WO2006061599A2

WO2006061599A2 - Method of manufacturing carbon nanotubes

Info

Publication number: WO2006061599A2
Application number: PCT/GB2005/004680
Authority: WO
Inventors: Takashi Uchino; Peter Ashburn; Cornelis H. De Groot; Konstantinos N. Bourdakos; David C. Smith
Original assignee: University Of Southampton
Priority date: 2004-12-07
Filing date: 2005-12-07
Publication date: 2006-06-15
Also published as: GB0426863D0; WO2006061599A3

Abstract

A method of manufacturing carbon nanotubes using a low-temperature growth process in a metal-catalyst free environment. The carbon nanotubes are grown using chemical vapour deposition on a material which contains the elements Ge and/or C. No metal catalysts are required to enable chemical vapour deposition of carbon nanotubes. The material containing Ge and/or C can be formed on a substrate by in situ growth by chemical vapour deposition or by any other method, such as a combination of vapour deposition of SiGe and C ion implantation. By not using metal catalysts for the CNT growth, the problems associated with these metals compromising CNT device performance are avoided. The method is capable of giving a high yield, and of producing predominantly single wall CNT. Using the method, a variety of CNT devices can be fabricated, such as transistors and field emitters. In an alternative embodiment, the carbon nanotubes are formed from Ge and/or C nanoparticles which are introduced into a C bearing gas stream.

Description

TITLE OF THE INVENTION

METHOD OF MANUFACTURING CARBON NANOTUBES

BACKGROUND OF THE INVENTION

The invention relates to a method of manufacturing carbon nanotubes (CNT) by chemical vapour deposition (CVD) or other deposition techniques, and to CNTs and CNT devices that can be made by this method.

Carbon nanotubes (CNTs) have interesting mechanical, thermal, electronic, optical, semiconducting and other properties which make them suitable for various nanotechnology applications. These applications include conductive polymers, advanced composites (aerospace), fibres, microscope probe tips, field emission devices including displays (Fan et al. Science 283, p.512 (1999) [1], US 6,333,016 [2]) and some membrane applications. Nanotubes can also be used as transistors for electronic circuits and as interconnects between different layers of integrated circuits.

In many applications of interest, such as advanced composites and electronics devices (displays and circuits), metal contaminants can have detrimental effects on the performance (e.g. mechanically, electronically or optically). These metal contaminants can arise from ferromagnetic or other metallic additives that are used as catalysts in the formation of the CNT. Metallic catalysts can be provided by doping the substrate or epitaxial layer on which the CNTs are grown, or through metallic species introduced in the gas phase during CVD.

Although the growth process of CNTs is still not well understood, it is thought that catalytic decomposition of the carbon precursor molecules on the surface of the catalyst particle is followed by diffusion of the released carbon atoms to form the carbon nanotubes. Catalyst particles used successfully in catalytic decomposition include a variety of metals and metal-containing precursors such as Fe, Co, Ni, Mo, Fe₂O₃, Fe₂(SO₄)₃, Fe(NO₃)₃, (NH₄)₆Mo₇O₂₄, Ru(acac)₃, MoO₂Cl₂, Ni(NO₃)₂, Co(NO₃)₂, FeCl₃ (see Moisala et al. J Phys. Cond. Matter 15, 3011 (2003) [3], and references in there) and Ru or a mixture of metals such as Ru/Fe and Fe/Pt (Chem. Mater. 2004, 16, 799-805 [4]). The size (diameter) of the catalyst particle has a strong influence on the size of the resulting nanotube diameter.

It is not easy to remove these impurities without significantly affecting the yield and/or adding to the costs. Alternative CNT growth methods which do not require metal catalysts do exist, but these are expensive, need high temperature process (>1300C), and/or require high vacuum (P<10^"7 Pa) (Kusonoki et al. Phil Mag. Lett. 79, 153 (1999) [5], US 2004/0035355 [6], US 6,740,224 [7]).

A detailed understanding of the growth of CNTs is still lacking, but the microstructure of the substrate is thought to play an important part. Methods have been developed in which the microstructured surface of the substrate assists the CNT growth, but these methods all require metal catalysts (US 2003/0124717 [8], US 6,181,055 [9], US 6,401,526 [1O]₅ US 2003/0124717 [11]).

It is therefore an object of the present invention to provide a convenient high-yield method of manufacturing CNT in a metal-catalyst free environment.

SUMMARY OF THE INVENTION

The invention in one aspect provides a method of manufacturing a carbon nanotube, comprising:

(a) providing a substrate;

(b) providing a basis material on the substrate, wherein the basis material includes at least one of Ge and C; and

(c) depositing C on the basis material to form a carbon nanotube.

It has been discovered that using such a method CNTs can be grown without needing to provide any metal catalyst in the basis material, for example none of the metal catalysts listed in the introduction above is required. In embodiments of the invention, the basis material on which the C is deposited is thus free of metal catalysts.

The method is successful in producing a high yield of CNTs. In particular, the CNTs produced by the method are in many cases entirely or predominantly single wall CNTs.

The Ge/C basis material presented for deposition of the carbon has the property that no metal catalyst particles are required for the carbon deposition to result in growth of CNTs. Although we do not fully understand the exact way the basis material is working as a catalyst for CNT growth, it is likely that the formation of initial carbon clusters is of prime importance and that the addition of Ge and/or C facilitates the segregation of C. Other measures such as oxidation and annealing of the basis material prior to carbon deposition may also promote segregation of the carbon.

