals in the SWNT samples are oxide free, as indicated by the lack of the metal-oxygen pre-edge peaks. Since the magnitude of the jumps in the absorption spectra of Ni and Co at the K edge are approximately the same (not shown), there should be approximately equal amounts of Ni and Co in the bi-MNP sample. The results of EXAFS measurements on the standards are similar to those obtained previously at room temperature,47 all showing that the modulation in Co EXAFS is slightly smaller than that of Ni.

Figure 7.5 shows the Fourier transforms (phase uncorrected) of those EXAFS profiles after data processing. Peaks in radial distribution functions (RDFs) represent different coordination shells, and the profiles of these peaks are determined by the local atomic arrangement. The lack of oxides is also clearly shown in the Fourier transforms because of the missing intensity near 1.5 Å (phase uncorrected). The peaks at 1.5 Å (phase uncorrected) in Figure 7.5 correspond to a side band of metal-metal scattering. The lack of the oxides may be due to the fact that all the particles are protected by multiple layers of graphite, as observed in other TEM investigations,16 suggesting that magnetic particles made in a similar manner should also be stable in air.

From Figure 7.5, it is apparent that the local environment around the Ni atoms in the bi-MNPs is different from that of Ni in the bulk. For Co, the RDF of the Co in bi-MNPs is almost identical to that of the bulk.

Using the single scattering theory together with Fourier filtering, fitting parameters are obtained for all the first coordination shells. The obtained fitting parameters are listed in Table 7.1.

FIGURE 7.4. EXAFS measurement on Co in the bi-MNP sample. The insert shows the near edge curves of Co in the bulk and bi-MNPs. The signal corresponding to Co oxides in the pre-edge region is below the detectable level. (Reprinted with permission from J. Phys. Chem.B 2002, 106, 5833-5839. Copyright 2002 American Chemical Society.)


0

1

2

3

4

5

6

1 2 3 4 5 6

Stad. CoCo in MNP

R (Å)

FT m

agni

tude

0

1

2

3

4

5

6

7

1 2 3 4 5 6

Stad. NiNi in MNP

R (Å)

FT m

agni

tude

FIGURE 7.5. Fourier transforms of EXAFS data of Co in the bi-MNP sample (solid line) and the standard (dashed line) (5a), and Ni in the bi-MNP sample (solid line) and the standard (dashed line) (5b). EXAFS data between 3.5 to 11 Å-1 is Fourier transformed to the RDFs shown here. (Reprinted with permission from J. Phys. Chem.B 2002, 106, 5833-5839. Copyright 2002 American Chemical Society.)

TABLE 7.1. Fitted parameters for the first coordination shells of Ni and Co in bi-MNPs. Materials S0 N R (Å) σ2 Ni standard 0.682 12.0 (0.6) 2.469 (0.002) 0.00538 (0.00027) Co standard 0.690 12.0 (0.6) 2.478 (0.003) 0.00590 (0.00037) Ni in bi-MNPs (Ni-Ni) 0.682 5.6 (0.4) 2.469 (0.005) 0.00532 (0.00054) Ni in bi-MNPs (Ni-Co) 0.682 4.4 (0.5) 2.473 (0.006) 0.00541 (0.00089) Co in bi-MNPs (Co-Co) 0.690 6.6 (0.5) 2.466 (0.005) 0.00538 (0.00059) Co in bi-MNPs (Co-Ni) 0.690 4.4 (0.5) 2.473 (0.006) 0.00541 (0.00089)

S0 is the scaling factor, R is the bond distance, σ is the Debye-Waller factor, and N is the coordination number. The standard deviations are shown in parentheses. (Reprinted with permission from J. Phys. Chem.B 2002, 106, 5833-5839. Copyright 2002 American Chemical Society.)

Results of Co and Ni in the bi-MNPs and the bulk are shown in Table 7.1. Since both

Ni and Co are in the bi-MNP, the backscattering atoms include both Ni and Co. Consequently, the first coordination-shell consists of both Ni-Ni and Ni-Co for an absorbing Ni atom, and Co-Co and Co-Ni for an absorbing Co atom. The optimal Ni-Co coordination number, which should be the same in the first shells for Ni and Co in the bi-MNPs, is obtained by scanning this coordination number from 1 to 9 while minimizing the combined error in the fitting for both Ni and Co in bi-MNPs. During the fitting, the disorders (σ), coordination numbers (N), bond distances (R), and E0 are varied. From Table 7.1, it is obvious that Ni atoms in bi-MNPs have fewer neighbors (5.6 Ni and 4.4 Co, totaling 10.0) than the Co atoms do (6.6 Co and 4.4 Ni, totaling 11.0). In addition, Co has more Co as nearest neighbors (6.6) than the Ni does (4.4), although in bulk these two


elements are totally miscible.48 The results from the Ge detector are similar to those shown in Table 7.1, except the coordination numbers are slightly lower. Based on the first coordination numbers of Co and Ni, the average size of these nanoparticles is determined to be 3.0 nm.

