Optimizing Carbon Nanotube Materials and Composites through Advanced Porosimetric Characterization

Background

Pore‑containing media are integral to advanced materials, with pore sizes classified by IUPAC as micropores (<2 nm), mesopores (2–50 nm), and macropores (>50 nm). Carbon nanotubes (CNTs), renowned for their high aspect ratios and specific surface areas, naturally form porous structures when they bundle and entangle into agglomerates. These porous CNT assemblies exhibit remarkable adsorption, separation, and conductive properties, making them attractive as electrode materials, filters, supports, and structural components.

Traditional N₂ adsorption techniques effectively probe micropores and mesopores but cannot capture macropores (>50 nm). Mercury intrusion porosimetry, leveraging mercury’s high surface tension, measures pore diameters from a few nanometers up to several hundred micrometers, encompassing both meso‑ and macropores. While mercury porosimetry has been applied to various carbon materials—carbon fibers, graphite, activated carbon—CNT agglomerates have not been systematically characterized across this full pore range.

To establish the value of mercury porosimetry for CNT agglomerates, we investigated (1) diverse CNT types, (2) multiple agglomerate morphologies, (3) dispersions in different solvents, and (4) various dispersion techniques. These parameters influence pore architecture and, ultimately, material performance. Our study confirms that porosimetry can reliably quantify pore distributions, distinguish agglomerate forms, and predict functional properties such as electrical conductivity in CNT‑rubber composites.

Methods

CNT Synthesis

Super‑Growth single‑walled CNTs (SG SWNTs) were produced via water‑assisted CVD on Fe‑Ni‑Cr alloy foils, employing He/H₂ carrier gases at 1 atm and 750 °C. The resulting CNT forests ranged from 100 µm to 1 mm in height.

Materials

Commercial HiPco SWNTs, CoMoCAT SWNTs, Bayer MWNTs, VGCFs, and fluorinated rubber (Daikin Daiel‑G912) were sourced from established suppliers.

CNT Dispersion

Dispersions were prepared at 0.03 wt% in MIBK, DMF, ethanol, or water using a 60 MPa high‑pressure jet‑mill (single pass). This technique preserves long CNTs while achieving uniform suspension.

Preparation of Buckypapers

0.01 wt% suspensions were filtered through 0.2–0.4 µm membranes, vacuum‑dried at 180 °C, and cut into 4 cm diameter, ~50 µm thick sheets.

Porosimetry of CNT Agglomerates

Mercury intrusion (Quantachrome PoreMaster 60) measured pore diameter (D) and volume via the Washburn equation: D = (–4γ cos θ)/P, where γ = 0.48 N m⁻¹ and θ = 140°. Buckypapers (~50–100 mg) were segmented into ~5 mm² pieces for analysis.

Preparation of CNT Rubber Composite Sheets

SG SWNT/MIBK dispersions (0.125 wt%) were mixed with fluorinated rubber/MIBK solutions, cast into 4 cm disks, and solvent‑evaporated at 25 °C for 16 h followed by 80 °C vacuum drying (6 h). Final composites were ~150 µm thick.

Structural Observation of CNT Agglomerates

Scanning electron microscopy (FE‑SEM S‑4800) examined CNT networks on spin‑coated Si substrates.

Electrical Conductivity Measurement of CNT Rubber Composite Sheets

Four‑point probe (MCP‑T610) measured surface resistance across ten points to calculate average conductivity and standard deviation.

Results and Discussion

Pore Characteristics of Various CNT Types

Mercury porosimetry revealed distinct pore‑size distributions for Buckypapers fabricated from different CNTs. Small‑diameter CNTs (CoMoCAT, HiPco, SG SWNT) produced peaks around tens of nanometers, whereas large‑diameter MWNTs and VGCFs generated broader peaks extending to ~1 µm. SEM confirmed that larger CNTs form sparser networks with larger pores. These findings demonstrate that CNT diameter and wall number directly influence pore breadth and volume.

Effect of Agglomerate Morphology

Three agglomerate forms—sparse SWNT forests, bundle‑network Buckypapers, and aligned, densely packed SWNTs—were compared. Porosimetry showed a decreasing pore volume from forest (0.03 g cm⁻³) to bundle network (0.4 g cm⁻³) to dense alignment (0.6 g cm⁻³). The densest structures exhibited the smallest pores, confirming that bulk density correlates with pore architecture. SEM images corroborated these results.

Solvent‑Dependent Dispersibility and Pore Size

Dispersions prepared in DMF, MIBK, ethanol, and water displayed a clear trend: good solvents (DMF) produced the smallest pores (max pore volume at 22 nm), whereas poor solvents (water) yielded the largest pores (95 nm). Total pore volume also increased with decreasing solvent quality. SEM imaging of spin‑coated networks aligned with the porosimetry data, confirming that solvent choice governs CNT bundle size and pore size.

Influence of Dispersion Technique on Pore Structure and Composite Conductivity

SG SWNTs dispersed by turbulent flow (nanomizer, star burst), cavitation (probe sonicator), and mechanical force (ball mill, bead mill, spin mixer, etc.) produced distinct pore distributions. Turbulent flow yielded the finest bundles with pore peaks at ~60–70 nm and the highest electrical conductivity (33 S cm⁻¹). Cavitation produced slightly larger pores (~56 nm) and moderate conductivity (20 S cm⁻¹). Mechanical force methods generated the largest pores (90 nm–10 µm) and the lowest conductivity (<16 S cm⁻¹). The data confirm that finer pore networks, achieved with gentle, high‑shear dispersion, enhance electron transport in rubber composites.

Conclusions

We have established mercury‑intrusion porosimetry as a comprehensive, quantitative method for characterizing meso‑ and macropores in CNT agglomerates. The technique distinguishes CNT types, packing density, and solvent‑dependent dispersibility, and it predicts functional properties such as electrical conductivity in CNT‑rubber composites. Although mercury poses environmental concerns, its use is justified by the unique pore‑size resolution it offers. This porosimetry platform provides a robust foundation for designing high‑performance CNT‑based materials and composites.