In particular, it is possible that during the anneal steps, the basis material undergoes movement of the atoms to produce islands, dots or other nanoparticles containing Ge and possibly other atoms. The segregation of Ge may well be aided by ion implantation and/or an oxidation step after the basis material is deposited. The nanoparticles then act as seeds for the growth of nanotubes via a vapour-liquid-solid mechanism. This mechanism involves the carbon feedstock decomposing on the nanoparticles, which under the growth conditions are liquid, causing the nanoparticles to become saturated with carbon at which point the nanotubes form spontaneously on the surface of the nanoparticle. The decomposition of further feedstock on the nanoparticle then leads to growth of the CNT.

More generally it is believed that the surface structure and composition variation of the basis material on the nanoscale determines the success of the CNT growth.

The deposition of C is advantageously carried out at a relatively low temperature, preferably at a temperature between 700 and 1100 Celsius, more preferably 800 to 1000 Celsius, still more preferably 800 to 900 Celsius. In one example we use a temperature of 850 Celsius.

The invention has been implemented using both Ge and C in the basis material. However, experiments indicate that the method will work successfully with only one of the elements Ge and C being present. More specifically, experiments show that CNTs may be grown with a Ge basis material without C, e.g. with introduction of C by C ion bombardment. Nevertheless, since C is of course added during the growth phase, it is possible that C is still important in some way for the CNT growth mechanism in addition to being the necessary feedstock for CNT growth.

The basis material can further include Si. For example the basis material may be a SiGeC alloy.

The basis material is preferably provided by deposition of at least one of its constituent elements on a substrate. The basis material may be deposited directly on the substrate, or more likely on an intermediate layer formed on the substrate.

In a preferred embodiment, the basis material and the carbon deposition are both formed by a vapour deposition process, such as chemical vapour deposition (CVD) or related methods such as plasma enhanced CVD (PECVD), microwave PECVD (MPECVD) or metallo-organic CVD (MOCVD) or other processes of catalytic decomposition of carbon such as floating catalyst (aerosol) method, or a combination of the above.

The basis material need not be provided by a deposition process. For example, the basis material may instead be provided by ion implantation in which Ge and/or C ions are implanted in a suitable material, such as a Si substrate or other Si layer. Another possibility would be C implantation in a Ge substrate, or C implantation in a SiGe layer, or Ge implantation in a Si substrate.

The basis material may advantageously by prepared to have a structured surface prior to deposition of C to form the CNT.

This may also be done by utilising strain effects caused by lattice mismatch between the basis material and another material bonded thereto, e.g. the basis layer may be SiGeC alloy formed on a Si substrate or buffer layer. These strain effects may be exploited so that Ge islands or dots form in the basis material. In this regard, it is noted that it is well known that the difference in lattice parameter between Si and Ge can cause inhomogeneous growth of Ge or SiGe on a Si substrate (or epitaxial layer). Ge quantum dot structures or other strain related micro structures have been produced in this way. It is also know that incorporation of C into the substrate can modify this inhomogeneous growth behaviour (Schmidt et al. Appl. Phys. Lett. 71, p.2340 (1997) [12] and Wakayama et al. J. Appl. Phys. 93, p.765 (2003) [13]).

A structured surface may also be induced by annealing the basis material, e.g. at between 900 - 1100 Celsius, either instead of or as well as exploiting lattice-mismatch induced strain effects. The anneal prior to C deposition is thought to further promote C segregation and thus assist the CNT growth.

The method may further include implanting the substrate with ions to at least partially amorphise its crystalline structure prior to forming the basis material. The basis layer may incorporate an oxide. An oxide layer may also advantageously be provided on top of the basis material prior to deposition of C to form the CNT. Oxide is thought to further promote C segregation and thus assist the CNT growth.

The basis material may be formed directly or indirectly on the substrate. The substrate may be a Si_xGeyC_1-x--y substrate, such as a Si substrate, a SiC substrate or a Ge substrate.

According to a second aspect of the invention, there is provided a method of manufacturing a carbon nanotube comprising: forming nanoparticles of at least one of Ge and C; and introducing the nanoparticles into a C bearing gas stream to form carbon nanotubes.

In this way the nanoparticles can be made free of metal catalysts and no metal catalysts need be provided in the gas stream.

The nanoparticles may include Ge and C, and optionally may further include Si. In one example, the nanoparticles are composed of a SiGeC alloy.

The nanoparticles can be formed in several different ways, including by gas evaporation, decomposition in organic solvent, and anhydrous solution synthesis.

The nanoparticles may include an oxide. The method may further comprise forming an oxide layer on the nanoparticles prior to introducing the nanoparticles into the C bearing gas stream. The oxide layer can be made of an oxide of Si, Ge or SiGe and can be formed by exposing the nanoparticles to hydrogen peroxide.

The nanoparticles may be carried on a substrate prior to introducing the nanoparticles into the C bearing gas stream, as may be the case if a gas evaporation method is used. The CNT formation may be carried out with the gas stream at a temperature between 700 and 1100 Celsius, more preferably 800 to 1000 Celsius, still more preferably 800 to 900 Celsius. In one example we use a temperature of 850 Celsius. The invention further relates to a device incorporating CNT manufactured by the method of the invention.