The squares of disorder (σ2) for the Co and Ni in the bi-MNPs and that for Ni in the bulk shown in Table 1 are very similar. The disorder for the Co in the bulk is slightly higher. This suggests that both Ni and Co in the bi-MNPs are in similarly ordered structures, which are also close to that of the Ni bulk. The slight decrease in disorder for Co in the bi-MNPs may be caused by the extensive mixing of Ni and Co, which makes Co atoms slightly more ordered than they are in the bulk. Although it is well known that the surface atoms in some of the small MNPs (1-2 nm) have larger disorders, this is apparently not the case for the Ni/Co bi-MNPs in the soot, based on the EXAFS and TEM measurements. It was observed in selected area diffraction (SAD) and dark field imaging measurements that the majority of the MNPs were highly ordered and crystalline nanoparticles. Micro-micro diffraction was performed on individual nanoparticles. The pattern fits well to a Ni or Co fcc (011) (a = 3.5 Å) structure.

The sizes of the particles in the bi-MNP/SWNT sample were estimated with TEM. We examined over 430 particles to obtain the size distribution. It is worth remarking that TEM is a sampling technique, i.e. only those which are visible to TEM can be measured. On the other hand, EXAFS measures all the material in a sample. It is possible that many particles, especially the small ones that were obscured by other materials, were not included in the TEM counting. The disparity between this result (8.0 ± 5.0 nm) and the value obtained based on EXAFS measurements (3.0 nm) is obvious. It shows that a large number of small particles may be missed in the TEM measurement.

We measured the composite samples with TEM dark field imaging because only the portions that have the (111) plane orientation in those crystalline particles are visible in the dark field. The average size of particles obtained from the dark field (5.0 ± 3.0 nm) is smaller than that obtained from the bright field measurements (8.0 ± 5.0 nm). The difference between the two average sizes may be explained by the multiple domains in those nanoparticles.

We also measured the composition of these particles on an individual nanoparticle basis. Samples were examined with EDX to determine the composition of individual particles. Figure 7.6a shows a typical image of the raw material. Fig. 7.6b displays the result of the EDX analysis of the single particle selected in Fig 7.6a, as pointed at by the arrow. We examined totally 20 nanoparticles of different sizes, and the average ratio of the amounts of Co to Ni in these nanoparticles was 50 ± 5% : 50 ± 5%.

7.3.2.2. Results on Co/Mo and Co/W Systems. Recently, Resasco et al. employed EXAFS and x-ray absorption near edge spectroscopy (XANES) to study both Co-Mo and Co-W catalysts in CO disproportion synthesis of SWNTs.34,49 Their findings indicate that monometallic catalysts (Co, Mo, or W) are much less effective than bi-MNPs because excessive amount of W or Mo help isolate small Co oxide nanoparticles for the synthesis of SWNTs. Even though their synthetic conditions were different from either the Ni/Co system in laser vaporization experiment or the HiPCO method in which Fe or other catalysts are also produced in situ and in space,14 the substrate effect was found negligible because the authors did not find any Si in the nanoparticles.


FIGURE 7.6. High resolution TEM image of Co/Ni nanoparticles (6a) and EDX results (6b) on one selected nanoparticle (pointed at by the arrow in 6a). More than 20 nanoparticles are examined by EDX, and the average ratio of Ni to Co is 50% ± 5%. (Reprinted with permission from J. Phys. Chem.B 2002, 106, 5833-5839. Copyright 2002 American Chemical Society.)

7.3.2.3. Other Binary Systems. Other binary systems have also been used in nanotube synthesis. Although many investigations have been carried out, only a brief review will be given here. It is well known that bimetallic nanoparticles are generally better catalysts in synthesizing SWNTs. As mentioned earlier, it is worth noting that this is true only if those catalysts are used without the interference from the substrate. On the other hand, in


at least one experiment bi-MNPs of Ni/Co were found to be no better than their mono-metallic counterparts in catalyzing the growth of SWNTs when they were used with alumina substrates.22 As shown in Sec. 7.3.1, it is apparent that the nanotube growth can be affected by the substrate even in the mono-MNPs case. Therefore, we will only consider the bimetallic catalysts that function as catalysts in the gas phase (tip growth).

