www.advmat.de www.MaterialsViews.com
COMMUNICATION
Isolation of Specific Small-Diameter Single-Wall Carbon Nanotube Species via Aqueous Two-Phase Extraction Jeffrey A. Fagan,* Constantine Y. Khripin, Carlos A. Silvera Batista, Jeffrey R. Simpson, Erik H. Hároz, Angela R. Hight Walker, and Ming Zheng Aqueous two-phase extraction (ATPE)[1,2] was recently demonstrated to enable the isolation of semiconducting and metallic single-wall carbon nanotube (SWCNT) populations through partitioning of the nanotubes across the spontaneously separating phases in a dextran – polyethylene glycol (PEG) aqueous two-phase system.[3] In this contribution, modification of the separation conditions, by changing the combination of nanotube dispersants, is shown to enable sequential and rapid isolation of multiple single small-diameter SWCNT species at high purity in an efficient manner. As realized in this contribution, separation of single species by ATPE has significant advantages over published separation methodologies for single SWCNT species, which have been extensively researched over the past decade. Methods including chromatographic and specific adsorption methods,[4–11] density gradient ultracentrifugation,[12–15,16] and electrochemical processing[17] for specific species isolation have enabled significant advances in characterization and property measurements, but suffer from an inability to separate all species (gel chromatography), cost and yield factors (DNA based ion-exchange chromatography), and scale/rate issues (density gradient ultracentrifugation). ATPE provides an alternative rapid process, utilizing inexpensive surfactants, which is highly tunable such that many individual species – including metallic species – in a small diameter population can be targeted and separated in a single experiment. Additionally, ATPE has potential for scaling via continuous processing methods,[18] the ability to accommodate a range of SWCNT dispersants, and available batch processing instrumentation such as counter current chromatography instruments[1,2,19] that can automate separations equivalent to
Dr. J. A. Fagan, Dr. C. Y. Khripin, Dr. C. A. Silvera Batista, Dr. M. Zheng National Institute of Standards and Technology (NIST) Materials Science and Engineering Division Gaithersburg, MD 20899, USA E-mail: [email protected] Dr. A. R. Hight Walker National Institute of Standards and Technology Semiconductor and Dimensional Metrology Division Gaithersburg, MD 20899, USA Prof. J. R. Simpson Towson University Department of Physics, Astronomy and Geosciences Towson, MD 21252, USA Dr. E. H. Hároz Los Alamos National Laboratory Center for Integrated Nanotechnologies Los Alamos, NM 87545, USA
DOI: 10.1002/adma.201304873
Adv. Mater. 2014, DOI: 10.1002/adma.201304873
Figure 1. Photograph showing (front row) the parent dispersion (far left) and vials of ATPE separated SWCNT enriched or single species including the (9,4), (7,5), (8,4) rich, (6,6), (7,6) rich, (9,2), (7,4), (6,5), and (6,4) with (7,3). Dispersions were diluted for the photograph, and do not indicate relative abundance in the parent dispersion.
several hundred theoretical plates. A set of photographs showing separated SWCNT (n,m) species is shown in Figure 1. For this contribution we have used PEG (molecular weight (MW) = 6000 Da) and dextran (MW ≈ 68 kDa) as the two phase separating polymers. Above a spinodal line of critical polymer concentrations (at room temperature), aqueous mixtures of these polymers undergo spontaneous decomposition into two phases.[1,2] This yields an upper phase that is PEG-rich and a lower dextran-rich phase. Extensive discussion of the partitioned phase compositions and other phase separating polymers systems are presented by Albertsson[1] and Zaslavsky.[2] Separation of a solute in ATPE occurs based on the differential preference of the solute (equilibrium affinity) for one phase or the other, not on buoyant density or rate of travel, differentiating ATPE from other separation schemes.[20] The chemical potential of the solute(s) in, and thus relative affinity for, the two phases can be tuned by temperature, polymer concentration, or the addition (quantity) of surfactant(s), certain salts or associating polymers. Ideally, the distribution of each species is governed by the energy difference it has for the two phases. A generalized schematic of the ATPE separation is shown in Figure 2. No significant capital equipment is required for ATPE, and it can be performed rapidly; even with hand shaking, equilibration of solutes between the mixing phases empirically occurs within a few seconds, and generally bulk phases (20 mL scale) are fully separated by centrifugation at ≈700g, g ≡ 9.8 m s−2, in (30 to 90) s. In this communication we report separation of SWCNTs initially dispersed in sodium deoxycholate (DOC) solution[21]
© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
1
www.advmat.de
COMMUNICATION
www.MaterialsViews.com
Figure 2. Schematic of the ATPE separation. When the constituent phases are mixed, the dispersed SWCNTs are exposed to gradients in polymer concentrations across which they have differential affinity. Each SWCNT species partitions independently across the two phases to a degree dependent on the magnitude of their differential affinity. Once each polymer-rich phase coalesces, the SWCNTs are separated on a macroscopic scale.
although other bile salt and non-bile salt surfactants can be used. Examples of the achieved phase separation in a Dextran-PEG ATPE system as a function of sodium dodecyl sulfate (SDS) concentration at a constant DOC concentration (0.0225%) are shown in Figure 3A. At low SDS concentrations all of the SWCNT species partition into the Dextran phase.
