Shear-SANS study of single-walled carbon nanotube suspensions

We report a combined shear small-angle neutron scattering (shear-SANS) and rheo-optical study of dilute aqueous suspensions of SWNT bundles dispersed using ionic surfactants. Both shear-SANS and flow birefringence reveal weak shear-induced alignment of SWNT bundles along the direction of flow. In terms of a nematic order parameter, the degree of alignment is found to increase with the shear rate, reaching ca. 0.08 at 2000 s . Addition of a soluble polymer to the SWNT suspensions diminishes shear-induced alignment. The factors limiting shear alignment in dilute SWNT suspensions are discussed. 2005 Elsevier B.V. All rights reserved. 1 Certain equipment, instruments or materials are identified in the Letter Single-walled carbon nanotubes (SWNTs) are structurally unique materials that offer great promise for novel applications due to their excellent mechanical, electrical, thermal, and optical properties [1,2]. Much fundamental research aimed at achieving practical SWNT-based technologies has been explored [3], whereas studies directed at the flow processing of SWNTs in suspension are still rather limited [4,5]. A better understanding of the flow response of carbon-nanotube suspensions is important to establishing efficient processing schemes tailored to specific SWNT applications. It is recognized that upon shearing, rod-like objects, such as multiwalled carbon nanotubes (MWNTs) [6–8] and tobacco mosaic virus (TMV) [9,10], can develop nematic order in suspensions. Much previous effort has focused on SWNT/polymer nanocomposites [11–14], and studies on concentrated aqueous suspensions of SWNT are possible due mostly to the recent discovery of an effective surfactant for SWNT dispersion [15]. Small-angle neutron scattering (SANS) is a suitable technique for characterizing SWNT aqueous suspensions. Previous SANS measure0009-2614/$ see front matter 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.cplett.2005.09.044 * Corresponding author. Fax: +1 906 487 2934. E-mail address: [email protected] (H. Wang). ments suggest both rigid-rod structures in dispersed SWNT suspensions [16,17], and an optimal surfactant concentration for dispersion [18]. This Letter reports details of the first neutron scattering experiment on the shear response of SWNTs dispersed in aqueous suspensions using ionic surfactants. Shear-induced alignment is quantified through an approximate nematic order parameter and compared with flow birefringence and rheological measurements. Single-walled carbon nanotubes (SWNTs) synthesized through high pressure catalytic decomposition of carbon monoxide (HiPco) were obtained from Carbon Nanotechnology Inc. As-received SWNTs were further purified through thermal oxidation at 260 C and 1 M HCl solution reflux, and processed by ultrasonication in dilute nitric acid [19]. After repeatedly rinsing with deionized water, dried SWNTs appear as continuous fibers (Fig. 1g). The SWNT suspension consists of 0.08% SWNT and 0.6% sodium dodecylbenzenesulfonate (SDBS) [15] by mass in D2O. in order to adequately specify the experimental details. Such identification does not imply recommendation by the National Institute of Standards and Technology, nor dose it imply the materials are necessarily the best available for the purpose. Fig. 1. 2D SANS intensity for the SWNT suspension at various shear rates (a)–(d) and for the 1% SDBS surfactant solution at the shear rate of 1000 s 1 (e). Anisotropic scattering in the CNT suspension is visible at high shear rates, whereas the surfactant solution remains isotropic at 1000 s . (f)–(h) AFM scans of the SWNT bundles dried from solution on Si substrates (image size: 5 · 5 lm). (f) Continuous and entangled SWNT fibers after purification but before dispersion with surfactant. Height analysis indicates bundle diameters of 20–30 nm. (g) Isolated SWNT bundles after 10 h ultrasonic dispersion in SDBS solution, with diameters of 12–17 nm and lengths of ca. 0.8 lm. (h) Fine SWNT bundles after 24 h ultrasonic dispersion, showing diameters of 4– 8 nm and lengths of ca. 0.5 lm. Only intermediate dispersion as in (g) results in detectable shear-induced alignment. H. Wang et al. / Chemical Physics Letters 416 (2005) 182–186 183 ARTICLE IN PRESS After ultrasonication in an ice bath for 10 h, the SWNTs exist in the suspension as bundles of diameter ranging from 12 to 17 nm, and lengths of about 0.5–1 lm (Fig. 1g). Better dispersion could be achieved by longer (48 h) ultrasonication, as shown in Fig. 1h. As a control, 1% SDBS surfactant solution in D2O was also prepared. In another suspension, premixed SWNT/SDBS suspension and a 4 % polyethylene oxide (PEO, Mw = 10 6 g/mol) in D2O solution were mixed to give 0.013% SWNT, 0.13% SDBS, and 2.67% PEO by mass. Shear small-angle neutron scattering (shear-SANS) measurements were performed using the 8 m SANS instrument at the NIST Center for Neutron Research. Incident