The device may be a transistor, e.g. a field effect transistor, or a field emitter. The device may also be a larger device incorporating CNT such as an electronic circuit comprising one or more CNT transistors or other CNT circuit elements, or a field emission display comprising a plurality of CNT field emitters. CNTs may also be used as interconnects, for example vertical interconnects, in place of conventional copper interconnects, between different layers in a multi-layer integrated circuit.

BRIEF DESCRIPTION OF THE DRAWINGS

For a better understanding of the invention and to show how the same may be carried into effect reference is now made by way of example to the accompanying drawings.

Figure 1 is a schematic drawing of a CVD system used to fabricate CNTs according to a first embodiment of the invention.

Figure 2 is a schematic drawing showing the principles of a CNT fabrication process using the CVD system of Figure 1.

Figure 3 is a Scanning Electron Microscopy (SEM) image of a SiGe layer on top of a Si substrate.

Figure 4 is an SEM image of Ge dots formed on a Si substrate.

Figure 5 is an SEM image of a SiGeC layer with C implantation followed by CNT deposition by CVD.

Figure 6 is an SEM image of the structure of Figure 5 after a vapour etch in HF has been applied, showing no adverse effect on the CNTs.

Figure 7 is an energy dispersive x-ray spectroscopy (EDS) image showing the constituents of a sample, showing the absence of metallic catalysts or contaminants.

Figure 8 is a Raman scattering image of the wafer after CNT growth showing a triple peak near 160cm^"1 that is characteristic of single walled CNTs.

Figures 9(a) to 9(c) show a sequence of schematic cross-sections of a device structure during fabrication of a field emission display. Figure 10(a) to 10(d) show a sequence of schematic cross-sections of a device structure during fabrication of a field effect transistor.

Figure 11 (a) to l l(f) show a sequence of schematic cross-sections of a device structure during fabrication of a vertical field effect transistor.

Figure 12 is a plan view of the vertical field effect transistor of Figures 1 l(a) to (f).

Figure 13 is a field-emission SEM image of CNTs fabricated according to a third example of the first embodiment.

Figure 14 shows Raman spectra of CNTs fabricated according to a third example of the first embodiment.

Figure 15 is a high resolution TEM image of CNTs fabricated according to a third example of the first embodiment.

Figure 16 shows field-emission SEM images taken at three process steps of fabrication according to the third example of the first embodiment. Figure 16(a) shows as-grown Ge dots. Figure 16(b) shows the sample after chemical oxidation with hydrogen peroxide. Figure 16(c) shows the sample after argon anneal at 1000 °C.

Figure 17 shows atomic force microscopy results of a substrate after CNT growth according to the third example of the first embodiment. Figure 17(a) is an atomic force microscopy (AFM) image, wherein to obtain the image the surface has been HF vapour etched to remove oxide fibres. The image shows single walled CNTs, crater- like remains of Ge dots, and some remaining Ge dots. Figure 17(b) is a topographical section along the dashed line in Figure 17(a) showing the profile of two of the crater- like features. Figure 18 shows a field-emission SEM image of single walled CNTs made according to a fourth example of the first embodiment.

Figure 19 is a schematic diagram of apparatus for producing CNTs from aerosol according to a second embodiment of the invention.

DETAILED DESCRIPTION

First Embodiment: CNT Deposition on Substrate

According to a first embodiment of the invention, CNTs are fabricated using Chemical Vapour Deposition (CVD). CVD refers to the formation of a non-volatile solid film on a substrate from the reaction of vapour phase chemical reactants containing the right constituents. A reaction chamber is used for this process, into which the reactant gases are introduced to decompose and react with the substrate to form the film.

A way of classifying CVD reactors is by their operating pressures. Atmospheric pressure CVD (APCVD) reactors operate at atmospheric pressure, and are therefore the simplest in design. Low-pressure CVD (LPCVD) reactors operate at medium vacuum (30-250 Pa) and higher temperature than APCVD reactors. Plasma Enhanced CVD (PECVD) reactors also operate under low pressure, but do not depend completely on thermal energy to accelerate the reaction processes, but rather also transfer energy to the reactant gases by using an RF-induced glow discharge.

The thermal APCVD process for the deposition of carbon nanotubes comprises the following steps:

1) a predefined mix of reactant methane (CH₄), hydrogen (H₂) and argon (Ar) is introduced at a specified flow rate into the reaction chamber;

2) the gas species move to the Si substrate with the SiGeC material;

3) the methane decomposes on the SiGeC material. The carbon nanotubes grow from the carbon atoms from the decomposed gas; 4) the gaseous by-products of the reactions are desorbed and evacuated from the reaction chamber.

Figure 1 shows a CVD system used to fabricate CNTs according to the first embodiment of the invention. The CVD system comprises: sources of and feed lines for methane and hydrogen 51, mass flow controllers 52 for metering the gases into the system, a reaction chamber 53, a heating system 54 for heating a wafer 55 placed in the reaction chamber 53.

In use, a wafer 55 is brought into the reaction chamber 53. The reaction chamber 53 is purged several times and the temperature is raised. The gases are allowed to enter the reaction chamber 53 and the decomposition of the methane starts. After a certain period of time, the methane gas flow is halted, the reaction is terminated, and the temperature is lowered.