However, with increasing SDS concentrations SWCNTs begin to selectively partition into the top phase; this is visible in the development of the green color in the top phase and the purple color of the bottom phase. At the greatest SDS concentrations shown, almost all of the SWCNTs partition into the upper phase. The shift in partitioning of each species is monotonic
Figure 3. A–C) Photograph (A) of a CoMoCat SWCNT dispersion separated by the ATPE method with increasing SDS concentrations. Absorbance spectra from the top-phase (B) and the bottom-phase (C); the polymer content and DOC content (225 mg/L) were not varied. With greater SDS concentration, more of the SWCNTs partition into the upper phase. From left, the SDS concentrations are: (0.25, 0.5, 0.65, 0.89, 1.1, and 1.44)%. As indicated by the presence/absence of absorbance features specific to individual SWCNT species, SWCNT species partition from the bottom to upper phase at different SDS concentrations. Each panel is scaled so that the peak value of the greatest spectra equals 1.5 (arbitrary units); spectra are also sequentially offset 1 arb. unit. D–F) Photograph (D) of the same SWCNT dispersion separated with decreasing DOC concentrations, and absorbance spectra from the top-phase (E) and the bottom-phase (F); the polymer content and SDS content (800 mg/L) were not varied. With decreased DOC concentration, more of the SWCNTs partition into the upper phase. From the left, the DOC concentrations are: (1000, 750, 500, 300, 200, 100, and 50) mg/L. G–I) Photograph (G) of the same SWCNT dispersion separated with increasing SDS concentrations with a constant SC content (900 mg/L) at 19 °C, and absorbance spectra from the top-phase (H) and the bottom-phase (I); the polymer content was not varied. From left, the SDS concentrations are: (0.5, 0.55, 0.6, 0.7, 0.9, and 1.1)%. Features from metallic species appear between the dashed lines, and are absent from top phase spectra.
2
wileyonlinelibrary.com
© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Mater. 2014, DOI: 10.1002/adma.201304873
www.advmat.de www.MaterialsViews.com
Figure 4) from semiconducting species (black text).[3] Because the overall mechanism is believed to be thermodynamic,[1] species will distribute into both phases over a finite range of SDSDOC or SDS-SC concentrations. The purity of a given fraction can thus typically be improved by “stepping back” top-phase fractions in SDS content for diameter sorting, i.e., by addition of bottom phase without SDS to lower the total SDS concentration, or by addition of clean opposite SC-SDS phase for metal-semi sorting,[3] followed by re-separation. SWCNT species partitioning in low ratios into the original top phase will partition down, improving the purity of the top-phase target SWCNT(s). In our investigations for diameter separation, the exact SDS concentration at which each SWCNT significantly partitions into the PEG-rich upper phase depends strongly on the choice of bile salt, the bile salt concentration, the co-surfactant and its concentration, and temperature. It is also known to vary mildly with the exact lot of each polymer used.[1] Detailed investigations of the potential mechanisms by which the interplay of DOC, SDS and polymer concentrations act to segregate specific SWCNT across the two phases are being pursued and will be reported in further contributions. Final isolation of specific species was typically accomplished by separation of the metallic from semiconducting SWCNTs using SC-SDS conditions (Figure 3G). This allowed separation of SWCNTs that partitioned similarly with DOC-SDS. Multiple single species can be isolated in approximately six steps, with many more by 8–10 steps. Some species, such as the (6,6), are obtainable at an estimated >85% purity in 3–4 steps; with single steps in our lab taking 40% (estimated by absorbance). Other semiconducting species not highlighted in Figure 5 were also enriched at certain conditions, but in general were present in too small of quantities to be isolated in these investigations. Improvement of the separation and the number of isolatable species is an area of continuing effort. Additional separated species and separation of a larger diameter parent population by the same methodology are shown in Figure S7 and S8 in the Supporting Information; the ability to resolve enantiomers of specific species has also been suggested by experimental
COMMUNICATION
