Carbon Nanotubes and Bucky Materials

Part of the Springer Handbooks book series (SPRINGERHAND)

Abstract

The chapter details the underlying phenomena that underpin electronic applications that have followed from the discoveries of C60 and carbon nanotubes. The reduced dimensionality of these self-organised structures, high electron mobility, weak electromigration, and the plethora of quantum electronic effects exhibited by these structures suggest they are serious candidates for molecular electronics. The detail of the surface chemistry and conditions of synthesis assume greater importance than for conventional electronic materials since all atoms are on the exterior of these structures, as is outlined with references to the wider literature. Essential electronic structure information is given with reference to the transport measurements that have contributed greatly to the evolution of the field with emphasis on the Coulomb blockade and ballistic transport phenomena. The major electronic applications are then outlined, giving the state-of-the-art figures of merit for performance and comments on prospects for realisation.

The discovery of C60 (Fig. 47.1) in 1986, and subsequently higher fullerenes, followed by that of carbon nanotubes in 1991 accelerated the interest in carbon-based molecules and compounds as electronically or optically active materials that had begun in the early 1970s. The award of the 2010 Physics Nobel Prize to Andre Geim and Konstantin Novoselov for the isolation and groundbreaking experiments on graphene emphazised the importance of carbon-based electronic materials. The emergence of the interdisciplinary field of nanotechnology has beneficially attracted attention and methodologies from other fields of science and engineering that have considerably enhanced the physical electronics research effort [47.1, 47.2].
Fig. 47.1

The C60 molecule, diameter 0.7 nm

47.1 Carbon Nanotubes

47.1.1 General

The electronic and structural description of carbon nanotubes and other sp2-bonded carbon nanostructures with curved surfaces (bucky materials, the term bucky is derived from buckminsterfullerene: the name given to C60 soon after discovery) will inevitably contain references to that of the familiar two-dimensional (2-D) planar counterpart, graphite. The strength of the in-plane covalent bond of the hexagonal graphitic network produces a short C–C distance of 0.142 nm and high mechanical stiffness.

Carbon nanotube is a generic term that includes both single- and multi-walled structures. Single-wall carbon nanotubes (SWNT s) are seamless cylindrical graphitic macromolecules of nanoscale diameter and micrometer length, capped by hemispherical ends. The closure of the cylinder is a result of pentagonal inclusions in the hexagonal carbon network of the nanotube walls, Fig. 47.2. For SWNTs, every carbon atom is a surface atom. The smallest SWNT diameter reported to date, 4 Å, corresponds to the predicted lower limit for stable SWNT formation from consideration of the stress energy built into the cylindrical structure [47.3]. SWNTs tend to form closely packed bundles, or ropes, of several tens of individual single-wall nanotubes.
Fig. 47.2

Single-wall carbon nanotube, length 1–10 μm, diameter typically 1.4 to > 10 nm

Multi-wall nanotubes (MWNT s) comprise several to tens of concentric cylinders of these graphitic shells with a layer spacing of 3.4 Å (the inter-planer spacing of graphite). SWNT diameters are typically ≈ 1 nm and a MWNT diameter can be in the range 2–100 nm with typical values of 10–20 nm. Nanotube lengths are typically 1–10 μm but can be ≈ 1 mm, giving rise to astonishing aspect ratios (length/diameter) of > 1000. MWNTs of similar dimensions can also be formed from a single graphite sheet rolled into a scroll structure [47.4]. Double-wall carbon nanotubes are a recent addition to the set of carbon nanostructures.

A consequence of the high C–C bond strength is that the in-plane Young’s modulus of a carbon nanotube is ≅ 1 TPa, making the carbon nanotube one of the stiffest known materials, while the tensile strength is ≅ 150 Gpa (nanotubes have 600 times the strength/weight ratio of steel). Nanotubes deform elastically by buckling, have very low defect density over hundreds to thousands of interatomic spacings, and the strain energy built into the cylindrical structure tends to promote self-repair. The stiff nanotube structure means that the optical-phonon population is minimal even at room temperature; the result can be ballistic electron transport over micrometer length scales (several orders of magnitude greater than that in conventional semiconductor nanostructures).

Another consequence of the short C–C distance is spatially extensive overlapping hybridised atomic orbitals – a π-electron system – with highly mobile delocalised electrons on the interior and exterior of the structure. One carbon atom contributes one unpaired π-electron to the nanotube. The promise of very high carrier mobility in the π-electron system is one motivation for carbon-based electronics; mobilities of circa \({\mathrm{10^{5}}}\,{\mathrm{cm^{2}V^{-1}s^{-1}}}\) have been measured in SWNTs at 50 K and suspended graphene flakes at 240 K [47.5, 47.6].

Another is the range of electronic attributes that are displayed by the carbon allotropes, e. g., room-temperature resistivity can range from 10−8 Ωm for SWNTs to > 1018 Ωm for crystalline C60.

