Selective Surface Charge Sign Reversal on Metallic Carbon

Selective Surface Charge Sign Reversal on Metallic Carbon Nanotubes for Facile Ultrahigh Purity Nanotube Sorting Jing Wang,† Tuan Dat Nguyen,† Qing Cao,*,‡ Yilei Wang,† Marcus Y.C. Tan,† and Mary B. Chan-Park*,† †

School of Chemical and Biomedical Engineering, Nanyang Technological University, Singapore 637459, Singapore IBM T.J. Watson Research Center, 1101 Kitchawan Road, Yorktown Heights, New York 10598, United States

‡

S Supporting Information *

ABSTRACT: Semiconducting (semi-) single-walled carbon nanotubes (SWNTs) must be purified of their metallic (met-) counterparts for most applications including nanoelectronics, solar cells, chemical sensors, and artificial skins. Previous bulk sorting techniques are based on subtle contrasts between properties of different nanotube/ dispersing agent complexes. We report here a method which directly exploits the nanotube band structure differences. For the heterogeneous redox reaction of SWNTs with oxygen/water couple, the aqueous pH can be tuned so that the redox kinetics is determined by the availability of nanotube electrons only at/near the Fermi level, as predicted quantitatively by the Marcus−Gerischer (MG) theory. Consequently, met-SWNTs oxidize much faster than semi-SWNTs and only met-SWNTs selectively reverse the sign of their measured surface zeta potential from negative to positive at the optimized acidic pH when suspended with nonionic surfactants. By passing the redox-reacted nanotubes through anionic hydrogel beads, we isolate semi-SWNTs to record high electrically verified purity above 99.94% ± 0.04%. This facile charge sign reversal (CSR)-based sorting technique is robust and can sort SWNTs with a broad diameter range. KEYWORDS: carbon nanotubes, sorting, purification, gel chromatography, charge sign reversal techniques, e.g., column chromatography19−21 density gradient ultracentrifugation (DGU),22,23 polymer-assisted aqueous twophase partition (ATP),24 selective dispersion,25−27 electrophoresis,28,29 etc., are less sensitive because they typically exploit subtle differences in adsorption and configuration of the dispersing agents on different SWNT species which result from differences in SWNT polarizability.19,21−24 Sorting of smaller diameter tubes is less affected by the sensitivity limitation due to the more distinct bandgaps for smaller diameter semiSWNTs. Single-chirality enrichment has been achieved for HiPco (diameter 0.8−1.2 nm) and CoMoCat (mean diameter 0.8 nm) tubes, with single semispecies’ purity reaching 94%.30−32 For larger diameter SWNTs, the sorting efficiency is limited more obviously by the separation mechanism. The highest electrically verified purity of enriched arc-discharge semi-SWNTs obtained after a single separation cycle has reached up to 98.3% via a hydrogel-based column chromatography method.21 To achieve higher purity, multiple iterations21

S

emiconducting (semi-) single-walled carbon nanotubes (SWNTs) exhibit superior performance and great potential in applications such as sub-7 nm logic transistors,1,2 solar cells,3−5 chemical sensors,6 and artificial skins7,8 due to their quasi ballistic transport of carriers, optimal electrostatics (even at extremely scaled device dimensions), near-infrared (NIR) bandgaps, strong optical absorptivity, and excellent mechanical properties.1,9−12 However, most SWNT synthetic methods produce mixtures of semi- and metallic (met-) nanotubes,13 and met-SWNTs must be nearly completely removed as they can cause catastrophic device/ circuit shorting failure. The dramatic and fundamental difference between met- and semi-SWNTs is in their band structures: met-SWNTs have a finite density of states (DOS) of electrons at the Fermi level, while semispecies have a bandgap devoid of electrons at the Fermi level. So far, among the numerous sorting approaches established,11,14,15 electrical breakdown16 and the recently developed thermocapillary enabled etching17,18 methods directly exploit this intrinsic difference in the SWNT band structure to achieve highly effective removal of met-SWNTs, but such methods are not easily scalable. Bulk sorting © XXXX American Chemical Society