with increasing SDS concentration, vide infra, and as visible by the color changes (each SWCNT species absorbs light at distinct wavelengths dictated by their specific (n,m) chiral vector) different species partition across the phases at different SDS concentrations. Spectra of the phases are reported in Figures 3B and 3C. The spectra show additional optical transitions appearing in the top phase with increasing SDS concentration; this is the hallmark of successive partitioning. Similar separations can also be achieved by varying the DOC content at a constant SDS concentration. This data is shown in Figure 3D–F. Separation in the same fashion as Figure 3A, but utilizing sodium cholate (SC) instead of DOC, is shown in Figure 3G–I. Using SC, primary separation is of metallic from semiconducting SWCNTs.[3] These results directly indicate how to fractionate a SWCNT sample into sub-populations by repetitive application of the ATPE method through variation of the total SDS content in sequential separations. We demonstrate the application of ATPE through a bulk, multi-stage, separation, combining both DOC-SDS and SC-SDS strategies to isolate single semiconducting and metallic species. In our scheme, SWCNTs were initially added to the two-phase system at a low SDS concentration and allowed to partition. The layers, containing different concentrations of each SWCNT species, were then separated by simple pipetting; a detailed discussion of the methods and facilitating methodologies is presented in the Supporting Information (SI). Aliquots of new top phase (with polymers concentrations approximating those of the actual top phase, but prepared separately) with a greater SDS concentration were then added to the separated bottom phase. Upon mixing, the greater SDS concentration in the new ATPE stage increases the number and strength of the SWCNT species partitioning into the top phase, as seen in Figure 3, thus generating a new separation. A generalized version of the overall separation strategy is shown in Figure S4 in the Supporting Information. Depending on the SWCNT species being targeted, the SDS concentration can be controlled as finely as desired. For the separation of CoMoCat[22] SWCNTs presented here, we typically choose to make four separations, generating four “top” fractions, and one “bottom” fraction ranging in nominal SDS concentration from ≈0.65% to ≈1.7%. This generated a rough diameter cut, with fractions rich in the (7,5)-(8,4)-(6,6), (7,5)-(7,6)-(9,2), (5,4)-(8,3)-(7,4), (6,5)-(7,4), and (6,4)-(7,3)-(5,5) groupings respectively. The order of species separation with increasing SDS content is reported graphically for the data in Figures 3A–C in Figure 4. After separation by (primarily) diameter in the DOC-SDS surfactant system, SC is added, and the DOC content reduced, while maintaining (generally) the SDS concentration and polymer concentrations; this drives separation of metallic (red text in
Figure 4. Approximate order of partition to the top phase with increasing SDS concentration at fixed DOC conc. (0.04%). Not all species are shown. As represented by the colored shapes, partition occurs over a range of SDS concentrations. The exact order, shape and breadth of the spreads are affected by separation conditions and enantiomer distribution.
Adv. Mater. 2014, DOI: 10.1002/adma.201304873
© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
3
www.advmat.de www.MaterialsViews.com
COMMUNICATION
Experimental Section[22] Single-wall carbon nanotube powder (Southwest Nanotechnologies SG65i grade, lot# SG65EX-002) was donated by the manufacturer. Sodium deoxycholate (98%, BioXtra), sodium cholate hydrate (>99%), sodium dodecyl sulfate (SDS), and dextran (D8821, 64 to 76 kDa) were acquired from Sigma–Aldrich. Polyethylene-glycol (MW 6 kDa) was purchased from Alfa-Aesar. A pre-purified dispersion was prepared from the parent powder via published methods. Detailed preparation is reported in the Supporting Information. ATPE separations were conducted at room temperature (24 ± 1 °C), unless noted otherwise, with a nominal polymer composition of 9% Dextran and 8% PEG (both mass basis) in the total volume. A low-speed benchtop centrifuge (Becton Dickinson, Dynac 420101, 4 hole–50 mL rotor, max acceleration ≈ 1500g) was used to speed most separations. Small-scale gradient separations, shown in Figure 3, were prepared by changing only the volumes of water and the varied surfactant across the experiment set.UV–vis–NIR absorbance spectra were collected on a Cary 5000 UV–vis–NIR spectrometer in 1 nm increments through a 1 mm or 2 mm quartz cuvette with an integration time of 0.1 s/nm (2 nm slit width). The spectra of the corresponding blank surfactant solution samples were collected separately and linearly subtracted during data analysis.
Supporting Information Supporting Information is available from the Wiley Online Library or from the author.
