9.2 Nanotubes

Structure of nanotubes

Carbon nanotubes (CNTs) are cylindrical nanostructures formed by rolling a single layer of graphene into a seamless tube. Their one-dimensional geometry produces quantum confinement effects that give rise to remarkable electronic, mechanical, and thermal properties. The specific way the graphene sheet is rolled determines nearly everything about a given nanotube's behavior.

Carbon nanotube types

Nanotubes are classified by the orientation of the graphene lattice relative to the tube axis:

• Zigzag nanotubes have carbon-carbon bonds running parallel to the tube axis.
• Armchair nanotubes have bonds oriented perpendicular to the tube axis.
• Chiral nanotubes have bonds at an intermediate angle, giving the lattice a helical twist.

Each type has a distinct atomic arrangement, which directly determines its electronic character (metallic vs. semiconducting).

Chirality and symmetry

The chiral vector $(n, m)$ specifies exactly how the graphene sheet is rolled to form the tube. It connects two crystallographically equivalent points on the graphene lattice, and the resulting tube's diameter and helical angle follow from this vector.

The chiral vector controls the electronic structure:

• Armchair nanotubes $(n = m)$ are always metallic.
• Zigzag $(m = 0)$ and chiral $(n \neq m, m \neq 0)$ nanotubes are metallic if $(n - m)$ is divisible by 3, and semiconducting otherwise.

Nanotube symmetry operations include rotations about the tube axis, reflections, and translations along the axis. These symmetries simplify the theoretical treatment of their electronic and vibrational properties.

Single-wall vs multi-wall

• Single-wall carbon nanotubes (SWCNTs) consist of one graphene cylinder, with typical diameters of 0.7–2 nm. Their well-defined chirality makes them ideal for studying fundamental quantum effects.
• Multi-wall carbon nanotubes (MWCNTs) contain multiple concentric graphene cylinders, with outer diameters reaching up to ~100 nm. The interlayer spacing is approximately 0.34 nm, matching the layer separation in graphite.

MWCNTs tend to be more thermally stable and easier to produce in bulk, but their electronic properties are harder to control because each wall can have a different chirality.

Electronic properties

The electronic structure of nanotubes arises from confining graphene's two-dimensional electron system onto a one-dimensional cylindrical surface. This confinement quantizes the allowed electron momenta around the circumference and produces the distinctive features of nanotube band structure.

Band structure

Nanotube band structure is derived from graphene's using the zone-folding method: periodic boundary conditions around the tube circumference select only certain allowed wavevectors from graphene's Brillouin zone, slicing the 2D bands into a set of 1D subbands.

A key consequence is the appearance of van Hove singularities (sharp peaks) in the density of states, characteristic of one-dimensional systems.

• Metallic nanotubes have subbands that cross the Fermi level, giving a continuous density of states there.
• Semiconducting nanotubes have a bandgap that scales inversely with diameter:

$E_g = \frac{2\gamma_0 \, a_{C\text{-}C}}{d}$

where $\gamma_0$ is the nearest-neighbor transfer integral (~2.7 eV), $a_{C\text{-}C}$ is the carbon-carbon bond length (~0.142 nm), and $d$ is the nanotube diameter.

Metallic vs semiconducting

The rule is straightforward: a nanotube with chiral indices $(n, m)$ is metallic if $(n - m) \mod 3 = 0$, and semiconducting otherwise. For a random distribution of chiralities, this means roughly 1/3 metallic and 2/3 semiconducting.

Metallic nanotubes can sustain current densities up to $\sim 10^9 \text{ A/cm}^2$, orders of magnitude beyond copper. Semiconducting nanotubes, with their tunable bandgap, are promising channel materials for field-effect transistors and chemical sensors.

Density of states

The 1D confinement produces a density of states (DOS) with sharp van Hove singularities at the onset of each subband.

• In metallic nanotubes, the DOS is nonzero at the Fermi level.
• In semiconducting nanotubes, the DOS is zero within the bandgap.