In a preferred embodiment, the deposition of carbon nanotubes is carried out using methane at a relatively low temperature, preferably at a temperature between a lower bound of 500, 600 or 700 Celsius and an upper bound of 900, 1000 or 1100 Celsius. In one example we use a temperature of 850 Celsius. The time of deposition in this embodiment is between 5 minutes and 25 minutes or between 1 minutes and 60 minutes. A typical time is 10 minutes. Other carbon precursors that can be used are acetylene (C₂H₂), ethylene (C₂H₄), carbon monoxide (CO), benzene, toluene, or many other carbon containing gases. The choice of gas will influence and change the favourable temperature for decomposition.

The basis material used in a preferred embodiment is a layer of SiGeC grown by LPCVD on a Si substrate. The LPCVD process for the deposition of the basis material involves the following steps:

1) a predefined mix of reactant gases and diluent gases is introduced at a specified flow rate into the reaction chamber; 2) the gas species move to the Si substrate;

3) the reactants are adsorbed on the surface of the substrate;

4) the reactants decompose on the substrate, and a continuous or discontinuous layer or dots of the basic material are grown on the substrate depending on the reaction gases, their concentrations and the temperature;

5) the gaseous by-products of the reactions are desorbed and evacuated from the reaction chamber.

The reaction gases used are silane (SiH₆) or dichlorosilane (SiH₂Cl₂), germane (GeH₄), and ethylene(C₂H₄) or methane (CH₄). The carrier gas is hydrogen (H₂). The deposition of the SiGe can be carried out at a relatively high temperature to create epitaxial growth, at a lower temperature to create polycrystalline growth, or at low temperature to create amorphous growth. Preferably the growth is carried out at a temperature between 700 and 1000 Celsius, or 550 and 700 Celsius, or 400 and 550 Celsius. In one example we use a temperature of 850 Celsius. The time of deposition is between 1 minute and 5 minutes, or between 0.1 minutes and 10 minutes. A typical time is 2 minutes, although gases can be turned on sequentially.

Figure 2 shows schematically the growth of the CNTs. A substrate 61 is placed in the growth chamber of a CVD system and a layer of basis material 62 containing Ge ^' and/or C and/or Si is deposited by CVD. (Optionally, this step may be followed by ion implantation.) The basis material layer 62 may undergo some additional treatment, such as oxidation and/or heat treatment. The wafer comprising the substrate 61 with the basis material layer 62 is then loaded into a CVD system for growth of CNTs 63. The same CVD system may be used both for growth of the basis material layer 62 and the CNTs 63. Depending on whether optional steps such as ion implantation are carried out, it may be possible to perform the whole process inside the CVD system, without wafer loading and unloading, so that the wafer does not have to leave the system between growth of the basis material and growth of the CNTs. Alternatively, different CVD systems may be used for the growth of the basis material and CNTs. The basis material may be¹ a layer or other structure containing Ge and/or C₅ and can be formed on a substrate by in-situ growth of a carbon doped (Si)Ge layer by CVD, or by a combination of vapour deposition of SiGe followed by C ion implantation. The substrate could be a Si wafer or any other suitable material, be it insulating, semiconducting or conducting.

Figure 3 is an SEM image showing the example of a basis layer of SiGe formed on a Si substrate.

Figure 4 is another SEM image showing an alternative example in which Ge dots are fabricated on a Si epitaxial layer formed on a Si substrate. In this example, the basis layer is formed from the Ge dot structure by the further step of ion implantation with C.

Figure 5 is a further SEM image showing another alternative example in which a basis layer of SiGeC is provided where the C has been incorporated by ion implantation. The image is taken after CNT growth by CVD with the CNTs being visible as the bright lines.

Some additional steps after or during the creation of the (Si)GeC layer are possibly beneficial to prepare the substrate for CNT growth. One additional step is an anneal, for example at a temperature of around 1000 C. The low melting point and vacuum pressure of Ge as compared to Si enables C segregation at this temperature, possibly forming carbon clusters. Another additional step is formation of a thin oxide layer on top of the deposited layer which may have a positive effect on the structure of the layer in respect of preparation for CNT growth. This oxide layer can be formed by addition of an oxygen containing gas or compound during the deposition of the layer or by chemically treatment afterwards, for instance with hydrogen peroxide. Subsequently, the CNT growth is carried out using a conventional CVD process. In this way, CNTs can be grown by CVD without metal catalysts. Using this process it has been found that a high proportion of the CNTs are single wall. Several variations for the preparation of the basis material have been implemented at the present time, as now illustrated.

Figure 6 shows with another SEM image a structure similar to that of Figure 5 after a vapour etch in HF showing no adverse effect on the CNTs as a result of this etch.

Figure 7 is an energy dispersive x-ray spectroscopy (EDS) image which evinces that the wafer is substantially free of metal catalysts and other contaminants.

Figure 8 is a Raman scattering image of the wafer after CNT growth. The triple peak near 160cm^'1 is characteristic of single walled CNTs. The strength of the peak is indicative of a high proportion of single wall CNTs in the sample.