Acknowledgements Figure 5. Absorbance spectra of selected SWCNT populations, rich in the identified (primary) species isolated via iterative ATPE separation. The spectra are scaled to 1 arb. unit by the peak of the primary SWCNT species in each fraction. All fractions were derived from the parent population whose spectra is shown at the top. The high concentration of spectral weight into specific peaks indicates isolation of single species at high purities.
results (bimodal extraction conditions, data not shown). Availability of these newly isolatable species is of interest for multiple application areas, as each species has unique properties and thus different applications will benefit from different species. Given the rapidity and quality of the separation, the potential for batch or continuous automated processing methods, and compatibility with other separation techniques such as size exclusion chromatography,[11] the ATPE technique should support significant advances in SWCNT use and technological development. In conclusion, the separation of specific small diameter SWCNT species at high purity using two-phase extraction is demonstrated. Isolation of specific species via the ATPE separation can be accomplished in minutes, requires no expensive capital equipment, and is highly tunable. Nanotube properties are well-maintained through the separation, and the technique is sufficiently flexible to encompass a wide variety of strategies and processing agents. Moreover, the technique is readily amenable to greater mass throughput via scaling, and for automation of single species extraction. 4
wileyonlinelibrary.com
C.Y.K. thanks funding from a National Research Council post-doctoral research fellowship.
Received: September 30, 2013 Published online:
[1] P. A. Albertsson, Partition of Cell Particles and Macromoleucles, 2nd ed., Wiley-Interscience, New York 1971. [2] B. Y. Zaslavsky, Aqueous Two-Phase Partitioning, Marcel Dekker, New York 1994. [3] C. Y. Khripin, J. A. Fagan, M. Zheng, J. Am. Chem. Soc. 2013, 135, 6822. [4] X. Tu, A. Manohar, A. Jagota, M. Zheng, Nature 2009, 460, 250. [5] X. Tu, A. R. Hight Walker, C. Y. Khripin, M. Zheng, J. Am. Chem. Soc. B 133, 12998. [6] H. P. Liu, D. Nishide, T. Tanaka, H. Kataura, Nat. Commun. 2011, 2, 309. [7] H. Liu, T. Tanaka, Y. Urabe, H. Kataura, Nano Lett. 2013, 13, 1996. [8] B. S. Flavel, M. M. Kappes, R. Krupke, F. Hennrich, ACS Nano 2013, 7, 3557. [9] K. Tvrdy, R. M. Jain, R. Han, A. J. Hilmer, T. P. McNicholas, M. S. Strano, ACS Nano 2013, 7, 1779. [10] C. A. Silvera-Batista, D. C. Scott, S. M. Mcleod, K. J. Ziegler, J. Phys. Chem. C 2011, 115, 9361. [11] C. Y. Khripin, X. Tu, J. M. Heddleston, C. Silvera-Batista, A. R. Hight Walker, J. Fagan, M. Zheng, Anal. Chem. 2013, 85, 1382. [12] M. S. Arnold, A. A. Green, J. F. Hulvat, S. I. Stupp, M. C. Hersam, Nat. Nanotechnol. 2006, 1, 60. [13] A. A. Green, M. C. Hersam, Adv. Mater. 2011, 23, 2185.
© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Mater. 2014, DOI: 10.1002/adma.201304873
www.advmat.de www.MaterialsViews.com
Adv. Mater. 2014, DOI: 10.1002/adma.201304873
COMMUNICATION
[14] S. Ghosh, S. M. Bachilo, R. B. Weisman, Nat. Nanotechnol. 2010, 5, 443. [15] J. A. Fagan, J. Huh, J. R. Simpson, J. L. Blackburn, J. M. Holt, B. A. Larsen, A. R. H. Walker, ACS Nano 2011, 5, 3943. [16] S. Cambré, W. Wenseleers, Angew. Chem. Int. Ed. 2011, 50, 2764. [17] K. Ihara, H. Endoh, T. Saito, F. Nihey, J. Phys. Chem. C 2011, 115, 22827. [18] L. A. Robbins, R. W. Cusack, in Perry’s Chemical Engineers’ Handbook, Vol. 7, (Eds: R. H. Perry, D. W. Green), McGraw-Hill, New York 1997, Ch. 15. [19] Y. Ito, Sep. Purif. Rev. 2005, 34, 131.
[20] O. Akbulut, C. R. Mace, R. V. Martinez, A. A. Kumar, Z. Nie, M. R. Patton, G. M. Whitesides, Nano Lett. 2012, 12, 4060. [21] W. Wenseleers, I. I. Vlasov, E. Goovaerts, E. Obraztsova, A. S. Lobach, A. Bouwen, Adv. Funct. Mater. 2004, 14, 1105. [22] Certain equipment, instruments or materials are identified in this paper in order to adequately specify the experimental details. Such identification does not imply recommendation by the National Institute of Standards and Technology nor does it imply the materials are necessarily the best available for the purpose. [23] R. B. Weisman, S. M. Bachilo, Nano Lett. 2003, 3, 1235. [24] J. Maultzsch, H. Telg, S. Reich, C. Thomsen, Phys. Rev. B 2005, 72, 205438.
© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
5
COMMUNICATION
Isolation of Specific Small-Diameter Single-Wall Carbon Nanotube Species via Aqueous Two-Phase Extraction Jeffrey A. Fagan,* Constantine Y. Khripin, Carlos A. Silvera Batista, Jeffrey R. Simpson, Erik H. Hároz, Angela R. Hight Walker, and Ming Zheng Aqueous two-phase extraction (ATPE)[1,2] was recently demonstrated to enable the isolation of semiconducting and metallic single-wall carbon nanotube (SWCNT) populations through partitioning of the nanotubes across the spontaneously separating phases in a dextran – polyethylene glycol (PEG) aqueous two-phase system.[3] In this contribution, modification of the separation conditions, by changing the combination of nanotube dispersants, is shown to enable sequential and rapid isolation of multiple single small-diameter SWCNT species at high purity in an efficient manner. As realized in this contribution, separation of single species by ATPE has significant advantages over published separation methodologies for single SWCNT species, which have been extensively researched over the past decade. Methods including chromatographic and specific adsorption methods,[4–11] density gradient ultracentrifugation,[12–15,16] and electrochemical processing[17] for specific species isolation have enabled significant advances in characterization and property measurements, but suffer from an inability to separate all species (gel chromatography), cost and yield factors (DNA based ion-exchange chromatography), and scale/rate issues (density gradient ultracentrifugation). ATPE provides an alternative rapid process, utilizing inexpensive surfactants, which is highly tunable such that many individual species – including metallic species – in a small diameter population can be targeted and separated in a single experiment. Additionally, ATPE has potential for scaling via continuous processing methods,[18] the ability to accommodate a range of SWCNT dispersants, and available batch processing instrumentation such as counter current chromatography instruments[1,2,19] that can automate separations equivalent to
Dr. J. A. Fagan, Dr. C. Y. Khripin, Dr. C. A. Silvera Batista, Dr. M. Zheng National Institute of Standards and Technology (NIST) Materials Science and Engineering Division Gaithersburg, MD 20899, USA E-mail: [email protected] Dr. A. R. Hight Walker National Institute of Standards and Technology Semiconductor and Dimensional Metrology Division Gaithersburg, MD 20899, USA Prof. J. R. Simpson Towson University Department of Physics, Astronomy and Geosciences Towson, MD 21252, USA Dr. E. H. Hároz Los Alamos National Laboratory Center for Integrated Nanotechnologies Los Alamos, NM 87545, USA
DOI: 10.1002/adma.201304873
Adv. Mater. 2014, DOI: 10.1002/adma.201304873
Figure 1. Photograph showing (front row) the parent dispersion (far left) and vials of ATPE separated SWCNT enriched or single species including the (9,4), (7,5), (8,4) rich, (6,6), (7,6) rich, (9,2), (7,4), (6,5), and (6,4) with (7,3). Dispersions were diluted for the photograph, and do not indicate relative abundance in the parent dispersion.
several hundred theoretical plates. A set of photographs showing separated SWCNT (n,m) species is shown in Figure 1. For this contribution we have used PEG (molecular weight (MW) = 6000 Da) and dextran (MW ≈ 68 kDa) as the two phase separating polymers. Above a spinodal line of critical polymer concentrations (at room temperature), aqueous mixtures of these polymers undergo spontaneous decomposition into two phases.[1,2] This yields an upper phase that is PEG-rich and a lower dextran-rich phase. Extensive discussion of the partitioned phase compositions and other phase separating polymers systems are presented by Albertsson[1] and Zaslavsky.[2] Separation of a solute in ATPE occurs based on the differential preference of the solute (equilibrium affinity) for one phase or the other, not on buoyant density or rate of travel, differentiating ATPE from other separation schemes.[20] The chemical potential of the solute(s) in, and thus relative affinity for, the two phases can be tuned by temperature, polymer concentration, or the addition (quantity) of surfactant(s), certain salts or associating polymers. Ideally, the distribution of each species is governed by the energy difference it has for the two phases. A generalized schematic of the ATPE separation is shown in Figure 2. No significant capital equipment is required for ATPE, and it can be performed rapidly; even with hand shaking, equilibration of solutes between the mixing phases empirically occurs within a few seconds, and generally bulk phases (20 mL scale) are fully separated by centrifugation at ≈700g, g ≡ 9.8 m s−2, in (30 to 90) s. In this communication we report separation of SWCNTs initially dispersed in sodium deoxycholate (DOC) solution[21]
© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
1
www.advmat.de
COMMUNICATION
www.MaterialsViews.com
Figure 2. Schematic of the ATPE separation. When the constituent phases are mixed, the dispersed SWCNTs are exposed to gradients in polymer concentrations across which they have differential affinity. Each SWCNT species partitions independently across the two phases to a degree dependent on the magnitude of their differential affinity. Once each polymer-rich phase coalesces, the SWCNTs are separated on a macroscopic scale.
although other bile salt and non-bile salt surfactants can be used. Examples of the achieved phase separation in a Dextran-PEG ATPE system as a function of sodium dodecyl sulfate (SDS) concentration at a constant DOC concentration (0.0225%) are shown in Figure 3A. At low SDS concentrations all of the SWCNT species partition into the Dextran phase.