Optical absorption and emission occur at energies corresponding to transitions between van Hove singularities in the valence and conduction bands (labeled $E_{11}$, $E_{22}$, etc.). Scanning tunneling spectroscopy (STS) can directly map the DOS of individual nanotubes, confirming these features experimentally.

Mechanical properties

The strong $sp^2$ covalent bonds in the graphene lattice give carbon nanotubes extraordinary mechanical performance, making them among the strongest and stiffest materials known.

Tensile strength

Carbon nanotubes have measured tensile strengths up to ~100 GPa, more than 100 times that of high-strength steel, at roughly one-sixth the density. This strength comes directly from the covalent carbon-carbon bonds. Defects and impurities reduce the tensile strength significantly, so real nanotubes fall short of the theoretical maximum for a perfect lattice.

Elastic modulus

The Young's modulus of SWCNTs reaches approximately 1 TPa, comparable to diamond. Their elastic response remains linear over a large strain range, meaning they can deform substantially and return to their original shape. MWCNTs generally show somewhat lower modulus because weak van der Waals interlayer coupling allows the walls to slide relative to each other. Nanotubes also exhibit notable radial elasticity, allowing them to withstand significant lateral compression without permanent deformation.

Thermal conductivity

Along the tube axis, individual nanotubes have measured thermal conductivities exceeding 3000 W/m·K at room temperature. For comparison, diamond reaches ~2000 W/m·K and copper ~400 W/m·K. Heat transport is dominated by phonons (quantized lattice vibrations) rather than electrons. Thermal conductivity depends strongly on tube length, diameter, and defect concentration; shorter tubes and higher defect densities reduce it.

Synthesis methods

Several techniques exist for producing carbon nanotubes, each with trade-offs in quality, yield, scalability, and control over nanotube properties.

Arc discharge

1. A high DC current is passed between two graphite electrodes in an inert gas atmosphere (typically He or Ar).
2. The intense heat vaporizes carbon from the anode.
3. Carbon condenses on the cathode and chamber walls, forming nanotubes along with amorphous carbon and catalyst residue.

Adding metal catalysts (Ni, Co, Fe) to the anode promotes SWCNT growth. Arc discharge produces nanotubes with relatively few structural defects, but yields a mixture of chiralities and requires post-synthesis purification.

Chemical vapor deposition

1. Metal catalyst nanoparticles (Fe, Co, Ni) are deposited on a substrate.
2. A hydrocarbon gas (methane, ethylene, acetylene) is introduced at 600–1200°C.
3. The gas decomposes on the catalyst surface, and carbon atoms assemble into nanotubes that grow from the catalyst particles.

CVD offers the best scalability and the most control over nanotube placement, diameter, and alignment. It can produce vertically aligned nanotube "forests" and patterned arrays suitable for device integration.

Laser ablation

1. A pulsed laser vaporizes a graphite target doped with metal catalysts inside a high-temperature furnace (~1200°C).
2. An inert gas carries the vaporized material downstream.
3. Nanotubes nucleate and grow in the gas flow, collecting on a cooled surface.

This method produces high-quality SWCNTs with a narrow diameter distribution and fewer defects than arc discharge. However, the high energy requirements and equipment costs limit its scalability.

Characterization techniques

Multiple complementary techniques are used to determine nanotube structure, chirality, and electronic properties.

Raman spectroscopy

Raman spectroscopy is a non-destructive optical technique that reveals both structural and electronic information through characteristic vibrational modes:

• Radial breathing mode (RBM): A low-frequency mode where all carbon atoms vibrate radially in phase. Its frequency relates to diameter by $\omega_{RBM} = A/d + B$, where $A$ and $B$ are empirically determined constants and $d$ is the diameter.
• G-band: Arises from in-plane $sp^2$ carbon stretching. In SWCNTs it splits into $G^+$ and $G^-$ peaks; the lineshape of $G^-$ differs between metallic and semiconducting tubes.
• D-band: Indicates structural disorder and defects. The ratio of D-band to G-band intensity ($I_D/I_G$) is a common measure of nanotube quality.

