Technical Note pubs.acs.org/ac
High Precision Fractionator for Use with Density Gradient Ultracentrifugation Yara Kadria-Vili, Griffin Canning, Sergei M. Bachilo, and R. Bruce Weisman* Department of Chemistry and Richard E. Smalley Institute for Nanoscale Science and Technology, Rice University, 6100 Main Street, Houston, Texas 77005, United States S Supporting Information *
ABSTRACT: The recent application of density gradient ultracentrifugation (DGU) for structural sorting of single-walled carbon nanotube samples has created a need for highly selective extraction of closely spaced layers formed in the centrifuged tube. We describe a novel computer-controlled device designed for this purpose. Through the use of fine needles, systematic needle motions, and slow flow rates, multiple sample layers can be aspirated under program control with minimal cross contamination between layers. The fractionator’s performance is illustrated with DGU-sorted samples of single-walled carbon nanotubes.
S
our experience, these approaches do not provide the fractionation convenience and performance needed to realize the separation potential of SWCNT DGU processing. We describe here the construction and operation of a novel, high precision fractionator designed specifically for automated extraction of layers from DGU tubes.
ingle-walled carbon nanotubes (SWCNTs) are a family of artificial nanomaterials widely studied for their unique physical, chemical, and biological properties and broad range of potential applications. Unfortunately, all current SWCNT growth methods produce nanotubes in a variety of structural forms that differ in diameter and roll-up angle. Because each structural form has distinct electronic, optical, and chemical properties, samples of as-produced SWCNTs are essentially mixtures of different chemical substances that must be separated before use in many incisive research studies or advanced applications. Since it was introduced into the SWCNT community in 2006, density gradient ultracentrifugation (DGU) has become widely established as an effective method for such sorting.1−14 In DGU processing, mixtures of individualized, surfactantsuspended SWCNTs are placed in a centrifuge tube filled with a medium that varies in density from top to bottom. Under prolonged centrifugation, the different nanotube structures migrate to form distinct layers at isopycnic points, positions where their specific buoyant density matches the density of the surrounding medium. The final step in DGU sorting is careful extraction of the separated layers from the centrifuge tube to obtain separated fractions. Several refinements have improved the resolution of the DGU layering process in separating SWCNTs, including the use of tailored density profiles, the addition of cosurfactants, and careful adjustment of surfactant concentrations.1,3,11,12,15 However, the extraction of highly resolved layers without cross-contamination from nearby layers requires specialized and precise fractionation methods. Currently available methods for DGU fractionation include aspiration by syringes or micropipettes (possibly mounted to vertical translation stages);11,12,15 controlled drainage after puncture of the tube bottom;16 slow upward displacement fractionation involving puncture of the centrifuge tube and careful injection of a chase solution;1,13,17−19 and piston gradient fractionation in which the tube contents are slowly forced through the hole of a lowered rubber piston.1,6,9,20,21 In © 2014 American Chemical Society
■
APPARATUS DESIGN AND OPERATION Our device, shown schematically in Figure 1, includes a motorized, reversible syringe pump for fraction extraction, small collection vials for receiving extracted fractions, a solenoid-actuated Y-valve to control the path of fluid flow, and three orthogonal motorized translation stages for precisely positioning the extraction needle in to the centrifuge tube. Operation of these components is controlled and coordinated by a custom LabVIEW program that also provides a convenient user interface for entering extraction parameters and monitoring the fractionation process. The apparatus was designed to meet several specific requirements. It is essential to minimize sample agitation and layer mixing, so we use small extraction needles (27 gauge: 0.41 mm OD, 0.20 mm ID) that travel through the sample in smooth, slow motions. For the same reason, the targeted layer is aspirated very slowly while the needle follows an Archimedean spiral path (constant separation between turns) in the x−y plane of the layer. This allows the fraction to be fully collected with little disturbance to nearby layers. When more than one layer is targeted, they are extracted in top-to-bottom order so that lower layers avoid mixing effects and retain their absolute positions. The liquid circuits use short lengths of small bore tubing to minimize dead volume in the system and allow collection of volumes smaller than 50 μL. This feature plus Received: June 27, 2014 Accepted: October 17, 2014 Published: October 17, 2014 11018
dx.doi.org/10.1021/ac502365w | Anal. Chem. 2014, 86, 11018−11023
Analytical Chemistry
Technical Note
Figure 1. Fractionator. (A) Schematic drawing (electrical cables not shown). (B) Photograph.
