Effect of the Catalyst on Structural and Electrophysical ... - Springer Link

0 downloads 0 Views 588KB Size Report
INTRODUCTION. The problems of microminiaturization and inte gration of microelectronic sensors with the means of electronic framing within silicon technology ...

ISSN 20751133, Inorganic Materials: Applied Research, 2010, Vol. 1, No. 2, pp. 110–114. © Pleiades Publishing, Ltd., 2010. Original Russian Text © V.V. Bolotov, V.E. Kan, N.A. Davletkil’deev, I.V. Ponomareva, O.V. Krivozubov, A.V. Okotrub, A.G. Kudashov, 2009, published in Perspektivnye Materialy, 2009, No. 3, pp. 24–29.

Effect of the Catalyst on Structural and Electrophysical Characteristics of the Layers of NitrogenContaining Carbon Nanotubes Obtained by Gas Phase Synthesis V. V. Bolotova, V. E. Kana, N. A. Davletkil’deeva, I. V. Ponomarevaa, O. V. Krivozubova, A. V. Okotrubb, and A. G. Kudashovb a

Rzhanov Institute of Semiconductor Physics (Omsk Branch), Siberian Branch, Russian Academy of Sciences, pr. Mira 55a, Omsk, 644077 Russia b Nikolaev Institute of Inorganic Chemistry, Siberian Branch, Russian Academy of Sciences, pr. Akad. Lavrent’eva 3, Novosibirsk, 630090 Russia Received April 3, 2008

Abstract—The morphological, structural, and electrophysical properties of the layers of nitrogencontaining carbon nanotubes synthesized by pyrolysis of acetonitrile on a silicon oxide–silicon substrate with a Ni layer catalyst, Fe volumetric catalyst obtained by the thermal decomposition of ferrocene ((C5H5)2Fe), and their combination were studied. It was found that the type of catalyst affects the morphology, thickness, homoge neity, structure, and defect content of the obtained films and individual nanotubes. The electrophysical char acteristics of the carbon nanotubes are similar and have an activation dependence on temperature. Key words: carbon nanotubes, structural properties, electrical properties, scanning electron microscopy, Xray diffraction, scanning probe microscopy, Raman scattering. DOI: 10.1134/S2075113310020073

INTRODUCTION The problems of microminiaturization and inte gration of microelectronic sensors with the means of electronic framing within silicon technology pose problems of using nanostructured materials with a developed surface as sensitive media. Carbon nano tube (CNT) layers which can be formed in the cycles of silicon technology, in particular, gas phase synthesis on the insulating SiO2 layers, belong to such materials. It was shown earlier [1, 2] that the pyrolysis of ace tonitrile (CH3CN) with the use of Fe as a catalyst results in the formation of layers of CNT of the mixed, multi and singlewall, type oriented perpendicular to the substrate. The CNT layers contained phases with both semiconductor and metallic conductivity. To use CNTs as sensitive elements in chemical sen sors of the resistive type, an increased yield in the syn thesis of the semiconductor CNTs is desirable [3, 4]. Creating the electric contacts to the CNT layers is an important physicotechnological problem. In this respect, it is promising to obtain and study the proper ties of structures on the basis of CNT films synthesized on the metal layer playing the role of the conducting contact and catalyst. Thus, it is necessary to study the morphological, structural, and electrophysical prop erties of the obtained CNT layers in the fields of syn thesis with different (volumetric, film) catalysts and in the regions of the CNT layers in their multifunctional usage.

This study is aimed at studying the morphology, structure, and electrophysical characteristics of nitro gencontaining CNT structures obtained with the use of a film catalyst (Ni) and volumetric catalyst (Fe) and their combination. EXPERIMENTAL METHODS The CNT films were obtained by the pyrolysis of acetonitrile on SiO2/Si substrates with a size of 1 × 1 cm according to the method described in [2, 5, 6]. As catalysts, iron nanoparticles formed as a result of the thermal decomposition of ferrocene and nickel films 30 nm thick obtained by magnetron evaporation were used. The samples were put in the central zone of the reactor, the synthesis zone [6]. Table 1 gives the description of the samples and synthesis parameters. In each series, samples of the same type were synthe sized. The concentration of the nitrogen atoms in the CNT walls was 1–2% according to the optimal condi tions of the synthesis [1, 2, 5]. The morphology and structures of the CNT layers were studied by optical microscopy on a Neophot2 microscope and raster electron microscopy (REM) on a Philips SEM 515 microscope. The morphology of the individual CNTs was studied and their type of con ductivity was determined by the methods of atomic force microscopy (AFM) in the semicontact mode and scanning tunneling spectroscopy (STS) on a Solver PRO probe microscope.

