CARBON NANOTUBES FUNCTIONALIZED WITH FULLERENES

Abstract

The present invention relates to covalently bonded fullerene-functionalized carbon nanotubes(CBFFCNTs), a method and an apparatus for their production and to their end products. CBFFCNTs are carbon nanotubes with one or more fullerenes or fullerene based molecules covalently bonded to the nanotube surface. They are obtained by bringing one or more catalyst particles, carbon sources and reagents together in a reactor.

Claims

1. A method for producing one or more fullerene functionalized carbon nanotubes, wherein the method comprises: bringing one or more catalyst particles and carbon sources and at least two reagents including CO.sub.2 and H.sub.2O, wherein the concentration of H.sub.2O is between 45 and 245 ppm and the concentration of CO.sub.2 is between 2000 and 6000 ppm, into contact with each other; heating in a reactor at a temperature of 250-2500 C. for catalytically decomposing the one or more carbon sources on the surface of the catalyst particles together with the reagents to produce one or more carbon nanotubes comprising one or more fullerenes and/or fullerene based molecules covalently bonded to the one or more carbon nanotubes; and collecting the produced one or more fullerene functionalized carbon nanotubes.

2. The method according to claim 1, wherein the carbon source is selected from the group consisting of methane, ethane, propane, ethylene, acetylene, benzene, toluene, xylene, trimethylbenzene, methanol, ethanol, octanol, tiophene and carbon monoxide.

3. The method according to claim 1, wherein the reagent is an etching agent.

4. The method according to claim 1, wherein the reagent is selected from the group, consisting of hydrogen, nitrogen, water, carbon dioxide, nitrous oxide, nitrogen dioxide, oxygen, ozone, carbon monoxide, octanol, thiophene and hydride.

5. The method according to claim 1, wherein the catalyst particle comprises a metal, a transition metal and/or a combination of metals and/or transition metals.

6. The method according to claim 1, wherein the catalyst particle comprises iron, cobalt, nickel, chromium, molybdenum and/or palladium.

7. The method according to claim 1, wherein the catalyst particle is produced using a chemical precursor and/or by heating a metal or metal containing substance.

8. The method according to claim 1, wherein the amount of fullerene and/or fullerene based molecules produced on the carbon nanotube is adjusted by adjusting the amount of one or more reagents used, by adjusting the heating temperature and/or by adjusting the residence time.

9. The method according to claim 1, wherein the heating is performed at a temperature of 600-1000 C.

10. The method according to claim 1, wherein the method further comprises the following step: introducing one or more additional reagents.

11. The method according to claim 1, wherein the method further comprises the following step: introducing one or more additives to produce a fullerene functionalized carbon nanotube composite material.

12. The method according to claim 1, wherein the method further comprises the following step: collecting the produced one or more fullerene functionalized carbon nanotubes and/or the fullerene functionalized carbon nanotube composite material in a solid, liquid, and/or gas dispersion, a solid structure, a powder, a paste, a colloidal suspension and/or as a film and/or surface deposition.

13. The method according to claim 1, wherein the method further comprises the following step: depositing a dispersion of produced fullerene functionalized carbon nanotubes and/or fullerene functionalized carbon nanotube composite material onto a surface and/or into a matrix and/or a layered structure and/or a device.

14. The method according to claim 1, wherein the fullerene functionalized carbon nanotubes are produced in a gas phase as an aerosol and/or on a substrate.

15. A functional material made using one or more fullerene functionalized carbon nanotubes produced with the method according to claim 1.

16. A thick or thin film, a line, a wire or a layered or three dimensional structure, wherein the thick or thin film, the line, the wire or the layered or three dimensional structure is made using one or more fullerene functionalized carbon nanotubes produced with the method according to claim 1.

17. A thick or thin film, a line, a wire or a layered or three dimensional structure, wherein the thick or thin film, the line, the wire or the layered or three dimensional structure is made using the functional material according to claim 15.

18. A device, wherein the device is made by using one or more fullerene functionalized carbon nanotubes produced with the method according to claim 1.