trajectory of the needle in the sample tube as determined from the specified flow rate, default motion speed, and layer volume. The number of rotations in the current computed spiral is displayed on the screen along with a plot of the planned needle trajectory. This spiral plot changes color from red to green to show the needle’s real-time position during extraction, as can be seen in Figure 2. The estimated time needed for the entire fractionation process (collecting all targeted layers from top to bottom) is also computed and displayed. Automated collection begins with the needle moving along the x−y spiral as the syringe aspirates liquid from the target layer until the column of extracted liquid reaches the valve position (see Figure 3). Then, needle motion halts, the valve is switched to deposit mode, and the water that originally filled the tubing is expelled into a waste vial. At this point, only ca. 3.9 μL of the water remains in the collection tubing to dilute the sample fraction, which is typically at least ten times larger in volume. Now, the valve is again switched to withdraw, the spiral needle trajectory resumes, and fraction collection continues until the layer is fully aspirated. Then, the needle moves to a sample region without SWCNTs and aspirates a volume
preplanned sequences of syringe motions and valve switching help to minimize dilution of the fractions during collection. Finally, automatic rinse cycles remove residues in the collection plumbing after a fraction has been collected, preventing crosscontamination of subsequent collected fractions. To prepare for a fractionation run, the user fills the tubes and syringe with water and fully depresses the syringe plunger using the control program’s manual mode. Then, the sample centrifuge tube is secured into its holder on the z-axis translation stage, and the tube is raised until the extraction needle is positioned just at the first target layer to be collected. The user enters parameters into the program control panel specifying the volume of syringe mounted on the syringe pump; the desired volumetric flow rates for collection and deposition; the locations and thicknesses of all targeted layers (found most easily from fluorescence mapping, as described below); the location of sample-free liquid in the centrifuge tube suitable for rinsing; and the number of rinse cycles to perform between the collection of different fractions. Figure 2 shows an image of the control screen. When the Start button is activated, the program computes for each targeted layer a spiral horizontal collection 11019
dx.doi.org/10.1021/ac502365w | Anal. Chem. 2014, 86, 11018−11023
Analytical Chemistry
Technical Note
Figure 2. Screen shot of fractionator control program set to extract three layers. The marked objects are (1) vertical positions of the targeted bands, relative to the topmost band; (2) thickness of the targeted bands; (3) display showing current stage of the extraction process; (4) user-entered flow rates for extract and expel modes; (5) user-entered parameters for washing system between fractions; (6) start control to begin automated fractionation; (7) planned (red) and actual (green) path of collection needle across the current targeted layer (white circle shows tube wall). Screens accessed by other tabs at top of screen provide additional control and read-out functions.
displayed distinct colored layers corresponding to separated structural forms of SWCNTs. To characterize these layers, we used a model NS3 NanoSpectralyzer (Applied NanoFluorescence) equipped with a vertical sample translator. This instrument focuses an excitation laser beam through the wall of the centrifuge tube and quickly captures the resulting shortwave IR emission spectrum of SWCNTs at that position. The tube height is automatically scanned in submillimeter steps to generate a set of spectra that form a 2-D map of fluorescence intensity as a function of wavelength and of depth in the tube.11 This map clearly conveys the identity, position, and width of each layer, providing information useful for guiding fractionation. A standard-range InGaAs spectrometer sensitive to ∼1600 nm can be used for spectral mapping of SWCNTs with diameters as large as ∼1.35 nm, if the H2O solvent is replaced by D2O for transparency beyond 1400 nm. Another approach to locating layers of large diameter nanotubes is to map their higher order absorption bands (e.g., E22), which appear at visible wavelengths. If a spectral mapping instrument is not available, one can carefully photograph the tube and an adjacent reference scale using a relatively long focal length camera lens and large camera-to-subject distance to avoid geometrical image distortions. Measurements on the photo can then provide positions and width estimates for layers of interest.
