Double-Walled Carbon Nanotubes: Synthesis, Structural Characteristics, And Advanced Applications In Nanoelectronics And Energy Systems

Molecular Composition And Structural Characteristics Of Double-Walled Carbon Nanotubes

Double-walled carbon nanotubes consist of two concentric cylindrical graphene sheets arranged in a nested configuration, where each layer comprises sp²-hybridized carbon atoms organized predominantly in hexagonal lattices with occasional pentagonal defects at tube terminations 13. The inner-layer carbon nanotube typically exhibits diameters between 0.5 and 2.5 nm, while the outer-layer nanotube ranges from 1.5 to 6 nm, with an interlayer spacing of approximately 0.34 nm consistent with the van der Waals gap in graphite 714. This dual-wall architecture provides DWCNTs with structural properties intermediate between SWCNTs and MWCNTs: the inner tube preserves the quantum confinement effects and ballistic electron transport characteristic of single-walled structures, while the outer tube enhances mechanical robustness and protects the inner tube from environmental degradation 210.

The electronic properties of DWCNTs depend critically on the chirality indices (n,m) of both constituent tubes, which determine whether each layer behaves as a metal or semiconductor. Theoretical calculations and experimental Raman spectroscopy studies reveal that the G-band shift under mechanical strain differs between inner and outer tubes, with strain transfer ratios (slope of inner-layer G-band shift to outer-layer G-band shift) ranging from 0.5 to 1.5 in high-quality DWCNT-polymer composites 25. This strain distribution indicates partial mechanical coupling between the two walls mediated by van der Waals interactions, enabling load transfer while maintaining independent electronic characteristics in each layer.

Structural characterization via transmission electron microscopy (TEM) and Raman spectroscopy confirms that high-purity DWCNTs (≥98 mass%) exhibit narrow diameter distributions and minimal defect densities compared to larger MWCNTs 14. The radial breathing mode (RBM) in Raman spectra provides a fingerprint for identifying DWCNTs, with characteristic frequencies inversely proportional to tube diameter: RBM frequencies for inner tubes typically appear at 250-350 cm⁻¹, while outer tubes show peaks at 150-250 cm⁻¹ 10. The intensity ratio of the D-band (defect-induced) to G-band (graphitic) serves as a quality metric, with I_D/I_G ratios below 0.1 indicating highly crystalline DWCNTs suitable for electronic applications 12.

Chemical Vapor Deposition Synthesis Routes For Double-Walled Carbon Nanotubes

Catalyst Design And Particle Size Control

The selective synthesis of DWCNTs via CVD requires precise control over catalyst particle size, as the nanotube wall number correlates directly with catalyst diameter 67. Iron-based catalysts with controlled thin-film thicknesses between 1.5 and 2.0 nm yield predominantly double-walled structures when subjected to thermal annealing at 600-900°C in reducing atmospheres 14. The catalyst preparation involves depositing transition metal precursors (Fe, Co, Ni, or bimetallic Fe-Mo combinations) onto oxide supports such as Al₂O₃, MgO, or SiO₂, followed by calcination and reduction steps that generate metal nanoparticles with narrow size distributions 813.

A critical innovation in DWCNT synthesis involves the use of bimetallic catalysts, particularly Fe-Mo systems, where molybdenum acts as a co-catalyst to suppress the formation of amorphous carbon and promote selective growth of double-walled structures 12. The optimal Fe:Mo molar ratio ranges from 1:0.02 to 1:0.1, with molybdenum modifying the catalyst surface energy and carbon diffusion kinetics to favor nucleation of two-layer tubes 8. Catalyst particle size distributions can be controlled through sol-gel methods, impregnation techniques, or physical vapor deposition of thin metal films, with particle diameters in the 2-5 nm range yielding the highest DWCNT selectivity (>50% of total nanotube population) 714.

CVD Process Parameters And Growth Mechanisms

The CVD synthesis of DWCNTs typically employs hydrocarbon precursors such as methane (CH₄), ethylene (C₂H₄), acetylene (C₂H₂), or liquefied petroleum gas (LPG) at temperatures between 600 and 1000°C 68. The reaction mechanism involves thermal decomposition of the carbon source on catalyst surfaces, followed by carbon dissolution into metal particles and precipitation as graphene layers that wrap around the catalyst to form tubular structures 13. The growth rate and wall number depend on the balance between carbon supply rate and carbon diffusion rate through the catalyst particle: conditions favoring carbon-limited growth (low precursor partial pressure, high hydrogen dilution) promote SWCNT and DWCNT formation, while carbon-rich conditions lead to MWCNT growth and catalyst encapsulation 1113.