CNT Growth Example 1.1

A lightly doped <100> silicon wafer is dipped in HF before loading into a CVD system. A 50 run layer of SiGe consisting of equal amounts of Si and Ge is deposited by chemical vapour deposition at 850 Celsius. The wafer is subsequently implanted with C at an energy of 20keV under zero degree angle with a dose of 3e¹⁶ cm^"2. This dose corresponds to a C concentration of around 5% in the SiGe layer. The wafer is dipped into hydrogen peroxide for 10s to create a thin oxide layer. The wafer is subsequently loaded into the furnace for carbon nanotube deposition. The furnace is ramped up to 1000 degrees Celsius in argon atmosphere and the wafer is annealed at this temperature for ten minutes, the temperature is now lowered to 850 degrees

Celsius and the carbon nanotubes are deposited by chemical vapour deposition for 10 minutes in a mixture of argon, hydrogen and methane. The resulting nanotubes are partly single wall carbon nanotubes. CNT Growth Example 1.2

A lightly doped <100> silicon substrate (wafer) is dipped in HF before loading into a CVD system. A layer of SiGeC is deposited by chemical vapour deposition. The layer is slightly oxidised by allowing oxygen-containing gas into the system. The same system is used for pre-anneal and carbon nanotube chemical vapour deposition, such that the vacuum is never broken between SiGeC growth and the carbon nanotube growth.

CNT Growth Example 1.3

A lightly doped <100> silicon wafer is dipped in HF before loading into a CVD system. A thin buffer layer of silicon is deposited followed by the deposition of Ge to form Ge Stransky-Krastinow dots by LPCVD. The wafer is subsequently implanted with C ions at an energy of 30keV under zero degree angle with a dose of 3el6 cm^"2. The wafer is then dipped in buffered HF solution to remove the native oxide and then dipped into 30% hydrogen peroxide solution for 20 minutes followed by immediate drying with water rinse. The wafer is subsequently loaded into the furnace for CNT deposition. The temperature of the furnace is ramped up to 1000 degrees Celsius in argon. The substrate is then annealed at this temperature for 10 minutes in a mixture of argon (1000 seem) and hydrogen (300 seem). The anneal is immediately followed by CNT growth at 850 deg Celsius in a mixture of methane (1000 seem) and hydrogen (300 seem).

Figures 13 to 17 show results of this example.

Figure 13 is a field-emission SEM image of as-grown single walled CNTs and SiGe oxide nanowires.

Figure 14 shows Raman spectra of as-grown fibres using 633 nm (1.96 eV) laser excitation: (a) G-band, (b) anti-Stokes spectrum of the radial breathing mode (RBM) of single walled carbon nanotubes, and (c) G' -band. The RBM modes indicate that the diameters of single walled nanotubes are in range from 1.6 to 2.1 nm.

Figure 15 is a high resolution TEM image of the grown nanofibres. A bundle of single walled CNTs were observed. The TEM image shows that the diameter of single walled CNTs is consistent with Raman results.

Figure 16 shows field-emission SEM images of Ge dots corresponding to the following process steps, (a) as-grown Ge dots, (b) after chemical oxidation with hydrogen peroxide, and (c) after argon anneal at 1000 °C. The as-grown substrates show that Ge dots in the form of cones with diameters in the range 20 to 250 nm and heights in the range 10 to 25 nm. The oxide formation changes the contrast of the dots as observed by SEM, however it makes little difference to the morphology of the dots. The anneal occurs at a temperature above the melting point of Ge (938 ⁰C) and the post-annealed substrates show that considerable changes due to the anneal and in particular the formation of the curly oxide fibres. In addition to the oxide fibres, the substrates also show features reminiscent of Ge dots and much smaller nanoparticles. At the end of the growth process, there is a clear association between the oxide fibres and the carbon nanotubes.

Figure 17(a) is an atomic force microscopy (AFM) image of a substrate after CNT growth and HF vapor etch to remove oxide fibres. The image shows single walled CNTs, crater-like remains of Ge dots, and some remaining Ge dots.

Figure 17(b) is a topographical section along the dashed line in Figure 17(a) showing the profile of two of the crater-like features. The AFM images of Figure 17(a) show that the diameters of the straighter fibres are consistent with bundles of single walled CNTs. The SEM and AFM images both show that within some of these craters are nanoparticles. Whilst these nanoparticles vary in size at least some have dimensions in the few nanometre range. For instance, the nanoparticle shown in the middle of the first crater shown in Figure 17(b) is 3.4 nm high; its lateral dimension cannot be determined due to limitations of the AFM tip size, but the diameter of nanoparticles are expected to be around 3 nm.

CNT Growth Example 1.4

This is the same as Example 1.3, except, the C ion implantation step is omitted.

Figure 18 shows results from this example in the form of a field-emission SEM image of single walled CNTs. The CNTs were only grown near the trenches. These CNTs showed Raman features characteristic of single walled carbon nanotubes.

Further CNT Growth Examples

Instead of depositing a thin layer of SiGe(C) on a separate substrate, such as a Si substrate, a Si(Ge)C substrate could be used on which the carbon nanotubes are directly grown. Li other words, the basis material carried by the substrate is the substrate itself. It will also be understood that the substrate may include epitaxially grown buffer layers of the same or different materials, as is conventional.

These and the other growth processes described in this document may be assisted by carbon segregation in the substrate. This can be accomplished by ion implantation of any species (e.g. Si, C, O, Ar) followed by an anneal which recrystallises the substrate prior to carbon deposition for the CNT growth.

Introduction of water (H₂O) or other oxidising agents intended to prolong the growth of carbon CNT during the deposition process may be introduced during the CVD growth of CNTs.

