Abstract Both single‐walled and multi‐walled carbon nanotubes (SWNTs and MWNTs) are produced using a solar reactor working under nitrogen atmosphere. Samples were studied by transmission electron microscopy (TEM), Raman spectroscopy and electron energy loss spectroscopy (EELS). The produced MWNTs exhibit the typical morphologies observed for nitrogen‐doped MWNTs synthesized with other methods and EELS measurements point out the presence of substituted nitrogen within the tubes. The doping level is not uniform and the maximum value reaches 11 at.%.

Single-walled carbon nanotubes were synthesized in gram quantities with a 50 kW solar reactor. Vaporization rate was in the range 6–15 g/h and two samples were collected for each run. The structure of the products was studied by SEM, TEM and Raman spectroscopy as a function of the buffer gas (helium or argon), the target length and the collected location. Good quality samples containing 1.2–1.6 nm diameter SWNTs were produced with helium at 450 hPa, 8 Nm3/h and a 15 cm target length using 2 at.% Ni and 2 at.% Co as catalyst. The material quality increased with the target length in helium. On the contrary, poor quality product was obtained whatever the target length in argon. TGA analysis showed that the best product purity was at least 35 wt%.

This paper aims at the optimum design of a solar reactor for carbon product processing by vaporization. The chosen reference case is the CNRS 1 MW solar furnace and graphite vaporization for fullerene synthesis. The method, which accounts for heat transfer and chemical reaction kinetics, is based on the combination of both experimental data obtained at laboratory scale and numerical simulation. For this very high temperature process (carbon vaporization is only significant at temperatures higher than 3200 K), the optimum diameter of the target is 22 cm and the soot production ranges between 80 and 150 g/h for an effective power of 325 kW (P=240 h Pa). The model is validated by experiments using a 6 cm o.d. graphite target.

Single-walled carbon nanotube samples were synthesized with solar energy. The products are characterized by electron microscopy, Raman spectroscopy and nitrogen isotherms. The maximum value of the hydrogen uptake is 1wt.% at 253K and 6MPa. The results are discussed by taking account of the structure and the BET surface of the samples. The hydrogen capacity is not proportional to the BET surface in these samples and does not seem to be linked to the amount of SWNT.

Fullerene mixture of C60 and C70 was produced in gram quantities in a solar reactor using partially (50 kW) the power of the 1 MW CNRS solar furnace at Odeillo. Fullerene yield was studied as a function of buffer gas (helium and argon), pressure (in the range 80–500 hPa) and gas flow rate. Mean Fullerene yield of 13.5% was measured with helium at 450 hPa and 10 sm3/h, at a carbon vaporization rate of 21 g/h. This mean value refers to three samples collected inside the experimental set up; the maximum yield for one single sample (inside the reactor) was 16%. The dramatic effect of pressure and gas flow rate on the process selectivity is correlated to the dilution number (number of carbon atoms versus number of buffer gas atoms) and to the results of a numerical simulation related to the temperature distribution in the annealing zone of the reactor.

C2 Swan band emission (d3πg→a3πu) near 517 nm is observed in a solar reactor for fullerene synthesis. On the basis of theoretical considerations and experimentals in the temperature range 3000–3400 K evidence is presented supporting the formation of excited C2 by absorption of solar photons. This phenomenon that we propose to name: solar-induced fluorescence (SIF) is described for the first time.

Heat and mass transport are obtained in a solar reactor using ‘in situ’ measurements linked to numerical simulation and allow the interpretation of the vaporization process as well as the determination of the cooling regime. Comparison with other processes (laser ablation or electric arc) point out some common behavior like the great influence of the cooling rate of vapors on the structure and yield of nanostructured carbon material. We also investigate the growth mechanisms of single wall carbon nanotubes (SWNTs) produced by the solar method as a function of the nature of catalysts and the temperature variation in the condensing area. The Raman spectra clearly show that the change of catalyst induces differences in the diameter of SWNT whereas TEM pictures enhance the change of both length and diameter of the bundles. All these results are explained considering that the key parameter is the temperature at which the SWNTs are formed. This temperature range can be related to the sublimation temperature of the target and to the eutectic temperature of the binary phase diagram. Finally we propose a new mechanism to explain the nucleation process and segregation rate which seems to depend on the capacity of catalyst to form carbide.