110

EFFECT OF THE CATALYST ON STRUCTURAL AND ELECTROPHYSICAL

111

Table 1. Samples of the films of oriented CNTs obtained on the SiO2 /Si substrates with the use of Fe and Ni catalysts Sample series 1 2 3

Catalyst

Synthesis temperature, °C

Synthesis time, min

Thickness of the CNT films (REM and optical microscopy data, μm)

850 850

30 30

30 15–20

850

30

4–6

Fe (ferrocene) Ni/Fe (Nifilm 30 nm on a substrate, Fedecomposition of ferrocene) Nifilm 30 nm on a substrate

To perform AFM and STS studies, the nanotubes were preliminarily deposited on a mica substrate or a thin gold film was deposited on a SiO2/Si substrate. To this end, a deposit of nanotubes was put in ethanol and was exposed to ultrasound treatment for 20 min. The solution obtained was put on a substrate which was placed in an exsiccator to remove alcohol and mois ture. The outer diameter of the nanotubes was deter mined from the analysis of the AFM images. The vol tammetric characteristics (VACs) were scanned and measured in air at the relative humidity of 30%. A tungsten needle made by electrochemical etching was used as a probe. The normalized differential conduc tivity was calculated on the basis of VACs obtained by averaging the repeatedly measured tunneling VAC at an individual scanning point. The phase composition of the layers and phase parameters were studied on Shimadzu XRD6000 Xray (CuKα, 30 kV, 30 mA) and D8 Advance Bruker AXS (CuKα, 40 kV, 40 mA) diffractometers. With the same aim, the Raman light scattering (RLS) spectra were recorded on an RFS 100/S Fourier spectrometer at room temperature under excitation with a Nd : YAG laser with the wavelength of 1.064 μm in quasiinverse geometry (resolution of 1 cm–1). The conductivity of the CNT films was measured by the fourprobe method in the temperature range from 77 to 293 K.

neously distributed and chaotically oriented CNT threads. CNT threads visually observed are twisted strips composed of several CNTs and individual CNTs with a large outer diameter on the order of ≈300 nm (Fig. 2a). The study of the morphology of individual CNTs showed that the tubes in the sample of series 2 have mainly curved form with a large number of elbow joints (Fig. 2b). The dominating part of the tubes has an outer diameter from 100 to 300 nm. In addition to the multilayer CNTs, in the samples of series 2, there are a large number of graphite and amorphous carbon particles. No expressed growth orientation of CNTs in this case is apparently determined by the coupling of the additional volumetric catalyst (Fe) to the CNT surfaces growing on the surface of the metallic catalyst layer. (a)

10 μm

RESULTS AND DISCUSSION The study of the morphology of the film split obtained with the use of the Fe catalyst (samples of series 1) shows that CNTs grow mainly perpendicular to the substrate surface; the thickness of the CNT layer is homogeneous over the sample area and is 30 μm (Fig. 1a). CNTs are mainly grouped in twisted strips with dimensions of 300–400 nm; individual CNTs have dimensions of 30–40 nm (Fig. 1b). The most probable outer CNT diameter was determined by averaging the heights of the AFM images of the nano tubes over the substrate. In the film obtained with the use of the multifunc tional Ni/Fe catalyst (samples of series 2), CNTs have no clearly expressed growth direction with respect to the substrate surface; the thickness of the CNT film is inhomogeneous over the sample area and is from 15 to 20 μm. The film is a friable structure with inhomoge INORGANIC MATERIALS: APPLIED RESEARCH