19. A device, wherein the device is made by using the functional material according to claim 15.

20. A device, wherein the device is made by using the thick or thin film, a line, a wire or a layered or three dimensional structure according to claim 16.

Description

LIST OF FIGURES

[0064] In the following section, the invention will be described in detail by means of embodiment examples with reference to accompanying drawings, in which

[0065] FIGS 1a-1e show a) a schematic representation of covalently bonded fullerene-functionalized carbon nanotube depicting covalent bonding and b)-e) low, intermediate and high resolution images of examples of CBFFCNTs;

[0066] FIG. 2 shows a block diagram of an arrangement for the method for production of CBFFCNTs, CBFFCNT composites, structures and devices;

[0067] FIGS. 3a-3c show preferred embodiments of the invention for aerosol production of CBFFCNTs, where the catalyst particles are formed by decomposing one or more catalyst particle precursors (a), where the catalyst particles are formed by a physical vapor nucleation method from a hot wire generator (b) separated in space from the reactor and (c) smoothly integrated with the reactor;

[0068] FIG. 4: Number size distribution of fullerenes measured from HR-TEM images;

[0069] FIG. 5: EELS spectra of different parts of CBFFCNTs showing the presence of oxygen in the covalent bond between CNTs and fullerenes;

[0070] FIG. 6: Comparison of ultraviolet-visible absorption spectra of CBFFCNTs and C.sub.60 and C.sub.70 standards;

[0071] FIG. 7: Comparison of Raman spectroscopy measurements of the samples carried out by using red (633 nm) blue (488 nm) lasers of samples prepared with high (lines 1 and 2) and low (lines 3 and 4) concentrations of functionalizing fullerenes. Inset shows details of the shift in the fullerene signal marked with arrows;

[0072] FIG. 8: MALDI-TOF spectrum, averaged over several solvents, evidencing the presence of C.sub.60H.sub.2 and C.sub.42COO as well as other fullerenes containing O and/or H atoms in the bridging groups;

[0073] FIG. 9: FT-IR spectra of CBFFCNTs demonstrating the presence of ethers (COC) and esters (COOC) in the sample;

[0074] FIGS. 10a-10b show Field emission properties of CBFFCNTs (synthesized in the ferrocene reactor without water vapour added) and CBFFCNTs (synthesized in the presence of 100 and 150 ppm of added water vapour): (a) Averaged current density against the electric field strength; (b) Fowler-Nordheim plot for the investigated samples; (c) Temporal behavior of the electron current at different field strengths;

[0075] FIG. 11: TEM image of CBFFCNTs produced through an aerosol Iron-octanol-thiophene system (t.sub.furn=1200 C., flow through bubbler Q.sub.CO=400 ccm and through an aerosol HWG Q.sub.N2/H2=400 ccm);

[0076] FIG. 12: FT-IR spectra obtained at the conditions of CNT synthesis in the aerosol HWG method: gas composition: CO.sub.2-120 ppm, H.sub.2O-10 ppm showing the in situ production of reagents on the reactor wall;

[0077] FIG. 13: TEM image of CBFFCNTs from in situ aerosol HWG and CO as carbon source, H.sub.2/N.sub.2 (7/93) mixture through HWG, t.sub.set=1000 C. and EELS measurements showing the presence of oxygen in the covalent bond between CNTs and fullerenes;

[0078] FIG. 14: TEM image of CBFFCNTs from in situ aerosol HWG and CO as carbon source, H.sub.2/N.sub.2 (0.07/99.93) mixture through HWG, t.sub.set=900 C. and EELS measurements showing the presence of oxygen in the covalent bond between CNTs and fullerenes;

[0079] FIG. 15: EELS spectra proving the presence of oxygen in the covalently bonded CBFFCNTs produced as an aerosol. H.sub.2/N.sub.2 (0.07/99.93) mixture through HWG, in the presence of water of 150 ppm, t.sub.set=900 C.; and

[0080] FIGS. 16a-16e show examples of bonding structures of fullerenes on nanotubes: (a) Equilibrium structure of C.sub.42 connected with a CNT via ester group. (b) Equilibrium structure of C.sub.60 weakly covalent bonded defect-free (8,8) CNT; (c) Equilibrium structure of a C.sub.60 weakly covalently bonded above a di-vacancy on a CNT; (d) and (e) Fullerene-molecules, reminiscent of buds, covalently attached to a CNT.