The π-electron system and stiff structure produce electrical conductivity, very high current-carrying capacity with weak electromigration. The best figures of merit for nanotube systems that have been measured since discovery are: electrical conductivity , \({\mathrm{10^{8}}}\,{\mathrm{{\Upomega}^{-1}m^{-1}}}\) (comparable with that of copper), carrier mobility : 104 cm2 ∕ Vs (comparable with that of high-quality GaAs), thermal conductivity 104 W ∕ mK (comparable with that of diamond), and sustainable current density of 1013 A ∕ m2 (> 1000 times greater than that of a metal nanowire).

Nanotube circumferential crystal momentum vectors are quantised due to the periodic boundary conditions imposed by the finite nanoscale diameter whereas those directed along the micrometer-scale axis show continuous variation; thus, one-dimensional (1-D) electronic transport is intrinsic for the SWNT. Electronic systems of reduced dimensionality – dots, wires and sheets of charge – created by top-down silicon processing have assumed importance in conventional microelectronics. Carbon nanostructures offer the possibility of self-organised molecular-level low-dimensional electronic systems, e. g.: zero-dimensional (0-D) (C60 and short SWNTs), 1-D (SWNTs), 2-D (graphene, large-diameter MWNT outer shell), and three-dimensional (3-D) (coupled-layer MWNTs).

Electronic modification of carbon nanotubes by insertion of C60 and other fullerenes into the central capillary to form a mixed-dimensionality hybrid structure called the peapod is a science promises a rich future for research and applications [47.7].

Carbon nanotubes are chemically stable in ambient conditions and structurally stable in vacuum for temperatures far greater than 1000C. Viewed in total, the above properties potentially make carbon nanotubes excellent electronic materials. These considerations place carbon nanotube electronics among a number of competing technologies poised to complement or replace silicon-based complementary metal–oxide–semiconductor (CMOS ) technology in the < 100 nm-feature-size domain. Individual SWNT-based field effect transistors, unipolar and bipolar, show potential for integrated optoelectronic applications. Much recently attention has been given to the production of large-area flexible transparent conducting SWNT networks with indium-doped tin oxide-matching electrical and optical performance [47.8].

47.1.2 Geometry

The unique feature of the electronic structure of carbon nanotubes is that it can range from the metallic to the semiconductor depending on the details of the microstructure. Essentially, the way the hexagonal network connects to itself to form a cylinder determines the electronic structure: the chiral vector usually used to describe the wrapping of the network is
$$\boldsymbol{C}=n\boldsymbol{a}_{1}+m\boldsymbol{a}_{2}\equiv(n,m)\;,$$
where a1 and a2 are the unit vectors of the hexagonal network, and n and m are integers. The resulting nanotube can then be described in the form (n, m), Fig. 47.3.
Fig. 47.3

A map of conductivity type for an individual SWNT indicating the n, and m indices of the chiral vector and the unit vectors a1 and a2

Electronic band-structure calculations show that an (n , m) nanotube is metallic at room temperature if 2n + m is a multiple of 3, otherwise it is a semiconductor with a band gap of Eg = 0.9 deV, where d is the nanotube diameter expressed in nanometers, i. e., typically Eg ≈ 0.5 eV. This remarkable property signals the possibility of band-gap engineering by control of the microstructure. The wrapping angle, or chiral angle, is given by the angle between a1 and C. From geometric and symmetry considerations,
$$\begin{aligned}\displaystyle&\displaystyle 0<|m|<n\;,\quad{\mathrm{0}}^{\circ}=\theta={\mathrm{30}}^{\circ}\;,\\ \displaystyle&\displaystyle\cos(\theta)=2n+\frac{m}{2\sqrt{(n^{2}+m^{2}+nm)}}\;,\end{aligned}$$
and nanotube diameter \(d={\mathrm{0.078}}\sqrt{(n^{2}+m^{2}+nm)}\) nanometers. The general case (n , m) is referred to as the chiral nanotube ; there are two special cases: (a) the zigzag nanotube (m = 0, \(\theta={\mathrm{0}}^{\circ}\)), and (b) the armchair nanotube (n = m, \(\theta={\mathrm{30}}^{\circ}\), all metallic) [47.2]. Assuming a random distribution of nanotube diameters in the reaction products, the relative populations of metallic to semiconducting electronic structure is 1 : 3. Single-symmetry (n , m) synthesis or post-synthesis sorting according to electronic structure are key fields for research effort. The task is proving rather difficult but there appear to be no fundamental barriers to progress.

A stable defect in the hexagonal network is the pentagon–heptagon pair (Stone–Wales or 5–7 defect). This defect will cause a sharp bend in an otherwise well-graphitised SWNT . Controlled introduction of defects allows the possibility of constructing Y- and T-junctions and other complex geometries from nanotubes, including rings and coils. Moreover, a 5–7 defect can connect a metallic to a semiconductor nanotube giving an Angstrom-scale heterojunction and hence a device density 104 times greater than present-day microelectronics [47.9].

47.1.3 Synthesis and Chemistry

The quality of carbon nanotubes in terms of crystallinity and impurity content depends on the conditions of synthesis by arc discharge using graphitic electrodes, laser vaporisation of graphite, or chemical vapour deposition (CVD ) using hydrocarbon gas and metal nanoparticle catalysts. Each method produces different nanotube samples and sample-to-sample variation from the same source. An extensive review of synthesis methods was published by Rakov [47.10]. The key factor for crystallinity, and therefore electronic quality, is the formation temperature: \(> {\mathrm{2000}}\,{\mathrm{{}^{\circ}\mathrm{C}}}\) is required for complete graphitisation. Little is know about the preferences the various growth processes have for certain nanotube symmetries over others.