Vol. 1

(b)

nm 400 300 200 100

0

100

200

300

400 nm

Fig. 1. (a) REM image of the CNT layer and (b) AFM image of the CNT twisted strips for the samples of series 1. No. 2

2010

112

BOLOTOV et al. (a)

Intensity, arb. units 250 200 150 100 50 0 160

(a)

(b)

120 20 μm

*

80 40

(b) 0

2

μm 4

0 nm 119

10

20

30 40 2θ, deg

50

60

Fig. 3. Xray diffractograms of the CNT layers for the sam ples of the series 1 (a) and 2 (b).

2

4 0 Fig. 2. (a) Image of the surface of the CNT film for the samples of series 2 obtained in an optical microscope; (b) form of the individual nanotube from the samples of series 2.

In the samples of series 3 obtained with the use of the Ni catalyst, the CNT growth is similar to that in the samples of series 2; however, unlike the samples of series 2, CNT threads are spaced more widely and have mainly rectangular form with a small amount of curvatures. The outer CNT diameter is also from 100 to 300 nm. Figure 3 shows typical diffractograms of the CNT layers for the sample of series 1 and 2. For the samples of series 1 (Fig. 3a), the analysis of the phase compo sition showed the presence of the following different structural carbon phases in the CNT layers: graphite, 34%; nanotubes and bulbs, 11%). Phases related to the catalyst constitute about 11%; the fraction of the amorphous carbon phase is about 44%. The low rela tive content of nanotubes and bulbs is apparently due to the preferentially perpendicular location of the nanotubes with respect to the substrate and the corre sponding small contribution of the CNT walls to the intensity of the Xray diffraction. For the samples of series 2 (Fig. 3b), the structural carbon phases are rep resented by graphite (about 6%) and nanotubes and bulbs (about 13%). A large amount of the amorphous carbon phase (51%) is present in the samples as well. Unlike the samples of series 1, in the samples of the

series 2, a noticeable amount of Fe catalyst particles as carbides and oxides (about 30%) is present. In the dif fractograms of the samples of series 2, a reflex from the substrate was observed (denoted with an asterisk in Fig. 3b). The large CNT contribution (compared to graphite) to the intensity of the Xray diffraction is most probably due to the fact that, in the samples of series 2, CNTs did not have a clearly expressed growth direction with respect to the substrate. The phase composition with respect to carbon for the samples of series 3 is similar to that for the samples of series 2. In the RLS spectra of the samples (Fig. 4) obtained in the presence of ferrocene (samples of series 1 and 2), radial breathing modes (RBMs) in the range from 180 to 287 cm–1 corresponding to singlewall CNTs with a diameter of 0.82–1.34 nm [7, 8] were discov ered. The CNT dimensions determined from the RLS spectra were averaged over the diameter of the scan ning spot with the area of 7.85 × 10–3 mm–3. In the sample spectra obtained in the presence of only Ni, no RBMs were discovered. In the RLS spectra of the samples of series 1, a band at 1591 cm–1 (Gband) typical of singlewall CNTs and the Dband (1250–1350 cm–1) [9] decomposable into three components (Table 2) were observed. The position of the Dband at 1267 and 1280 cm–1 are attributed to the CNT structure, and the band at 1315 cm–1 is closer to the amorphous phase. The Dband has a lower intensity than that of the Gband. This indicates that, in this sample the CNT structure is more highly ordered. In the RLS spectra of the samples of series 2 obtained with the use of the multifunctional Ni/Fe catalyst, the Dband has a larger halfwidth and inten sity, indicating the larger content of the disordered microcrystalline graphite in the sample. The samples of series 3 grown without the presence of iron have the

INORGANIC MATERIALS: APPLIED RESEARCH

Vol. 1

No. 2

2010

EFFECT OF THE CATALYST ON STRUCTURAL AND ELECTROPHYSICAL Intensity, rel. units 0.05 0.04 0.03 0.02 0.01 0

113

(dI/dU)/(I/U), rel. units 3.0

D Ni

G

2.5 2.0 1.5 1.0

G

0.006 0.004 RBM

0.5 Fe

D

0

−0.5

0.002

0

0.5

U, V

Fig. 5. Spectrum of the normalized differential conductiv ity of an individual CNT.