DETAILED DESCRIPTION OF THE INVENTION

[0081] FIG. 1a is a diagram of the structure of the new composition of matter (CBFFCNTs) showing the covalent bonding of fullerenes to CNTs. FIGS. 1b-1e are TEM images of the new CBFFCNT material, wherein one or more fullerenes are covalently bonded to the outer surface of CNTs.

[0082] FIG. 2 shows a block diagram of one embodiment of the method according to the present invention for CBFFCNT production. The first step of the method is to obtain aerosolized or substrate supported catalyst particles from a catalyst particle source. These particles can be produced as part of the process or can come from an existing source. In the reactor, the catalyst particles are heated together with one or more carbon sources and with one or more reagents. The carbon source catalytically decomposes on the surface of catalyst particles together with the reagents to form CBFFCNTs. During and/or after the formation of CBFFCNTs, the entire product or some sampled portion of the product can be selected for further processing steps such as further functionalization, purification, doping, coating and/or mixing. All or a sampled part of the resulting CBFFCNT product can then be collected directly, or incorporated into a functional product material which can further be incorporated in devices.

[0083] FIG. 3(a) shows one embodiment of the method to realize the present invention for the continuous production of CBFFCNTs wherein catalyst particles are grown in situ via decomposition of a catalyst particle precursor. The precursor is introduced from source (4) via carrier gas from a reservoir (2) into the reactor (6). Subsequently, the flow containing the catalyst particle precursor is introduced into the high temperature zone of the reactor (6) through a probe (5) and mixed with additional carbon source flow (1). One or more reagents for CBFFCNT-growth are supplied from reservoir (3) and/or produced catalytically on the reactor wall (7) if the wall is composed of a suitable material which, in combination with one or more carrier gases, precursors and/or carbon sources leads to the catalytic production of suitable reagents.

[0084] FIG. 3(b) shows one embodiment of the method according to the present invention for continuous production of CBFFCNTs, where the catalyst particles are formed by the physical vapor nucleation method from a hot wire generator (HWG) (9) separated in space from the reactor used for the production of one or more CBFFCNTs. In said embodiment, a carbon source and reagents are supplied by a carrier gas passing through a saturator (8). The saturator can also be used to introduce additional reagents for CBFFCNT doping, purification and/or further functionalization. The reagent for CBFFCNT growth can also be produced catalytically on the reactor wall (7) if the wall is composed of a suitable material which, in combination with one or more carrier gases, precursors and/or carbon sources leads to the catalytic production of suitable reagents. Another carrier gas is supplied from a carrier gas reservoir (2) to the HWG (9), which is operated with the help of an electric power supply (10). As the carrier gas passes over the heated wire, it is saturated by the wire material vapor. After passing the hot region of the HWG, the vapor becomes supersaturated, which leads to the formation of particles due to the vapor nucleation and subsequent vapor condensation and cluster coagulation. Inside the CBFFCNT reactor (6) or before, when needed, the two separate flows containing the catalyst particles and the carbon source and reagent(s) are mixed and subsequently heated to the reactor temperature. The carbon source can be introduced through the HWG if it does not react with the wire. Other configurations are possible according to the invention.

[0085] In order to avoid diffusion losses of the catalyst particles and to better control their size, the distance between the HWG and the location where the formation of CBFFCNT occurs, can be adjusted.

[0086] FIG. 3(c) shows one embodiment of the method according to the present invention, wherein the catalyst particles are formed by a physical vapor nucleation method from a hot wire generator smoothly integrated with the reactor. Here, the HWG is located inside the first section of the reactor.