However, with increasing SDS concentrations SWCNTs begin to selectively partition into the top phase; this is visible in the development of the green color in the top phase and the purple color of the bottom phase. At the greatest SDS concentrations shown, almost all of the SWCNTs partition into the upper phase. The shift in partitioning of each species is monotonic
Figure 3. A–C) Photograph (A) of a CoMoCat SWCNT dispersion separated by the ATPE method with increasing SDS concentrations. Absorbance spectra from the top-phase (B) and the bottom-phase (C); the polymer content and DOC content (225 mg/L) were not varied. With greater SDS concentration, more of the SWCNTs partition into the upper phase. From left, the SDS concentrations are: (0.25, 0.5, 0.65, 0.89, 1.1, and 1.44)%. As indicated by the presence/absence of absorbance features specific to individual SWCNT species, SWCNT species partition from the bottom to upper phase at different SDS concentrations. Each panel is scaled so that the peak value of the greatest spectra equals 1.5 (arbitrary units); spectra are also sequentially offset 1 arb. unit. D–F) Photograph (D) of the same SWCNT dispersion separated with decreasing DOC concentrations, and absorbance spectra from the top-phase (E) and the bottom-phase (F); the polymer content and SDS content (800 mg/L) were not varied. With decreased DOC concentration, more of the SWCNTs partition into the upper phase. From the left, the DOC concentrations are: (1000, 750, 500, 300, 200, 100, and 50) mg/L. G–I) Photograph (G) of the same SWCNT dispersion separated with increasing SDS concentrations with a constant SC content (900 mg/L) at 19 °C, and absorbance spectra from the top-phase (H) and the bottom-phase (I); the polymer content was not varied. From left, the SDS concentrations are: (0.5, 0.55, 0.6, 0.7, 0.9, and 1.1)%. Features from metallic species appear between the dashed lines, and are absent from top phase spectra.
2
wileyonlinelibrary.com
© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Mater. 2014, DOI: 10.1002/adma.201304873
www.advmat.de www.MaterialsViews.com
Figure 4) from semiconducting species (black text).[3] Because the overall mechanism is believed to be thermodynamic,[1] species will distribute into both phases over a finite range of SDSDOC or SDS-SC concentrations. The purity of a given fraction can thus typically be improved by “stepping back” top-phase fractions in SDS content for diameter sorting, i.e., by addition of bottom phase without SDS to lower the total SDS concentration, or by addition of clean opposite SC-SDS phase for metal-semi sorting,[3] followed by re-separation. SWCNT species partitioning in low ratios into the original top phase will partition down, improving the purity of the top-phase target SWCNT(s). In our investigations for diameter separation, the exact SDS concentration at which each SWCNT significantly partitions into the PEG-rich upper phase depends strongly on the choice of bile salt, the bile salt concentration, the co-surfactant and its concentration, and temperature. It is also known to vary mildly with the exact lot of each polymer used.[1] Detailed investigations of the potential mechanisms by which the interplay of DOC, SDS and polymer concentrations act to segregate specific SWCNT across the two phases are being pursued and will be reported in further contributions. Final isolation of specific species was typically accomplished by separation of the metallic from semiconducting SWCNTs using SC-SDS conditions (Figure 3G). This allowed separation of SWCNTs that partitioned similarly with DOC-SDS. Multiple single species can be isolated in approximately six steps, with many more by 8–10 steps. Some species, such as the (6,6), are obtainable at an estimated >85% purity in 3–4 steps; with single steps in our lab taking 40% (estimated by absorbance). Other semiconducting species not highlighted in Figure 5 were also enriched at certain conditions, but in general were present in too small of quantities to be isolated in these investigations. Improvement of the separation and the number of isolatable species is an area of continuing effort. Additional separated species and separation of a larger diameter parent population by the same methodology are shown in Figure S7 and S8 in the Supporting Information; the ability to resolve enantiomers of specific species has also been suggested by experimental
COMMUNICATION
