H2 storage on single- and multi-walled carbon nanotubes

https://doi.org/10.1016/j.ijhydene.2009.11.029Get rights and content

Abstract

The adsorption of hydrogen on single-walled and multi-walled carbon nanotubes (CNTs) was investigated at 77 and 298 K, in the pressure range of 0–1000 Torr. The adsorption isotherms indicate that adsorption follows the Langmuir model. Hydrogen uptakes were found to depend strongly on the nature of the CNTs. Single-walled CNTs adsorb significantly higher quantities of hydrogen per unit mass of the solid, while the opposite is true on a per unit surface area basis. This observation implies that adsorption takes place selectively on specific sites on the surface. The hydrogen uptake capacity of CNTs was also found to be affected by the purity of the materials, increasing with increasing purity. Temperature programmed desorption indicated that relatively strong adsorption bonds develop between adsorbent and adsorbate and that a single type of adsorption site exists on the solid surface.

Introduction

Hydrogen is considered to be the most promising alternative energy carrier in the global energy balance of the future. The use of carbon-based fossil fuels for over a century seems to have caused measurable and catastrophic alterations to the earth's climate. It is widely hoped that the use of carbon-free energy carriers could reverse or decelerate the “greenhouse” phenomenon. However, although hydrogen possesses significant advantages, it also exhibits major drawbacks in its utilization. The most important one being its storage characteristics, which are primarily associated with its use in transportation applications.

There are four major technologies for hydrogen storage: (1) compressed gas, (2) cryogenic liquid, (3) in the form of metal hydrides, and (4) adsorbed on high surface area materials. The first two alternatives are hampered with problems related to tank volume, compression requirements, safety and loss due to evaporation. On the other hand, metal hydrides posses the disadvantages of large weight, excessive cost of manufacture and high temperatures of decomposition.

Adsorption of hydrogen on the surface of porous materials of high surface area could be a viable alternative for hydrogen storage with the potential to meet the capacity goals set by DOE (6.5 wt.%) as well as the advantages of low weight and ease of desorption. Among the materials examined in this respect, carbon nanotubes (CNTs) [1] possess a prominent position.

It has been proposed that hydrogen can be absorbed on carbon nanotubes by physisorption and/or chemisorption. Physisorption occurs when hydrogen maintains its molecular structure and “it is trapped” in the CNTs with Van der Waals forces. In chemisorption, atoms of hydrogen create chemical bonds with the carbon of the nanotubes. Many theoretical studies of the mechanisms of hydrogen adsorption on CNTs have appeared in the literature [2], [3], [4], [5], [6], [7], [8]. However, the precise mechanism of hydrogen adsorption on carbon nanotubes is not ascertained. Consequently, it is difficult to determine if hydrogen is adsorbed exclusively by physisorption or chemisorption also takes place.

Hydrogen can be stored on the inside area of CNTs, shaping a cylindrical monolayer form, or at the outer surface, or between the nanotubes in the case of bundles of carbon nanotubes. Three different places of bonding have been proposed [9]: above the carbon atom (top), in the middle of the C–C bonds (bridge) or in the centre of a hexagonal of carbons (hollow). Moreover, various configurations of hydrogen with respect to its interaction with the walls of the nanotubes are considered, namely perpendicular, longitudinal, and transversal.

Many publications have been devoted to the theoretical study of hydrogen adsorption on CNTs. Monte Carlo simulations [10], [11], [12] and other calculations [13], [14], [15] have been carried out to predict the hydrogen storage capacity of CNTs, based on the assumption of physical adsorption. The most important factor for these approaches is the choice of the intermolecular potential function that describes the molecular interaction between hydrogen and carbon atoms. Darkrim and Levesque [16], employing a Monte-Carlo numerical simulation in the Grand Canonical Ensemble estimated a maximum storage capacity of CNTs of 11 wt.% at 7.5 × 104 Torr (10 MPa) and 77 K, corresponding to a ratio of hydrogen to carbon atoms of 2.