Received: September 15, 2015 Accepted: February 22, 2016

A

DOI: 10.1021/acsnano.5b05795 ACS Nano XXXX, XXX, XXX−XXX

Article

www.acsnano.org

Article

ACS Nano or more time and energy consuming approaches (e.g., DGU22,23) are required. A new class of sorting mechanism that directly exploits the band structure difference between semi- and met-SWNTs may likely achieve much higher sorting efficiency. Redox reactions involve the transfer of electrons so that they may potentially exploit the different band structures of metversus semi-SWNTs for achieving high-selectivity nanotube sorting.33−36 In general, met-SWNTs are most easily oxidizable (to donate electrons), followed by semi-SWNTs with smaller bandgap (large diameter), while semi-SWNTs with large bandgap (small diameter) are the most inert.37,38 With the exploitation of redox reactions, small diameter HiPco tubes have been chirality-fractionated via complementary chromatography36 and ATP34 methods. The sorting of large diameter arcdischarge SWNTs by tuning the solution pH and with DGU method also exploits the differential redox reactions of metversus semi-tubes.35 However, in these previous redox-based sorting studies, it is also the resulting differential adsorption and reorganization of the dispersing agent that is used for the separation.33−36 No work has been reported that can realize high-purity sorting by restricting the redox reaction to nanotube electrons at/near the Fermi level and thus directly exploiting the differential band structures of metversus semiSWNTs that may result in opposite nanotube surface charges. We herein report a novel separation strategy that exploits the differential transient redox reaction rates between metversus semi-SWNTs with oxygen/water couple (eq 1) at the optimized acidic pH to achieve opposite nanotube surface charge signs. The transient heterogeneous redox reaction of SWNTs with O2/H2O couple can be tuned by pH so that electrons transferring out of the nanotubes are restricted to those at the Fermi level, in accordance to the Marcus− Gerischer (MG) theory;39 we can then differentiate met- from semi-SWNTs which differ dramatically in their absence/ presence of bandgaps. We also calculated the amount of charge transfer from semi- versus met-SWNTs based on the MG theory. We show that by optimizing the pH of the O2/H2O couple and using noninterfering nonionic surfactant (e.g., Triton X-405), met-SWNTs can selectively invert their overall surface charge sign to positive due to the more extensive pdoping; at neutral pH, the SWNTs are anionic because of surface defects. Semi-SWNTs, on the other hand, undergo significantly less oxidation and remain negatively charged. A synergistic chromatographic system has been designed using negatively charged agarose beads, nonionic surfactants, and acidic pH (Scheme 1A). These oppositely charged distinct electronic types (i.e., metversus semispecies) can be sorted efficiently utilizing their electrostatic attraction/repulsion with negatively charged hydrophilic agarose beads (beads functionalization shown in Scheme 1B). This novel charge sign reversal (CSR) sorting technique is highly selective and robust, achieving record high purity of semi-SWNTs above 99.94% after a single pass through the column and is able to sort SWNTs with a wide diameter distribution.

Scheme 1. (A) Schematic of the CSR Chromatographic Sorting Process (Not Drawn to Scale); (B) the Modification Reactions for Sepharose4B-CR and Sepharose4B-SANS Beadsa

a

The font in red shows the reaction site in the CR and SANS molecules.

4(e−h+)SWNT + O2 + 4H+ ⇌ 4(h+)SWNT + 2H 2O

(1)

where e− and h+, respectively, represent an electron and a hole in a SWNT. Figure 1 depicts the actual relative positions of the SWNT occupied DOS (left side of each subfigure) and the unoccupied states of the O2/H2O (“Dox” curve in each subfigure) for metand semi-SWNTs at different pH values. The Fermi level of undoped SWNTs (EFi) is around −4.7 eV and is higher than the electrochemical potential of O2/H2O couple (in the absolute scale, EO2/H2O) in the whole neutral to acidic pH range, which varies from −5.24 to −5.66 eV as pH decreases from 7 to 0 in accordance with the Nernst equation (Figure 1 and Figure S1 in Supporting Information). On the basis of thermodynamics, the heterogeneous electron transfer from a SWNT to the O2/H2O couple can proceed until the new Fermi level of the doped SWNT equals the electrochemical potential of the O2/H2O couple;44 the amount of electrons transferrable due to thermodynamics is indicated by the black-bordered areas “I” in Figure 1. Comparing the area I for met-SWNT with that for semi-SWNT at neutral pH (Figure 1, panel A versus C), we observe similar amounts of electrons are thermodynamically transferable from both SWNT types. Over most of the acidic pH range, semi-SWNTs actually have the tendency to lose

RESULTS AND DISCUSSION SWNTs suspended in solution at neutral pH are negatively charged because of defects generated during the nanotube purification and suspension steps.40,41 They may obtain positive charges through the pH-mediated redox reaction with dissolved oxygen/water redox couple as described by eq 1:37,38,42,43 B