Reaction products can contain significant quantities of residual metal catalyst particles and non-tubular forms of carbon (typically 30 wt% of the raw material). Purification processes ranging from the simple to complex are commonly used: oxidation and acid washing [47.10], solvent treatment followed by ultrafiltration [47.11], or flocculation using aqueous surfactants [47.12]. Controlled growth of MWNTs perpendicular to a substrate has been widely explored due to its technological importance [47.13]

Surface functionalisation of SWNTs is required for several applications as a means of modifying charge exchange with the ambient. Oxidation of nanotubes with the acid mixture H2SO4–HNO3 leads to high concentrations of carboxylic, carbonyl, and hydroxyl groups on the surface and removal of the tips to expose the interior surface. Carboxyl (−COOH) groups are then readily derivatised by a variety of reactions. Covalent functionalisation, however, necessarily disrupts the rigid structure and π-electron system with consequent degradation of mechanical and electrical properties but provides the best stability and accessibility [47.14]. Non-covalent routes to nanotube functionalisation offer ease of synthesis and minimum disruption of the tubular structure [47.15].

Exposure to ambient humidity, oxygen and other gases profoundly affects the measured electronic properties of nanotubes through charge exchange and the quality of metallic electrical contacts made to nanotubes. The electronic properties of carbon nanotube systems are extremely sensitive to the presence of molecular oxygen due to the formation of the charge-transfer complex C p +δ –O 2 δ , i. e., oxygen-exposed nanotubes are p-type [47.16]. Nanotubes exhibit ultrahigh sensitivity at room temperature to O2, N2O, NH3 but not H2 [47.17]. Nanotubes must be elevated to temperatures \(> {\mathrm{700}}\,{\mathrm{{}^{\circ}\mathrm{C}}}\) in ultrahigh vacuum to completely remove absorbents.

47.1.4 Electronic Structure and Transport

The electronic structure of SWNTs is predicted by tight-binding calculations extended from the 2-D approach for graphite to the cylindrical 1-D nanotube system [47.2], Fig. 47.4 . The tight-binding modelling follows from the original band structure calculations of graphite by Wallace in 1947 [47.18].
Fig. 47.4

(a) Electron energy E versus wavevector k showing the intersection of the dispersion relations of two bands at the Fermi energy EF; electron occupancy is indicated by shading. The essentially zero-gap semiconductor structure is common to those sp2-bonded carbon allotropes for which k can vary in two dimensions resulting in a conical shape of the E-k plots (graphene, carbon nanotubes). (b) The case of semiconducting SWNT; a band-gap appears owing to the combination of the conical E-k shape and quantisation of the circumferential crystal momentum vector

In the case of nanotubes circumferential crystal momentum vectors are quantised due to the periodic boundary conditions imposed by a finite nanoscale diameter. This simple approach has been proven, mainly by low-temperature scanning tunnelling spectroscopy, to be a good description of the electronic structure of carbon nanotubes [47.19]. The curvature of the nanotubes is ignored in this model but can introduce some important modifications [47.20]. However, the original model is, nevertheless, adequate for the classification of nanotubes into metallic and semiconducting types according the simple rule outlined in Sect. 47.1.1 . In common with other 1-D conductors with half-filled bands the expectation was that carbon nanotubes will undergo a Peierls distortion, i. e., and a gap will open in the band structure at the Fermi energy due to atomic displacement modifying the unit cell. However, it has been shown theoretically that the effect very quickly diminishes as a function of nanotube radius [47.2].

Electrical transport measurements on individual SWNTs , SWNT ropes, and MWNTs have produced diverse results in the decade since discovery and are still the focus of intense research effort. The picture is complicated by several factors:
  1. 1.

    The current pathway through MWNTs is greatly complicated by interlayer coupling.

     
  2. 2.

    Two-terminal resistances in the range 100 Ω–100 MΩ were first reported [47.21].

     
  3. 3.

    The processing, contacting, and manipulation methods used undoubtedly introduced non-intrinsic resistances and potential barriers.

     
Nevertheless, the best transport parameters that have emerged since discovery, quoted in Sect. 47.1, are highly desirable from many device points of view. The quality of lithographically defined contacts is critical for electrical transport measurements. There are three possible outcomes of the contacting process:
  1. 1.

    The contacts are of good quality; then the total resistance will be the diffusive or quantum resistance for that structure (6.5 kΩ for a ballistic SWNT) [47.22] .

     
  2. 2.

    No contact is made and the nanotube is isolated by tunnel barriers; in this case the nanotube will act as a Coulomb island and blockade phenomena will be observed at low temperature [47.23].

     
  3. 3.

    The contacts are of an intermediate type that allows the nanotube to act as a cavity and Fabry–Perot interference phenomena are observed [47.24].

     
The exact physics of contacting to carbon nanotubes is poorly understood; the exact atomic arrangement at the interface is likely to play a role as are the ambient gases.