Zhou et al. used various precursors to grow nanotubes in free space.24 Other groups, including that of Smalley, employed bi-MNPs to grow SWNTs in the gas phase.8 Eklund et al. also employed Fe/Ni and Ni/Co particles in free space nanotube production.50 They found that the diameter of SWNTs increases with increasing temperature, although no explanations were given. In another study, Takizawa et al. investigated Ni/Co in comparison with Ni and Co. They saw 10-nm diameter bundles of nanotubes and an enhancement factor of two to a few hundred. Ni/Co was especially better than Co alone, although the enhancement factor from Ni to Ni/Co was only a few times.51

Kiang used Co and Fe as primary catalysts, and added promoters such as S, Y, Pb, and Bi.9 As a result, larger diameter nanotubes were produced. Yield was also increased by 3 folds in comparison with that using Co alone.9 Kiang et al. also tested the addition of W, Bi, and Pb to Co pure metal on the growth of SWNTs, and found that the features of SWNTs were altered when the second metal was added.52

Iijima et al. investigated Y and Ni on the generation of SWNTs using an arc-discharged method.26 Y was found to be able to accelerate the rate of soot generation, and more importantly, to control the diameter of the SWNTs. Y alone did not catalyze the growth of SWNTs, and no structural explanation was given for the accelerated rate and diameter control effect.

In all these studies, binary catalysts behave similarly to the Ni/Co system described above, which is that the binary catalysts exhibit superior catalytic abilities than their pure elemental counterparts.

To date, no studies have yielded the connection between nanotube catalysis and electronic promoters such as those systems investigated by Goodman et al., implying that much more work is needed in terms of understanding the catalytic process of the growth of SWNTs.53

7.3.2.4. Discussion. Although it is not critical to connect the nanotube-sized nanoparticles to nanotube growth, it is easier to understand the growth mechanisms if the two species are structurally linked. Increasingly MNPs have been directly linked to the tips of nanotubes by TEM imaging, and it is now generally believed that these nanoparticles are responsible for the growth of SWNTs for the following reasons. First, many kinds of mono-MNPs have been found at the end of SWNTs synthesized in slow-growing environments, including CVD and a specially designed, two-step laser vaporization process.42,54 Second, under otherwise identical conditions, the yield of SWNTs is generally much higher when two metals instead of one are used.42 If the two metals stay apart from each other in individual nanoparticles or catalytic entities, the observed enhancement phenomena in those experiments would be difficult to explain. Thus, the two metals must stay in close vicinity, most likely next to each other, to enable the enhancement.

One may argue that since no nanotubes are observed without MNPs, the role of catalysis might be ill-defined. Recently, nanotubes have been found in a non-catalyst


environment,55 which confirms the validity of regarding MNPs as catalysts, not as a part of the nanotubes.

In most cases where bi-MNPs have the highest enhancement effect such as in laser vaporization and arc discharge of metal-graphite composite targets, however, direct evidence to link SWNT-diameter sized bi-MNPs (~ 1.6 ± 0.5 nm) to the end of nanotubes is still lacking, possibly due to the fast rates of SWNT formation under these conditions. These are the cases where the nanotubes are grown in free space or gas phase.

For the Ni/Co system, the results shown above can be explained in three possible ways. First, assuming that all Ni and Co are in the form of nanoparticles, and the average size of the particles is 3 nm, then because the disorders are similar for Ni and Co atoms in those nanoparticles (see Table 7.1), the lower coordination number for Ni indicates that more Ni atoms are in the surface. For example, nanoparticles of 3.0-nm diameter containing equal amounts of Ni and Co have on average 10.5 nearest neighbors.56 On the other hand, this number is 9.0 for Ni and 12.0 for Co for the same size bi-MNPs with all the surface atoms being Ni. Clearly, the real structure (with N=10.0 for Ni) lies somewhere in between the full segregation model (N=9.0) and the totally uniform bi-MNPs (10.5).

The second explanation can be that there are more isolated Ni atoms in the soot since such an arrangement can effectively lower the coordination number for Ni. However, this is unlikely for two reasons. First, if many Ni atoms are in the atomic form, they could be easily oxidized, and those oxides should be visible in the EXAFS measurements Although one may argue that it should be possible to tell whether carbides exist by examining the metal-carbon peak in EXAFS, we must point out that it is difficult to tell whether the surface is passivated with carbon based on EXAFS data since the backscattering amplitude of carbon is weak. Furthermore, the overlap between the main peak in Ni-C RDF curve and that of the side band from Ni-Ni at ~1.4 Å (phase uncorrected, see Figure 7.5) makes it difficult to predict the amount of metal carbide in the sample. Secondly, TEM data shows that there are equal amounts of Co and Ni in all the particles examined, as shown in Figure 7.6. This microscopic finding agrees with the x-ray absorption measurements shown earlier, which probe several micrograms of the similar material that contains billions of nanoparticles. All these results support the notion that all Ni and Co are in the bi-MNP form.

The last possible scenario is that the Ni portion in the bi-MNPs has more defects than that of Co. However, as the disorders and dark field images have indicated, both metals are similarly structured. Therefore, it is unlikely that the Ni portion has more defects.