with increasing SDS concentration, vide infra, and as visible by the color changes (each SWCNT species absorbs light at distinct wavelengths dictated by their specific (n,m) chiral vector) different species partition across the phases at different SDS concentrations. Spectra of the phases are reported in Figures 3B and 3C. The spectra show additional optical transitions appearing in the top phase with increasing SDS concentration; this is the hallmark of successive partitioning. Similar separations can also be achieved by varying the DOC content at a constant SDS concentration. This data is shown in Figure 3D–F. Separation in the same fashion as Figure 3A, but utilizing sodium cholate (SC) instead of DOC, is shown in Figure 3G–I. Using SC, primary separation is of metallic from semiconducting SWCNTs.[3] These results directly indicate how to fractionate a SWCNT sample into sub-populations by repetitive application of the ATPE method through variation of the total SDS content in sequential separations. We demonstrate the application of ATPE through a bulk, multi-stage, separation, combining both DOC-SDS and SC-SDS strategies to isolate single semiconducting and metallic species. In our scheme, SWCNTs were initially added to the two-phase system at a low SDS concentration and allowed to partition. The layers, containing different concentrations of each SWCNT species, were then separated by simple pipetting; a detailed discussion of the methods and facilitating methodologies is presented in the Supporting Information (SI). Aliquots of new top phase (with polymers concentrations approximating those of the actual top phase, but prepared separately) with a greater SDS concentration were then added to the separated bottom phase. Upon mixing, the greater SDS concentration in the new ATPE stage increases the number and strength of the SWCNT species partitioning into the top phase, as seen in Figure 3, thus generating a new separation. A generalized version of the overall separation strategy is shown in Figure S4 in the Supporting Information. Depending on the SWCNT species being targeted, the SDS concentration can be controlled as finely as desired. For the separation of CoMoCat[22] SWCNTs presented here, we typically choose to make four separations, generating four “top” fractions, and one “bottom” fraction ranging in nominal SDS concentration from ≈0.65% to ≈1.7%. This generated a rough diameter cut, with fractions rich in the (7,5)-(8,4)-(6,6), (7,5)-(7,6)-(9,2), (5,4)-(8,3)-(7,4), (6,5)-(7,4), and (6,4)-(7,3)-(5,5) groupings respectively. The order of species separation with increasing SDS content is reported graphically for the data in Figures 3A–C in Figure 4. After separation by (primarily) diameter in the DOC-SDS surfactant system, SC is added, and the DOC content reduced, while maintaining (generally) the SDS concentration and polymer concentrations; this drives separation of metallic (red text in
Figure 4. Approximate order of partition to the top phase with increasing SDS concentration at fixed DOC conc. (0.04%). Not all species are shown. As represented by the colored shapes, partition occurs over a range of SDS concentrations. The exact order, shape and breadth of the spreads are affected by separation conditions and enantiomer distribution.
Adv. Mater. 2014, DOI: 10.1002/adma.201304873
© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
3
www.advmat.de www.MaterialsViews.com
COMMUNICATION
Experimental Section[22] Single-wall carbon nanotube powder (Southwest Nanotechnologies SG65i grade, lot# SG65EX-002) was donated by the manufacturer. Sodium deoxycholate (98%, BioXtra), sodium cholate hydrate (>99%), sodium dodecyl sulfate (SDS), and dextran (D8821, 64 to 76 kDa) were acquired from Sigma–Aldrich. Polyethylene-glycol (MW 6 kDa) was purchased from Alfa-Aesar. A pre-purified dispersion was prepared from the parent powder via published methods. Detailed preparation is reported in the Supporting Information. ATPE separations were conducted at room temperature (24 ± 1 °C), unless noted otherwise, with a nominal polymer composition of 9% Dextran and 8% PEG (both mass basis) in the total volume. A low-speed benchtop centrifuge (Becton Dickinson, Dynac 420101, 4 hole–50 mL rotor, max acceleration ≈ 1500g) was used to speed most separations. Small-scale gradient separations, shown in Figure 3, were prepared by changing only the volumes of water and the varied surfactant across the experiment set.UV–vis–NIR absorbance spectra were collected on a Cary 5000 UV–vis–NIR spectrometer in 1 nm increments through a 1 mm or 2 mm quartz cuvette with an integration time of 0.1 s/nm (2 nm slit width). The spectra of the corresponding blank surfactant solution samples were collected separately and linearly subtracted during data analysis.
Supporting Information Supporting Information is available from the Wiley Online Library or from the author.