Article

ACS Nano

height) of the Gaussian curves of availability of energy states of the O2/H2O couple can be tuned by pH. From Figure 1, we see that only the low energy states in the Dox curve of O2/H2O overlap the occupied SWNT electron states at/near the SWNT Fermi level (EFi). These overlapping regions are represented by the shaded areas “II” in Figure 1, whose area determines the reaction rate. At even lower energy levels (lower than the electrochemical potential of the O2/H2O couple, e.g., −6 V), the O2/H2O couple has essentially no unoccupied states to receive the nanotube electrons, and thus, the electron transfer rate from these energy levels is effectively zero (according to eq M4 in Methods). At neutral pH, the electron transfer kinetics for both met- and semi-nanotubes is slow (illustrated by the small shaded regions of II in Figure 1A,C). As the pH is reduced to acidic values, the redox potential of the O2/H2O couple decreases so that the overlap of SWNTs with O2/H2O states increases (Figure 1, from left to right, the solid-shaded regions II increase), resulting in faster transient electron transfer kinetics for both electronic types of SWNTs. However, the absence of a bandgap in met-SWNTs leads to a significantly larger overlap between the nanotube occupied and the O2/H2O unoccupied states compared to that of semi-SWNTs (Figure 1, panels A and B have larger shaded regions II compared to panels C and D, respectively). The larger overlap for met-SWNTs, particularly at acidic pH, results in a higher rate of transient electron transfer from met-SWNTs to the dissolved O 2 /H 2 O couple and thus a higher concentration of cationic charges on met-SWNTs surfaces, as illustrated by Figure 2A. On the basis of the analysis above, we speculate that, although both types of SWNTs have anionic charges at neutral pH which are caused by carboxylic groups on their defect sites,40 at sufficiently acidic pH, met-SWNTs may quickly gain enough positive charge to compensate the initial anionic surface charge and even acquire cationic net surface charge, a phenomenon that has previously been observed in diamonds as well.42 On the other hand, the more slowly oxidized semi-SWNTs will acquire less positive charge than the met-SWNTs. With suitable adjustment of pH, it could be possible to cause met- and semi-SWNTs to bear opposite overall surface charge polarity. The next step is to experimentally verify this expected differential redox effect on carbon nanotubes. To measure changes of surface charges on SWNTs, confounding effects due to charges on ionic surfactants (typically used for dispersing nanotubes) must be eliminated. We accomplished this by suspending and debundling the nanotubes in water with the help of nonionic surfactants, rather than the charged surfactants such as sodium dodecylsulfonate (SDS) or sodium cholate (SC) commonly used for sorting. Triton X-405 is a perfect choice since it disperses SWNTs efficiently (especially large diameter SWNTs)48 and is stable in acidic to basic pH. Consequently, we are able for the first time to directly observe, by means of zeta potential measurements of bulk nanotube samples (Figure 2B), the pH-dependent differential redoxinduced surface charging of met- and semi-SWNTs. As the pH is reduced, the zeta potential of both types of SWNTs increases (Figure 2B) because of increased p-doping of the SWNTs; this change is also reflected by the narrowing of the G− band and the upshifting of the 2D band in the Raman spectra (Figure 2C).49,50 In addition, the zeta potential of met-SWNTs is always higher than that of semi-SWNTs (Figure 2B), corroborating the greater degree of electron transfer from met-SWNTs. Crucially, over a substantial pH range of around

Figure 1. Illustration of the Marcus−Gerischer theory for differential charging of met- vs semi-SWNTs by bandgap-dependent heterogeneous electron transfer from nanotubes to O2/H2O couple. (A and B) Electron transfer from met-SWNTs to O2/ H2O couple in (A) neutral and (B) acidic solution. (C and D) Electron transfer from semi-SWNTs to O2/H2O couple in (C) neutral and (D) acidic solution. Region I (black-bordered areas): thermodynamically transferable electrons. Region II (solid-shaded areas in Dox): actual transferable electrons in a short time scale, limited by kinetics. Top to bottom: Faster electron transfer kinetics (black arrows) from met-SWNTs than from semi-SWNTs to O2/ H2O couple. Left to right: As solution pH decreases from neutral to acidic, the electrochemical potential of O2/H2O couple becomes more negative (blue outlined arrows) so that area II (overlap of nanotubes’ occupied states with unoccupied O2/H2O couple states) of both electronic types increases. EFi is intrinsic Fermi level, EBG is the bandgap energy of semi-SWNTs, Eox is electrochemical potential of oxidizing species of O2/H2O couple (i.e., O2), Dox is the DOS of the oxidizing species of O2/H2O couple (unoccupied states).

more electrons thermodynamically (see Figure S2, Supporting Information). However, the approach to thermodynamic equilibrium is controlled by the reaction kinetics, and the redox reaction 1 is known to be rate-limited by the heterogeneous electrontransfer process (eq 2):45,46 O2 + e−SWNT → O−2

(2)