EXAMPLE 1

CBFFCNT Synthesis from Carbon Monoxide as Carbon Source Using Ferrocene as Catalyst Particle Source and Water Vapor and/or Carbon Dioxide as Reagent(s)

[0087] Carbon source: CO.

[0088] Catalyst particle source: ferrocene (partial vapor pressure in the reactor of 0.7 Pa).

[0089] Operating furnace temperatures: 800, 1000, and 1150 C.

[0090] Operating flow rates: CO inner flow (containing ferrocene vapor) of 300 ccm and CO outer flow of 100 ccm.

[0091] Reagent: water vapor at 150 and 270 ppm and/or carbon dioxide at 1500-12000 ppm.

[0092] This example was carried out in the embodiment of the present invention shown in FIG. 3(a). In this embodiment, catalyst particles were grown in situ via ferrocene vapor decomposition. The precursor was vaporized by passing room temperature CO from a gas cylinder (2) (with a flow rate of 300 ccm) through a cartridge (4) filled with the ferrocene powder. Subsequently, the flow containing ferrocene vapour was introduced into the high temperature zone of the ceramic tube reactor through a water-cooling probe (5) and mixed with additional CO flow (1) with a flow rate of 100 ccm.

[0093] Oxidation etching agents, for example water and/or carbon dioxide, were introduced together with the carbon source.

[0094] The partial vapour pressure of ferrocene in the reactor was maintained at 0.7 Pa. The reactor wall set temperature was varied from 800 C. to 1150 C.

[0095] The aerosol product was collected downstream of the reactor either on silver disk filters or on transmission electron microscopy (TEM) grids.

EXAMPLE 2

CBFFCNT Synthesis from a Plurality of Carbon Sources and Reagents and Using Hot Wire Generator as Catalyst Particle Source

[0096] Carbon source: CO, thiophene and octanol.

[0097] Catalyst particle source: hot wire generator.

[0098] Catalyst material: iron wire of 0.25 mm in diameter.

[0099] Operating flow rates: CO flow of 400 ccm through thiophene-octanol (0.5/99.5) solution and hydrogen/nitrogen (7/93) flow of 400 ccm through the HWG.

[0100] Reagent: H.sub.2, octanol and thiophene.

[0101] Operating furnace temperature: 1200 C.

[0102] This example illustrating the synthesis of CBFFCNTs was carried out in the embodiment of the present invention shown in FIG. 3(b). Catalyst particles were produced by vaporizing from a resistively heated iron wire and subsequent cooling in a H.sub.2/N.sub.2 flow. Next the particles were introduced into the reactor. Octanol and thiophene vapor was used as both carbon sources and reagents and were introduced via a saturator (6). Partial pressures for the octanol and thiophene vapours were 9.0 and 70.8 Pa, respectively. Carbon monoxide was used as a carrier gas, carbon source and reagent precursor and was saturated by passing it through the octanol-thiophene solution at the flow rate of Q.sub.CO=400 ccm at room temperature. The reactor walls, saturated with iron, also served as a reagent precursor since CO.sub.2 (about 100 ppm) and water vapor (about 30 ppm) were formed on the walls of the reactor in the heating zone. The products formed with octanol-thiophene in CO are shown in FIG. 11 clearly demonstrating the coating of CNTs with fullerenes.

EXAMPLE 3

CBFFCNT Synthesis from Carbon Monoxide as Carbon Source Using Hot Wire Generator as Catalyst Particle Source and Reagent Introduced or Formed on the Walls of the Reactor

[0103] Reactor tube: stainless steel with a composition of Fe 53, Ni 20, Cr 25, Mn 1.6, Si, C 0.05 weight %.

[0104] Carbon source: CO.

[0105] Catalyst particle source: hot wire generator.

[0106] Catalyst material: iron wire of 0.25 mm in diameter.

[0107] Operating furnace temperature: 928 C.

[0108] Operating flow rates: CO outer flow of 400 ccm and hydrogen/nitrogen (7/93) inner flow of 400 ccm.