Second Embodiment: CNT formation in aerosol A second embodiment of the invention has been derived from analysis of experimental results which suggest that the active part of the basis material is in some examples nanoparticles of Ge formed before the CNT growth step. Consequently, according to a second embodiment of the invention, instead of holding the nanoparticles on a substrate for the CNT growth as in the first embodiment, the nanoparticles are distributed in an aerosol for the CNT growth, similar to conventional CNT growth techniques. In the second embodiment, the nanoparticles are created in solution using known methods prior to introduction into the aerosol gas stream, or otherwise produced in a way in which they can subsequently be introduced into an aerosol gas stream.

Accordingly, it is possible to fabricate CNTs without metal catalysts in a hybrid method which uses a conventional aerosol approach for formation of the CNTs, with nanoparticle starting material created as described herein.

Figure 19 is a schematic diagram of apparatus for producing CNTs from aerosol according to the second embodiment of the invention.

An inlet duct 88 is connected to receive one or more carrier gases and one or more carbon source gases from a plurality of inlets 72 controlled by respective mass flow controllers 70. Also connected to the inlet duct 88 is a reservoir 74 for containing in solvent the nanoparticles, for example Ge nanoparticles. This may be in colloid for example. The nanoparticles may be introduced into the inlet duct by means of an aerosol injector 76 comprising a fine nozzle or plurality of nozzles. The injector 76 may include ultrasonic source to agitate the nanoparticles in the solvent and/or a heater to pre-heat the solvent. The inlet duct is connected at its output end to a reaction chamber 80 of a furnace 78 with a heating system 86 for heating the gases inside the chamber to induce CNT formation from the input gases and nanoparticles.

The reaction chamber 80 is connected on its output side to a nanotube trap 82 which in turn on its output side has a vent 84. The method of producing an aerosol of particles from a solvent is now described.

The nanoparticles, for example Ge nanoparticles, are held in a suitable solvent, e.g. methanol, ethanol, water, toluene, benzene, in the reservoir 74. The nanoparticles are then injected through the injector 76 into a flow of gas in the inlet duct 88. The gas flow will include the process gas or gases, including a carbon-containing gas, an inert carrier gas. The introduced nanoparticles will then be carried in this gas flow either within droplets of the solvent or as individual particles in a gas containing evaporated solvent. The gas flow with the nanoparticles then enters into the reaction chamber 80 of the furnace 78. If the nanoparticles enter the furnace in solvent droplets, the heat generated by the heating system 86 will cause the solvent droplets to evaporate. The injector 76 may use ultrasonic agitation and/or heating to assist the process of producing the aerosol.

In an alternative method, the nanoparticles are formed in the reaction chamber 80 from Ge containing reagents introduced into the reactor.

For example, GeH₄ or tetramethyl germanium could be introduced in the feed into the furnace and decomposed, either within the main heated zone of the furnace or in a pre-treatment zone at either a lower or higher temperature, to form germanium particles. Another example would be a source of evaporated or sputtered germanium atoms, dimers or small clusters introduced into the gas feed which coalesce in the furnace to form germanium particles.

Examples of the second embodiment are now described.

Example 2.1

Form Ge nanoparticles as described above in the first embodiment Example 1.3

Introduce the nanoparticles into a stream of a mixture of gases including a carbon source as well as a carrier gas. The carbon source could, for example, be methane, ethane or ethylene. The carrier gas could, for example, be argon, hydrogen, oxygen, water or ethanol, or a suitable mixture of these.

Heat this mixture of gases to initiate CNT growth at a temperature of around 850 degrees Celsius.

Example 2.2

Same as Example 2.1 except that the first step of forming the Ge nanoparticles is different. In this example, the Ge nanoparticles are formed by a known gas evaporation method of Bostedt et al [14] in which Ge nanocrystals with sizes ranging from 1 to 5 nm were evaporated on Si substrate at 1200 ⁰C.

Example 2.3

Same as Example 2.1 except that the first step of forming the Ge nanoparticles is different. In this example, the Ge nanoparticles are formed by a known method of Gerion et al [15] based on decomposition of source gas in organic solvent. Solution synthesis was used to form Ge nanocrystals with sizes ranging from 5 to 30 nm. The source gas (tetraethylgermane) was decomposed at 400 ⁰C into organic solvents such as methanol, hexane, and toluene.

Example 2.4

Same as Example 2.1 except that the first step of forming the Ge nanoparticles is different. In this example, the Ge nanoparticles are formed by a known method of Gerung et al [16] using anhydrous solution synthesis of germanium nanocrystals from the germanium(II) precursor Ge[N(SiMe₃)2]₂.

Device Fabrication Example 1 : Emitter Tips for Field Emission Displays A method of forming emitter tips for use in a field emission display. A carbon-doped SiGe layer, an insulating layer, and a conductor layer are formed on a Si substrate in sequence. An annular groove is formed in the conductive layer and the insulating layer. A cavity is formed by isotropic wet etching. Carbon nanotubes are grown on the carbon-doped SiGe layer using CVD.

Figures 9(a) to 9(c) show a process flow and cross-section diagrams of a field emission display. A conductive cathode 13, an insulating layer 12, and a carbon doped SiGe layer 11 are formed on a substrate 10. A plurality of CNTs 2 are formed on the SiGe layer 11, connecting to a tip cavity 3 and an opening 4 nearby.