According to MG theory, the transient rate at which SWNT electrons of a specific energy are transferred out depends on the availability of nanotube electrons at that energy as well as the availability of unoccupied states of the O2/H2O couple at the same energy level. SWNT electrons with energies overlapping with those of unoccupied states in the O2/H2O couple are transferred out. The density of occupied states of the solid SWNT is determined by the nanotube DOS, which is fixed, and also by the instantaneous Fermi level, which declines with the progressive oxidation of the SWNT (Figure 1, left side of each subfigure). The DOS of the O2/H2O couple is taken to be fixed (as the oxidizing bath is effectively infinite so that the O2 concentration near the SWNT does not decline significantly with time). The energy states of the O2/H2O couple have a Gaussian distribution in which the most populous species is not at the electrochemical potential (also called the electrolyte Fermi level, and is a function of pH) but is offset from it by a reorganization energy penalty (λ)45−47 (Figure 1, “Dox” curve in each subfigure). λ is an energy penalty that relates to the reorganization of the polar water solvent molecules around the ions/molecules involved in the rate-limiting electron transfer reaction 2 and is approximately 1 eV for the O2/H2O couple.45−47 As importantly, the energy levels (i.e., vertical C


Article

ACS Nano

SWNTs and reinforces the sorting force caused by the opposing charge signs between metversus semi-SWNTs. The sulfonate groups have pKa of around −2.8 and were chosen because they remain negatively charged over the full pH range we used for sorting optimization. The use of nonionic surfactant avoids the external charges induced to SWNTs and allows the electrostatic adsorption/repulsion to be the dominant separation force. We shall describe various optimization efforts starting with the effect of pH on CSR SWNT sorting using congo red (CR, compound 1, Scheme 1B) functionalized beads, hereafter referred to as Sepharose4B-CR beads. In the pH range of ∼3.2 to ∼3.6, where met-SWNTs have cationic net charges and semi-SWNTs are anionic (Figure 2B), high purity (>99%) sorted semi-SWNTs can be obtained (Figure 3 and Figure S3

Figure 2. pH dependence of charge transfer from SWNTs. (A) Schematics illustrating that, at sufficiently acidic pH, metallic species acquire positive net surface charge while semiconducting species remain anionic. (B) Measured zeta potential vs pH for unsorted (black circles), 99% semi-SWNTs (red triangles), and 70% met-SWNTs (olive squares). The mean and standard error are calculated from 3 runs of each sample. (C) Raman spectroscopy (633 nm, red laser) showing G bands (1560−1590 cm−1) and 2D bands (near 2630 cm−1) of unsorted P2-SWNTs at different pHs. All spectra are normalized at the G+ peak intensity but staggered. (D) Calculated concentrations of holes in semi-SWNTs (red triangles) and met-SWNTs (olive squares) vs pH based on the Marcus−Gerischer model.

3−4, the measured zeta potentials of met- and semi-SWNTs indeed have opposite signs (Figure 2B), as predicted by our analysis above based on the MG theory; there is no charge on Triton X-405 at this condition (Table S1). At even more acidic pH lower than the isoelectric points, the zeta potentials of both semi- and met-SWNTs start to decrease likely due to the adsorption of negative counterions (Cl−) in the Helmholtz diffuse layer (Scheme S1, Supporting Information).51 We quantitatively calculated the concentrations of holes acquired by SWNTs (Figure 2D; see Methods for details). Our result indicates that in the pH range of 3−4, met-SWNTs are substantially more p-doped (∼0.005 to ∼0.007 holes per carbon-atom, a severalfold higher degree of p-doping than that of semi-SWNTs). This contrast is large enough to impart a differentiating charge sign between the met- and semi-SWNT populations. To exploit this significant charge sign contrast for sorting, we developed a novel electrostatic chromatography method (Scheme 1A) to sort out the semi-SWNTs from an input mixture with met-SWNTs. The chromatography beads used are agarose beads (Sepharose 4B, GE) functionalized with anionic naphthalene−sulfonate compounds by a two-step reaction52 (Scheme 1B). The coverage of the surface functionalization is calculated to be ∼0.33 mmol anionic naphthalene−sulfonate groups on 1 mL functionalized beads (details shown in Supplementary Note 2, Supporting Information). The hydrophilicity of the agarose hydrogel beads minimizes the possible confounding nonspecific adsorption between the beads and

Figure 3. UV−vis-NIR spectra of unsorted and sorted semi-SWNTs collected at sorting pH of 2.01 to 7.00 using Sepharose4B-CR beads. All the absorption spectra are normalized to the absorbance of S22 peaks.

in Supporting Information) as the electrostatic interaction between the anionic beads and nanotubes retains met-SWNTs but repels semi-SWNTs which are eluted first (Scheme 1A). Below pH = 2.5, some nanotubes start to bundle as the zeta potential decreases, making the semi-SWNTs purity decrease to around 94% (Figure 3 and Figure S3 in Supporting Information). At the optimized sorting pH of 3.24, UV−visNIR spectra suggest that the purity of the semi-SWNT fraction is >99% (Figure 3, purple curve). Our CSR method is highly versatile and the operational parameters can be easily optimized for different beads, surfactants and SWNT sources. We hypothesize that although the electrostatic adsorption/repulsion between the negatively charged beads and the oppositely charged SWNTs is the main D