[0109] Reagents: H.sub.2, CO.sub.2 and H.sub.2O formed on the reactor walls.

[0110] This example illustrating the synthesis of CBFFCNTs was carried out in the embodiment of the present invention shown in FIG. 3(c), wherein CO was used as both a carbon source and a reagent precursor. The reactor walls, composed of mostly iron, also served as a reagent precursor since CO.sub.2 and water vapor were formed on the walls of the reactor in the heating zone. FIG. 12 shows typical FT-IR spectra obtained at the conditions of CBFFCNTs growth at reactor temperatures of 924 C. The main gaseous products were H.sub.2O and CO.sub.2 with concentrations of 120 and 1540 ppm. It was experimentally found that the effluent composition did not change considerably when the iron particle source was turned off, i.e. when the current through the HWG was off. Accordingly, CO.sub.2 and H.sub.2O formed at the reactor walls. FIGS. 13-15 are examples of CBFFCNTs and their EELS spectra showing the presence of oxygen in covalent bonds between the CNT and fullerene and/or fullerene based molecule.

EXAMPLE 4

Effect of Reagents and Temperature

[0111] This example illustrating the effect of the reagents and/or the temperature on the amount of fullerenes and/or fullerene based molecules formed on the carbon nanotube was carried out using a ferrocene reactor and water vapor and carbon dioxide as reagents. It was found out that the optimal reagent concentrations were between 45 and 245 ppm, preferably between 125 and 185 ppm, for water and between 2000 and 6000 ppm, preferably about 2500 ppm, for carbon dioxide with the highest fullerene density above 1 fullerene/nm.

[0112] When almost no water vapor was used then the carbon nanotubes contained only a small number of fullerenes and/or fullerene based molecules. Further, it was noticed that when using high concentrations of water vapor (>365 ppm) or carbon dioxide (>6250 ppm), the main product contained only few fullerene-functionalized carbon nanotubes.

[0113] Further the effect of the reactor temperature on the product was studied with 145 ppm water vapor introduced in the reactor. At temperatures 1100 and 1150 C. only particles were produced. The maximum fullerene coverage was found at 1000 C. and the amount of fullerenes decreased with decreasing temperature down to 800 C.

[0114] Results

[0115] FIG. 1 shows the typical material produced with the method according to the present invention. HR-TEM images revealed that the coating comprised fullerenes. Their spherical nature has been confirmed by tilting the samples. Statistical measurements performed on the basis of HR-TEM images revealed that the majority of bonded fullerenes comprises C.sub.42 and C.sub.60 (FIG. 4). Importantly, a substantial fraction is C.sub.20 fullerenes, the smallest possible dodecahedra. Such structures have never been seen in samples produced by prior art fullerene production methods.

[0116] Electron Dispersive X-ray Spectroscopy (EDX) and Electron Energy Loss Spectroscopy (EELS) measurements revealed the presence of oxygen in fullerene-functionalized CNT structures. The chemical elemental analysis of the as-produced sample of fullerene-functionalized CNTs was carried out with a field emission transmission electron microscope (Philips CM200 FEG). EELS spectra of the sample synthesized by using pure hydrogen gas through the HWG are shown in FIG. 5. One can see the presence of oxygen in the fullerene-functionalized CNTs indicating a covalent bond via oxygen and/or oxygen containing bridges.

[0117] For an independent characterization of the structures in question, Matrix-Assisted Laser Desorption Ionization Time-of-Flight (MALDI-TOF) mass spectrometric, Ultraviolet-visible (UV-vis) absorption, Fourier Transform Infrared (FT-IR) and Raman spectroscopic measurements on the samples were performed. The UV-vis absorption spectra of a sample in n-hexane are consistent with the presence of both nanotubes and fullerenes (FIG. 6). The characteristic ripple structure at wavelengths above 600 nm is due to van Hove singularities known for CNTs. In addition to characteristic C.sub.60 fullerene peaks (e.g., a weak peak at 256 nm), other bands at 219, 279 and 314 nm appeared shifted or different from 212 and 335 nm fullerene peaks. That can be explained by the presence of various fullerenes as well as strong asymmetry induced by covalent attachment to the nanotube. This asymmetry may remove degeneracy of the electron spectrum to reveal additional bands, i.e. the broadening of existing peaks or the appearance of new ones.