The Kataura plot maps optical transition energies ($E_{ii}$) against nanotube diameter, allowing assignment of $(n, m)$ indices from combined Raman and optical absorption data.

Electron microscopy

• Transmission electron microscopy (TEM) provides atomic-resolution images. It can resolve individual walls in MWCNTs and, with high-resolution imaging, reveal the atomic lattice and defects directly. Electron diffraction patterns can be used to determine chirality.
• Scanning electron microscopy (SEM) is useful for studying nanotube morphology, alignment, and the overall structure of nanotube films and arrays at lower magnification.

Atomic force microscopy

AFM maps the 3D surface topography of nanotubes deposited on substrates, providing diameter, length, and information about bundle formation. Tapping mode minimizes tip-induced damage. Beyond imaging, AFM can mechanically probe individual nanotubes (measuring stiffness and friction) and even manipulate them. Kelvin probe force microscopy, a variant of AFM, measures local work function variations along a nanotube.

Applications

Electronics and sensors

• Nanotube field-effect transistors (CNT-FETs): Semiconducting SWCNTs serve as the channel material, offering high carrier mobility and potential for sub-10 nm scaling.
• Transparent conductive films: Random nanotube networks can replace indium tin oxide (ITO) in flexible displays and touch screens.
• Chemical and biosensors: Nanotube conductivity is highly sensitive to adsorbed molecules, enabling detection of specific gases or biomolecules at very low concentrations.
• Memory devices: Charge storage on individual nanotubes has been explored for non-volatile memory elements.

Energy storage

• Battery electrodes: CNTs added to lithium-ion battery electrodes improve electrical conductivity and accommodate volume changes during cycling, increasing capacity and rate capability.
• Supercapacitors: High surface area and conductivity make nanotube electrodes well-suited for high-power-density energy storage.
• Photovoltaics: Nanotubes serve as electron acceptors or transparent electrodes in organic and hybrid solar cells.
• Thermoelectrics: The combination of high electrical conductivity and (in certain architectures) suppressed thermal conductivity makes nanotubes interesting for thermoelectric energy conversion.

Composite materials

Adding even small weight fractions of CNTs to polymers can significantly enhance mechanical strength, stiffness, and electrical conductivity. Applications include aerospace structural components, sporting goods (bicycle frames, tennis rackets), electromagnetic shielding, and antistatic coatings. Research into self-healing composites uses nanotubes both as reinforcement and as conductive pathways to detect and respond to damage.

Quantum effects

At the nanoscale, quantum phenomena dominate nanotube transport and optical behavior.

Confinement in nanotubes

Electrons are confined to the cylindrical surface of the nanotube. Their momentum around the circumference is quantized, while they remain free to move along the tube axis. This produces discrete energy subbands. Confinement effects grow stronger as diameter decreases, leading to larger subband spacing and larger bandgaps in semiconducting tubes. One notable consequence is the formation of excitons (bound electron-hole pairs) with unusually high binding energies (hundreds of meV), which dominate the optical response of semiconducting SWCNTs.

Ballistic transport

In a defect-free nanotube, electrons can travel micrometers without scattering, even at room temperature. This ballistic transport regime gives metallic nanotubes a quantized conductance:

$G = \frac{4e^2}{h}$

The factor of 4 comes from two spin channels and two degenerate subbands at the Fermi level. Ballistic transport means minimal resistive heating, which is why nanotubes can carry such extreme current densities.

Coulomb blockade

When a short nanotube segment is isolated between tunnel barriers (forming a quantum dot), single-electron charging effects appear. The condition for observing Coulomb blockade is:

$k_B T \ll \frac{e^2}{C}$

where $C$ is the total capacitance of the nanotube quantum dot. When this condition holds, electrons tunnel onto the dot one at a time, producing a stepwise (staircase) increase in current with applied voltage. This effect is the basis for single-electron transistors and has relevance to quantum information processing. The temperature dependence of the blockade also provides a way to extract the electronic level spacing of the nanotube.

Defects and doping

Controlled introduction of defects and dopants allows tailoring of nanotube properties for specific applications.