Conventional methods for the synthesis of fullerenes and carbon nanotubes such as laser or electric arc ablation have failed when the process is scaled up. Our ultimate goal is to scale a solar process up from 2 to 250 kW; this paper shows that our method for achieving this scale-up is valid because we were able to predict process performance variables at the 50 kW level from preliminary experimental results from 2 kW experiments. The key parameters that characterize this process are the carbon soot mass flow rate and the desired product yield. The carbon soot production rate is a function of the target temperature and this can be predicted in a straightforward way from a heat transfer model of the larger system. The yield is a more complicated function of specific reactor variables such as patterns of fluid flow, residence times at various temperatures, and the reaction chemistry, but we have found that for fullerenes it depends primarily on the concentration of carbon vapor in the carrier gas, the target temperature and the temperature distribution in the cooling zone. Using these parameters, we scaled our process up to 50 kW and compared the predicted results to the measured performance. A graphite target 6 cm in diameter was vaporized in an argon atmosphere and a reduced pressure of 120–240 hPa with a solar flux density in the range 600-920W/cm2. Vaporization rates as high as 50 g/h were measured with a fullerene production rate equal to about 2 g/h, i.e., the expected results.

Fullerene synthesis was performed at 2 kW and 50 kW scales by vaporization of graphite at the focus of solar furnaces. We obtained a fullerene yield of about 10% with a 50-mg/h production at small scale for target temperature ranging from 3000 K to 3700 K and pressure between 50 hPa and 250 hPa. We achieved the reactor scale up on the basis of numerical simulations and of pertinent parameters such as pressure and carbon atom fraction in the buffer gas. The production of 1 g/h fullerene was still obtained choosing pertinent pressure and carbon atom dilution in argon. New experiments are in progress to improve this result.

Abstract Conventional methods for the synthesis of fullerenes and carbon nanotubes such as laser or electric arc ablation have failed when the process is scaled up. Our ultimate goal is to scale a solar process up from 2 to 500 kW; this paper shows that our method for achieving this scale up is valid because we were able to predict process performance variables at the 50 kW level from preliminary experimental results from 2 kW experiments. The key parameters that characterize this process are the carbon soot mass flow rate and the desired product yield. The carbon soot production rate is a function of the target temperature and this can be predicted in a straightforward way from a heat transfer model of the larger system. The yield is a more complicated function of specific reactor variables such as patterns of fluid flow, residence times at various temperatures and the reaction chemistry, but we have found that for fullerenes it depends primarily on the concentration of carbon vapor in the carrier gas, the target temperature and the temperature distribution in the cooling zone. Using these parameters, we scaled our process up to 50 kW and compared the predicted results to the measured performance. A graphite target 6 cm in diameter was vaporized in an argon atmosphere and a reduced pressure of 120–240 hPa with a solar flux density in the range 600–900 W/cm2. Vaporization rates of 20 g/h were measured with a fullerene production rate equal to or greater than 1 g/h, i.e. the expected results.

We investigate the growth mechanisms of single wall carbon nanotubes (SWNTs) produced by the solar method as a function of the nature of catalysts and the temperature variation in the condensing area. The Raman spectra clearly show differences in SWNT diameters induced by the change of catalyst whereas TEM pictures enhance the change of both length and diameter of the bundles. All the results are explained considering that the formation temperature range of SWNTs is the key parameter related to the target sublimation temperature and to the eutectic temperature of the binary phase diagram.

Experimental results with a 2 kW solar furnace, in a wide range of vaporization rates (0.1–4 g/h), under variable pressure and flow rate of argon, are used with numerical simulation to define key parameters for large scale synthesis of fullerenes with solar energy. The vaporization process is controlled by diffusion in the temperature and pressure ranges 3000-3700 K and 70-250 hPa respectively. In the solar reactor, fullerene yield is governed by the dilution of carbon vapor in argon and the temperature gradient in the cooling zone. Criteria for both parameters are suggested. Consequently, these data, combined with a validated numerical model of the reactor, may be used for the design of large-scale solar process.