sufficient to move all of the collected sample fraction beyond the valve position. At this point, the collected fraction is deposited into a collection vial by setting the syringe pump to expel mode with the valve in its deposit position. Before collecting the next targeted fraction in the centrifuge tube, the syringe extracts and expels more SWCNT-free liquid to rinse the system (typically three times) and prevent crosscontamination. Then, the tube is moved vertically to access the next layer, and the process repeats until all the targeted fractions have been collected. To aid in extracting layers that are very close to each other, we have made a simple apparatus modification for dual-channel collection. This involves adding a second syringe to the existing syringe pump frame and connecting it through a second Y-valve to a separate deposit tubing branch and to a second collection needle mounted 1 mm above the first. The two channels can share the same pinch valve.
■
RESULTS AND DISCUSSION Instrument Performance and Extracted Sample Characterization. To test fractionator performance, we used a HiPco SWCNT sample that had been processed by nonlinear DGU.11 When it was removed from the centrifuge, the tube 11020
dx.doi.org/10.1021/ac502365w | Anal. Chem. 2014, 86, 11018−11023
Analytical Chemistry
Technical Note
Figure 3. (A) Schematic diagram of the fractionator plumbing, showing fluid paths for extract and expel modes. (B) Simplified flowchart of the fractionation process.
after fluorescence mapping, we consider that these findings demonstrate good fractionation performance in recovering two closely spaced layers with high efficiency and only minor crosscontamination.
Figure 4 shows such a photograph of the centrifuge tube before fractionation along with the fluorescence map (measured with 638 nm excitation) of the small region marked by the rectangular box. (See Supporting Information for the full fluorescence map.) The emissive layers near depths of 36.5 and 38.3 mm are identified from their peak wavelengths as the (8,3) and (7,5) SWCNT species.22,23 They correspond to the adjacent reddish and greenish layers in the photo. We fractionated this sample using the dual needle configuration described above, extracting 70 μL through each needle. Figure 5 shows the normalized fluorescence spectra (with 645 nm excitation) of the two collected fractions. We have quantified the compositions of these fractions using two independent methods (details in Supporting Information): multiwavelength fluorimetric analysis24,25 and absorption spectral analysis with recently determined (n,m)-specific absorptivity values.26 The (7,5)/(8,3) concentration ratio is found to be 0.11 in the upper fraction and 7.8 in the lower fraction. Considering that some mixing likely occurred through diffusion and unintended agitation when the tube was transported to the fractionator
■
CONCLUSION We have developed a novel automated fractionation apparatus that precisely extracts layers formed in centrifuge tubes during density gradient ultracentrifugation. This device enhances the capability for structural sorting of single-walled carbon nanotube mixtures. The apparatus is designed to fully collect targeted layers with minimal mixing of the tube contents. It causes no damage to the centrifuge tube, uses no consumable parts, and selectively collects only regions of interest rather than the entire contents of the tube. It can be operated with a single extraction needle for common samples or with dual needles for the effective collection of very close-lying layers. All functions are performed by coordinated actions of a motorized syringe pump, valve, and translation stages under control of a custom LabVIEW computer program. 11021
dx.doi.org/10.1021/ac502365w | Anal. Chem. 2014, 86, 11018−11023
Analytical Chemistry
Technical Note
Figure 4. (A) Image of a centrifuge tube containing a SWCNT sample processed by nonlinear density gradient ultracentrifugation. Colored layers contain specific nanotube structural species. (B) Fluorescence spectral map of the marked region (35 to 39 mm from top) of the centrifuge tube, showing strong emission from two species. Data were measured with 638 nm excitation as described in the text.