Structural defects

• Stone-Wales defects: A 90° rotation of a C-C bond converts four hexagons into two pentagons and two heptagons. These defects introduce localized electronic states and reduce mechanical strength.
• Vacancies: Missing carbon atoms create dangling bonds and act as scattering centers for electrons and phonons.
• Interstitials: Extra carbon atoms incorporated between walls (in MWCNTs) or adsorbed on the surface.

Defects can also serve as reactive sites for chemical functionalization or as nucleation points during growth.

Chemical functionalization

Functional groups can be attached to nanotube sidewalls or open ends through covalent chemistry (e.g., carboxylation, amidation, polymer grafting). This improves solubility and processability but disrupts the $sp^2$ network, which degrades electronic and mechanical properties. Non-covalent functionalization (through $\pi$-$\pi$ stacking of aromatic molecules or wrapping with polymers/surfactants) preserves the intrinsic electronic structure while still modifying surface chemistry and dispersibility.

Substitutional doping

Replacing carbon atoms with heteroatoms alters the electronic character:

• Nitrogen doping introduces extra electrons, producing n-type behavior and shifting the Fermi level into the conduction band.
• Boron doping creates electron-deficient sites, producing p-type behavior.

Doping can also enhance catalytic activity (e.g., N-doped CNTs for oxygen reduction in fuel cells). The main challenges are achieving uniform dopant distribution and precise control of dopant concentration.

Nanotube interactions

Van der Waals forces

Van der Waals interactions, arising from fluctuating dipole moments in the electron clouds of adjacent tubes, are the dominant force between nanotubes. The interaction strength depends on tube diameter and separation. For two parallel nanotubes, the van der Waals potential per unit length scales as:

$U(r) = -\frac{A}{12\pi d^2 r^5}$

where $A$ is the Hamaker constant, $d$ is the nanotube diameter, and $r$ is the center-to-center distance. These forces drive bundling, aggregation, and adsorption onto surfaces.

Nanotube bundles

Nanotubes spontaneously aggregate into aligned bundles due to van der Waals attraction. Bundling modifies the electronic properties: intertube coupling in metallic bundles can open a pseudogap at the Fermi level, reducing conductivity. Separating bundles into individual tubes is a persistent challenge. Common approaches include ultrasonication in surfactant solutions and polymer wrapping.

Nanotube-substrate interactions

When deposited on a substrate, nanotubes adhere primarily through van der Waals forces. The substrate's roughness and surface chemistry influence nanotube alignment and orientation. Substrate-induced strain can locally modify the electronic structure of a nanotube, shifting its bandgap or altering its transport properties. Controlling these interactions is essential for reproducible device fabrication.

Theoretical models

Tight-binding approximation

The tight-binding model describes nanotube electronic structure by expressing electron wavefunctions as linear combinations of atomic $p_z$ orbitals. Only nearest-neighbor interactions are included, parameterized by the hopping integral $\gamma_0$. This approach correctly predicts whether a given $(n, m)$ tube is metallic or semiconducting and captures the overall shape of the band structure. It is computationally efficient but misses effects like curvature-induced rehybridization, which matter for small-diameter tubes.

Zone-folding method

Zone folding starts from graphene's known 2D band structure and imposes the periodic boundary condition set by the chiral vector. This quantizes the allowed wavevectors around the circumference into a discrete set, each producing a 1D subband. The method gives accurate low-energy electronic structure for tubes with diameters above ~1 nm. For very small tubes, curvature effects cause significant deviations from the zone-folding predictions, and more sophisticated methods are needed.

Density functional theory

DFT is an ab initio approach that solves for the ground-state electron density using the Kohn-Sham equations, accounting for electron-electron interactions through exchange-correlation functionals. It provides accurate predictions of nanotube geometry, energetics, and electronic structure without empirical parameters. The trade-off is computational cost: DFT calculations are typically limited to unit cells containing a few hundred atoms, restricting studies to small-diameter or short nanotube segments. DFT is especially valuable for studying defects, doping, and functionalization where empirical models lack reliability.