Open Access
Add to BibSonomyAbstract
To understand the reaction mechanisms taking place by growing carbon nanotubes via the catalytic chemical vapor deposition process, a strategy to monitor in situ the gas phase at reaction conditions was developed applying linear Raman spectroscopy. The simultaneous determination of the gas temperature and composition was possible by a new strategy of the evaluation of the Raman spectra. In agreement to the well-known exothermic decomposition of acetylene, a gas temperature increase was quantified when acetylene was added to the incident flow. Information about exhaust gas recirculation and location of the maximal acetylene conversion was derived from the composition measurements.
©2010 Optical Society of America
1. Introduction
The unique chemical and physical properties of carbon nanotubes (CNTs), which make them suitable for a wide range of applications, have stimulated interest for research since their discovery by Iijima in 1991 [1]. CNTs exhibit specific electronic properties and depending on their geometry, they can be either metallic or semiconducting [2]. In the field of mechanics, they are considered a potential composite material due to their extraordinary strength and low density. Additionally, CNTs are characterized by their high thermal stability up to 1673 K in vacuum [3–5].
Among the developed methods for growing CNTs, the catalytic chemical vapor deposition (CCVD) of hydrocarbon gases turns out to be very promising because of its comparative simplicity, ease of control and low cost [6–8]. The way how certain operation conditions and catalyst properties influence the characteristics of the resulting CNTs has been so far analyzed via trial and error investigations using only input and output data. Based on this, some models have been derived, which intend to describe the functioning chain of the CNTs deposition [9–14]. Nevertheless, the CNTs nucleation and growth mechanism is not fully understood yet when achieved by such ex situ methods not taking into consideration the intermediate steps of the reaction process.
On this account, we here introduce a novel in situ measurement strategy, which is based on linear Raman spectroscopy and which allows to analyze also the intermediate processes, which take place inside the CNT reactor and are not accounted for conventionally. This paper reports about the investigations carried out to monitor in situ the gas phase at CCVD operation conditions using acetylene as carbon source. Iron catalyst nanoparticles, which are typically deposited on a substrate, were not implemented in order to exclude their effect in the first instance. The simultaneous determination of the acetylene decomposition and the gas temperature at different defined locations apart from the substrate surface, on which the CNTs are to be grown, was possible via a new approach for the evaluation of the Raman spectra and taking advantage of the incident gas flow composition.
2. Experimental setup
A schematic of the home-made CNT reactor in which the investigation has been accomplished is shown on the left-hand side of Fig. 1 . It consists of a vertically oriented cold wall flow reactor of 600 cm3 inner volume, made of stainless steel and provided with four optical accesses in crosswise arrangement.
Fig. 1 Schematic of the optically accessible cold wall flow reactor and location of the measurement positions A, B and C.
The CCVD process is operated at steady-state conditions, at atmospheric pressure and at a constant substrate temperature of 953 K. The required gases (nitrogen, hydrogen and acetylene) enter the reactor at the bottom with room temperature and a flow rate in the range of 50-100 SCCM, whereas the hot exhaust gases leave the reactor at the top. A resistive heating cartridge with a maximum heating power of 400 W is built inside the reactor together with NiCrNi-thermocouples, which measure and control the substrate temperature. The substrate consists of a Si/SiO2-wafer and it is fixed to the heating cartridge in such a way that the incident gas flow impinges on it. The distance between the inlet pipe exit and the Si/SiO2-wafer was set to be 5 mm. On the right-hand side of Fig. 1, the gas flow field developed between the inlet pipe and the Si/SiO2-wafer is shown. Detailed information about the flow field investigations will be presented in a future publication.
In Fig. 1 three different locations are labeled as A, B and C, where composition and temperature measurements were carried out. A and B are located 1 mm in front of the heating cartridge positioned in the centre and in the border below the substrate surface, respectively. C is located 4 mm below B. These different locations have been selected in order to track the trajectory of a fluid element at different conditions departing from the knowledge of the flow field. In Fig. 1 the locations A, B and C are not labeled as measurement points but as measurement lines of 3.2 mm length, as the Raman signals were integrated along these lines.