Carbon single-wall nanotubes (SWNTs) can be produced from carbon sublimation by various methods; we have compared electric arc and laser techniques with solar energy processing. Generally, produced SWNTs are not significantly different (on the basis of mean diameter and diameter distribution) for a given catalyst as shown by TEM and Raman spectroscopy characterisation. However, the double-pulse Nd: YAG laser method produces a high purity material but with a poor production rate. Finally, we have demonstrated that the solar processing permit to set the SWNTs mean diameter in the range 0.9 nm-2 rim by changing the catalyst and to control, in a certain extent, the production yield which depend of the temperature distribution inside the reactor.

We investigate the laser-energy dependence of the Raman profile of single-wall carbon nanotube (SWNT) samples with various distributions of diameters. We show that resonant Raman is an efficient tool for the study of the structure and electronic properties of SWNT. The tube diameter distribution is derived from the comparison between the experimental frequencies of the radial A1g breathing mode range (RBM) and the calculated RBM frequency of SWNT bundles. Metallic or semi-conducting tubes are identified in the light of calculations of allowed optical transitions. The assignments are confirmed by the observation (absence) of a Breit–Wigner–Fano-like lineshape for the tangential graphite-like modes of metallic (semiconducting) nanotubes.

We present in this paper a new original process for fullerenes production consisting of treating carbon powders through a 3-phase thermal plasma. The main difference with the classical arc process is that the input carbon rate, not limited to the electrodes erosion, can be independently controlled. As expected, the experiments carried out using ten different industrial carbon grades have shown a high sensitivity depending on the carbon precursor. These results have confirmed that the new process is a very promising route for producing bulk quantities of fullerenes. Before optimization, 3.5% of extractable fullerenes were obtained at atmospheric pressure with acetylene black.

We report here some studies of the growth mechanisms of single-wall carbon nanotubes (SWNTs) produced by the solar method as a function of the experimental conditions and the nature of catalysts. A large set of transmission electron microscopy (TEM) pictures seems to confirm the existence of one dominant growth mechanism, close to the model proposed by Saito et al., whatever the used catalyst might be. Nevertheless, the Raman spectra clearly show that the change of catalyst induces differences in diameter, structure, and electronic properties of SWNTs.

In this work we report the production of carbon nanotubes by direct vapourization of graphite targets containing different couples of catalysts using a 2 kW solar furnace. With this small-scale experimental set-up we observed the evolution of the structure of the produced nanotubes with experimental conditions, pressure and flow rate of the inert gas (argon) and target composition. High pressure favours the production of bundles of single-wall nanotubes (SWNTs) and the purity of the produced material depends on the target temperature and the cooling rate of the vapour. We also discuss the influence of the target composition.

The Raman spectroscopy have allowed us to perform studies on singlewalled nanotubes (SWNT's) produced by following methods: electric arc, laser ablation and solar energy. As this characterization method provides a great deal of informations, we will present a comparison between the nanotubes produced by all these processes and the influence of some synthesis parameters. By using spin casting, we have produced thin films of PMMA-SWNT's for different concentrations. Then, we have characterized these new materials by Raman spectroscopy. The aim of these investigations is to get information on the possible interactions between these two materials. In particular, we have studied the evolution of the composites films spectra as a function of the nanotubes concentration in the polymer.

After the promising results obtained with a 2 kW solar furnace for fullerenes and nanotubes synthesis, a large scale production project using the 1 MW Odeillo solar furnace started in 1997. This paper presents the first experimental results obtained with a concept-validation vessel and the comparison with a numerical simulation of the target thermal behavior. It is shown that a 6 mm i.d. graphite rod heated by a 500 W/cm(2) incident solar flux density (I-s) reaches a front temperature of 2800 K, in agreement with the thermal model. On this basis, accurate prediction of maximum working temperature of the 1 MW reactor is proposed : 3400 K for I-s = 900 W/cm(2).

We discuss here the influence of some synthesis parameters in the production of single wall carbon nanotubes (SWNT) with a 2 kW solar furnace at the Odeillo Institute. We first study the specific role of catalysts like nickel or cobalt on the SWNT formation process. SEM and TEM pictures shaw significant differences in the yield of production and suggest different growth mechanisms. Furthermore, we report Raman spectra for samples produced using lanthanium or sulfur as catalysts. These samples provide very striking results in the low frequency range (in term of distribution in diameter) with a good correlation with the graphite-like modes region.