Successful transport measurements have yielded a plethora of quantum-electronic and low-dimensional transport phenomena . Tans et al. [47.23] demonstrated true quantum wire behaviour; the remarkable feature of this work was to identification of an electron phase-coherence length on the micrometer scale at 4 K for nanotubes: several orders of magnitude greater than that in conventional semiconductor nanostructures . With an extraordinarily skilful experiment Bachtold et al. [47.25] demonstrated the Aharonov–Bohm effect by circulating circumferential current in a contacted MWNT with the axis parallel to a magnetic field. The periodicity of the magneto-conductance variation with applied field corresponded to an interferometer diameter equal to that of the outer MWNT shell. This finding gave weight to an earlier conclusion from quantum conductance measurements of Frank et al. [47.26], namely that current is only carried in the outer shell for MWNTs. This was posed as a general statement but was contradicted by the controlled layer-by-layer conductivity analysis of Collins et al. [47.27] in which the current-carrying contribution of each MWNT layer was determined by vaporisation of successive MWNT layers. This work concluded that at least eight layers contribute to the current-carrying capacity of a MWNT.

The landmark paper of Frank et al. identified ballistic conductance in MWNT systems at room temperature by controlled lowering of a MWNT bundle into a liquid mercury contact [47.26]. This was a remarkable result since the implication is that the electron mean free path is on the micrometer scale at room temperature. A ballistic conductor has a conductance given by the Landauer formula
$$G=2\left(\frac{e^{2}}{h}\right)\sum_{i}^{N}T_{i}\;,$$
where N is the number of conducting channels, T is the transmission factor (experimental result: T = 1 for most nanotube systems), and \(G_{0}=2e^{2}/h\) is the quantum conductance (the factor of 2 is due to spin degeneracy). Frank et al. measured a conductance of 1G0 >  per MWNT, implying that current is carried in the outer shell only and there is a missing 1G0 since a metallic carbon nanotube has two crossing 1-D energy bands at the Fermi level, which yields a conductance of 2G0 for a single nanotube shell. However, subsequent results have shown the expected 2G0 result [47.28]. Therefore, the controversy over the experiment by Frank et al. remains unsolved.

SWNTs were expected to exhibit the predicted electron–electron interaction effects in 1-D systems, the so called Luttinger liquid (LL ) state. In a strictly 1-D system the independent electron approximation breaks down and the normal 3-D Fermi gas picture is replaced by the LL. In the LL description electron–electron interactions dominate and the electrons are in a highly correlated state. The LL state requires a high degree of order since disorder destroys the correlated ground state. The main features of the LL are suppression of the density of states at the Fermi energy according to a power law and separation of spin and charge excitations. In transport measurements, the LL state is manifested as a tunnelling conductance from a normal electrode according to G ( T )  ≈ T α , for eV ≪ kBT, and at large voltages \(G(V)=\mathrm{d}I/\mathrm{d}V\approx V^{\alpha}\) for eV ≫ kBT, where the parameter α is determined by the strength of the interaction between the electrons. Such power laws have been observed in contacted SWNTs, although stronger evidence, such as a clear demonstration of spin–charge separation, in necessary to identify the LL state unambiguously [47.29].

Coulomb blockade phenomena have been observed in low-temperature transport measurements on carbon nanotube systems often by unintentional incorporation of tunnel junctions between the metal electrode and the nanotube. The effect occurs when a mesoscopic island with capacitance C (self-capacitance of the nanotube plus the barrier capacitance at the nanotube–electrode junction) is isolated by tunnel barriers from two electrodes. If the charging energy for a single electron on the island, e2 ∕ C, is larger than the thermal energy kBT, then the system will not conduct due to electron–electron repulsion. Coulomb blockade can be modulated by a gate electrode from full blocking to a conducting state limited only by the tunnel barriers. Such a system forms a single-electron transistor (SET). The capacitance of a nanotube is \(\approx{\mathrm{10^{-17}}}{F}\), thus SET operation conditions are fulfilled at temperatures below 10 K when the entire nanotube defines the Coulomb island. To minimise the capacitance and therefore elevate the operating temperature inter-tube islands can be defined by crossing nanotubes or intra-tube islands bending with a scanning probe microscope (kinks and bends in SWNTs can create tunnel barriers). The latter approach has been used to create a SET that operates close to room temperature [47.30]. Graphene is likely to be the carbon allotrope to produce usable SET devices owing to the ease with which flakes can be lithographically processed [47.31].

Superconductivity in nanotube systems was first observed as the proximity effect with SWNTs or SWNT ropes connecting two superconducting electrodes [47.32]. The first report signalling intrinsic superconductivity in SWNT ropes was by Kociak et al. in 2001 [47.33]. Subsequently, 1-D superconductivity has been observed in arrays of small-diameter SWNTs at 15 K and in indvidual MWNTs up to 12 K [47.34, 47.35].