We have simulated different Ni/Co bi-MNPs with carbon on the surface, and have compared them with the experimental data. Figure 7.7a shows the results of the simulation for Co. It contains the Fourier transforms of the EXAFS data of the Co bulk (fcc), pure Co MNPs, uniform Co/Ni bi-MNPs, and surface segregated Co/Ni bi-MNPs. In the simulation, we assume the disorders for Co in all those samples are the same with the exception of Co bulk, for which a higher value based on the measurements is used. The average size of MNPs is 3 nm. The results shown in Figure 7.7a suggest that only the surface segregated Co/Ni bi-MNPs can duplicate the measured result, which was identical to the EXAFS data of the bulk. Similar calculations have been performed on Ni, and the correct model is again the surface segregated Co/Ni bi-MNPs. The results are shown in Figure 7.7b.


FIGURE 7.7 (a) Simulation results of Co bulk (solid line), pure Co MNPs (dotted line), uniformly mixed Co/Ni MNPs (dash line), and surface segregated Co/Ni MNPs (solid line). Note that the Co bulk and surfaced segregated model produce similar RDF profiles. (b) Simulation results for Ni. Note that the amplitude of RDF for surfaced segregated model is ~ 20% lower than that of the bulk, reproducing the experimental results. The size of the MNPs is 3 nm. (Reprinted with permission from J. Phys. Chem.B 2002, 106, 5833-5839. Copyright 2002 American Chemical Society.)

Based on the experimental results and simulation, we propose that bi-MNPs have more Ni atoms in the surface and more Co atoms on the inside. This model allows us to explain all the observations and the fitted parameters. The reason that the Co in the sample produces a similar EXAFS to that for Co bulk is because the Co are slightly more ordered in the bi-MNPs, and the majority of Co are on the inside of Ni/Co bi-MNPs. By the same token, the modulation in the EXAFS of Ni is lower because of the large number of surface atoms, which have fewer nearest neighbors.

The observed segregation effect can be partially attributed to the lower surface energy of Ni atoms in the surface.57 If this is the dominant cause, then nanoparticles of smaller sizes should experience an even stronger surface segregation effect due to the higher percentage of surface atoms. For example, the percentage of surface atoms increases to above 50% when the size of bi-MNPs becomes smaller than 2.5 nm. Since the average diameter of SWNTs synthesized in this manner is usually 1.6 ± 0.5 nm, the Co/Ni bi-MNPs in the 1-2 nm diameter range should be responsible for their growth. In this case, the percentage of the surface atoms can be over 75%. This means that all the core atoms are Co, and both Co and Ni are in the surface, although Ni dominates. Such a segregation arrangement may create an optimized environment for the growth of SWNTs. If this is true, then the size of the nanoparticles controls the surface composition and consequently, the chemical properties of those metal nanoparticles.

Another possibility of explaining the surface segregation is that the formation of graphene sheets outside these nanoparticles influences their structure.

One may further speculate on the roles of surface segregation in assisting the growth of SWNTs. It is possible to view Ni atoms as electron rich in comparison to Co.

0

1

2

3

4

5

6

7

8

2 3 4

FT m

agni

tude

R (Å)

BulkUniPureSeg.

0

1

2

3

4

5

6

7

8

2 3 4

FT m

agni

tude

R (Å)

BulkUniPureSeg.


Therefore, more carbons are needed to totally satisfy the bonding requirements for Co. Consequently, Co holds on to carbon more tightly. It is also possible that Co tends to form large size particles. With Ni on its corners, the smaller nanoparticles may be maintained for an extended period of time, which is good for SWNT production.

In our analysis, we have assumed that MNPs are spherical, which may not be true as other geometries of MNPs have been found.22 Furthermore, it is possible that these Ni/Co bi-MNPs may adopt special shapes that favor the growth of SWNTs as other bi-MNPs have undergone similar transitions.58

7.3.3. Other Multifunctional Catalytic Systems

As discussed above, we consider binary catalysts deposited on substrate to be tertiary systems due to the interaction between the two components. A good indication that these systems are tertiary is that only a small portion of the nanoparticles have participated in the growth of SWNTs. For example, Co-Mo systems on alumina should be considered tertiary systems since Al plays a part in the catalytic process: it deteriorates the quality of Co-Mo catalysts.34 Among other studies, Dai et al. used Fe/Mo and Fe/Ru systems deposited on alumina nanoparticles to synthesize SWNTs,59 although no data was given on whether alumina participated in the catalytic growth process. Ru was claimed to be able to release the grown SWNTs after a certain length, and Mo was credited for promoting aromatization reactions of methane. Liu and colleagues reported using Fe/Mo systems on substrate,60 and they have also used Fe/Mo in alumina aerogel.61 It is worth mentioning that Fe/Mo catalysts have been investigated by others and surface segregation has also been found in this system.62 Thus it is interesting to apply EXAFS on these systems.

