The formation of stacked-cup carbon nanotubes using chemical vapor deposition from ethanol over silica

CARBON

4 8 ( 2 0 1 0 ) 3 1 7 5 –3 1 8 1

available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/carbon

The formation of stacked-cup carbon nanotubes using chemical vapor deposition from ethanol over silica Alicja Bachmatiuk a,*, Felix Bo¨rrnert a, Franziska Scha¨ffel b, Mujtaba Zaka b, Grazyna Simha Martynkowa c, Daniela Placha c, Ronny Scho¨nfelder a, Pedro M.F.J. Costa a,d, Nicholas Ioannides e, Jamie H. Warner b, Ru¨diger Klingeler a, Bernd Bu¨chner a, Mark H. Ru¨mmeli a,f a

IFW Dresden, P.O. Box 270116, 01069 Dresden, Germany Department of Materials, University of Oxford, Parks Road, Oxford OX1 3PH, United Kingdom c Nanotechnology Center, VSB Technical University of Ostrava, 17. listopadu 15, 708 33 Ostrava – Poruba, Czech Republic d CICECO, University of Aveiro, 3810-193 Aveiro, Portugal e London Metropolitan University, 166-220 Holloway Road, London N7 8DB, United Kingdom f Technische Universita¨t Dresden, 01062 Dresden, Germany b

A R T I C L E I N F O

A B S T R A C T

Article history:

Processes involved in using SiO2 particles as catalysts for stacked-cup carbon nanotube for-

Received 17 March 2010

mation in a spray pyrolysis chemical deposition method from ethanol were investigated. In

Accepted 30 April 2010

addition, the recyclability of the SiO2 substrates is investigated. The SiO2 particles are

Available online 10 May 2010

shown to reduce to SiC. Moreover, the addition of triethylborate to produce boron species, and extended reactions, through recycling, leads to higher yields by improving the availability of SiC species. The formation of the carbon nanostructures is best explained through a carbon dissolution mechanism. Ó 2010 Elsevier Ltd. All rights reserved.

1.

Introduction

There are numerous chemical vapor deposition methods to produce carbon nanotubes using metal catalysts [1–8] which remain as undesired impurities after synthesis. These can be removed post-synthesis through aggressive chemical treatments, e.g. by acids [8,9]. However, purification usually does not fully eliminate metal particles and introduces defects to the carbon nanotubes. Metal free carbon nanostructures are desirable materials for wide potential applications in composites, drug delivery, electronic circuits, especially for the silicon industry. The general requirement for the silicon industry for metal free carbon nanotubes is well known. Metals reduce chip lifetime because they react unfavourably with many materials found in circuits. Hence, the use of non-metallic catalysts is desirable for silicon compatibility

(and also composites). Recently various investigations have successfully implemented ceramic catalyst particles, for example, SiO2, ZrO2, SiC, MgO or Al2O3 [10–17]. The use of SiO2 as a catalyst is particularly attractive for integration into Si based technology and its use as a carbon nanotube catalyst has been reported by various groups including, Liu et al. [12,18], Huang et al. [15], and Bachmatiuk et al. [19]. Common to all these studies is the need to pre-treat the substrates before synthesis, e.g. by scratching the surface, or by pre-heating the substrates in air or hydrogen, or through chemical treatment. In this article we present studies investigating metal free carbon nanostructure synthesis from SiC particles formed through the carbothermal reduction of quartz. Dense mats of carbon nanotubes were formed directly on the surface of quartz substrates at 900 °C. We also investigated the

* Corresponding author: Fax: +49 3514 659313. E-mail address: [email protected] (A. Bachmatiuk). 0008-6223/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.carbon.2010.04.055

3176

CARBON

4 8 ( 2 0 1 0 ) 3 1 7 5 –3 1 8 1

recyclability of the quartz substrates to gain insight into the carbothermal reduction process.

2.