Two broad categories of exploitation of the above electronic attributes have arisen; nanoelectronic devices, with an individual SWNT or MWNT as the acive channel in a source-drain configuration deposited on the top SiO2 layer of a heavily doped silicon substrate (which serves as a backgate), and applications which exploit networks of ropes or bundled SWNTs, Fig. 47.5. Although, the intermediate approach of using highly aligned SWNT arrays as the active channel has had some success [47.36].
Fig. 47.5

(a) Example of a contacted individual MWNT on the insulating SiO2 layer of a conducting silicon substrate. (b) Example of a network comprising SWNT ropes depostied on a quartz substrate

47.1.5 Nanoelectronic Devices

Semiconducting nanotubes are especially important for nanoelectronic device applications. Semiconducting behaviour has been observed in individual SWNTs; in SWNT ropes and MWNTs mixed metallic and semiconducting current pathways greatly complicate the IV characteristic. Field-effect transistors (FET s) with individual SWNT channels have been demonstrated and found to have higher transconductance than can be achieved with state-of-the-art metal–oxide–semiconductor field-effect transistors (MOSFET s) [47.37]. The conducting state is attained with negative gate voltages, implying that the carbon nanotube forms a normally off p-type conduction channel (probably due to unintentional doping by exposure to oxygen). Logic gates made from nanotube FETs have been demonstrated [47.38]. The nature of the metal electrode–SWNT Schottky barrier is still not fully understood.

The prospects for highly integrated circuits consisting of nanotube elements must be considered remote at this time for several reasons (some previously mentioned):
  1. 1.

    Current synthesis techniques produce mixtures of metallic or semiconductor nanotubes and these tend to form ropes or bundles.

     
  2. 2.

    The precise nature of the metal electrode–nanotube contact is poorly understood.

     
  3. 3.

    Scanning probe manipulation techniques used to fabricate prototype individual devices are not scalable.

     
Nevertheless, carbon nanotube-based FETs, both unipolar and bipolar, show some promise for intergrated optoelectronic applications [47.39].

47.1.6 Networks

This work is largely driven by the necessity to find an alternative to the conventional transparent conducting electrode material, indium-doped tin oxide (ITO ), owing to the terrestrial rarity of indium and the desirability of flexibility for some emerging technologies. Individual SWNTs constitute ideal elements for flexible transparent conducting networks with ITO-matching electrical and optical performance, namely, that is, sheet resistance much less than \(<{\mathrm{100}}\,{\mathrm{{\Upomega}/sq}}\) optical transmission of 90% at 550 nm wavelength. The basic network element is unavoidably a bundle rather than an individual SWNT. Electrical transport within the bundle and the inter-bundle contact resistance has greatest influence on the electrical transport within the network; contact resistances can be minimised by careful removal of residual surfactant introduced by solution processing [47.11, 47.5, 47.6]. SWNT networks have been modelled as metallic filamentary pathways interrupted by thin tunnel barriers, backscattering by zone-boundary phonons, fluctuation-assisted tunnelling, and variable-range hopping (VRH ). The contribution of each component is a function of network thickness; VRH dominates in the thinnest networks and metallic conductivity increases with thickness [47.16, 47.17]. Approaches to transparent conducting electrode production include chemical doping and electrical modification by various methods, including acid treatment [47.11, 47.12], decoration by metal nanoparticles [47.13, 47.3, 47.40], conducting polymer composites, graphene-SWNT hybrids [47.14, 47.15], metallic enrichment, and purification through ultracentrifugation [47.7].

Many approaches for the production of such hybrid networks have been reported, including, physical evaporation of metal nanoparticles onto the SWNT network, nanoparticle attachment by chemical reaction with functionalised SWNTs, and electroless nanoparticle deposition methods [47.24, 47.29, 47.30, 47.31, 47.32, 47.33].

The room temperature sheet resistance of the resultant hybrid networks has been reported to be both increased and decreased relative to the un-decorated SWNT network [47.29, 47.34, 47.35]. A 1 ∕ 1500 reduction of room temperature sheet resistance of SWNT network by Au-nanoparticle decoration with an optical transmittance of 90% at 550 nm wavelength has been reported [47.40].

Metal nanoparticle-decorated SWNT networks show promise for selective gas sensing based on differential analyte-nanoparticle reactivity and charge exchanges with the network. There are also useful plasmonic and thermoelectric benefits from such hybrid systems [47.20, 47.21, 47.22]. The metal nanoparticle-network interaction is not well understood. Decoration of individual SWNTs with gold nanoparticles results in electron transfer from the nanotube to the nanoparticle and the formation of a potential barrier at the SWNT-metal interface; the barrier characteristics depend on the work function of the metal [47.13, 47.23, 47.24, 47.26, 47.27, 47.28, 47.36].

47.1.7 Other Electronic Applications

Although nanoelectronics has been the driving force for carbon nanotube research, many other electronic applications are being explored; this section outlines some of them.

Interconnects

Ballistic transport on the micrometer scale at room temperature and high current-carrying capacity suggests carbon nanotubes are good candidates for high-bandwidth dissipation-less interconnect for nanoscale circuit elements. Since chemical processing tends to degrade nanotube electrical properties, methods of directed growth are being explored as a means of wiring integrated circuits, e. g., guiding through via holes [47.41], by electric field [47.42], or surface modification [47.43].