7.4. FUTURE WORK

Because the yield of SWNTs is usually strongly temperature dependent, it is desirable to measure the structures of bi-MNPs produced at different temperatures, including room temperature. Further investigations are also necessary to learn more about the structures of these nanoparticles at high temperatures to determine whether these nanoparticles retain the surface segregated structure at the growth temperature of > 800 °C.

Since EXAFS and XANES are both bulk methods, it is only valid to employ them to study either uniform catalysts or the catalysts on an individual nanoparticle basis. It is thus clear that current characterization methods used to understand the structure-function relationship for those MNPs are limited. First, the catalysts are in a heterogeneous group, and even in the Ni/Co case, there is a possibility that the accuracy of the measurements might be affected by the presence of nanoparticles of different sizes. Similarly, the conclusion that Co oxides in Co-Mo or Co-W systems are the catalysts is both significant and worth further investigation because to date all catalysts responsible for the growth of SWNTs are made of pure metals. Therefore, in order to understand all aspects about the catalytic processes involving MNPs and SWNTs, new methodologies and instruments are needed to characterize nanomaterials under extreme reaction conditions and in small quantities. Single nanoparticle analysis seems most necessary. Along this direction,


several pieces of work have been done. For example, Alivisatos et al. have used XANES to analyze single nanoparticles.30 Atomic resolution TEM may also play a major role in the future.5

An alternative is to produce MNPs with a total control. This requires the synthetic methods that are able to produce MNPs of known composition, size, shape, and known interfacial structure connecting the substrate.

7.5. CONCLUSION

We have discussed different catalysts that catalyze the growth of SWNTs. Mono- and bi-MNPs are investigated with EXAFS spectroscopy, SEM, and TEM. It is found that special structures associated with those catalysts are responsible for the increased catalytic abilities. Surface segregation or cluster segregation is important for those MNPs to remain highly catalytic throughout the growth period.

We also conclude that MNPs are highly non-uniform, especially when the yield of SWNT production is low, and bulk spectroscopic methods may not be applicable to studying those catalysts. Even for the Ni/Co system which can catalyze the growth of SWNTs at high yields, the sample still shows some nonuniformity. For example, the average size measured by EXAFS is 3 nm, although this value is different from that obtained from TEM data, which is 8 nm. This implies that many particles are too small to be visible to even high resolution TEM. Our findings suggest that the structure factor in multicomponent catalysts may play an important role in the growth mechanisms of SWNTs as the multifunctional effects in other established catalyst systems.

ACKNOWLEDGMENTS

I thank my graduate student J. Carter and G. Cheng and undergraduate student D. Masiel for their assistance. I am grateful for Professor P. Alivisatos, director of the Molecular Foundry, which is supported by the Director, Office of Science, Office of Basic Energy Sciences, Division of Materials Science and Engineering, of the U.S. Department of Energy under Contract No. DE-AC03-76SF00098. I thank Dr. V. Puntes for his experimental assistance in making metal nanoparticles. I also thank C. Echer at the National Center for Electron Microscope (NCEM), and the excellent staff at Stanford Synchrotron Radiation Laboratory (SSRL) for experimental support. I am also grateful to DOE for support of both facilities. Acknowledgement is made to the Donors of The Petroleum Research Fund, administrated by the American Chemical Society, for partial support of this research. This work is also partially supported by the Camille and Henry Dreyfus Foundation and National Science Foundation (Grant CHE-0135132).

REFERENCES

1. J. L. Rousset, Fjcs Aires, F. Bornette, M. Cattenot, M. Pellarin, L. Stievano and A. J. Renouprez, Characterization and reactivity of Pd-Pt bimetallic supported catalysts obtained by laser vaporization of bulk alloy, Appl. Surf. Sci. 164, 163-168 (2000).


2. G. A. Somorjai and K. McCrea, Roadmap for catalysis science in the 21st century: a personal view of building the future on past and present accomplishments, Appl. Cata. A-Gen. 222 (1-2), 3-18 (2001).

3. John H. Sinfelt, Bimetallic catalysts : discoveries, concepts, and applications (Wiley, New York, 1983). 4. R.T. Baker, Carbon 27, 315 (1989). 5. T. W. Hansen, J. B. Wagner, P. L. Hansen, S. Dahl, H. Topsoe and C. J. H. Jacobsen, Atomic-resolution in

situ transmission electron microscopy of a promoter of a heterogeneous catalyst, Science 294 (5546), 1508-1510 (2001).