Figure 9(a) shows the process step of depositing a carbon-doped SiGe layer on a Si substrate.

Figure 9(b) shows the process step of depositing an insulating layer, and a conductive layer on a Si substrate sequentially.

Figure 9(c) shows the process step of forming an annular groove in the conductive layer and the insulating layer. The annular opening on the conductive layer is formed by photolithography and etching. Then annular groove is formed by isotropic wet etching the conductive layer and the insulating layer through the annular opening. After removing the photoresist, the carbon nanotube is grown by CVD to form an emitter tip.

Device Fabrication Example 2: Field Effect Transistor

A self-aligned carbon nanotube field effect transistor comprises a SiGe source and drain formed on a substrate, and a carbon nanotube deposited on a source and drain, and a gate formed over a substantial portion of the carbon nanotube, separated from the carbon nanotube by a dielectric film. Figures 10(a) to 10(d) show a process flow and cross-section diagrams of a field effect transistor according to this example. An insulating layer 21 and a carbon doped SiGe layer 11 are formed on a substrate 10. Source 5 and drain 6 are defined by photolithography and dry etching. CNTs 22 are grown on the SiGe layer 11 using CVD, connecting to source 5 and drain 6. Gate dielectric 23 is formed to wrap around carbon nanotube channel 22. Gate poly-Si electrode 24 is formed on the channel. An insulating SiO₂ layer 25 is deposited over the entire wafer using CVD. Then the contact windows are defined and the SiO₂ layer in the contact windows is etched away to expose the SiGe 5, 6 or poly-Si 25. Metal is deposited over the entire wafer using sputtering and the metal lines 26 are patterned through photolithography and dry etching.

Figure 10(a) shows the source and drain region formed by carbon-doped SiGe. After a thermal oxide formation over a silicon substrate, a carbon-doped SiGe layer is deposited and patterned on the oxide layer.

Figure 10(b) shows channel formation using a carbon nanotube. The carbon nanotube is grown on the carbon doped SiGe to connect the source and drain.

Figure 10(c) shows gate formation. The poly-Si gate is separated from the carbon nanotube by a dielectric layer.

Figure 10(d) shows metal electrode formation. The source, drain, and gate connect to the metal lines, individually.

Device Fabrication Example 3: Vertical Field Effect Transistor

A vertical self-aligned carbon nanotube field effect transistor comprises a SiGe source and metal drain, and poly-Si gate. A carbon-doped SiGe layer, an insulating layer, a conductor layer, and another insulating layer are formed on a Si substrate in sequence.

An opening is formed to expose the SiGe layer. CNTs are grown on the carbon-doped SiGe layer in the opening by CVD. Metal is then deposited over the entire wafer using sputtering. An insulating layer is deposited to form a gate dielectric. Then the drain pad is patterned. Subsequently drain and gate electrodes are formed.

Figures 11 (a) to ll(f) show a process flow and cross-section diagrams of a vertical field effect transistor according to this example.

Figure 12 shows a plan view of the vertical field effect transistor comprising source, drain, and gate electrodes.

The device is fabricated as follows. A SiO₂ layer 27, a doped poly-Si layer 28, and a Si₃N₄ layer 29 are formed on a substrate 10. After the formation of an opening 40, a single-walled carbon nanotube 30 is grown on the exposed SiGe layer 11. The other side of the carbon nanotube reaches to the outside of the opening. The length of the carbon nanotube is around 1 μm. Thus, the width of the opening should be less than lμm. A metal layer 31 is deposited over the entire wafer using sputtering. Metal is not deposited on the sidewall of the opening and carbon nanotube because of the poor coverage of the sputtering technique. An insulating layer 32 is deposited and the drain pad is patterned through photolithography and dry etching. Subsequently drain and gate electrodes 33 are formed.

Figure 11 (a) shows the carbon-doped SiGe layer on Si substrate.

Figure ll(b) shows a stack of SiO₂, poly-Si and Si₃N₄ layer.

Figure ll(c) shows channel formation using a carbon nanotube. The carbon nanotube is grown on the carbon doped SiGe in the window.

Figure ll(d) shows metal contact formation for source and drain.

Figure ll(e) shows drain pad formation. After a gate dielectric formation, the oxide and metal are patterned.

Figure ll(f) shows metal electrode formation. The drain and gate connect to the metal lines. The source electrode is formed on the backside of the substrate.

REFERENCES

1. Fan et al. Science 283, p.512 (1999)

2. US 6,333,016 3. Moisala et al. J Phys. Cond. Matter 15, 3011 (2003)

4. Chem. Mater. 2004, 16, 799-805

5. Kusonoki et al. Phil Mag. Lett. 79, 153 (1999)

6. US 2004/0035355

7. US 6,740,224 8. US 2003/0124717

9. US 6,181,055

10. US 6,401,526

11. US 2003/0124717

12. Schmidt et al. Appl. Phys. Lett. 71, p.2340 (1997) 13. Wakayama et al. J. Appl. Phys. 93, p.765 (2003)

14. C. Bostedt, T. Buuren, J. M. Plitzko, T. Moller, and L. J. Terminello, J. Phys. Cond. Matt. 15, 1017 (2003)

15. D. Gerion, N. Zaitseva, C. Saw, M. Casula, S. Fakra, T. Buuren, and G. Galli, Nano Lett. 4, 597 (2004) 16. H. Gerung, SD Bunge, TJ Boyle, CJ Brinker and SM Han, Chemical Communications page 1914 (2005)

Claims

1. A method of manufacturing a carbon nanotube comprising:

(a) providing a substrate;

(c) depositing C on the basis material to form a carbon nanotube.