translation stages. Each is a PI miCos model VT-80 stage driven by a Kollmorgen SMC-50X stepper motor controller with RS232 computer interface. Positioning resolutions down to 0.2 μm can be attained in microstepping mode. The solenoid pinch valve is actuated by an auxiliary circuit of one Kollmorgen SMC-50X under control of the apparatus computer. A custom LabVIEW (National Instruments) program coordinates operations of the syringe pump, pinch valve, and stepper motors during fractionation. This LabVIEW program is available free of charge by request to the authors. Samples of SWCNTs produced in the Rice University HiPco reactor (batch HPR 188.4) were processed using the nonlinear DGU method described by Ghosh et al.11 We dispersed 5 mg of raw SWCNTs in a 10 mL, 2% (w/v) aqueous solution of sodium cholate by applying 1 h of bath sonication (Sharpertek model Stamina XP) followed by 30 min of tip sonication with a 3 mm probe at a power setting of 7 W (Misonix model Microson XL). The resulting suspension was centrifuged for 1 h at 13 300g (Baxter Scientific model Biofuge-13) in order to remove nanotube bundles and residual iron catalyst. The supernatant was collected and used as the starting material for DGU processing. Our DGU density gradient was prepared by pipetting a series of aqueous iodixanol (Sigma-Aldrich OptiPrep) solutions into a 13 mm diameter, 5 mL capacity polyclear open-top ultracentrifuge tube (Seton Scientific model 7022). From tube bottom to top, the layered volumes and iodixanol concentrations (w/v) were: 500 μL of 30%, 420 μL of 27.5%, 540 μL of 25%, 660 μL of 22.5%, 660 μL of 20%, and 720 μL of 17.5%. A 625 μL portion of the SWCNT dispersion was mixed with 375 μL of a solution containing 60% iodixanol and 2% sodium cholate surfactant and then injected into the DGU tube near the 22.5% iodixanol layer before centrifugation at 250 000g (max) for 22 h.
Figure 5. Fluorescence spectra (645 nm excitation) of the two fractions collected from the centrifuge tube shown in Figure 4. The targeted layers, containing (8,3) and (7,5) SWCNTs, were extracted simultaneously using the dual needle fractionator configuration.
■
APPARATUS AND METHOD DETAILS The fractionator is built around a computer-controlled syringe pump (Harvard Apparatus 11 Plus) that can drive two syringes in parallel. We typically install 100 μL syringes with cemented needles (Hamilton Gastight 1700 series), which allow for flow rates between 0.0049 and 79 μL/min. Each syringe is connected through a 40 mm length of 0.25 mm ID Tygon microbore tubing to one port of three-port Y-connector. Tubes connected to the connector’s other two ports lead immediately to a two-way solenoid pinch valve that selectively closes one tube while leaving the other open. Each of those two tubes leads to a 40 mm long blunt end needle made from 27 gauge steel tubing. One needle is used to withdraw fluid from the centrifuge tube; the other leads to a collection vial. When the apparatus is configured in dual extraction mode, the two withdrawal needles (leading to separate syringes) are mounted next to each other with a 1 mm vertical offset. Motion of the extraction needles relative to the centrifuge tube is achieved by holding the needle stationary and moving the tube with an assembly of three orthogonally mounted
■
ASSOCIATED CONTENT
S Supporting Information *
Additional information including a full spectral map of the DGU tube and spectral characterization of the extracted fraction compositions. This material is available free of charge via the Internet at http://pubs.acs.org. 11022
dx.doi.org/10.1021/ac502365w | Anal. Chem. 2014, 86, 11018−11023
Analytical Chemistry
■
Technical Note
(26) Streit, J. K.; Bachilo, S. M.; Ghosh, S.; Lin, C. W.; Weisman, R. B. Nano Lett. 2014, 14 (3), 1530−1536.
AUTHOR INFORMATION
Corresponding Author
*Tel: 713-348-3709. Fax: 713-348-5155. E-mail: weisman@ rice.edu. Notes
The authors declare the following competing financial interest(s): R.B.W. has a financial interest in Applied NanoFluorescence, LLC, which manufactures one of the instruments used in this study.
■
ACKNOWLEDGMENTS This research was supported by grants from the National Science Foundation (CHE-1112374) and the Welch Foundation (C-0807).