[0118] Since fullerenes are located on the surfaces of CNTs, the fullerene Raman scattering may be similar to surface enhanced Raman scattering (SERS), where metallic CNTs act as an enhancing substrate. The signal from fullerenes was strong for red laser (633 nm) irradiation (the red laser resonantly excites mostly metallic CNTs) as compared to green (514 nm) and blue (488 nm) lasers for which the signal from exclusively semiconducting CNTs can be distinguished. FT, Raman (1064 nm), though out of the metallic CNT resonance wavelength (therefore only a small fraction of sufficiently thick metallic CNTs can respond), still retains very weak fullerene feature at 1400 cm.sup.1 between the D- and G-bands along with a strong fullerene feature from the H.sub.g(1) mode at 265 cm.sup.1. This may occur because the enhancement factor for SERS increases with the wavelength even though the signal itself decreases. Raman spectra of the studied structures show a pronounced G-band at 1600 cm.sup.1, associated with CNTs, and a weak dispersive D-band at 1320-1350 cm.sup.1, depending on the excitation energy. In addition, characteristic features at 1400 cm.sup.1 and 1370 cm.sup.1, may be associated with fullerenes even though they are considerably shifted compared to the 1469 cm.sup.1 peak of the A.sub.g(2) pentagonal mode and 1427 cm.sup.1 peak of the first-order Raman H.sub.g(2) mode for pure C.sub.60. In the case of C.sub.60 modified CNTs of one prior art there was almost no shift in the fullerene signal, which demonstrates that simple mechanical milling of fullerenes with CNTs produces structures fundamentally different from those described in this patent application. Such a dramatic softening of the A.sub.g(2) and H.sub.g(2) modes may correlate with the reconstruction in the electron spectra found in UV due to strong interaction with the CNTs.

[0119] Importantly, the Raman spectrum of C.sub.60-CNT nanocomposites produced by the prior art mechanical milling of fullerenes with CNTs did not show a similar shift in the position of the C.sub.60 peak indicating the fundamental difference between the compared structures.

[0120] The MALDI-TOF spectrum obtained from the fullerene-functionalized CNT sample with dichloromethane as a matrix (FIG. 8) shows peaks of different ionized and hydrogenated fullerenes containing up to three oxygen atoms. The main MALDI-TOF spectrum peaks are attributed to C.sub.60 (C.sub.60H.sub.2, C.sub.60H.sub.2O) and C.sub.42 (C.sub.42COO). Therefore on the basis of the MALDI-TOF measurements one can see that fullerenes are attached to CNTs via either ether (preferable for fullerenes larger than C.sub.54) or ester (for smaller fullerenes) bridges. In order to confirm this, FT-IR measurements were performed (FIG. 9). One can see from the presence of both ether and ester groups in the samples.

[0121] In order to confirm that the fullerenes observed on the CNTs are covalently bonded, it was attempted both to evaporate and to dissolve the attached fullerenes. The presence of fullerenes on the tubes after the heat and solvent treatments would indicate the covalent nature of the attachment between the fullerenes and CNTs. Thermal treatment of the samples in inert helium or argon/hydrogen atmospheres showed no changes in the observed fullerene-CNT structures. A careful washing of the FFCNTs in different solvents (hexane, toluene and decaline) did not result in any significant alteration of the examined samples. Moreover, a mass-spectrometric investigation of the solvent after the CNT washing did not reveal the presence of any dissolved fullerenes further supporting the conclusion that the fullerenes were covalently bonded to the nanotubes.