Article

ACS Nano force for the electronic type sorting, the other groups from the functionalizing compounds may also affect the sorting purity. Besides CR, the Sepharose 4B beads were modified with another molecule, sodium 4-amino-1-naphthalenesulfonate (SANS, compound 2, Scheme 1B), hereafter referred to as Sepharose4B-SANS beads. The commonality between the CR and SANS molecules is that they both contain naphthalene− sulfonate groups. The naphthalene rings enhance the adsorption of met-SWNTs onto the beads due to the π−π stacking between the free electrons of the aromatic rings and extended π electrons in the more polarizable met-SWNTs.53,54 The SANS modification results in higher-purity sorting than the CR modification (Figure 4A) because the simple naphthalenesulfonate structure of SANS avoids the complicated and unknown effects from the conjugated azo bonds and the primary amine groups of the CR molecules. The commercially available anionic Sepharose-based beads, SP Sepharose (functionalized with sulfopropyl groups, instead of naphthalene−sulfonate), were also tested for comparison. As expected, the SP Sepharose beads do not perform as well as our Sepharose4B-CR and Sepharose4B-SANS beads (Figure 4A) due to the confounding effect of the alkyl chain which has nonselective hydrophobic interaction with both semi- and metSWNTs. The nonspecific hydrophobic interaction between the commercially available anion exchange beads and SWNTs may also be a reason why only limited purity was achieved for the HiPco SWNTs sorting in a pioneer work reported by Jagota et al. (with nonionic surfactant and pH adjustment).55 We also tested the pristine Sepharose 4B, which has been reported for small diameter tube sorting (but not for large diameter tubes) in the SDS-based hydrogel chromatography methods. The pristine Sepharose 4B beads perform effective sorting for the arc-discharge tubes with Triton X-405 at acidic condition even without any surface modification (Figure 4A). Although agarose is electrically neutral according to its chemical structure, the Sepharose beads are slightly anionic due to the small amount of sulfate and carboxyl groups introduced in the manufacturing process, as indicated in the product data sheet. The zeta potential of −26.38 ± 1.48 mV of the Sepharose 4B beads confirmed their negative surface charge. This negative surface charge might be the reason the unmodified Sepharose 4B beads also can sort the SWNTs by the CSR strategy. However, with the modification of the selected naphthalene− sulfonate compounds (CR and SANS), the beads’ surface charge becomes more uniform and stable, and thus the purity of the sorted tubes is further improved. We also explored Sephacryl S-200HR beads which are made from allyl dextran and N,N′-methylenebis(acrylamide). As expected, although the Sephacryl S-200HR beads have been widely reported for effective sorting of large diameter tubes via the SDS-based hydrogel chromatography methods,21,56,57 they do not show much sorting effect in the CSR system with Triton X-405 as surfactant (due to their electrically neutral surface), as shown in Figure 4A. Figure 4B,C shows the effect of type and concentration of the nonionic surfactants on the sorting of the arc-discharge SWNTs from two manufacturers, P2-SWNTs from Carbon Solutions (P2 for short in this study) and ASP-100F SWNTs from Hanwha Chemical (Hanwha for short in this study). The individual dispersion of SWNTs is a key issue for the high purity sorting, as shown in Figure 4B. Higher concentration of surfactants (e.g., 2.8% Triton X-405 versus 0.7% Triton X-405) in the eluting solution keeps the SWNTs from bundling during

Figure 4. Optimization of sorting conditions. (A) Beads: UV−visNIR spectra of unsorted P2-SWNTs (black) and semi-P2-SWNTs sorted by various beads, unmodified Sephacryl S-200HR (red), commercial anionic beads SP Sepharose (modified with sulfopropyl) (navy), unmodified Sepharose 4B (olive), Sepharose4B-CR (blue), and Sepharose4B-SANS (magenta). All the semi-fractions were eluted by 2.8% Triton X-405 solution. (B) Eluting surfactants: UV−vis-NIR spectra of unsorted P2-SWNTs (black) and semi-P2SWNTs eluted by 0.7% Triton X-405 (blue), 1% Brij L23 (red) and 2.8% Triton X-405 (magenta) solution. The olive spectrum was eluted under the same condition as the magenta one, and then the sorted semi-SWNT suspension was centrifuged again at 122 000g for 1 h to further remove the small bundles which may contain metSWNTs. (C) SWNT sources: UV−vis-NIR spectra of unsorted Hanwha-SWNTs (black) and semi-Hanwha-SWNTs eluted by 2.8% Triton X-405 (olive), 0.7% Triton X-405 solution (red), and 1% Brij L23 (blue) solution. All the unsorted SWNTs (P2 and Hanwha) were dispersed in 2.8% Triton X-405 solution. All the SWNT dispersion and elution were done at pH = 3.24. All the spectra are normalized to the absorbance of the S22 peaks.

