Figure 1

Schematic section of a nanotube bundle containing 37 nanotubes with a diameter of $(17\pm 1)\mathrm{\AA }$. The dispersion of the nanotube diameter gives rise to an imperfect lateral ordering and heterogeneous interstitial channels (IC) providing possible adsorption sites for atoms and small molecules (black dots drawn to the size of a particle with $4\mathrm{\AA }$ diameter). Other adsorption sites are the grooves (G) and the rounded outer surface of the bundle (S).Reuse & Permissions

Figure 2

Adsorbed amounts of ${\mathrm{H}}_2$, ${\mathrm{D}}_2$, ${\mathrm{O}}_2$, Ar, $\mathrm{C}{\mathrm{H}}_4$, and $\mathrm{C}{\mathrm{O}}_2$ on SWNTBs, normalized to the respective monolayer density on graphite and to $1\mathrm{g}$ of the SWNTB sample, as a function of the chemical potential relative to the bulk chemical potential at the same temperature. Two risers are observed for all gases except for $\mathrm{C}{\mathrm{O}}_2$. The first riser corresponds to high energy adsorption sites (grooves and the widest interstitial channels, see Fig. 1). The second riser (lower binding energy) is assigned to adsorption on the curved graphene sheets forming the outer surface of the bundles.Reuse & Permissions

Figure 3

Isosteric heat of adsorption $q_{\mathit{st}}$ as a function of coverage (total adsorbed amount). The data for $\mathrm{C}{\mathrm{H}}_4$ and ${\mathrm{O}}_2$ were obtained by isothermal calorimetry and for the other adsorbates by adsorption isotherm measurements. Solid lines are guides to the eye. The two plateaus for ${\mathrm{H}}_2$, ${\mathrm{D}}_2$, ${\mathrm{O}}_2$, Ar, and $\mathrm{C}{\mathrm{H}}_4$ correspond to the two risers in Fig. 2. For large doses $q_{\mathit{st}}$ tends towards the latent heat of vaporization $q_{3D}$ (see Table ).Reuse & Permissions

Figure 4

(a) Neutron diffraction spectra of the bare SWNTB sample (broken line) and upon adsorption of $2.7\mathrm{mmol}∕\mathrm{g}$ (solid line) of ${}^{36}\mathrm{Ar}$. The dotted lines indicate the positions of the Bragg diffraction peaks expected for a hexagonal packing of the tubes within the bundles with a lattice spacing of about $17\mathrm{\AA }$. (b) Difference between the diffraction spectra in (a), revealing the changes induced upon adsorption. The vertical lines indicate the peak positions expected for Ar condensed in linear chains (dashed lines) and in 2D hexagonal arrays (dotted lines).Reuse & Permissions

Figure 5

Diffraction difference spectra (after subtraction of the bare SWNTB spectrum) for several doses (as given in mmol/g in the legends) for Ar (a), $\mathrm{C}{\mathrm{D}}_4$ (b), ${\mathrm{D}}_2$ (c), ${\mathrm{O}}_2$ (d), and $\mathrm{C}{\mathrm{O}}_2$ (e). Sharp double-spiked features, particularly apparent in (e), arise from the contributions of catalyst, cell, and amorphous carbon. Dotted lines indicate the positions where spurious peaks arising from a poor background subtraction of these contributions have been removed from the spectra in (a)–(d).Reuse & Permissions

Figure 6

(a) Neutron diffraction spectrum of the bare SWNTB sample in the vicinity of the (10) bundle lattice peak fitted by the sum of a polynomial of order three (background) and a Gaussian peak (solid lines). (b) Diffraction spectra (after background removal) of the bare SWNTB sample (open circles) and after adsorption of 0.7 and $2.7\mathrm{mmol}∕\mathrm{g}$ (squares and upper triangles, respectively) of ${}^{36}\mathrm{Ar}$ (filled symbols) and ${}^{40}\mathrm{Ar}$ (open symbols). The solid lines are Gaussian fits to the data. Note that a major shift and a concomitant intensity increase is observed only for the highest dose of ${}^{36}\mathrm{Ar}$. (c) (10) diffraction peak obtained from a simulation of a bare seven (10,10)-SWNT bundle (open circles) and after ${}^{36}\mathrm{Ar}$ adsorption. The two ${}^{36}\mathrm{Ar}$ coverages correspond to the filling of the grooves (solid squares) and the complete coverage of the outer surface layer (solid triangles) and can thus be compared to the spectra (same symbols) in (b). Solid lines are guides to the eye. Again, an important shift is only observed for the largest dose, i.e., after complete coverage of the outer surface of the bundle.Reuse & Permissions

Figure 7

Diffraction spectra (after polynomial background removal) in the vicinity of the (10) bundle lattice peak for the bare SWNTB sample (filled squares) and after adsorption of the several doses given in Table (open symbols) of $\mathrm{C}{\mathrm{D}}_4$ (a), ${\mathrm{O}}_2$ (b), and $\mathrm{C}{\mathrm{O}}_2$ (c). The solid lines are Gaussian fits to the data. With increasing adsorbed amount the peak intensity increases, whereas a noticeable peak shift is only observed for the highest doses in (a) and (b), corresponding to the complete coverage of the outer surface of the bundles.Reuse & Permissions

Figure 8

Pair correlation functions for Ar on SWNTBs (a) for some IC and all G sites populated with linear chains and (b) for IC, G, and S sites populated. The peak at $6.7\mathrm{\AA }$ in (b) is characteristic of hexagonal Ar packing on the outer bundle surface (S sites).Reuse & Permissions

Figure 9

Calculated diffraction patterns for Ar adsorbed on SWNTBs at two doses: some IC and all G sites populated (broken lines) and some IC, all G and S sites populated (solid lines). (a) Diffraction from the adsorbed Ar atoms, only (no C skeleton). The spectrum reveals a pronounced rise at $1.7{\mathrm{\AA }}^{-1}$ caused by the linear chains formed on the G and IC sites, whereas the hexagonal packing on S sites gives a well-defined peak at $2.0{\mathrm{\AA }}^{-1}$. (b) Difference diffraction patterns obtained by subtracting the bare SWNTB spectrum from the total diffraction spectrum $(\mathrm{Ar}+\mathrm{SWNTB})$. (c) Additional smoothing of the spectrum in (b) with a Gaussian function for comparison with the experimental data shown in Fig. 5a.Reuse & Permissions