6. S. Iijima and T. Ichihashi, Single-Shell Carbon Nanotubes of 1-Nm Diameter, Nature 364 (6439), 737-737 (1993).

7. D. S. Bethune, C. H. Kiang, M. S. Devries, G. Gorman, R. Savoy, J. Vazquez and R. Beyers, Cobalt-Catalysed Growth of Carbon Nanotubes With Single-Atomic-Layerwalls, Nature 363 (6430), 605-607 (1993).

8. T. Guo, P. Nikolaev, A. Thess, D. T. Colbert and R. E. Smalley, Catalytic Growth of Single-Walled Nanotubes By Laser Vaporization, Chem. Phys. Lett. 243 (1-2), 49-54 (1995).

9. C. H. Kiang, Growth of large-diameter single-walled carbon nanotubes, J. Phys. Chem. A 104 (11), 2454-2456 (2000).

10. L. Fiermans, R. De Gryse, G. De Doncker, P. A. Jacobs and J. A. Martens, Pd segregation to the surface of bimetallic Pt-Pd particles supported on H-beta zeolite evidenced with X-ray photoelectron spectroscopy and argon cation bombardment, J. Cata 193 (1), 108-114 (2000).

11. H. Yasuda, N. Matsubayashi, T. Sato and Y. Yoshimura, Confirmation of sulfur tolerance of bimetallic Pd-Pt supported on highly acidic USY zeolite by EXAFS, Catalysis Letters 54 (1-2), 23-27 (1998).

12. Y. Kumeda, Y. Fukuhiro, T. Taketsugu and T. Hirano, Theoretical study of nanotube growth in terms of frontier density distribution, Chem. Phys. Lett. 333 (1-2), 29-35 (2001).

13. C. F. Cornwell and L. T. Wille, Critical strain and catalytic growth of single-walled carbon nanotubes, J. Chem. Phys. 109 (2), 763-767 (1998).

14. M. J. Bronikowski, P. A. Willis, D. T. Colbert, K. A. Smith and R. E. Smalley, Gas-phase production of carbon single-walled nanotubes from carbon monoxide via the HiPco process: A parametric study, J. Vac. Sci. & Tech A-Vac Surf. and Films 19 (4 PT2), 1800-1805 (2001).

15. N. Kitamura and A. Oshiyama, Open edge growth mechanisms of single wall carbon nanotubes, Journal of the Physical Society of Japan 70 (7), 1995-2011 (2001).

16. J. Jiao and S. Seraphin, Single-walled tubes and encapsulated nanoparticles: comparison of structural properties of carbon nanoclusters prepared by three different methods, J. Phys. and Chem Sol. 61 (7), 1055-1067 (2000).

17. P. Nikolaev, M. J. Bronikowski, R. K. Bradley, F. Rohmund, D. T. Colbert, K. A. Smith and R. E. Smalley, Gas-phase catalytic growth of single-walled carbon nanotubes from carbon monoxide, Chem. Phys. Lett. 313 (1-2), 91-97 (1999).

18. F. Kokai, K. Takahashi, M. Yudasaka, R. Yamada, T. Ichihashi and S. Iijima, Growth dynamics of single-wall carbon nanotubes synthesized by CO2 laser vaporization, J. Phys. Chem. B 103 (21), 4346-4351 (1999).

19. J. H. Hafner, M. J. Bronikowski, B. R. Azamian, P. Nikolaev, A. G. Rinzler, D. T. Colbert, K. A. Smith and R. E. Smalley, Catalytic growth of single-wall carbon nanotubes from metal particles, Chem. Phys. Lett. 296 (1-2), 195-202 (1998).

20. Y. H. Lee, S. G. Kim and D. Tomanek, Catalytic growth of single-wall carbon nanotubes: An ab initio study, Phys. Rev. Lett. 78 (12), 2393-2396 (1997).

21. A. Maiti, C. J. Brabec and J. Bernholc, Kinetics of metal-catalyzed growth of single-walled carbon nanotubes, Phys. Rev. B-Cond. Mat 55 (10), R6097-R6100 (1997).

22. H. J. Dai, A. G. Rinzler, P. Nikolaev, A. Thess, D. T. Colbert and R. E. Smalley, Single-Wall Nanotubes Produced By Metal-Catalyzed Disproportionation of Carbon Monoxide, Chem. Phys. Lett. 260 (3-4), 471-475 (1996).

23. C. J. Brabec, A. Maiti, C. Roland and J. Bernholc, Growth of Carbon Nanotubes - a Molecular Dynamics Study, Chem. Phys. Lett. 236 (1-2), 150-155 (1995).

24. S. Seraphin and D. Zhou, Single-Walled Carbon Nanotubes Produced At High Yield By Mixed Catalysts, Appl. Phys. Lett. 64 (16), 2087-2089 (1994).