the elution, and thus, higher semi-SWNT purity is achieved. In addition, the sorted semi-SWNTs can be centrifuged to remove the remaining small bundles which may contain met-tubes to further improve the purity (olive curve, Figure 4B). The precipitant after the postchromatography centrifugation was redispersed in 2.8% Triton X-405 solution and also characterized by UV−vis-NIR absorption (Figure S4). The higher M11 absorbance of the precipitant, compared to that of the as-sorted tubes and the supernatant, indicates the higher content of the met-tubes in the precipitant. Brij L23, another E


Article

ACS Nano

The redox reaction between SWNTs and the O2/H2O couple is mild and reversible, generating no/few defects on the reacted SWNTs, particularly semi-SWNTs with lower reaction extent. The almost nil extent of damage of semi-SWNTs during CSR sorting is evident from the low (equal to or even lower than the as-purchased raw SWNT samples) Raman D band (∼1350 cm−1) which reveals the disorder of the graphitic structure of SWNTs (Figure S7, Supporting Information). Figure S8A−C (Supporting Information) shows the X-ray spectroscopy (XPS) spectra of raw P2-SWNTs (starting material, without probe sonication) and the sorted semi- and met-P2-SWNTs, respectively. The binding energies of C 1s and O 1s of met-SWNTs are significantly downshifted, respectively, by 2 eV compared to the raw SWNTs C 1s and O 1s (Figure S8C versus Figure S8A), confirming that met-SWNTs are pdoped by the redox process (reaction 1) above.58−63 The binding energy of C 1s and O 1s for semi-SWNTs does not show obvious downshift compared to the raw SWNTs (Figure S8B versus Figure S8A) since the semi-tubes are much less reacted as we explained in Figure 1. More importantly, the high-resolution C 1s peak of met- and semi-SWNTs show no additional subpeaks compared to raw SWNT C 1s peak (Figure S8D−F), confirming no significant addition of new functionality/defects. The absence of additional functional groups on SWNTs confirms our hypothesis that the positive zeta potential of met-SWNTs likely arises from the depletion of electrons and such phenomenon has also been observed for electron-depleted bulk and nanodiamonds.42,51,60,64 With optimization of various parameters, we achieved ultrahigh purity (>99.0%) for sorted semi-SWNTs with just one pass through the column, as indicated by the UV−vis-NIR spectra (Figure 4). Since common optical methods cannot precisely measure met-SWNTs content below the 1.0% level, we precisely quantified the purity of the sorted semi-HanwhaSWNTs sample (Figure 6A) via direct electrical testing of thousands of single-nanotube transistors as illustrated in Figure 6B. The semi-SWNTs sample was sorted through a Sepharose4B-SANS column. The raw Hanwha-SWNTs were dispersed in 2.8% Triton X-405 solution and eluted with 1% Brij L23 solution, both at pH 3.24. The density of deposited nanotubes was carefully controlled and the device channel length was set at 100 nm, which ensured that in most cases only one nanotube bridged the source/drain electrodes, as illustrated in the Figure 6B inset. The transfer characteristics for all devices made on a representative individual chiplet are depicted in Figure 6C; no met-SWNTs were present in this collection of 320 devices. We measured a total of 3328 connected devices, and identified only 2 devices possibly constructed on metSWNTs, whose transfer curves are shown in Figure 6D. The direct electrical measurement results give a lower bound of the purity level of 99.94 ± 0.04%, representing the highest purity of sorted semi-SWNTs that has ever been reported in the literature. And the purity still has great potential to be further improved by centrifuging the eluted semi- fraction, as demonstrated in Figure 4B. The one-pass yield of semi-SWNTs sorted by this CSR method is dependent mainly on the yields of two process steps: the prechromatography centrifugation step (Yield 1) and the chromatographic step (Yield 2), as illustrated in Scheme S2 (Supporting Information). Other steps of the process do not result in any SWNT loss. The prechromatography centrifugation step removes the bundled nanotubes before the actual chromatography sorting. For P2-SWNTs, Yield 1 varies from

effective nonionic surfactant for SWNTs, can be used as eluting solution as well (red curve, Figure 4B). For Hanwha-SWNTs, which are also produced by arc-discharge method but with different diameter/chirality distribution, Brij L23 shows better eluting efficiency than Triton X-405 (Figure 4C). When we examine the Raman spectra (Figure 5) of unsorted versus semi-SWNTs sorted via different beads at three laser

Figure 5. Raman spectra of unsorted and sorted nanotubes with laser wavelength (A) 785 nm, (B) 633 nm, and (C) 514 nm. Black, unsorted P2-SWNTs; red, semi-P2-SWNTs sorted by Sepharose4BCR beads; blue, semi-P2-SWNTs sorted by Sepharose4B-SANS beads; olive, supernatant of the semi-P2-SANS sample after 122 000g postchromatography centrifugation for 1 h. All the SWNT dispersion and elution were done with 2.8% Triton X-405 at pH = 3.24. All the spectra are normalized to the intensity of the G+ bands.