The experimental Raman setup is shown in Fig. 2 . The excitation source is a frequency-doubled Nd:YAG laser cluster, which is explained in detail elsewhere [15]. The laser cluster emits pulses of 8 ns (FWHM) duration with a repetition rate of 10 Hz and a single-pulse energy of 360 mJ at 532 nm. Using a beam splitter and two mirrors the laser pulse is temporally stretched in a single-loop pulse stretcher to reduce the pulse intensity [16]. Moreover, by means of a plano-concave (f = −100 mm) and a plano-convex lens (f = 400 mm) the laser beam is expanded by a factor 4 in order to decrease the fluence. With these measurements, laser-induced damage of the optical accesses and laser-induced breakdown in the gas phase can be avoided. By a further spherical lens (f = 250 mm), the laser beam is focused to the desired measurement location inside the CNT reactor.
Perpendicularly to the laser propagation, the scattered light is collected and collimated with an achromatic lens (f = 100 mm). A long pass filter with high transmission for wavelengths longer than 535 nm is used to suppress the elastically scattered light, which is much more intensive than the Raman signal itself [17,18]. The Raman signals are transmitted and focused onto the entrance slit of a spectrometer (f = 250 mm, 600 lines mm−1) by a second achromatic lens (f = 200 mm). The signals are detected with an electron multiplying charge-coupled device (EMCCD) camera (Andor Newton 971, 1600 x 400 pixels of size 16 x 16 μm2) with a quantum efficiency of more than 90% in the detected spectral range. The EMCCD camera was mounted onto the spectrometer with the 400-pixel axis representing the spatial resolution of the line-shaped probe volume. In order to increase the signal-to-noise ratio two measures were taken. On the one hand, the 400-pixel axis was binned completely to obtain an average spectrum of the 3.2 mm long probe volume and on the other hand, 200 single spectra were added together in computer memory to achieve an accumulated spectrum. Any EM gain factor was applied by acquiring the Raman spectra. Four single laser shots temporally shifted by 1 μs and with a repetition rate of 10 Hz, were used to acquire a single spectrum. Thus, 800 single laser shots were used to acquire the resulting accumulated spectrum. The exposure time of the camera was set to the minimum possible value of 10 μs in order to overcome the interfering black body radiation of the heating cartridge. A pulse generator was used to synchronize the EMCCD camera and the laser.
3. Results and discussion
Figure 3 shows three Raman spectra, which were recorded from the gas phase inside the cold wall reactor at location A (see Fig. 1, right). One Raman spectrum was acquired at room temperature before heating up the gas flow of 85 SCCM. At these cold conditions, the composition of the gas flow was 98 vol.-% nitrogen and 2 vol.-% hydrogen. The second Raman spectrum was recorded when steady-state hot conditions inside the CNT reactor were reached. At this condition, the fed gas composition was the same as for the cold condition and the substrate temperature was kept constant at 953 K. The third spectrum corresponds to a measurement, when acetylene was added to the incident gas flow without changing the overall flow rate at hot conditions (Twafer = 953 K). Only the composition of the gas flow was changed to 97 vol.-% nitrogen, 2 vol.-% hydrogen and 1 vol.-% acetylene.
Fig. 3 Raman spectra for the determination of the gas temperature and composition. Acquired under CCVD operation conditions at location A (see Fig. 1). Overall volumetric flowrate = 85 SCCM. (a) Complete spectral range. (b) Zoom-out of the rotational lines J1, J2 and J3 of hydrogen.
The temperature of the gas flow was evaluated by analyzing the peak integrals of the pure rotational Raman lines of hydrogen. The population of the according molecular rotational energy levels is known to be a function of temperature, which follows Boltzmann statistics [19]. Thus, the rotational temperature was achieved by performing a least squares linear fit to the rotational lines J1, J2 and J3 on a Boltzmann plot, since they are temperature dependent.
The composition of the gas flow was determined by analyzing the peak integrals of the Stokes vibrational Raman Q-branches of the nitrogen and hydrogen molecules and of the C-C vibration band of acetylene. The peak integrals are directly proportional to the number density of the nitrogen, hydrogen and acetylene molecules, respectively. Consequently, the acetylene conversion XC2H2, which describes how much acetylene has already been decomposed, can be monitored indirectly by probing the gas composition. Calibration measurements were carried out at different adjusted gas compositions at room temperature. The effect of temperature onto the Raman scattering cross section under operation conditions was accounted for [20].