Field-Emission Displays

The nanoscale diameter, aspect ratio of ≈ 1000, and high conductivity make carbon nanotubes ideal candidates for field emitters [47.44]. Field-emission currents from single nanotubes and aligned or randomly oriented nanotube thin films have been extensively studied. The promise is of low-threshold-field electron emission with a current density sufficient to drive a phosphor screen for display purposes. SWNTs and MWNTs have proved to be remarkably good field emitters with threshold emission fields in the range 1–10 V ∕ μm and capable of carrying very high current density. The exact emission mechanism, the role played by surface absorbents, and the cause of emission current saturation are under debate. Nevertheless, major manufacturers have recently produced prototype field-emission displays with carbon nanotube sources; with further development, this is likely to be the first major application of carbon nanotubes to reach the marketplace [47.45].

Electron-Beam Lithography

The present minimum feature size for silicon microelectronics is 130 nm, which is achieved using extreme-UV optical lithography . Alternative technologies are being explored for the future production of < 100 nm devices. Conventional electron-beam lithography can achieve a 5-nm line width but is limited by writing time because a single beam is used to write the entire pattern. Thus arrays of electron guns operating in parallel are being considered as a route to reduction of the writing time while maintaining the high resolution of electron-beam lithography. Carbon nanotubes are currently the most promising candidates for use as the emission source [47.46].

Electro-Optic Materials

Polymer–nanotube composite materials have been studied from the point of view of applications in electro-optics. Loading the layers of organic light-emitting diodes (LED s) with low concentrations of nanotubes effectively increased the lifetime of the devices by preventing the build up of local hot spots thorough the high thermal conductivity that can be achieved in a nanotube percolation network [47.47]. Nanotube-induced local ordering of the matrix polymer suggests efficiency improvements may also be possible [47.48]. Using nanotube mats as an electrode in solar-cell applications apparently gave no improvement in device performance; however, these early measurement were performed with defective MWNTs so there may be a case for further investigation using the high-quality SWNTs available today.

Nanofabricators

Carbon-nanotube atomic force microscopy tips for fabricating oxide nanostructures in Si and Ti by anodisation is a rapidly expanding field aimed at Tbit ∕ cm2 data storage [47.49]. The technique utilises ambient moisture in the oxidation process with the tip biased negative relative to the surface. Line widths of 5–10 nm suitable for antidotes and tunnel junctions can be achieved.

Actuators

Electromechanical actuators based on SWNT sheets have been shown to generate higher stresses than natural muscle when operating in physiological conditions and higher strains than high-modulus ferroelectrics [47.50]. The actuation mechanism is a geometrical expansion of the carbon–carbon covalent bond caused by electrochemical double-layer charging . Work densities per cycle substantially higher than any previously known technology are predicted for SWNT sheets with mechanical properties close to those of individual SWNTs. This can be achieved by inter-tube binding and alignment optimisation.

Sensors

The conductivity of nanotube systems is highly sensitive to gaseous ambients, which affect the sign and amount of injected charge. The dimensions of a nanotube sensing element is such that very low quantities of analyte species will produce a measurable response. Nanotube gas sensors certainly have prospects to challenge conventional gas sensors for certain uses [47.19]. In addition to transduction, bio-sensors require a bio-receptor (e. g., enzyme or cell) immobilisation matrix. Carbon-nanotube-based bio-sensors meet both requirements and have been found to promote homogeneous electron-transfer reactions.

47.2 Bucky Materials

Over the last decade the field of bucky materials has been dominated by carbon nanotubes, as the number of publications and patent submissions testify. After intense research activity since their discovery, the most noteworthy of C60-based electronic materials are superconducting heterofullerides of composition K2MC60, where M = Fe, Ni, Cu, Ag, Co …, with Tc ≈ 15 K. However the future for encapsulation of species in the C60, and higher fullerene, cage appears promising, in particular as the basis of devices for quantum information processing. Other non-carbon-nanotube structures are now under intense investigation; these include BN, W, MoS2, heterogeneous nanotubes and the ternary compound Mo6C9−xH x (C = chalcogen, H = halogen, \(3<x<6\)). The science of these structures is still in its infancy but, given the rich experience of over a decade of carbon nanotube research, there is reason to be optimistic of exciting new science and technology.