25. R. Sen, Y. Ohtsuka, T. Ishigaki, D. Kasuya, S. Suzuki, H. Kataura and Y. Achiba, Time period for the growth of single-wall carbon nanotubes in the laser ablation process: evidence from gas dynamic studies and time resolved imaging, Chem. Phys. Lett. 332 (5-6), 467-473 (2000).


26. M. Takizawa, S. Bandow, M. Yudasaka, Y. Ando, H. Shimoyama and S. Iijima, Change of tube diameter distribution of single-wall carbon nanotubes induced by changing the bimetallic ratio of Ni and Y catalysts, Chem. Phys. Lett. 326 (3-4), 351-357 (2000).

27. C. F. Cornwell and L. T. Wille, Proposed growth mechanism of single-walled carbon nanotubes, Chem. Phys. Lett. 278 (4-6), 262-266 (1997).

28. A. Maiti, C. J. Brabec, C. M. Roland and J. Bernhole, Growth Energetics of Carbon Nanotubes, Phys. Rev. Lett. 73 (18), 2468-2471 (1994).

29. J. C. Charlier, A. DeVita, X. Blase and R. Car, Microscopic growth mechanisms for carbon nanotubes, Science 275 (5300), 646-649 (1997).

30. V. F. Puntes, K. M. Krishnan and A. P. Alivisatos, Colloidal nanocrystal shape and size control: The case of cobalt, Science 291 (5511), 2115-2117 (2001).

31. C. B. Murray, S. H. Sun, H. Doyle and T. Betley, Monodisperse 3d transition-metal (Co, Ni, Fe) nanoparticles and their assembly into nanoparticle superlattices, Mrs Bulletin 26 (12), 985-991 (2001).

32. G. Cheng, J. Carter, D. Masiel and T. Guo, Synthesis of metal nanoparticles and carbon nanotubes, to be submitted (2003).

33. A. Thess, R. Lee, P. Nikolaev, H. J. Dai, P. Petit, J. Robert, C. H. Xu, Y. H. Lee, S. G. Kim et al., Crystalline Ropes of Metallic Carbon Nanotubes, Science 273 (5274), 483-487 (1996).

34. W. E. Alvarez, B. Kitiyanan, A. Borgna and D. E. Resasco, Synergism of Co and Mo in the catalytic production of single-wall carbon nanotubes by decomposition of CO, Carbon 39 (4), 547-558 (2001).

35. J. M. Thomas and G. Sanker, The Role of XAFS in the in situ and ex situ Elucidation of Active Sites in Designed Solid Catalysts, J. Synchrotron Rad. 8, 55 (2001).

36. A. L. Ankudinov, B. Ravel, J. J. Rehr and S. D. Conradson, Real-space multiple-scattering calculation and interpretation of x-ray-absorption near-edge structure, Phys. Rev. B-Cond. Mat 58 (12), 7565-7576 (1998).

37. H. Hosokawa, H. Fujiwara, K. Murakoshi, Y. Wada, S. Yanagida and M. Satoh, In-Situ Exafs Observation of the Surface Structure of Colloidal Cds Nanocrystallites in N,N-Dimethylformamide, J. phys. Chem. 100 (16), 6649-6656 (1996).

38. F.W. Lytle, D.E. Sayers and E.A. Stern, Reprot of the International Workshop on Standards and Criteria in X-ray Absorption Spectroscopy, Physica B 158, 701-722 (1989).

39. K. P. De Jong and J. W. Geus, Carbon nanofibers: Catalytic synthesis and applications [Review], Cata. Rev.-Sci. & Eng. 42 (4), 481-510 (2000).

40. M. S. Dresselhaus, G. Dresselhaus, Phaedon Avouris and LINK (Online service), Carbon nanotubes : synthesis, structure, properties, and applications (Springer, Berlin ; New York, 2001).

41. Y. Saito, M. Okuda, M. Tomita and T. Hayashi, Extrusion of Single-Wall Carbon Nanotubes Via Formation of Small Particles Condensed Near an Arc Evaporation Source, Chem. Phys. Lett. 236 (4-5), 419-426 (1995).

42. Y. Zhang, Y. Li, W. Kim, D. Wang and H. Dai, Imaging as-grown single-walled carbon nanotubes originated from isolated catalytic nanoparticles, Appl. Phys. A-Mat. Sci. & Proc. 74 (3), 325-328 (2002).

43. C. L. Cheung, A. Kurtz, H. Park and C. M. Lieber, Diameter-controlled synthesis of carbon nanotubes, J. Phys. Chem. B 106 (10), 2429-2433 (2002).

44. Y. Saito, M. Okuda, N. Fujimoto, T. Yoshikawa, M. Tomita and T. Hayashi, Single-Wall Carbon Nanotubes Growing Radially From Ni Fine Particles Formed By Arc Evaporation, Jap. Appl. Phys. P2-Lett. 33 (4A), L526-L529 (1994).