0 0.004

D

G

0.003 RBM

lower yield of microcrystalline graphite compared to the samples synthesized with the use of Ni/Fe; the lat ter had a higher defect structure (cf. the ratio of the intensities of the D and Gbands in the RLS spec trum in Fig. 4 and data of Table 2). RLS data agree well with REM, AFM, and Xray diffraction data.

Ni/Fe

0.002 0.001 0 500

1000 1500 RLS shift, cm−1

2000

STS data for all samples show that the type of spec trum of the differential conductivity of individual CNTs is characteristic of nanotubes with the metallic type of conductivity. It should be noted that the tun neling VAC could be measured for sufficiently long nanotubes with the diameter not less than 60 nm. The tubes with larger diameters in most cases could not be scanned since they moved on the substrate during scanning. Figure 5 shows the characteristic depen dence of the normalized differential conductivity of an individual CNT on the applied voltage between the nanotube and the STS probe. The strong modulation of the spectrum of the differential conductivity is due to the noise in the tunneling nanotube–probe contact.

Fig. 4. RLS spectra of the CNT films obtained from differ ent catalysts.

Gband at 1586 cm–1 and Dband at 1318 cm–1. The intensity of the Dband is larger than that of the Gband, indicating the larger content of amorphous phase in the samples. Thus, when studying the structural characteristics of the CNT films, from the RLS, it was found that the catalyst affected the defect content of the structure: samples synthesized with the use of only Fe had a

Table 2. Position and nature of the main RLS peaks of the CNT films Samples 1 (Fecatalyst) RLS shift, halfwidth cm–1 of peak, cm–1 180 200 253 259 276 287 1267 1282 1315 1592

10 8 10 9 5 7 22 20 32 28

Samples 2 (Ni/Fecatalyst)

Samples 3 (Nicatalyst)

nature of peak

RLS shift, cm–1

halfwidth of peak, cm–1

nature of peak

RLS shift, cm–1

halfwidth of peak, cm–1

nature of peak

RBM RBM RBM RBM RBM RBM D1 D2 D3 G

253 262 277 287 1280 1330 1590

7 9 7 5 48 89 52

RBM RBM RBM RBM D1 D2 G

1318 1586

147 60

D G

INORGANIC MATERIALS: APPLIED RESEARCH

Vol. 1

No. 2

2010

114

BOLOTOV et al.

to CNTs obtained in the second and the third manner. CNTs grown on a thin Ni layer on the SiO2/Si sub strate with the participation of Fe and without it had the dimension of 100–300 nm and no clearly expressed growth directions with respect to the sub strate surface, and the film thickness was inhomoge neous over the area. In the study of the temperature dependences of the conductivity of the CNT bulk, it was found that the dependence as a whole has an acti vation nature.

lnσ −0.24

EA = 6.6 meV

−0.28 −0.32 −0.36 −0.40 40

EA = 1.7 meV

REFERENCES 50

60 70 1/kT, eV−1

80

90

Fig. 6. Temperature dependence of the conductivity of the CNT layers for the samples of series 2.

Figure 6 shows the temperature dependence of the conductivity of the CNT layers obtained from the samples of series 2 in the temperature range from 100 to 300 K. It is seen that the conductivity increases with the increase in the temperature. One can distinguish (Fig. 6) two regions with the activation energy of 1.7 and 6.6 meV. Thus, the films which have a mixed com position with respect to CNT (with semiconductor and metallic conductivity, graphite and amorphous inclusions) have an activation nature of the tempera ture dependence of the conductivity. Two different mechanisms of the conductivity can correspond to two activation energies (lower and higher) for the regions of low (100–200 K) and high temperatures (200– 300 K), respectively. As a supposition, one can attribute the values of the activation energy to the bar riers between individual nanotubes at the transverse transport of the charge carriers as well as to the pres ence of two main carbon phases in the synthesized films (except for nanotubes, amorphous carbon is present). Thus, in spite of the individual CNTs having metallic conductivity, the conductivity of the CNT layer as a whole has an activation nature. CONCLUSIONS Films of nitrogencontaining CNTs synthesized in different ways were studied: (1) with the use of a volu metric catalyst (Fe) on SiO2/Si; (2) with the use of a mixed catalyst, film catalyst (Ni layer) and volumetric catalyst (Fe), on SiO2/Si; (3) with the use of a film cat alyst (Ni layer) on SiO2/Si. The comparative analysis of the morphology and structures these CNT films by REM, AFM, STS, Xray diffraction, and RLS meth ods showed the following: films synthesized in the first manner had oriented growth and homogeneous thick ness, the dimension of individual CNTs was 30– 40 nm, and CNTs had less defect structure compared