Experimental

Aerosol assisted chemical vapor deposition (AACVD) experiments were performed in a horizontal tube furnace, in which carbon nanomaterials were grown on quartz substrates. The purity of the utilized substrates was checked by energy dispersive X-ray spectroscopy (EDX) which shows that no metal impurities were involved in the syntheses. Only signals from Si, O and C along with Cu (from the TEM grid) were detected. Both fresh (unused) and recycled (used) substrates were explored. The carbon feedstock (ethanol or an ethanol/triethylborate mixture) was mixed with argon and carried to the reaction center of the oven with a flow rate of 7200 ml/min. The temperature of the processes was 900 °C and the reaction time was 30 min. The obtained products were characterized using various techniques. The morphology was investigated using aberration corrected transmission electron microscopy (TEM) on a FEI Titan3 operated at 80 kV and a scanning electron microscope (SEM) on a FEI Nova-Nanosem. The electronic properties were studied using IR optical absorption spectroscopy on a Bruker 113 Fourier transform spectrometer, and Raman spectra collected using a Thermo Scientific DXR SmartRaman spectrometer (excitation laser k = 532 nm). Conductance measurements of individual stacked-cup carbon nanotubes were conducted in a FEI Tecnai T20 TEM with a scanning tunnelling microscopy (STM) holder (Nanofactory).

3.

Results and discussion

Fig. 1 presents scanning electron microscopy (SEM) images of the materials produced on fresh quartz substrates (A) and (B), using as a carbon feedstock ethanol and ethanol/triethylborate mixture, respectively. In the case of ethanol as the carbon feedstock with a fresh substrate, the surface of the quartz substrate after the reaction was mostly covered with an amorphous carbon layer (e.g. Fig. 1A), with only a few carbon nanofibres being observable in TEM studies as shown in Fig. 1A (inset). When the ethanol/triethylborate mixture was used as a carbon source, dense mats of carbon nanofibres are obtained (e.g. Fig. 1B). Hence, the addition of triethylborate to ethanol during CVD experiments enhances the reaction. In part this can be attributed to a lowering of the quartz melting temperature [20]. The potential to recycle the substrates was also investigated. Prior to reusing a substrate they were first burnt in air (open air oven 800 °C, 15 min.), washed in hydrochloric acid (5 M) and then thoroughly rinsed in distilled water. After the reaction, visual inspection of the samples shows an increase in coverage of black soot-like material on the support by about 50% (±10%). SEM studies confirm this material consists of dense mats of carbon nanotubes/fibres were formed. Fig. 1C and D present examples of material synthesized on reused quartz substrates after ethanol and ethanol/triethylborate experiments, respectively. In the recycling experiments ethanol served as the feedstock. In addition to obtaining an enhanced surface area coverage, the black material on the surface

appears darker. Hence we conducted transmitivity studies using a white light source (3000–12,000 cm 1) to evaluate if the nanotube/fibre film forming on the surface was denser after recycling. The transmission percentage is determined from the transmission intensity of the sample against that from a blank pristine quartz substrate. The data are provided in Table 1 and show a transmission reduction of 9% (±6%) for the substrate recycled after an initial reaction using ethanol as the feedstock. This value is clearly a massive under-estimate of the true change in fibre formation since the initial the material consisted of amorphous carbon and only a few fibres, whereas after recycling only dense mats of fibres are obtained. In the case of a recycled substrate after an initial reaction using a ethanol/triethylborate feedstock the reduction in transmission was 18% (±8%). We also checked that the cleaning treatment before using a recycled treatment does not influence the CNT yield. To do this we applied the same air burn and acid treatment (see above) as used with recycled substrates to a pristine substrate and ran the CVD reaction with ethanol as the feedstock. The produced sample consisted of amorphous carbon and only a few fibrous structures, exactly the same as is obtained without pre-treatment. This result, along with the reduced transmission and increased material coverage on the supports after recycling, confirm recycled substrates are more efficient at producing the fibrous structures. The obtained carbon nanostructures were studied in greater detail using transmission electron microscopy. Fig. 2 highlights typical structures found in the samples. In panel A, an overview image shows the type of structures which dominate the samples. These fibre/tubular structures comprise stacked hemispherical structures. Often the hemispherical caps are incomplete at their poles and so when stacked provide a hollow core. The structures, to some degree, resemble bamboo like carbon nanotubes. The inter-layer spacing of these layers is around 0.35 nm, typically that found between graphene layers in graphite. In the case of recycled substrates dense mats of fibres are obtained although differences could be observed. The recycled sample, obtained from a substrate initially used with an ethanol carbon source, yielded similar structures to those formed using fresh substrates. However, when using a recycled substrate, initially used with an ethanol/triethylborate feedstock mixture, many of the nanofibres did not show clearly defined stacked caps, but appear to be much more disordered. This is easily observed by comparing the left and right panels in Fig. 2B (this is also reflected in Raman spectroscopy studies through their relatively poor G/D ratio and is discussed later). In all samples, Y junctions are found on the fibres. Many of these Y junctions contain particles at or near the junction within the fibre, as illustrated in panel C of Fig. 2. Such encapsulated particles were more frequently observed in the recycled samples, in particular the recycled sample initially used with an ethanol/triethylborate feedstock mixture. At times similar particles were found at the roots of the fibre/tube structures. Energy dispersive X-ray spectroscopy (EDX), for example panel D, and diffraction data obtained from the fast Fourier transform of the images indicate these particles are SiC [19]. No SiO2 particles within the structures were ever observed. This suggests SiO2 from