From the three spectra given in Fig. 3b it is noticeable that the temperature for all three measurements is different, as the peak integrals of the Raman lines J1, J2 and J3 vary. The measured temperatures are 297.0 K, 576.2 K and 599.3 K for the cold and hot conditions, without and with acetylene, respectively. This temperature difference was expected for the measurements carried out at cold and hot conditions. However, the comparison of the two measurements at hot conditions, one without acetylene and one with 1 vol.-% acetylene in the gas flow, also indicate a temperature increase when acetylene was added, which must be attributed to the exothermic thermal decomposition of acetylene [3,21].
Table 1 summarizes some results from the simultaneous measurement of the gas temperature and acetylene conversion XC2H2 ( = 1 - [c/c0] where c and c0 are the acetylene concentrations in the measurement and in the inlet flow, respectively).
Table 1. Measured gas temperature and acetylene conversion XC2H2 at locations A, B and C before and after C2H2 (acetylene) addition into the incident flow
The measured temperature is presented before and after start of acetylene addition at measurement locations A, B and C. Start of acetylene addition defines the moment when the acetylene flow was started and simultaneously the nitrogen flow was reduced in order to keep the overall gas flow rate constant at 85 SCCM and to achieve an overall inflow gas phase composition of 97 vol.- % nitrogen, 2 vol.- % hydrogen and 1 vol.- % acetylene. The pipe system around the CNT reactor was projected to realize response times to composition changes of less than 30 s.
From Table 1 can be seen that the temperature at all measurement locations rises significantly already 1 min after acetylene addition. The three different measurement locations A, B and C show three different temperature levels. A represents the coldest location, as here the fresh cold gas is introduced into the cold wall flow reactor. At location B the gas flow has been in contact with the substrate surface, which has a temperature of 953 K, and therefore shows a higher temperature than location A. The hottest location is C, where recirculating exhaust gas is present. The temperature at location C is higher than at location B, as also gas, which had been in contact with the hot heating cartridge further upstream, can travel to location C, whereas at location B dominantly the gas which is freshly impinging on the hot Si/SiO2-wafer is present.
The acetylene conversion, which exactly corresponds to the given temperature, is also shown in Table 1. At location A, where fresh gas is introduced towards the hot Si/SiO2-wafer, the conversion is nearly 32%. The temperature value at this location, which is 599.3 K, is still below the decomposition temperature of acetylene [22]. Therefore, the measured conversion of acetylene at location A must be due to mixing of the fresh gas, which is entering from the pipe, and exhaust gas, which may be entrained from the outer regions (for example from location C). As there is no significant conversion difference between location B and C, one may assume that the conversion is taking place mainly on the hot Si/SiO2-wafer surface. If there were significant conversion of the carbon source further upstream, when the gas is passing the hot heating cartridge, the recirculation of this exhaust gas to location C would result in conversions higher than conversions at locations B. At the given operation conditions, acetylene conversions larger than 82% have not been quantified.
4. Conclusion
For the first time to the best of our knowledge, in situ optical measurements based on linear Raman spectroscopy were successfully applied to monitor simultaneously both the gas temperature and the acetylene decomposition at operation conditions favorable for the CCVD of CNTs. The measured gas temperatures at different locations and before and after acetylene addition are in agreement to the well-known exothermic decomposition of acetylene. The composition measurements reveal the presence of exhaust gas recirculation and the location of the maximal acetylene conversion, mainly on the substrate surface. Thus, this approach qualifies for monitoring in situ the gas phase during the CCVD of CNTs and consequently, offers the possibility to establish correlations between the decomposition mechanism, reaction conditions and the microstructure of the resulting CNTs.
Acknowledgment
The authors gratefully acknowledge financial support of parts of this work by the German Research Foundation (DFG) and the funding of the Erlangen Graduate School in Advanced Optical Technologies (SAOT) by the DFG in the framework of the German excellence initiative.