Acknowledgements Figure 5. Absorbance spectra of selected SWCNT populations, rich in the identified (primary) species isolated via iterative ATPE separation. The spectra are scaled to 1 arb. unit by the peak of the primary SWCNT species in each fraction. All fractions were derived from the parent population whose spectra is shown at the top. The high concentration of spectral weight into specific peaks indicates isolation of single species at high purities.
results (bimodal extraction conditions, data not shown). Availability of these newly isolatable species is of interest for multiple application areas, as each species has unique properties and thus different applications will benefit from different species. Given the rapidity and quality of the separation, the potential for batch or continuous automated processing methods, and compatibility with other separation techniques such as size exclusion chromatography,[11] the ATPE technique should support significant advances in SWCNT use and technological development. In conclusion, the separation of specific small diameter SWCNT species at high purity using two-phase extraction is demonstrated. Isolation of specific species via the ATPE separation can be accomplished in minutes, requires no expensive capital equipment, and is highly tunable. Nanotube properties are well-maintained through the separation, and the technique is sufficiently flexible to encompass a wide variety of strategies and processing agents. Moreover, the technique is readily amenable to greater mass throughput via scaling, and for automation of single species extraction. 4
wileyonlinelibrary.com
C.Y.K. thanks funding from a National Research Council post-doctoral research fellowship.
Received: September 30, 2013 Published online:
[1] P. A. Albertsson, Partition of Cell Particles and Macromoleucles, 2nd ed., Wiley-Interscience, New York 1971. [2] B. Y. Zaslavsky, Aqueous Two-Phase Partitioning, Marcel Dekker, New York 1994. [3] C. Y. Khripin, J. A. Fagan, M. Zheng, J. Am. Chem. Soc. 2013, 135, 6822. [4] X. Tu, A. Manohar, A. Jagota, M. Zheng, Nature 2009, 460, 250. [5] X. Tu, A. R. Hight Walker, C. Y. Khripin, M. Zheng, J. Am. Chem. Soc. B 133, 12998. [6] H. P. Liu, D. Nishide, T. Tanaka, H. Kataura, Nat. Commun. 2011, 2, 309. [7] H. Liu, T. Tanaka, Y. Urabe, H. Kataura, Nano Lett. 2013, 13, 1996. [8] B. S. Flavel, M. M. Kappes, R. Krupke, F. Hennrich, ACS Nano 2013, 7, 3557. [9] K. Tvrdy, R. M. Jain, R. Han, A. J. Hilmer, T. P. McNicholas, M. S. Strano, ACS Nano 2013, 7, 1779. [10] C. A. Silvera-Batista, D. C. Scott, S. M. Mcleod, K. J. Ziegler, J. Phys. Chem. C 2011, 115, 9361. [11] C. Y. Khripin, X. Tu, J. M. Heddleston, C. Silvera-Batista, A. R. Hight Walker, J. Fagan, M. Zheng, Anal. Chem. 2013, 85, 1382. [12] M. S. Arnold, A. A. Green, J. F. Hulvat, S. I. Stupp, M. C. Hersam, Nat. Nanotechnol. 2006, 1, 60. [13] A. A. Green, M. C. Hersam, Adv. Mater. 2011, 23, 2185.
© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Mater. 2014, DOI: 10.1002/adma.201304873
www.advmat.de www.MaterialsViews.com
Adv. Mater. 2014, DOI: 10.1002/adma.201304873
COMMUNICATION
[14] S. Ghosh, S. M. Bachilo, R. B. Weisman, Nat. Nanotechnol. 2010, 5, 443. [15] J. A. Fagan, J. Huh, J. R. Simpson, J. L. Blackburn, J. M. Holt, B. A. Larsen, A. R. H. Walker, ACS Nano 2011, 5, 3943. [16] S. Cambré, W. Wenseleers, Angew. Chem. Int. Ed. 2011, 50, 2764. [17] K. Ihara, H. Endoh, T. Saito, F. Nihey, J. Phys. Chem. C 2011, 115, 22827. [18] L. A. Robbins, R. W. Cusack, in Perry’s Chemical Engineers’ Handbook, Vol. 7, (Eds: R. H. Perry, D. W. Green), McGraw-Hill, New York 1997, Ch. 15. [19] Y. Ito, Sep. Purif. Rev. 2005, 34, 131.
[20] O. Akbulut, C. R. Mace, R. V. Martinez, A. A. Kumar, Z. Nie, M. R. Patton, G. M. Whitesides, Nano Lett. 2012, 12, 4060. [21] W. Wenseleers, I. I. Vlasov, E. Goovaerts, E. Obraztsova, A. S. Lobach, A. Bouwen, Adv. Funct. Mater. 2004, 14, 1105. [22] Certain equipment, instruments or materials are identified in this paper in order to adequately specify the experimental details. Such identification does not imply recommendation by the National Institute of Standards and Technology nor does it imply the materials are necessarily the best available for the purpose. [23] R. B. Weisman, S. M. Bachilo, Nano Lett. 2003, 3, 1235. [24] J. Maultzsch, H. Telg, S. Reich, C. Thomsen, Phys. Rev. B 2005, 72, 205438.
© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
5