CARBON

3177

4 8 ( 20 1 0 ) 3 1 7 5–31 8 1

Fig. 1 – Scanning electron micrographs providing overviews of the different samples: (A) sample produced on a fresh quartz substrate using ethanol as the carbon source, inset: TEM image confirming the presence of carbon nanofibres. (B) Sample produced on a fresh quartz substrate using an ethanol/triethylborate mixture as the feedstock. (C) and (D) Samples produced on recycled quartz substrates with ethanol as the feedstock after initially using ethanol and an ethanol/triethylborate mixture source, respectively.

Table 1 – Data showing the transmission percentage of the supports after each reaction relative to pristine quartz support. The reduction in transmission of recycled samples as a ratio to non-recycled supports (recycling gain) is also provided. Run 1 2 3 4

Feedstock Ethanol Ethanol Ethanol/tri-ethylborate Ethanol

Support

Transmission (%)

Error (%)

Fresh quartz Recy. quartz (after run 1) Fresh quartz Recy. quartz (after run 3)

87 78 62 44

3 3 3 5

the substrate reduces to SiC during the reaction. The TEM studies also provide the opportunity to derive statistical data on the fibre/tube diameters. The diameter distribution of the samples produced with fresh substrates using ethanol as the feedstock was in the range 35–45 nm and with ethanol/triethylborate, 30–55 nm for the samples produced with reused substrates. The diameter distributions were 40–55 nm for the samples produced using ethanol, and 15–70 nm for the samples formed on reused substrates using ethanol/triethylborate. This data shows that increased diameters of the fibres are obtained when using recycled substrates. In order to better comprehend the processes occurring in the reaction and the presence of SiC, spectroscopic studies were conducted. Fig. 3 presents the IR optical absorption spectra for the various samples. In panel A the samples prepared in ethanol are presented whilst in panel B the samples in which an ethanol/triethylborate feedstock mixture was used are shown. In all cases the spectra show peaks corresponding to SiO2 (between

Recycling gain (%)

Error (%)

9

6

18

8

1000 and 1300 cm 1) and SiC (between 670 and 900 cm 1) [21]. This confirms the TEM data indicating SiC in the sample. The width and structure of the peaks vary between the samples. In the case of samples formed using pure ethanol as the C source (panel A) the SiC responses are relatively broad, in particular when compared to the peaks in panel B. This might be attributed to finite size effects which would red-shift the Si–C stretching response. However, this is unlikely as the particle size data obtained from the TEM investigations show the greatest size distributions and the smallest particle sizes are obtained when using the ethanol/triethylborate feedstock mixture. These samples (panel B) exhibit narrower SiC peaks. This indicates finite size effects are not responsible for the broader SiC peaks. Studies by Sun and Miyasato [22] show the width of the peak increases with increasing amorphous SiC content. In addition, the intensity of the SiC peak decreases with increasing amorphous content. Between 1000 and 1300 cm 1 lie peaks corresponding to SiO2 which arise due to various infrared active longitudinal

3178

CARBON

4 8 ( 2 0 1 0 ) 3 1 7 5 –3 1 8 1

Fig. 2 – (A) Overview of typical fibres found in all samples except when using a recycled support initially formed using the ethanol/triethylborate mix. (B) Comparison of a recycled nanofibre formed from a recycled reaction (left panel – first feedstock ethanol/triethylborate followed by a pure ethanol reaction) and right panel – a fibre form from a single ethanol reaction. (C) Typical nanofibre/tube Y junction with SiC particle embedded at the junction. (D) Typical EDX spectrum of nanoparticles in the structures. n.b. the Cu signals are from the Cu TEM grid. Inset: high angular annular dark field (HAADF) image of the investigated sample (see panel C). The square highlights the region from which the EDX spectrum was collected.

