Single-walled carbon nanotubes in superacid: X-ray and calorimetric evidence for partly ordered H2SO4

Liquid anhydrous sulfuric acid forms a partly ordered structure in the presence of single-walled carbon nanotubes (SWNTs). X-ray scattering from aligned fibers immersed in acid shows the formation of molecular shells wrapped around SWNTs. Differential scanning calorimetry of SWNT-acid suspensions exhibits concentration-dependent supercooling/melting behavior, confirming that the partly ordered molecules are a new phase. We propose that charge transfer between nanotube π electrons and highly oxidizing superacid is responsible for the unique partly ordered structure. Comments Copyright American Physical Society. Reprinted from Physical Review B, Volume 72, Issue 4, Article 045440, July 2005, 5 pages. Publisher URL: http://dx.doi.org/10.1103/PhysRevB.72.045440 Author(s) Wei Zhou, John E. Fischer, P. A. Heiney, H. Fan, Virginia A. Davis, M. Pasquali, and Richard E. Smalley This journal article is available at ScholarlyCommons: http://repository.upenn.edu/mse_papers/65 Single-walled carbon nanotubes in superacid: X-ray and calorimetric evidence for partly ordered H2SO4 W. Zhou and J. E. Fischer* Department of Materials Science and Engineering, University of Pennsylvania, Philadelphia, Pennsylvania 19104, USA P. A. Heiney Department of Physics and Astronomy, University of Pennsylvania, Philadelphia, Pennsylvania 19104, USA H. Fan, V. A. Davis, M. Pasquali, and R. E. Smalley Center for Nanoscale Science and Technology, Rice University, Houston, Texas 77005, USA Received 7 February 2005; published 18 July 2005 Liquid anhydrous sulfuric acid forms a partly ordered structure in the presence of single-walled carbon nanotubes SWNTs . X-ray scattering from aligned fibers immersed in acid shows the formation of molecular shells wrapped around SWNTs. Differential scanning calorimetry of SWNT-acid suspensions exhibits concentration-dependent supercooling/melting behavior, confirming that the partly ordered molecules are a new phase. We propose that charge transfer between nanotube electrons and highly oxidizing superacid is responsible for the unique partly ordered structure. DOI: 10.1103/PhysRevB.72.045440 PACS number s : 61.48. c, 81.07.De Many theories and simulations predict “structured water” shells surrounding dissolved cations due to short-range hydration effects. Experimental evidence is sparse and indirect, and the concept is still debated.1,2 Here we report unambiguous structural and thermodynamic evidence for partly ordered protic solvent molecules surrounding a molecular scale solute. Our solvent is anhydrous 102% sulfuric acid 2 wt. % excess SO3 and the solute is single-walled carbon nanotubes SWNTs .3 We show that H2SO4 molecules readily intercalate SWNT ropes,4 forming a partly ordered solvent structure wrapped around the tubes and tube ropes. Moreover, the solvent molecules associated with the nanotube solute exhibit different phase behavior compared to pure acid. We propose that charge transfer between nanotube electrons and H2SO4 molecules is responsible for the formation of “structured acid” around the nanotubes. We used well-aligned nanotube fibers5 spun from purified HiPco SWNT for the x-ray scattering experiments. The major impurity was 1.2 at. % Fe residual catalyst particles. Most of the residual Fe was encapsulated in graphitic shells. Very little amorphous carbon was observed in transmission electron microscopy of similarly purified samples.6 Graphite impurities are ubiquitous in SWNT produced by arc or laser ablation from graphite targets, while the HiPco process uses CO and Fe CO 5 as feedstock and catalyst, respectively. Nanotube alignment within the oriented one-dimensional 1D fiber assemblies was characterized by a Gaussian mosaic dispersion full width at half-maximum FWHM of 31.5°,5 thus providing an excellent system in which to look for ordered solvent molecules around a solute. Three types of sample were studied. Dry fibers were vacuum-annealed at 1100 °C for 2 h to drive off impurities. About 25 fibers, 5 mm in length and 40 m in diameter, were packed into 0.5 mm glass capillaries, carefully maintaining the fibers parallel to the axis of the capillary. Swollen fibers were prepared as above by adding anhydrous sulfuric acid to the capillary. The density of dry fiber is 1.1 g/cm3 and thus contains 30 vol. % voids5 since the ideal density of an 8, 8 nanotube crystal is 1.5 g/cm3 we take 8, 8 as representative of the 11 Å mean diameter of HiPco tubes . These dry fibers immersed in sulfuric acid swell by 30%–60% in diameter upon reimmersion.5 A control sample of pure acid was also measured. All sample preparations involving acid were carried out in a dry box, with only brief exposure to air when the capillaries were sealed off. X-ray experiments employed Cu K radiation from a rotating anode source, doubly focusing optics, evacuated flight path, and two-dimensional 2D wire detector.7 All samples were measured in transmission for 2 h with the fiber axis perpendicular to the incident beam. Scattering from an empty capillary was subtracted. Figure 1 a shows the detector image for the dry fiber inset along with the scattering profiles intensity vs wave vector Q obtained by azimuthal integration of the 2D data over the full circle. Figure 1 b presents the same information for the swollen fiber and the control sample, and Fig. 1 c gives azimuthal variation of intensity at fixed Q for dry and swollen fibers. These were obtained for the dry sample top by averaging pixels in the range 0.4 Q 0.6 Å−1 which straddles the 10 reflection of the triangular lattice,4 and for the swollen sample bottom by averaging pixels in the range 1.1 Q 1.9 Å−1 which straddles the first peak in the acid scattering profile. Mosaic FWHM’s are both 32°, a surprising result discussed in detail below. The 2D pattern from dry fibers has two components. Intense Bragg reflections with azimuthal preferred orientation are attributed to the aforementioned aligned SWNTs in semicrystalline ropes, while strong small-angle x-ray scattering SAXS diffuse intensity is assigned to the structural inhomogeneity on large length scales aligned ropes vs voids between ropes ,8 discussed in detail below. The dotted line in Fig. 1 a is a simulated profile of just the Bragg scattering,4 assuming an average tube diameter of 11.2 Å and a Gaussian diameter distribution with 2 Å FWHM. It reproduces the Bragg reflections very well. PHYSICAL REVIEW B 72, 04544