[0122] Our atomistic density-functional-theory based calculations showed that systems composed of fullerenes covalently bonded through ester groups with single vacancy nanotubes can exist, although the assumed configurations are metastable with respect to forming perfect tubes together with oxidized fullerenes (FIG. 16a). Calculations with a model Hamiltonian that has been successfully applied to describe the formation of peapods and the melting of fullerenes showed that, in addition to oxygen-based bridges, i.e. oxygen containing bridging groups, some fullerenes are directly covalently bonded to CNTs or even make hybrid structures. Results for the different attachments of fullerenes on an (8, 8) nanotube are presented in FIG. 16b-e. One of the viable hybrid geometries involves imperfect fullerenes, for example hemisphere-like fullerenes, covalently bonded to defective nanotubes. Such structures covalently bonded, reminiscent of buds on a branch, are depicted in FIGS. 16d and 16e and can be recognised in HR-TEM images. The local binding energies in these structures suggest that none of the atoms is less stable than those in a C.sub.60 molecule.

[0123] As for the mechanism of the hybrid material formation, HR-TEM observations suggest that both fullerenes and CNTs originate from graphitic carbon precipitated at the surface of, for example, Fe nanoparticles catalysing CO disproportionation. This is supported by Molecular Dynamics simulation results predicting that various carbon nanostructures are formed at the surface of such catalysts. One mechanism for single-walled CNT formation is at steady-state conditions wherein carbon continually precipitates to the catalyst particle surface to form an uninterrupted layer, partially covering the catalyst particle. The presence of heptagonal carbon rings in this layer is a prerequisite for the negative Gaussian curvature found at the location where the nanotube grows from the Fe nanoparticle. This negative curvature, together with instabilities in the forming carbon structure, induced by oxidation etching curling carbon layers, can cause a spontaneous restructuring of the incipient carbon sheet to form fullerenes.

[0124] The uniqueness of this method to produce fullerenes is strongly supported by two facts. First, although C.sub.60 fullerene synthesis is typically not favoured in the presence of abundant hydrogen (since it can damage incipient cages), hydrogen can quickly terminate available dangling bonds and thus stabilise the smaller fullerenes. It is worth noting that hydrogen was either introduced or in situ formed in the described experimental setups. Second, the smallest C.sub.20 fullerenes have not been observed in conventional prior art processes, because, unlike C.sub.60, they are not formed spontaneously in carbon condensation or cluster annealing processes.

[0125] Fullerene-functionalized CNTs are interesting for cold electron field emission (FE) due to the large number of highly curved surfaces acting as emission sites on conductive CNTs. In the material according to the present invention the fullerenes can act as electron emission sites and can lower the FE threshold voltage and increase the emission current. This was confirmed by measuring the FE from a mat of in-plane deposited non-functionalized CNTs and fullerene-functionalized CNTs. The measurements were done using 450 m and 675 m spacer between the cathode and anode, and a 2 mm hole. The averaged current density versus the electric field is shown in FIG. 10a together with the results obtained from the best known field emitters. The FFCNTs demonstrate a low field threshold of about 0.65 V/m and a high current density compared to non-functionalized CNTs. Note that the non-functionalized CNTs synthesised at similar conditions but without adding etching agents had a field threshold for FE as high as 2 V/m. The Fowler-Nordheim plot in the inset of FIG. 10a has a characteristic knee at low currents that corresponds to temporal current pulses which are a manifestation of the discrete nature of electron emission sites (see FIG. 10b). Research demonstrated similar FE performance from the as-produced CoMoCAT sample of single-walled CNTs.

[0126] The chemical nature of the bonding between CNTs and fullerenes can also be confirmed by two additional experimental observations. First it is known that non-covalently attached fullerenes are highly mobile on the surface of CNTs under exposure to a TEM beam, while our TEM observations showed fullerenes to be stationary. Second, FE measurements demonstrated very stable and reproducible electron emission from the CBFFCNT-samples. If the fullerenes were not strongly bonded to CNTs, the effect of their detachment would be experimentally observed as a change in the shape of the current via field strength curve over time.

[0127] The invention is not limited merely to the embodiment examples referred to above; instead many modifications are possible within the scope of the inventive idea defined by the claims.