Fig. 3 – Optical absorption spectra of the samples: (A) solid line – sample obtained from a fresh quartz substrate, dashed line – sample from a recycled substrate after ethanol experiment. The feedstock in both experiments was ethanol. (B) Dotted line – sample from a fresh quartz substrate using ethanol/triethylborate feed source, dashed–dotted line –– sample obtained using ethanol on a recycled substrate after a CVD reaction with ethanol/triethylborate. and transverse optical pairs (LO and TO, respectively) formed via Si–O asymmetric stretching [23]. The sharp peaks at 1380 cm 1 also correspond to SiO2. The relative intensity of the various bands varies between the samples and this, in part, can be attributed to differences in the heating treatment, namely, recycling (extended heating) and in the case of an ethanol/triethylborate feedstock,

boron lowering melting points. This can lead to LO–TO splitting [21]. In addition, a Si–O–C peak at 1080 cm 1 is visible [24], which even dominates in the samples formed using the ethanol/triethylborate mix. The Si–O–C peak is stronger in both samples formed on recycled substrates. This pattern correlates nicely with the SEM data, and suggests that the amount of carbon nanofibres/tubes formed is related to the

CARBON

4 8 ( 20 1 0 ) 3 1 7 5–31 8 1

3179

Fig. 4 – (A) Raman spectra showing presence of SiO2 and SiC and sp2 carbon (G and D modes) from the various samples. Solid line – on a fresh quartz substrate using ethanol, dashed line – using ethanol on a recycled substrate after an ethanol reaction, dotted line – on a fresh quartz substrate using ethanol/triethylborate, dashed–dotted line –– using ethanol on a recycled substrate after ethanol/triethylborate experiment. (B) G to D (area) ratios for the samples produced with ethanol (fresh and recycled) (left panel), solid line – on a fresh quartz substrate with a mixture of ethanol/triethylborate and then recycled using an ethanol feedstock (right panel).

efficiency of the SiO2 transition to Si–O–C or SiC. The TEM studies indicate SiC is the more likely candidate. In panel B of Fig. 3, a weak peak to the right of the SiC peak 640 cm 1 is visible. This corresponds to Si–O–B units. Si–O–B responses also occur around 900 cm 1 [25,26]. We also used Raman spectroscopy to further investigate the samples as it can usefully complement the IR spectroscopy. Fig. 5A shows clear peaks related to SiO2 (between 200 and 800 cm 1) and SiC (between 750 and 1000 cm 1) in the spectra for all analyzed samples, confirming the IR spectroscopic data [21]. Raman spectroscopy is also a powerful technique to detect sp2-hybridized carbon signals which are clearly visible around 1600 cm 1 through the so-called G modes and the defect induced (D) band at 1350 cm 1 [27,28]. Both these modes are present in all samples and confirm the presence of sp2 carbon. In addition, the ratio of the G/D modes is often used as a measure of sp2 carbon nanostructure crystal purity of the samples [27]. Fig. 4B presents a comparison of the G/D ratios for the samples produced only with ethanol (left panel) and with the ethanol/triethylborate mixture (right panel). The highest G/D ratio was obtained from the sample using ethanol with a fresh substrate. The next best G/D ratios occur for the

samples produced on a recycled (ethanol) quartz substrate, and fresh substrate with an ethanol/triethylborate mixture. The sample produced on a recycled substrate (initial use ethanol/triethylborate feedstock mixture) showed a significant degradation in the G/D ratio. This is G/D is in agreement with the TEM observations showing rather disordered structures (i.e. defective graphitic structures). We also investigated the electrical properties of various individual tubes in TEM using an electrical probing STM holder. Fig. 5A shows a typical example of a contacted tube. Panel B provides the I–V characteristics for the tube shown in panel A. The curve shows a linear dependence between 500 and 500 mV after which the curve departs from the linear behaviour. This same shape of the I–V curve was found for all examined tubes. The electrical resistance (obtained from the low voltage linear section) was found to vary between 0.5 and 1 MX (linear resistance: 0.3 and 0.5 MX/lm). These are similar to values found for multiwalled carbon nanotubes (MWCNT) [29]. In MWCNT the conduction path occurs along the tube walls, however in our tubes the stacked structure might at first suggest no such conduction path exists. Close examination of the structure in TEM (e.g. Fig. 5C) shows the edges of

Fig. 5 – (A) TEM image of a contacted stacked-cup carbon nanotube in TEM in a STM holder during I–V measurements arrows: 1 – tube attached to the rim, 2 – carbon nanotube, 3 – nanotube attached to the tungsten STM tip. (B) Characteristic I–V curves (forwards and reverse) from the contacted tube in panel A (C) high resolution TEM image of the stacked-cup carbon nanotube showing the inter-joining of graphitic layers at their edges.