1. Bolotov, V.V., Kan, V.E., Ponomareva, I.V., Krivo zubov, O.V., Davletkil’deev, N.A., Sten’kin, Yu.A., Kudashov, A.G., Danilevich, V.S., and Okotrub, A.V., Structural Characteristics for the Films Consisting of Aligned Carbon Nanopipes Generated by Gas–Core Synthesis, Perspekt. Mater., 2007, no. 1, pp. 5–11. 2. Kudashov, A.G., Okotrub, A.V., Yudanov, N.F., Romanenko, A.I., Bulusheva, L.G., Abrosimov, A.G., Chuvilin, A.L., Pazhetov, E.M., and Boronin, A.I., Gas–Core Synthesis of Nitrogen–Containing Carbon Nanopipes and Their Electronic Properties, Fiz. Tverd. Tela, 2002, vol. 44, issue 4, pp. 626–629. 3. Kong, J., Franklin, N.R., Zhou, C., Chaplin, M.G., Peng, S., Cho, K., and Dai, H., Nanotube Molecular Wires as Chemical Sensors, Science, 2000, vol. 287, pp. 622–625. 4. Suehiro, J., Zhou, G., and Hara, M., Fabrication of a Carbon Nanotube–Based Gas Sensor Using Dielec trophoresis and Its Application for Ammonia Detec tion by Impedance Spectroscopy, J. Phys. D: Appl. Phys., 2003, vol. 36, pp. L109–L114. 5. Kudasov, A.G., Okotrub, A.V., Bulusheva, L.G., Asanov, I.P., Shubin, Yu.V., Yudanov, N.F., Yudanova, L.I., Danilovich, V.S., and Abrosimov, O.G., Influence of Ni–Co Catalyst Composition on Nitrogen Content in Carbon Nanotubes, J. Phys. Chem. B, 2004, vol. 108, pp. 9048–9053. 6. Kudashov, A.G., Kurenya, A.G., Okotrub, A.V., Gusel’nikov, A.V., Danilovich, V.S., and Bulusheva, L.G., Synthesis and Structure of the Films Consisting of Carbon Nanopipes Aligned Normally to the Substrate, Zh. Tekhn. Fiz., 2007, vol. 77, issue 12, pp. 96–100. 7. Sauvajol, J.L., Angalaert, E., Rols, S., and Alvarez, L., Photons in Single Wall Carbon Nanotube Bundles, Carbon, 2002, vol. 40, pp. 1697–1714. 8. Lafi, L., Cossement, D., and Chahine, R., Raman Spectroscopy and Nitrogen Vapor Adsorption for the Study of Structural Changes during Purification of Sin gle–Wall Carbon Nanotubes, Carbon, 2005, vol. 43, pp. 1347–1357. 9. Rao, A.M., Richter, E., Bandow, S., Chase, B., Eklund, P.C., Williams, K.A., Fang, S., Sub baswamy, K.R., Menon, M., Thess, A., Smalley, R.E., Dresselhaus, G., and Dresselhaus, M.S., Diameter– Selective Raman Scattering from Vibrational Modes in Carbon Nanotubes, Science, 1997, vol. 75, pp. 187–191.

INORGANIC MATERIALS: APPLIED RESEARCH

Vol. 1

No. 2

2010

Suggest Documents