References

  1. 47.1
    M. S. Dresselhaus, G. Dresselhaus, P. Avouris (eds): Carbon Nanotubes: Synthesis, Structure, Properties, and Applications, Topics Appl. Phys. Ser., Vol. 80 (Springer, Berlin Heidelberg New York 2000) Google Scholar
  2. 47.2
    R. Saito, G. Dresselhaus, M.S. Dresselhaus: Physical Properties of Carbon Nanotubes (Imperial College Press, London 1998)CrossRefGoogle Scholar
  3. 47.3
    N. Wang, Z.K. Tang, G.D. Li, J.S. Chen: Nature 408, 50 (2000)CrossRefGoogle Scholar
  4. 47.4
    L.M. Viculus, J.J. Mack, R.B. Kaner: Science 299, 1361 (2003)CrossRefGoogle Scholar
  5. 47.5
    S. Ilani, P.L. McEuen: Annu. Rev. Condens. Matter Phys. 1, 1 (2010)CrossRefGoogle Scholar
  6. 47.6
    K.I. Bolotin, K.J. Sikes, J. Hone, H.L. Stormer, P. Kim: Phys. Rev. Lett. 101, 096802 (2008)CrossRefGoogle Scholar
  7. 47.7
    X. Liu, T. Pichler, M. Knupfer, M.S. Golden, J. Fink, H. Kataura, Y. Achiba, K. Hirahara, S. Iijima: Phys. Rev. B 65, 45419 (2002)CrossRefGoogle Scholar
  8. 47.8
    G. Grüner: J. Mater. Chem. 16, 3533 (2006)CrossRefGoogle Scholar
  9. 47.9
    L. Chico, V.H. Crespi, L.X. Benedict, S.G. Louie, M.L. Cohen: Phys. Rev. Lett. 76, 971 (1996)CrossRefGoogle Scholar
  10. 47.10
    E.G. Rakov: Russ. Chem. Rev. 69, 25 (2000)CrossRefGoogle Scholar
  11. 47.11
    A.C. Dillon, T. Genett, K.M. Jones, J.L. Alleman, P.A. Parilla, M.J. Heben: Adv. Mater. 11, 1354 (1999)CrossRefGoogle Scholar
  12. 47.12
    K. Tohji, H. Takahashi, Y. Shinoda, N. Shimizu, B. Jeyadevan, I. Matuoka, Y. Sato, A. Kasuya, S. Ito, Y. Nishina: J. Phys. Chem. B 101, 1974 (1997)CrossRefGoogle Scholar
  13. 47.13
    J.-M. Bonard, T. Stora, J.-P. Salvetat, F. Maier, T. Stoeckli, C. Duschul, L. Forro, W.A. de Heer, A. Chatelain: Adv. Mater. 9, 827 (1997)CrossRefGoogle Scholar
  14. 47.14
    A. Huczko: Appl. Phys. A 74, 617 (2002)CrossRefGoogle Scholar
  15. 47.15
    S.E. Baker, W. Cai, T.L. Lasseter, K.P. Weidkamp, R.J. Hamers: Nano. Lett. 2, 1413 (2002)CrossRefGoogle Scholar
  16. 47.16
    R.J. Chen, Y. Zhang, D. Wang, H. Dai: J. Am. Chem. Soc. 123, 3838 (2001)CrossRefGoogle Scholar
  17. 47.17
    G.U. Sumanasekera, C.K.W. Adu, S. Fang, P.C. Eklund: Phys. Rev. Lett. 85, 1096 (2000)CrossRefGoogle Scholar
  18. 47.18
    P.R. Wallace: Phys. Rev. B 71, 622 (1947)CrossRefGoogle Scholar
  19. 47.19
    J. Kong, N.R. Franklin, C. Zhou, M.G. Chapline, S. Peng, K. Cho, H. Dai: Science 287, 5453 (2000)CrossRefGoogle Scholar
  20. 47.20
    J.W.G. Wildöer, L.C. Venema, A.G. Rinzler, R.E. Smalley, C. Dekker: Nature 391, 59 (1998)CrossRefGoogle Scholar
  21. 47.21
    M. Ouyang, J.L. Huang, C.L. Cheung, C.M. Lieber: Science 292, 702 (2001)CrossRefGoogle Scholar
  22. 47.22
    T.W. Ebbesen, H.J. Lezec, H. Hiura, J.W. Bennett, H.F. Ghaemi, T. Thio: Nature 382, 54 (1996)CrossRefGoogle Scholar
  23. 47.23
    S.J. Tans, M.H. Devoret, H. Dai, A. Hess, R.E. Smalley, L.G. Geerlings, C. Dekker: Nature 386, 474 (1997)CrossRefGoogle Scholar
  24. 47.24
    M. Bockrath, D.H. Cobden, P.L. McEuen, N.G. Chopra, A. Zettl, A. Thess, R.E. Smalley: Science 275, 1922 (1997)CrossRefGoogle Scholar
  25. 47.25
    A. Bachtold, C. Strunk, J.-P. Salvetat, J.-M. Bonard, L. Farrro, T. Nussbaumer, C. Schonenberger: Nature 397, 673 (1999)CrossRefGoogle Scholar
  26. 47.26
    S. Frank, P. Poncharal, Z.L. Wang, W.A. de Heer: Science 280, 1744 (1998)CrossRefGoogle Scholar
  27. 47.27
    P.C. Collins, M.S. Arnold, P. Avouris: Science 292, 1331 (2001)CrossRefGoogle Scholar
  28. 47.28
    W. Liang, M. Bockrath, D. Bozovic, J.H. Hafner, M. Tinkham, H. Park: Nature 411, 665 (2001)CrossRefGoogle Scholar