2. The method of claim 1, wherein the basis material is free of metal catalysts.

3. The method of claim 1 or 2, wherein no metal catalysts are provided in gas ^' phase during deposition of the C.

4. The method of claim 1 , 2 or 3, wherein the basis material includes Ge and C.

5. The method of any preceding claim, wherein the basis material further includes Si.

6. The method of any preceding claim, wherein the basis material is a SiGeC alloy.

7. The method of any preceding claim, wherein the C is deposited by a vapour deposition process.

8. The method of any preceding claim, wherein the basis material is provided by deposition of at least one of Ge and C.

9. The method of claim 8, wherein the basis material is provided by a vapour deposition process.

10. The method of claim 8, wherein the basis material is provided by a chemical vapour deposition process.

11. The method of claim 10, wherein the chemical vapour deposition process is carried out with a gas containing oxygen.

12. The method of any one of claim 1 to 7, wherein the basis material is provided by implantation of Ge and/or C ions.

13. The method of any preceding claim, wherein the basis material is prepared to have a structured surface prior to the deposition of C to form the CNT.

14. The method of claim 13, wherein the structured surface is induced at least in part by strain in the basis material.

15. The method of claim 14, wherein the strain is induced by a lattice mismatch between the basis material and another material bonded thereto.

16. The method of claim 13, 14 or 15, wherein the structured surface is induced at least in part by annealing the basis material.

17. The method of claim 16, wherein the anneal is performed between 900 - 1100 Celsius.

18. The method of any one of claims 13 to 17, wherein the structured surface is induced at least in part by implanting the substrate with ions to at least partially amorphise its crystalline structure prior to forming the basis material.

19. The method of any preceding claim, wherein the basis material includes an oxide.

20. The method of any preceding claim, further comprising forming an oxide layer on the basis material prior to the deposition of C.

21. The method of claim 20, wherein the oxide layer is made of an oxide of Si, Ge or SiGe.

22. The method of claim 20 or 21 , wherein the oxide layer is formed by exposing the basis material to hydrogen peroxide.

23. The method of any one of claims 1 to 22, wherein the basis material is integral with the substrate.

24. The method of any one of claims 1 to 22, wherein the basis material is formed as a layer carried by the substrate.

25. The method of any preceding claim, wherein the substrate is formed of Si_xGeyC_1-x--y.

26. The method of claim 25, wherein y=0 and the substrate is formed of Si_xC)_-x.

27. The method of any one of the preceding claims, wherein the deposition of C is carried out at by CVD at a temperature between 700 and 1100 Celsius.

28. The method of any one of the preceding claims, wherein Ge islands or dots form in the basis material.

29. A method of manufacturing a carbon nanotube comprising:

(a) forming nanoparticles of at least one of Ge and C; and

(b) introducing the nanoparticles into a C bearing gas stream to form carbon nanotubes.

30. The method of claim 29, wherein the nanoparticles are free of metal catalysts.

31. The method of claim 29 or 30, wherein no metal catalysts are provided in the gas stream.

32. The method of claim 29, 30 or 31 , wherein the nanoparticles include Ge and C.

33. The method of any one of claims 29 to 32, wherein the nanoparticles further include Si.

34. The method of any one of claims 29 to 33, wherein the nanoparticles are composed of a SiGeC alloy.

35. The method of any one of claims 29 to 33, wherein the nanoparticles are formed by gas evaporation.

36. The method of any one of claims 29 to 33, wherein the nanoparticles are formed by decomposition in organic solvent.

37. The method of any one of claims 29 to 33, wherein the nanoparticles are formed by anhydrous solution synthesis.

38. The method of any one of claims 29 to 37, wherein the nanoparticles include an oxide.

39. The method of any one of claims 29 to 38, further comprising forming an oxide layer on the nanoparticles prior to introducing the nanoparticles into the C bearing gas stream.

40. The method of claim 39, wherein the oxide layer is made of an oxide of Si, Ge or SiGe.

41. The method of claim 39 or 40, wherein the oxide layer is formed by exposing the nanoparticles to hydrogen peroxide.

42. The method of any one of claims 29 to 41 , wherein the nanoparticles are carried on a substrate prior to introducing the nanoparticles into the C bearing gas stream.

43. The method of any one of claims 29 to 42, wherein the CNT formation is carried out with the gas stream at a temperature between 700 and 1100 Celsius.

44. A device incorporating CNT manufactured by the method of any preceding claim.

45. The device of claim 44, wherein the device is an electronic component.

46. The device of claim 44, wherein the device is a transistor.

47. The device of claim 44, wherein the device is a field emitter.

48. An electronic circuit comprising CNT electronic components manufactured by the method of any one of claims 1 to 43.

49. A field emission display comprising a plurality of CNT field emitters manufactured by the method of any one of claims 1 to 43.

50. A multi-layer integrated circuit comprising a plurality of CNTs manufactured by the method of any one of claims 1 to 43, wherein the CNTs are arranged to form interconnects between different layers of the integrated circuit.