■
REFERENCES
(1) Arnold, M. S.; Green, A. A.; Hulvat, J. F.; Stupp, S. I.; Hersam, M. C. Nat. Nanotechnol. 2006, 1 (1), 60−65. (2) Kim, W.-J.; Nair, N.; Lee, C. Y.; Strano, M. S. J. Phys. Chem. C 2008, 112 (19), 7326−7331. (3) Yanagi, K.; Iitsuka, T.; Fujii, S.; Kataura, H. J. Phys. Chem. C 2008, 112 (48), 18889−18894. (4) Tsuchiya, K.; Uchida, K.; Kaminosono, Y.; Shimizu, K.; Ishii, T.; Yajima, H. Jpn. J. Appl. Phys. 2013, 52 (3 PART 1), 035102. (5) Niyogi, S.; Densmore, C. G.; Doorn, S. K. J. Am. Chem. Soc. 2008, 131 (3), 1144−1153. (6) Shastry, T. A.; Morris-Cohen, A. J.; Weiss, E. A.; Hersam, M. C. J. Am. Chem. Soc. 2013, 135 (18), 6750−6753. (7) Sundramoorthy, A. K.; Mesgari, S.; Wang, J.; Kumar, R.; Sk, M. A.; Yeap, S. H.; Zhang, Q.; Sze, S. K.; Lim, K. H.; Chan-Park, M. B. J. Am. Chem. Soc. 2013, 135 (15), 5569−5581. (8) Posseckardt, J.; Battie, Y.; Fleurier, R.; Lauret, J.-S.; Loiseau, A.; Jost, O.; Mertig, M. Phys. Status Solidi B 2010, 247 (11−12), 2687− 2690. (9) Green, A. A.; Hersam, M. C. Adv. Mater. 2011, 23 (19), 2185− 2190. (10) Arnold, M. S.; Stupp, S. I.; Hersam, M. C. Nano Lett. 2005, 5 (4), 713−718. (11) Ghosh, S.; Bachilo, S. M.; Weisman, R. B. Nat. Nanotechnol. 2010, 5 (6), 443−450. (12) Zhao, P.; Einarsson, E.; Xiang, R.; Murakami, Y.; Maruyama, S. J. Phys. Chem. C 2010, 114 (11), 4831−4834. (13) Bonaccorso, F.; Hasan, T.; Tan, P. H.; Sciascia, C.; Privitera, G.; Di Marco, G.; Gucciardi, P. G.; Ferrari, A. C. J. Phys. Chem. C 2010, 114 (41), 17267−17285. (14) Green, A. A.; Duch, M. C.; Hersam, M. C. Nano Res. 2009, 2 (1), 69−77. (15) Zhao, P.; Einarsson, E.; Lagoudas, G.; Shiomi, J.; Chiashi, S.; Maruyama, S. Nano Res. 2011, 4 (7), 623−634. (16) Englund, P. T.; Smith, H. O.; Sandbek, E. Anal. Biochem. 1971, 40 (2), 490−493. (17) Brakke, M. K. Anal. Biochem. 1963, 5 (4), 271−283. (18) Crochet, J.; Clemens, M.; Hertel, T. J. Am. Chem. Soc. 2007, 129 (26), 8058−8059. (19) Hopkins, T. R. Anal. Biochem. 1973, 53 (1), 339−341. (20) Antaris, A. L.; Seo, J. W. T.; Green, A. A.; Hersam, M. C. ACS Nano 2010, 4 (8), 4725−4732. (21) Tyler, T. P.; Shastry, T. A.; Leever, B. J.; Hersam, M. C. Adv. Mater. 2012, 24 (35), 4765−4768. (22) Bachilo, S. M.; Strano, M. S.; Kittrell, C.; Hauge, R. H.; Smalley, R. E.; Weisman, R. B. Science 2002, 298 (5602), 2361−2366. (23) Weisman, R. B.; Bachilo, S. M. Nano Lett. 2003, 3 (9), 1235− 1238. (24) Weisman, R. B. Anal. Bioanal. Chem. 2010, 396, 1015−1023. (25) Rocha, J.-D. R.; Bachilo, S. M.; Ghosh, S.; Arepalli, S.; Weisman, R. B. Anal. Chem. 2011, 83, 7431−7437. 11023
dx.doi.org/10.1021/ac502365w | Anal. Chem. 2014, 86, 11018−11023