3180

CARBON

4 8 ( 2 0 1 0 ) 3 1 7 5 –3 1 8 1

the stacked graphitic layers are for the most part joined. This joining of the graphitic layers provides a conduction pathway and can account for the linear section in the I–V curves. At higher voltages where the curve is no longer linear, a reduction in resistance is observed. This is due to an enhanced contribution to conduction by inter-layer tunnelling [30]. The various data clearly show the formation of SiC in the reaction. Furthermore, transition species, like Si–O–C are also found. This points to SiO2 undergoing a carbothermal reduction process. Once the SiC forms, we argue the excess carbon species supersaturate the SiC particles, which eventually precipitate C forming hemispherical graphitic caps. The use of triethylborate provides boron species which reduce melting points and also help reduce SiO2 species; Si–O–B units are observed, which, act as intermediaries for SiC formation. Hence boron accelerates SiC formation which in turn increases the carbon nanofibre production. The I–V behaviour of the structures is consistent with metal-type electrical conduction, possibly based on diffusive conduction via the edge links between stacked graphitic layers.

4.

Conclusions

We have shown that it is possible to synthesize carbon nanotubes using a non-metal catalyst, SiO2, in an aerosol assisted chemical vapor deposition process. The SiO2 particles are reduced to SiC via a carbothermal reduction process. This process can be accelerated by the introduction of triethylborate to the ethanol feedstock. This acceleration process occurs through the formation of intermediary species (Si–O–B) which are easily transformed to SiC. The recycling of the substrates is feasible and yields higher carbon nanostructure yields. Again this is due to greater SiC formation. Further studies are required to more fully comprehend the mechanisms behind this process.

[6]

[7]

[8]

[9]

[10]

[11] [12]

[13]

[14]

[15]

[16]

[17]

Acknowledgments The research was supported in part by the European Network CARBIO, Contract MRTN-CT-2006-035616. FS acknowledges funding from the Alexander von Humboldt Foundation.

[18]

[19] R E F E R E N C E S

[20] [1] Lin M, Tan JPY, Boothroyd C, Loh KP, Took ES, Foo YL. Dynamical observation of bamboo-like carbon nanotube growth. Nano Lett 2007;7(8):2234–8. [2] Baker T. Catalytic gasification of graphite. Chem Ind (Lond) 1982;18:698–702. [3] Yoshida H, Takeda S, Uchiyama T, Kohno H, Homma Y. Atomic-scale in situ observation of carbon nanotube growth from solid state iron carbide nanoparticles. Nano Lett 2008;8(7):2082–6. [4] Costa S, Borowiak-Palen E, Bachmatiuk A, Ru¨mmeli MH, Gemming T, Kalenczuk RJ. Filling of carbon nanotubes for bio-applications. Phys Status Solidi B 2007;244:4315–8. [5] Bachmatiuk A, Borowiak-Palen E, Ru¨mmeli MH, Gemming T, Kalenczuk RJ. Influence of the substrate loading on the

[21]

[22]

[23]

[24]