neutrons of wavelength k = 10 Å and a sample-to-detector distance of 3.84 m yielded a range of scattering wavevector, 0.006 Å 1 < Q < 0.1 Å . The neutron wavelength dispersion, Dk/k = 0.15, is mainly responsible for the resolution of the SANS measurement. For example, DQs are 0.0016 Å 1 and 0.003 Å 1 for Qs of 0.007 and 0.05 Å , respectively. However, this finite Q-resolution has no effect on the data analysis because of the diffuse features of the SANS spectra in this study. A Couette-type shear instrument was used in this study. The outer diameter of the inner quartz stator and the inner diameter of the outer quartz rotor are 60 and 61 mm, respectively, giving a 0.5 mm gap between the two cylinders. This instrument covers a range of shear rates, _ c, from 0.05 to 2000 s . The sample temperature was maintained at 25 C, which was controlled by a bath circulating through the stator. SANS measurements were made at various _ c, with the incident neutron beam in the gradient direction, perpendicular to the flow-vorticity plane. The scattered neutrons were counted with a 2D detector. The X and Y coordinates of 2D spectra represent the flow and vorticity directions, respectively. After correction for background and detector efficiency, and conversion to an absolute scale using the direct beam intensity, the 2D intensity was either circularly, annularly, or rectangular-section averaged to yield the scattering cross section of the sample in the corresponding geometries. Rheo-optical measurements were carried out using a Rheometrics rheometer with plate-plate geometry and an optical attachment. The birefringence at k = 670 nm and shear visocity were recorded simultaneously as a function of _ c, from 2 to 8000 s . At each _ c, data were acquired at a rate of 1 Hz and were averaged over 20 measurements. Fig. 1 shows the 2D SANS intensity for the SWNT suspension at various _ cs and for the SDBS surfactant solution at _ c = 1000 s . Anisotropic scattering in the SWNT suspension is visible at high rates of strain, whereas the surfactant solution remains isotropic at 1000 s . As the SDBS forms micelles in solution, the intensity maxima at Q = 0.052 Å 1 indicates the average spacing between micelles, about 120 Å. However, this value is not sensitive to the surfactant concentration. For example, as the surfactant concentration increases by 10-fold, from 0.5% to 5%, the peak position slightly shifts to higher Q, with a change of ca. 15%. It is understood that in this ionic surfactant system, the dominant interactions between micelles are due to long range Coulombic forces rather than short range hard-wall interactions; as the force determines the average spacing, a change in overall concentration would reflect a change in the molecular packing density in individual micelles rather than the number density of micelles. This invariance is important in the analysis below for the SWNT suspension, in which the scattering contribution from surfactant molecules not adsorbed onto SWNT bundles is approximated as proportional to that from the pure SDBS solution. Fig. 1 also shows AFM scans of the SWNT bundles dried from solution on Si substrates [(f)–(h), image size: 5 · 5 lm]. Fig. 1f shows continuous and entangled SWNT fibers after purification but before dispersion with surfactant. Height analysis indicates bundle diameters of 20– 184 H. Wang et al. / Chemical Physics Letters 416 (2005) 182–186 ARTICLE IN PRESS 30 nm. This is typical SWNT morphology before debundling. Fig. 1g shows isolated SWNT bundles after 10 h of ultrasonic dispersion in SDBS solution, with diameters of 12–17 nm and lengths of ca. 0.8 lm. Fig. 1h shows fine SWNT bundles after 48 h of ultrasonic dispersion, with diameters of 4–8 nm and lengths of ca. 0.5 lm. Only intermediate dispersion as in Fig. 1g results in a measurable degree of shear-induced alignment as shown in (a)–(d). Neither suspension with brief disruption of the bundle networks (in either room-temperature ionic liquids or surfactant solutions with brief ultrasonication) nor finely dispersed SWNT in solution [similar to (h) or better] results in detectable shear-induced anisotropy. To analyze the effect of shear on the SWNT suspension, rectangular sections across the beam center in both the flow and vorticity directions were averaged to yield a 1D projection. Fig. 2 shows the SANS spectra obtained in this manner along the flow direction at 0 s 1 (filled circles) and 1000 s 1 (open squares). The intensity at low-Q is due to the surfactant-dispersed SWNTs and shows power-law dependence. The high-Q scattering comes from SDBS not adsorbed onto SWNT bundles and is invariant under shear, which is clearly revealed from radial averaging the entire 2D intensity. The reduced SDBS intensity compared to that of the neat surfactant solution is due both to the lower SDBS concentration after adsorption and poor statistics from rectangular sectioning. For comparison, the SANS spec-