A key process innovation involves the introduction of oxidizing agents, particularly water vapor (H₂O), into the CVD reactor atmosphere 714. Water vapor concentrations of 100-1000 ppm selectively etch amorphous carbon and defective nanotube structures while preserving high-quality DWCNTs, resulting in enhanced purity (>90%) and increased growth rates (1-10 g/h) 8. The water-assisted CVD mechanism involves hydroxyl radicals that oxidize weakly bonded carbon species and maintain catalyst activity by preventing excessive carbon coating 7. Typical process conditions for high-purity DWCNT synthesis include: temperature 750-850°C, hydrocarbon flow rate 50-200 sccm, hydrogen flow rate 200-500 sccm, water vapor partial pressure 0.01-0.1 kPa, and growth time 10-60 minutes 614.

Vertically Aligned DWCNT Forests And Bulk Structures

Advanced CVD methods enable the growth of vertically aligned DWCNT arrays (forests) with heights ranging from 0.1 μm to several millimeters on planar substrates 714. These aligned bulk structures are produced by depositing catalyst thin films on silicon wafers with intermediate buffer layers (Al₂O₃, TiN) that prevent catalyst-substrate alloying and promote uniform particle formation 7. During CVD growth, the high density of nucleation sites and crowding effects between adjacent nanotubes enforce vertical alignment perpendicular to the substrate surface. The resulting DWCNT forests exhibit exceptional properties including: areal densities of 10¹⁰-10¹² tubes/cm², packing fractions of 10-30%, specific surface areas exceeding 400 m²/g, and electrical conductivities of 10²-10⁴ S/m along the alignment axis 714.

The growth kinetics of aligned DWCNT forests follow a time-dependent height evolution described by h(t) = h₀ + k·t^n, where h₀ is the initial height, k is a growth rate constant (typically 0.1-10 μm/min), and n is an exponent (0.5-1.0) reflecting the growth mechanism 7. Catalyst deactivation occurs through particle coarsening, carbon encapsulation, or oxidation, limiting forest heights to 0.1-5 mm under standard conditions 14. Post-synthesis purification involves oxidative treatments with CO₂ or air at 400-600°C to remove residual catalyst particles and amorphous carbon, followed by acid washing (HCl, HNO₃) to dissolve metal oxides, yielding DWCNT materials with purities exceeding 98 mass% 814.

Separation And Enrichment Methodologies For High-Purity Double-Walled Carbon Nanotubes

Thermal Oxidation And Selective Etching

As-synthesized DWCNT materials typically contain mixtures of SWCNTs, DWCNTs, and MWCNTs with three or more walls, necessitating post-synthesis separation to obtain enriched DWCNT populations 10. High-temperature oxidation in air or oxygen atmospheres at 400-550°C preferentially removes SWCNTs due to their higher surface curvature and lower thermal stability, while DWCNTs and MWCNTs exhibit similar oxidation resistance 10. The oxidation kinetics follow an Arrhenius relationship with activation energies of 120-150 kJ/mol for SWCNTs and 180-220 kJ/mol for DWCNTs, enabling selective SWCNT removal while preserving 70-90% of the initial DWCNT content 10. However, thermal oxidation introduces structural defects and functional groups (carboxyl, hydroxyl) that can degrade electronic properties, requiring careful optimization of temperature and duration 10.

Density Gradient Ultracentrifugation

Density gradient ultracentrifugation (DGU) provides a powerful method for separating nanotubes by wall number, diameter, and electronic type 10. The technique involves dispersing nanotube mixtures in surfactant solutions (sodium dodecyl sulfate, sodium cholate) and layering the dispersion over density gradients formed by iodixanol or sucrose solutions with densities ranging from 1.0 to 1.4 g/cm³ 10. Ultracentrifugation at 100,000-300,000 g for 4-24 hours causes nanotubes to migrate to equilibrium positions where their buoyant density matches the surrounding medium, with SWCNTs (density ~1.3 g/cm³), DWCNTs (density ~1.4 g/cm³), and MWCNTs (density ~1.5-2.0 g/cm³) forming distinct bands 10. Fractionation of these bands followed by surfactant removal yields DWCNT-enriched samples with purities exceeding 80%, though the method is limited to small-scale separations (milligram quantities) and requires extensive optimization of surfactant type and concentration 10.

Chemical Functionalization And Chromatographic Separation

Selective chemical functionalization exploits reactivity differences between nanotube types to enable chromatographic separation 10. Metallic nanotubes exhibit higher reactivity toward diazonium salts and other electrophiles compared to semiconducting tubes, enabling electronic-type separation via ion-exchange or size-exclusion chromatography 10. For DWCNT enrichment, functionalization strategies target the outer tube while preserving the inner tube's pristine structure, leveraging the differential reactivity of the two walls 10. Gel permeation chromatography using agarose or sephacryl columns can separate nanotubes by effective diameter, with DWCNTs (outer diameter 2-6 nm) eluting between SWCNTs (diameter 0.5-2 nm) and larger MWCNTs (diameter >6 nm) 10. These methods achieve DWCNT purities of 60-90% but require multi-step procedures and may introduce residual functional groups that affect subsequent applications 10.