  29. 47.29
    M. Bockrath, D.H. Cobden, L. Jia, A.L. Rinzler, R.E. Smalley, L. Balents, P.L. McEuen: Nature 397, 598 (1999)CrossRefGoogle Scholar
  30. 47.30
    H.W. Postma, T. Teepen, Z. Yao, M. Grifoni, C. Dekker: Science 293, 76 (2001)CrossRefGoogle Scholar
  31. 47.31
    T. Ihn, J. Güttinger, F. Molitor, S. Schnez, E. Schurtenberger, A. Jacobsen, S. Hellmüller, T. Frey, S. Dröscher, C. Stampfer, K. Ensslin: Mater. Today 13, 44 (2010)CrossRefGoogle Scholar
  32. 47.32
    A.F. Morpurgo, J. Kong, C.M. Marcus, H. Dai: Science 286, 263 (1999)CrossRefGoogle Scholar
  33. 47.33
    M. Kociak, A.Y. Kasumov, S. Gueron, B. Reulet, L. Vaccarrini, I.I. Khodos, Y.B. Gorbatov, V.T. Volkov, H. Bouchiat: Phys. Rev. Lett. 86, 2416 (2001)CrossRefGoogle Scholar
  34. 47.34
    Z. Wang, W. Shi, H. Sie, T. Zhang, N. Wang, Z. Tang, X. Zhang, R. Lortz, P. Sheng, I. Sheikin: A. Demuer, Phys. Rev. B, 81, 174530 (2010)CrossRefGoogle Scholar
  35. 47.35
    I. Takesue, J. Haruyama, N. Kobayashi, S. Chiashi, S. Maruyama, T. Sugai, H. Shinohara: Phys. Rev. Lett. 96, 057001 (2006)CrossRefGoogle Scholar
  36. 47.36
    S.J. Kang, C. Kocabas, T. Ozel, M. Shim, N. Pimparkar, M.A. Alam, S.V. Rotkin, J.A. Rogers: Nat. Nanotechnol. 2, 230 (2007)CrossRefGoogle Scholar
  37. 47.37
    S.J. Wind, J. Appenzaller, R. Martel, V. Derycke, P. Avouris: Appl. Phys. Lett. 80, 3817 (2002)CrossRefGoogle Scholar
  38. 47.38
    A. Bachtold, P. Hadley, T. Nakanishi, C. Dekker: Science 294, 1317 (2001)CrossRefGoogle Scholar
  39. 47.39
    P. Avouris, Z. Chen, V. Perebeinos: Nat. Nanotechnol. 2, 605 (2007)CrossRefGoogle Scholar
  40. 47.40
    C.F. McAndrew, M. Baxendale: Nanotechnology 24, 305202 (2013)CrossRefGoogle Scholar
  41. 47.41
    F. Kreupl, A.P. Graham, G.S. Duesberg, W. Steinhögl, M. Liebau, E. Unger, W. Hönlein: Microelectron. Eng. 64, 399 (2002)CrossRefGoogle Scholar
  42. 47.42
    Y. Zhang, A. Chang, J. Cao, Q. Wang, W. Kim, Y. Li, N. Morris, E. Yenilmez, J. Kong, H. Dai: Appl. Phys. Lett. 79, 3155 (2001)CrossRefGoogle Scholar
  43. 47.43
    J. Liu, M.J. Cavasant, M. Cox, D.A. Walters, P. Boul, L. Wei, A.J. Rimberg, K.A. Smith, D.T. Colbert, R.E. Smalley: Chem. Phys. Lett. 303, 125 (1999)CrossRefGoogle Scholar
  44. 47.44
    J.-M. Bonard, M. Croci, C. Klinke, R. Kurt, O. Noury, N. Weiss: Carbon 40, 1715 (2002)CrossRefGoogle Scholar
  45. 47.45
    N.S. Lee, D.S. Chung, I.T. Han, J.H. Kang, Y.S. Choi, H.Y. Kim, S.H. Park, Y.W. Jin, W.K. Yi, M.J. Yun: Diamond Rel. Mater. 10, 265 (2001)CrossRefGoogle Scholar
  46. 47.46
    W.I. Milne, K.B.K. Teo, M. Chhowalla, G.A.J. Amaratunga, J. Yuan, J. Robertson, P. Legagneux, K. Pirio, K. Bouzehouane, D. Pribat, W. Bruenger, C. Trautmann: Curr. Appl. Phys. 1, 317 (2001)CrossRefGoogle Scholar
  47. 47.47
    B. McCarthy, J.N. Coleman, R. Czerw, A.B. Dalton, M. in het Panhius, A. Maiti, A. Drury, P. Bernier, J.B. Nagy: J. Phys. Chem. B 106, 2210 (2002)CrossRefGoogle Scholar
  48. 47.48
    H. Ago, K. Petritsch, M.S.P. Shaffer, A.H. Windle, R.H. Friend: Adv. Mater. 11, 1281 (1999)CrossRefGoogle Scholar
  49. 47.49
    H. Dai, N. Franklin, J. Han: Appl. Phys. Lett. 73, 1508 (1998)CrossRefGoogle Scholar
  50. 47.50
    R.H. Baughman, C. Cui, A.A. Zakhidov, Z. Iqbal, J.N. Barisci, G.M. Spinks, G.G. Wallace, A. Mazzoldi, D. De Rossi, A.G. Rinzler, O. Jaschinski, S. Roth, M. Kertesz: Science 284, 1340 (1999)CrossRefGoogle Scholar

Copyright information

© Springer International Publishing AG 2017

Authors and Affiliations

  1. 1.School of Physics and AstronomyQueen Mary University of LondonLondonUK

Personalised recommendations