Experimental studies have reported widely different assessments of hydrogen adsorption capacity of CNTs. Dillon et al. [17], using Temperature Programmed Desorption (TPD), reported that non-purified SWCNTs could store significant amounts of hydrogen, ranging between 5 and 10 wt.%, at room temperature under a pressure of 300 Torr. The TPD experiment suggested that physisorption of hydrogen mainly occurred, as the activation energy of hydrogen desorption was estimated to be 19.6 kJ/mol. Ye et al. [18] measured hydrogen adsorption on purified SWCNTs by a volumetric method, using a Sieverts apparatus, at low temperature (80 K), in the range of 3 × 104–6 × 104 Torr (40–80) bar. The hydrogen storage capacity of those samples was evaluated up to 8 wt.%. Liu et al. [19], using SWCNTs of low purity (50–60%), produced by a semi-continuous hydrogen arc discharge technique and treated with chlorohydric acid, reported a hydrogen storage capacity ranging between 2 and 4 wt.%, at room temperature, under a pressure of 9 × 104 Torr (12 MPa). The high hydrogen uptake is attributed to the presence of cavities or defects which originate from the acid treatment. Similarly, Zhu et al. [20] determined the hydrogen adsorption capacity of MWCNTs (produced by catalytic decomposition of an acetylene-hydrogen mixture at 900 °C) to be up to 5 wt.% at room temperature under the pressure of 7.5 × 104 Torr (100 atm). Finally, Wu et al. [21] measured the hydrogen storage capacity of MWCNTs (obtained by decomposition of CO and CH4 on powder Co/La2O3 catalyst) to be up to 0.25 wt.%, under ambient conditions.

Discrepancies concerning hydrogen storage capacity of CNTs are mainly attributed to: (a) The interaction potential models used to describe the gas-solid interaction, in the case of theoretical analysis. (b) Different experimental conditions employed, primarily temperature. (c) Different production methods of CNTs and consequently different specific surface areas. For example, the CVD method seems to result in CNTs of higher surface areas, as discussed by Basca et al. [22]. (d) Different pre-treatment of CNTs prior to adsorption (heating, vacuum). (e) Different configurations of CNTs, such as tube diameters, tube lengths and inter tube spacing, which permit hydrogen either to move into the tube or to be confined in the interstitial pores. Also, if CNTs are opened in the edge, they are able to adsorb hydrogen on both the inside and outside walls, whereas closed CNTs do not. (f) Different means of purification of CNTs.

Aim of the present study is to provide a better understanding of the parameters which affect surface reactivity and hydrogen storage on CNTs, such as specific surface area and wall thickness, as expressed by single-walled, thin-walled and multi-walled carbon nanotubes, and to investigate desorption characteristics.

Section snippets

Production and characterization of carbon nanotubes

The CNTs used in the present study were produced by the chemical vapor deposition (CVD) method, employing C2H4 (for MWCNTs) or C2H2 (for SWCNTs) as carbon source. Mixed oxides of Fe2O3 and Al2O3 were used as catalytic materials. Further details of preparation of CNTs or catalysts can be found in [23], [24], [25].

CNTs were characterized with respect to their crystallinity, wall thickness (single or multi-walled), diameter and length by SEM and Raman spectroscopy. Specific surface area was

Characterization of CNTs

The CNTs prepared in the framework of the present study were characterized using SEM, Raman spectroscopy and measurement of total surface area by the BET method. Prior to characterization and use, the CNT samples were pre-treated with HNO3, following the procedure described earlier. This treatment results in elimination of at least part of the catalyst particles which are encapsulated into the CNTs as well as other materials such as amorphous carbon. As the catalyst particles are removed,

Summary and conclusions

CNTs of variable wall thickness (multi walled, thin walled and single walled) and of variable purity were prepared by CVD of C2H4 or C2H2 and tested for their hydrogen adsorption capacity at 77 and 298 K, in the pressure range of 0–1000 Torr. The quantity of hydrogen adsorbed at equilibrium was found to be significantly affected by the type of CNT employed or by the wall thickness and by the purity of the CNTs. Maximum adsorption capacity per unit mass of the solid was observed over SWCNTs,

Acknowledgements

This work was funded by the General Secretariat of Research and Technology (GSRT) Hellas and the Commission of the European Community, under the PENED 2003 Program (03ED516) and Dr. Stephanos Nitodas for the Raman spectra.

References (26)

  • S.M. Lee et al.

    A hydrogen storage mechanism in single-walled carbon nanotubes

    J Am Chem Soc

    (2001)
  • S.M. Lee et al.

    Hydrogen storage in single-walled carbon nanotubes

    Appl Phys Lett

    (2000)
  • M. Rzepka et al.

    Physisorption of hydrogen 0n microporous carbon and carbon nanotubes

    J Phys Chem B

    (1998)
  • Cited by (114)

    View all citing articles on Scopus
    View full text