Carbon 41 (2003) 2377–2383 Physisorption of hydrogen in single-walled carbon nanotubes * ¨ P. Sudan , A. Zuttel, Ph. Mauron, Ch. Emmenegger, P. Wenger, L. Schlapbach Physics Department, University of Fribourg, Perolles, CH-1700 Fribourg, Switzerland Received 12 December 2002; accepted 18 June 2003 Abstract The interaction of hydrogen with single-walled carbon nanotubes (SWNTs) was analysed. A SWNT sample was exposed to D or H at a pressure of 2 MPa for 1 h at 298 or 873 K. The desorption spectra were measured by thermal desorption 2 2 spectroscopy (TDS). A main reversible desorption site was observed throughout the range 77 to 320 K. The activation energy of this peak at about 90 K was calculated assuming first-order desorption. This corresponds to physisorption on the surface of the SWNTs (19.261.2 kJ / mol). A desorption peak was also found for multi-walled carbon nanotubes (MWNTs), and also for graphite samples. The hydrogen desorption spectrum showed other small shoulders, but only for the SWNT sample. They are assumed to originate from hydrogen physisorbed at sites on the internal surface of the tubes and on various other forms of carbon in the sample. The nanosized metallic particles (Co:Ni) used for nanotube growth did not play any role in the physisorption of molecular hydrogen on the SWNT sample. Therefore, it is concluded that the desorption of hydrogen from nanotubes is related to the specific surface area of the sample.  2003 Elsevier Ltd. All rights reserved. Keywords: A. Carbon nanotubes; C. Thermal analysis; D. Gas storage, Activation energy 1. Introduction was 0.3 mass% for a SWNT sample and 2.15 mass% for a high surface area Saran carbon. At high H pressure (7 2 Since 1997 many research groups have studied the MPa) and 80 K the hydrogen-to-carbon ratio for the storage of hydrogen in carbon graphite samples of various SWNT sample reached 7.7 mass% in the initial absorption. shapes (tubular, platelet, herringbone) as well as in carbon Nijkamp et al. [3] performed H adsorption at 77 K in the 2 nanotubes (CNTs). This research was stimulated by an pressure range 0–0.1 MPa on a large number of carbonace- article by Dillon et al. [1], who estimated the hydrogen ous sorbents. The amount of adsorbed hydrogen correlated storage capacity of SWNTs to be 5–10 mass%. However, well with the specific surface area. The samples were no other laboratory could independently confirm this chosen to represent a large variation in surface areas. In promising experimental result. those two studies [2,3], the major part of the uptake was Scientists working on the topic of H absorption in due to physisorption of hydrogen on the sample. 2 solids have reported their findings for various carbon In a review paper, Browning et al. [4] claimed that it is nanostructures. The absorption results are based on several generally accepted that the majority of uptake can be types of bonding between carbon and hydrogen. The explained by physisorption at subambient temperatures on authors concluded that the absorption is due to physisorp- the exterior and interior surfaces. tion, and either chemisorption or an intermediate, weakly In the above study of Dillon et al., the excellent chemisorbed state between physisorption and chemisorp- hydrogen storage property of SWNTs was deduced from tion. TDS spectra [1], which showed two hydrogen peaks, the Hydrogen gas adsorption isotherms ( T 5 80 K) on major one at 150 K and the second at 300 K. No enthalpy carbon samples were measured by Ye et al. [2]. The was calculated for the low-temperature (150 K) peak. On hydrogen adsorption obtained at a pressure of 0.32 MPa the other hand, the desorption activation energy of the high-temperature sites (second peak at 300 K) was found to be 19.6 kJ / mol. This heat of adsorption corresponds to *Corresponding author. Tel.: 141-26-300-9101; fax: 141-26- physisorbed H within the cavities of SWNTs. It was 300-9747. 2 E-mail address: patrick.sudan@unifr.ch (P. Sudan). assumed that only the SWNTs in the sample contributed to 0008-6223 / 03 / $ – see front matter  2003 Elsevier Ltd. All rights reserved. doi:10.1016 / S0008-6223(03)00290-2