wavelengths (785, 633, and 514 nm), there is a significant general reduction/disappearance of the metallic peaks in the radial breathing mode (RBM) region with the laser excitation wavelengths of 633 and 785 nm, which are more informative for larger diameter tubes, indicating high purity of the semisamples. Figures S5 and S6 (Supporting Information) show the UV−vis-NIR and Raman spectra of unsorted, semi-, and metP2-SWNTs (sorted by Sepharose4B-CR beads) and HanwhaSWNTs (sorted by Sepharose4B-SANS beads), respectively. For the met-SWNTs samples, the RBM met- peaks greatly increase and the G+ bands broaden accordingly, for both P2 (Figure S5B−D) and Hanwha tubes (Figure S6B−D), corroborating the higher met-SWNT content in the sorted P2 and Hanwha met-SWNTs samples, which agrees well with the UV−vis-NIR characterizations (Figures S5A and S6A, respectively). F


Article

ACS Nano

fractions (0.70 and 0.68 μm, respectively), as shown in Figure S13 (Supporting Information). The purity of the third fraction is a bit lower than the first two fractions (Figure S14, Supporting Information), perhaps due to the elution of some incompletely charge-reversed nonarmchair metallic tubes which have small but nonvanishing bandgaps.34 The CSR method is applicable to SWNTs with a wide diameter range. In addition to arc-discharge SWNTs, we have also successfully sorted HiPco SWNTs, which require lower sorting pH (∼2) due to their smaller diameters (Figure S15, Supporting Information). It is worthy of emphasis that, in our CSR method, the nonionic surfactants do not directly contribute toward sorting but merely act to individually suspend the nanotubes. Hence, the CSR strategy distinguishes itself clearly from the previous SDS-based hydrogel chromatography sorting approaches. It relies on the opposing charges induced directly on the surface of different electronic types of nanotubes, in accord with theoretical calculation. We have focused our sorting efforts on a complementary column chromatography-based technique for its low cost and high throughput to realize the final physical separation of met-/semiSWNT species from solution. However, we would like to mention that other techniques, such as selective adsorption of these differently charged nanotubes on chemically functionalized surfaces,54 could also be utilized to achieve high selectivity separation by the same mechanism.

Figure 6. Direct electrical characterizations to obtain the purity of sorted nanotubes. (A) UV−vis-NIR spectrum of the as-sorted semiHanwha-SWNTs used for electrical characterization. (B) Optical image of an array of single-tube devices made for purity testing with the SEM image of a single device shown in the inset (white scale bar = 200 μm, red scale bar = 100 nm). (C) Transfer characteristics of 320 devices made on a single multidevice chip with the sourcedrain current (IDS) plotted in logarithmic scale. The applied sourcedrain bias is −0.5 V with the gate-source bias (VGS) swept between −3 and 2 V. (D) Transfer curves for the 2 devices connected by possible met-SWNTs.

CONCLUSION In conclusion, by dispersing the SWNTs with nonionic surfactants and subjecting them to a simple but carefully chosen redox reaction with O2/H2O couple at optimized pH, we selectively invert the measured surface charge sign of metSWNTs to cationic while keeping semi-SWNTs anionic. Our calculation of the extent of p-doping, employing the MG theory, verifies that the redox reaction of SWNTs with O2/H2O couple withdraws severalfold more electrons from met-SWNTs than from semi-SWNTs. By exposing the oppositely charged SWNT species to anionic agarose chromatography beads and optimizing pH as well as bead functionalization, we have achieved record-high purity (99.94 ± 0.04%) sorting of semiSWNTs in a single chromatography pass; the purity was verified with direct electrical measurements of more than 3000 single tube devices and has great potential to be further improved. This sorting process is also fast and convenient, is compatible with nanotubes with a wide diameter distribution (0.8−1.7 nm, using arc-discharge and HiPco SWNTs), and is robust against variations in the raw materials and experimental conditions. The discovery of a new sorting parameter (charge sign) and the associated system needed to effectively exploit this makes it possible to produce ultrahigh-purity sorted nanotubes in large scale, which overcomes a major obstacle to the large-scale application of SWNTs in practical electronic devices, sensors, and photovoltaics.