Mechanical Properties And Strain Transfer Mechanisms In DWCNT Composites

Double-walled carbon nanotubes exhibit exceptional mechanical properties including Young's moduli of 0.5-1.5 TPa, tensile strengths of 50-150 GPa, and failure strains of 10-30%, positioning them among the strongest known materials 25. These properties derive from the strong in-plane covalent bonding of sp² carbon atoms and the seamless cylindrical structure that distributes loads uniformly 3. In polymer composites, the mechanical reinforcement efficiency depends critically on stress transfer from the matrix to the nanotubes, which occurs through interfacial shear at the outer tube surface and interlayer shear between the inner and outer walls 25.

Raman spectroscopy under applied strain provides quantitative insights into load transfer mechanisms in DWCNT composites 25. When a composite is subjected to tensile strain, the G-band peak (associated with C-C stretching vibrations) shifts to lower wavenumbers due to bond elongation, with shift rates of -10 to -30 cm⁻¹/% strain for the outer tube and -5 to -20 cm⁻¹/% strain for the inner tube 25. The ratio of inner-to-outer tube G-band shift slopes ranges from 0.5 to 1.5 in well-dispersed DWCNT-resin composites, indicating partial mechanical coupling between the walls 25. Ratios approaching 1.0 suggest strong interlayer coupling and efficient load transfer to the inner tube, while ratios below 0.5 indicate sliding between walls due to weak van der Waals interactions 25.

The mechanical performance of DWCNT composites depends on nanotube dispersion quality, alignment, and interfacial bonding with the matrix 25. Composites containing aligned DWCNTs exhibit anisotropic mechanical properties with Young's moduli of 20-100 GPa along the alignment direction and 5-20 GPa perpendicular to alignment 2. Functionalization of the outer tube with carboxyl, amine, or silane groups enhances interfacial adhesion and stress transfer efficiency, increasing composite tensile strength by 50-200% compared to pristine DWCNT composites 25. Optimal DWCNT loading fractions range from 0.1 to 5 wt%, with higher loadings leading to nanotube aggregation and reduced reinforcement efficiency 2.

Electronic Properties And Applications In Field-Effect Transistors

Band Structure And Electronic Type Distribution

The electronic properties of DWCNTs arise from quantum confinement of electrons in the radial direction, resulting in one-dimensional electronic band structures 310. Each constituent tube can be metallic (when n-m is divisible by 3) or semiconducting (otherwise), leading to four possible DWCNT configurations: metal-metal (M-M), metal-semiconductor (M-S), semiconductor-metal (S-M), and semiconductor-semiconductor (S-S) 310. The electronic type distribution in as-synthesized DWCNTs typically follows statistical predictions with approximately 33% metallic and 67% semiconducting tubes for each wall, yielding ~11% M-M, ~44% M-S/S-M, and ~44% S-S DWCNTs 10. The band gap of semiconducting tubes scales inversely with diameter according to E_g ≈ 0.8 eV·nm / d, resulting in band gaps of 0.3-0.8 eV for typical DWCNT diameters 3.

Field-Effect Transistor Performance Parameters

DWCNTs demonstrate superior performance in field-effect transistor (FET) applications compared to SWCNTs and MWCNTs due to their intermediate diameter, reduced defect density, and enhanced current-carrying capacity 310. DWCNT-FETs fabricated with semiconducting S-S tubes exhibit on/off current ratios of 10⁴-10⁷, subthreshold swings of 70-150 mV/decade, and carrier mobilities of 1000-10,000 cm²/V·s at room temperature 3. The presence of the outer tube provides mechanical protection and reduces environmental sensitivity (hysteresis, threshold voltage shifts) compared to SWCNT-FETs, while maintaining the high mobility and electrostatic control of single-walled structures 310.

The fabrication of DWCNT-FETs involves depositing individual or small bundles of DWCNTs onto SiO₂/Si substrates, followed by photolithographic patterning of source and drain electrodes (Pd, Ti/Au) with channel lengths of 0.1-10 μm 3. The silicon substrate serves as a back gate electrode, enabling electrostatic modulation of the nanotube conductance 3. Device performance depends critically on the electronic type of both DWCNT walls: M-M tubes exhibit metallic behavior with no gate modulation, M-S and S-M tubes show ambipolar characteristics with moderate on/off ratios (10²-10