45. Y. Li, J. Liu, Y. Q. Wang and Z. L. Wang, Preparation of monodispersed Fe-Mo nanoparticles as the catalyst for CVD synthesis of carbon nanotubes, Chem. Mat. 13 (3), 1008-1014 (2001).

46. G. Cheng and T. Guo, Surface segregation in Ni/Co bimetallic nanoparticles produced in single-walled carbon nanotube synthesis, J. Phys. Chem. B 106 (23), 5833-5839 (2002).

47. ftp://xafs.chem.msu.su/data/Farrel_Lytle_data/, . 48. T. B. Massalski, H. Okamoto and ASM International, Binary alloy phase diagrams, 2nd ed. (ASM

International, Materials Park, Ohio, 1990). 49. W. E. Alvarez, F. Pompeo, J. E. Herrera, L. Balzano and D. E. Resasco, Characterization of single-walled

carbon nanotubes (SWNTs) produced by CO disproportionation on Co-Mo catalysts, Chem. Mat. 14 (4), 1853-1858 (2002).

50. S. Bandow, S. Asaka, Y. Saito, A. M. Rao, L. Grigorian, E. Richter and P. C. Eklund, Effect of the growth temperature on the diameter distribution and chirality of single-wall carbon nanotubes, Phys. Rev. Lett. 80 (17), 3779-3782 (1998).


51. M. Yudasaka, R. Yamada, N. Sensui, T. Wilkins, T. Ichihashi and S. Iijima, Mechanism of the effect of NiCo, Ni and Co catalysts on the yield of single-wall carbon nanotubes formed by pulsed Nd : YAG laser ablation, J. Phys. Chem. B 103 (30), 6224-6229 (1999).

52. CH. Kiang, W. A. Goddard III, R. Beyers, J. R. Salem and D. S. Bethune, Catalytic Effects of Heavy Metals on the Growth of Carbon Nanotubes and Nanoparticles, J. Phys. Chem. Solids 57 (1), 35 (1995).

53. M. Valden, X. Lai and D. W. Goodman, Onset of catalytic activity of gold clusters on titania with the appearance of nonmetallic properties, Science 281 (5383), 1647-1650 (1998).

54. D. B. Geohegan, H. Schittenhelm, X. Fan, S. J. Pennycook, A. A. Puretzky, M. A. Guillorn, D. A. Blom and D. C. Joy, Condensed phase growth of single-wall carbon nanotubes from laser annealed nanoparticulates, Appl. Phys. Lett. 78 (21), 3307-3309 (2001).

55. V. Derycke, R. Martel, M. Radosvljevic, F. M. R. Ross and P. Avouris, Catalyst-free growth of ordered single-walled carbon nanotube networks, Nano Letters 2 (10), 1043-1046 (2002).

56. R. E. Benfield, Mean Coordination Numbers and the Non-Metal Metal Transition in Clusters, J. Chem. Soc.-Farad. Trans. 88 (8), 1107-1110 (1992).

57. Vladimír Ponec and G. C. Bond, Catalysis by metals and alloys (Elsevier, Amsterdam ; New York, 1995). 58. C. Pan, F. Dassenoy, M. J. Casanove, K. Philippot, C. Amiens, P. Lecante, A. Mosset and B. Chaudret, A

new synthetic method toward bimetallic ruthenium platinum nanoparticles; composition induced structural changes, J. Phys. Chem. B 103 (46), 10098-10101 (1999).

59. H. J. Dai, J. Kong, C. W. Zhou, N. Franklin, T. Tombler, A. Cassell, S. S. Fan and M. Chapline, Controlled chemical routes to nanotube architectures, physics, and devices, J. Phys. Chem. B 103 (51), 11246-11255 (1999).

60. B. C. Liu, S. H. Tang, Z. L. Yu, B. L. Zhang, T. Chen and S. Y. Zhang, Catalytic growth of single-walled carbon nanotubes with a narrow distribution of diameters over Fe nanoparticles prepared in situ by the reduction of LaFeO3, Chem. Phys. Lett. 357 (3-4), 297-300 (2002).

61. B. Zheng, Y. Li and J. Liu, CVD synthesis and purification of single-walled carbon nanotubes on aerogel-supported catalyst, Applied Physics a (Materials Science Processing) 74 (3), 345-348 (2002).

62. S. H. Huh, H. K. Kim, J. W. Park and G. H. Lee, Surface segregation and surface composition in binary Fe-Mo nanoclusters, J. Kor. Phys. Soc. 38 (6), 791-793 (2001).

Documents

Multifunctional Catalysts for Singlewall Carbon Nanotube ...nanofast.ucdavis.edu/Publications/Ch7.pdf137 7 Multifunctional Catalysts for Singlewall Carbon Nanotube Synthesis T. Guo*