quality and diameter distribution of SWCNT in alcohol-CVD. Phys Status Solidi B 2007;244:3925–9. Audier M, Oberlin A, Oberlin M, Coulon M, Bonnetain L. Morphology and crystalline order in catalytic carbons. Carbon 1981;19:217–24. Carneiro OC, Kim MS, Yim JB, Rodriguez NM, Baker RTK. Growth of graphite nanofibers from the iron–copper catalyzed decomposition of CO/H2 mixtures. J Phys Chem B 2004;107:4237–44. Ru¨mmeli MH, Scha¨ffel F, Bachmatiuk A, Adebimpe D, Trotter G, Bo¨rrnert F, et al. Investigating the outskirts of Fe and Co catalyst particles in alumina-supported catalytic CVD carbon nanotube growth. ACS Nano 2010;4:1146–52. Bachmatiuk A, Borowiak-Palen E, Ru¨mmeli MH, Kramberger C, Hu¨bers H-W, Gemming T, et al. Facilitating the CVD synthesis of seamless double-walled carbon nanotubes. Nanotechnol 2007;18:275610 (5pp). Takagi D, Hibino H, Suzuki S, Kobayashi Y, Homma Y. Mechanism of gold-catalyzed carbon material growth. Nano Lett 2007;7(8):2272–5. Takagi D, Kobayashi Y, Homma Y. Carbon nanotube growth from diamond. JACS 2009;131(20):6922–3. Liu B, Ren W, Gao L, Li S, Pei S, Liu C, et al. Metal-catalyst-free growth of single-walled carbon nanotubes. JACS 2009;131(6):2082–3. Ru¨mmeli MH, Borowiak-Palen E, Gemming T, Pichler T, Knupfer M, Kalbac M, et al. Novel catalysts, room temperature, and the importance of oxygen for the synthesis of single-walled carbon nanotubes. Nano Lett 2005;5(7):1209–15. Ru¨mmeli MH, Kramberger C, Gru¨neis A, Ayala P, Gemming T, Bu¨chner B, et al. On the graphitization nature of oxides for the formation of carbon nanostructures. Chem Mater 2007;19(17):4105–7. Huang S, Cai Q, Chen J, Qian Y, Zhang L. Metal-catalyst-free growth of single-walled carbon nanotubes on substrates. JACS 2009;131(6):2094–5. Hirsch A. Growth of single-walled carbon nanotubes without a metal catalyst—a surprising discovery. Angew Chem Int Ed 2009;48:5403–4. Steiner SA, Baumann TF, Bayer BC, Blume R, Worsley MA, MoberlyChan WJ, et al. Nanoscale zirconia as a nonmetallic catalyst for graphitization of carbon and growth of singleand multi-wall carbon nanotubes. JACS 2009;131(34):12144–54. Liu H, Takagi D, Chiashi S, Homma Y. The growth of singlewalled carbon nanotubes on a silica substrate without using a metal catalyst. Carbon 2010;48:114–22. Bachmatiuk A, Bo¨rrnert F, Grobosch M, Scha¨ffel F, Wolf U, Scott A, et al. Investigating the graphitization mechanism of SiO2 nanoparticles in chemical vapor deposition. ACS Nano 2009;3:4098–104. Hlavac J. Glass science and technology, the technology of glass and ceramics, vol. 12. Elsevier: Czechoslovakia; 1983. Ru¨mmeli MH, Adebimpe D, Borowiak-Palen E, Gemming T, Ayala P, Ioannides N, et al. Hydrogen activated axial interconversion in SiC nanowires. J Solid State Chem 2009;182:602–7. Sun Y, Miyasato T. Infrared absorption properties of nanocrystalline cubic SiC films. Jpn J Appl Phys 1998;37:5485–9. Kamitos EI, Patsis AP, Kordas G. Infrared-reflectance spectra of heat-treated sol–gel-derived silica. Phys Rev B 1993;48:12499–505. Primeau N, Vatutey C, Langlet M. The effect of thermal annealing on aerosol–gel deposited SiO2 films: a FTIR deconvolution study. Thin Solid Films 1997;310:47–56.

CARBON

4 8 ( 20 1 0 ) 3 1 7 5–31 8 1

[25] Parshar VK, Orhan JB, Sayah A, Cantoni M, Gijs MAM. Borosilicate nanoparticles prepared by exothermic phase separation. Nat Nanotechnol 2008;3:589–94. [26] Sigoli FA, Kawano Y, Davolos MR, Jafelicci M. Phase separation in pyrex glass by hydrothermal treatment: evidence from micro-Raman spectroscopy. J Non-Cryst Solids 2001;284:49–54. [27] Jorio A, Pimenta MA, Souza Filho AG, Saito R, Dresselhaus G, Dresselhaus MS. Characterizing carbon nanotube samples with resonance Raman scattering. N J Phys 2003;5:139.1–.17.

3181

[28] Dresselhaus MS, Dresselhaus G, Saito R, Jorio A. Raman spectroscopy of carbon nanotubes. Phys Report 2005;409:47–99. [29] Dai H, Wong EW, Lieber CM. Probing electrical transport in nanomaterials: conductivity of individual carbon nanotubes. Science 1996;272:523–6. [30] Kim T-H, Wendelken JF, Li A-P, Du G. Probing electrical transport in individual carbon nanotubes and junctions. Nanotechnology 2008;19:485201–6.