References and links
1. S. Iijima, “Helical microtubules of graphitic carbon,” Nature 354(6348), 56–58 (1991). [CrossRef]
2. T. Belin and F. Epron, “Characterization methods of carbon nanotubes: a review,” Mater. Sci. Eng. B 119(2), 105–118 (2005). [CrossRef]
3. K. B. K. Teo, C. Singh, M. Chhowalla, and W. I. Milne, “Catalytic synthesis of carbon nanotubes and nanofibers,” in Encyclopedia of Nanoscience and Nanotechnology (American Scientific Publishers, Stevenson Ranch, CA, USA, 2003), pp. 665–686.
4. T. C. Schmitt, A. S. Biris, D. W. Miller, A. R. Biris, D. Lupu, S. Trigwell, and Z. U. Rahman, “Analysis of effluent gases during the CCVD growth of multi-wall carbon nanotubes from acetylene,” Carbon 44(10), 2032–2038 (2006). [CrossRef]
5. S. J. Tans, A. R. M. Verschueren, and C. Dekker, “Room-temperature transistor based on a single nanotube,” Nature 393(6680), 49–52 (1998). [CrossRef]
6. C. H. See and A. T. Harris, “A Review of Carbon Nanotube Synthesis via Fluidized-Bed Chemical Vapor Deposition,” Ind. Eng. Chem. Res. 46(4), 997–1012 (2007). [CrossRef]
7. M. Escobar, M. S. Moreno, R. J. Candal, M. C. Marchi, A. Caso, P. I. Polosecki, G. H. Rubiolo, and S. Goyanes, “Synthesis of carbon nanotubes by CVD: Effect of acetylene pressure on nanotubes characteristics,” Appl. Surf. Sci. 254(1), 251–256 (2007). [CrossRef]
8. C. Singh, M. S. P. Shaffer, and A. H. Windle, “Production of controlled architectures of aligned carbon nanotubes by an injection chemical vapour deposition method,” Carbon 41(2), 359–368 (2003). [CrossRef]
9. H. 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). [CrossRef]
10. S. Esconjauregui, C. M. Whelan, and K. Maex, “The reasons why metals catalyze the nucleation and growth of carbon nanotubes and other carbon nanomorphologies,” Carbon 47(3), 659–669 (2009). [CrossRef]
11. R. T. K. Baker, “Catalytic growth of carbon filaments,” Carbon 27(3), 315–323 (1989). [CrossRef]
12. A.-C. Dupuis, “The catalyst in the CCVD of carbon nanotubes–a review,” Prog. Mater. Sci. 50(8), 929–961 (2005). [CrossRef]
13. G. G. Tibbetts, “Why are carbon filaments tubular?” J. Cryst. Growth 66(3), 632–638 (1984). [CrossRef]
14. K. Liu, K. Jiang, C. Feng, Z. Chen, and S. Fan, “A growth mark method for studying growth mechanism of carbon nanotube arrays,” Carbon 43(14), 2850–2856 (2005). [CrossRef]
15. A. Braeuer, S. R. Engel, R. F. Hankel, and A. Leipertz, “Gas mixing analysis by simultaneous Raman imaging and particle image velocimetry,” Opt. Lett. 34(20), 3122–3124 (2009). [CrossRef] [PubMed]
16. J. Kojima and Q. V. Nguyen, “Laser pulse-stretching with multiple optical ring cavities,” Appl. Opt. 41(30), 6360–6370 (2002). [CrossRef] [PubMed]
17. D. A. Long, Raman Spectroscopy (London, 1977).
18. B. Schrader, Infrared and Raman Spectroscopy (Weinheim, 1995).
19. A. Braeuer and A. Leipertz, “Two-dimensional Raman mole-fraction and temperature measurements for hydrogen-nitrogen mixture analysis,” Appl. Opt. 48(4), B57–B64 (2009). [CrossRef] [PubMed]
20. A. C. Eckbreth, Laser Diagnostics for Combustion Temperature and Species (Cambridge, MA, 1988).
21. R. T. K. Baker, M. A. Barber, P. S. Harris, F. S. Feates, and R. J. Waite, “Nucleation and growth of carbon deposits from the nickel catalyzed decomposition of acetylene,” J. Catal. 26(1), 51–62 (1972). [CrossRef]
22. T. Tanzawa and W. C. Gardiner Jr., “Reaction mechanism of the homogeneous thermal decomposition of acetylene,” J. Phys. Chem. 84(3), 236–239 (1980). [CrossRef]