∼71% to ∼32% as the precentrifugation speed increases from 50 000 to 122 000g (Table S2). Yield 2 for P2-SWNTs keeps at ∼30% regardless of the prechromatography centrifugation speed. Therefore, the one-pass yield for the 99% to 99.9% purity semi-P2 SWNTs ranges from about 22% to 9%, respectively, which is tuned by varying the prechromatography centrifugation speed from 50 000 to 122 000g; the purity of the SWNTs is estimated by the UV−vis-NIR spectra (Figure S9, Supporting Information). For Hanwha-SWNTs, when prechromatography centrifuged at 122 000g, the purity reaches up to 99.94 ± 0.04% when tested by single nanotube devices while the yield is also ∼9% (Table S2). The prechromatography centrifugation step ensures the individual dispersion of the unsorted SWNTs, without any sorting effect, either by electronic types (Figure S10, Supporting Information) or by length (Figure S11, Supporting Information). The average length of the sorted semi-P2SWNTs is ∼0.65 μm, regardless of the prechromatography centrifugation speed ranging from 50 000 to 122 000g (Figure S11, Supporting Information); the as-purchased P2-SWNTs are with bundle length 0.5−1.5 μm (∼1 μm) according to the supplier’s information. Semi-Hanwha tubes are slightly longer than the semi-P2 tubes, with the average length of 0.82 μm with precentrifugation speed of 122 000g (Figure S12, Supporting Information). The CSR sorting shows slight size exclusion effect, due to the intrinsic size exclusive function of the Sepharose beads: the semi-SWNTs elution was collected into 3 fractions (0.5 mL each), and the average length of the first fraction (0.61 μm) is slightly shorter than the other two

METHODS Materials. A solution of 70% Triton X-405 in H2O was obtained from Dow Chemical. Sepharose 4B, Sephacryl S-200HR and SP Sepharose beads (GE Healthcare) were purchased from SigmaAldrich. Sodium hydroxide (NaOH, ≥98%), hydrochloric acid (HCl, 36−37%), (±)-epichlorohydrin (epichlorohydrin, ≥99%), sodium dodecylbenzenesulfonate (SDBS, technical grade), sodium cholate (SC, ≥99%), Brij L23, congo red (CR, 85%), sodium 4-amino-1naphthalenesulfonate (SANS, technical grade) and all the solvents G


Article

ACS Nano

ultrahigh-vacuum environment of 10−9 Torr. The length of SWNTs was measured with a Dimension Icon (Bruker) Atomic Force Microscope. Device Fabrication for Purity Testing. The as-separated semiSWNTs solution was diluted 200-fold with a 1% sodium cholate solution and then, after a brief sonication, was drop-casted onto a heavily doped silicon substrate with 10 nm thermal oxide. The deposited nanotubes were washed with ethanol and then annealed in vacuum ( 10) was added into the column in order to elute the metallic-enriched fraction. The column was then flushed thoroughly with D.I. water after which it was ready to be reused for further separation. The three stages of sorting through Sepharose4B-SANS beads were photographed and shown in Figure S16 in Supporting Information. The sorted semi-SWNTs can be postchromatography centrifuged at 122 000g for 1 h to remove the remaining small bundles and thus further improve the purity. The sorted SWNTs were characterized using a Varian Cary 4000 UV−vis-NIR spectrophotometer and a Renishaw inVia Raman microscope equipped with 514 nm wavelength (2.41 eV), 633 nm wavelength (1.96 eV), and 785 nm wavelength (1.58 eV) laser sources in a backscattering configuration. X-ray photoelectron spectroscopy (XPS) measurements were made using a Kratos Axis Ultra DLD X-ray photoelectron spectroscope with a monochromatic Al Kα X-ray source (SPECS, 1486.7 eV) and electron analyzer (Omicron, EA125) in an

O2 → h+SWNT + O−2

(M1)

The rate equation for this elementary reaction is r=

d[h+] dt

(M2) −1

where r is the electron transfer rate (number of electrons s per carbon-atom), [h+] is the concentration of holes generated in SWNTs (number of holes per carbon-atom). Equation M2 for semi- and metSWNTs can be combined to give:

d[h+]semi =

rsemi + d[h ]met rmet

(M3) 39

Following Bard and Faulkner, the electron transfer rate can be derived from the MG model with the following equation: ∞

r(λ , E F , pH) = νACox

∫−∞ εWox(λ , E , pH)·f (E , EF)·ρ(E) dE (M4) −1

where r is the electron transfer rate (number of electrons s per carbon-atom), ν is the vibrational frequency (s−1), A is the SWNT sidewall area occupied by a single carbon-atom in the SWNT lattice (cm2), Cox is the surface oxygen concentration inside this area (number of oxygen molecules per cm2), ε is the proportionality function (eV), Wox is the probability density function (eV−1) of vacant states for the oxidizing species (O2), which is a function of pH for O2/ H2O couple, λ is the reorganization energy in the MG model (eV), f(E,EF) is the Fermi−Dirac distribution function (dimensionless), and ρ is the DOS of SWNTs (states eV−1 per carbon-atom) (here we use semi-SWNTs species (12,8) and met-SWNTs species (12,9) with diameter 1.4 nm for calculation 65) (Figure S17, Supporting Information). The integration variable E is the SWNT electron state energy.

Wox(λ , E , pH) =

2⎞ ⎛ − (E − (E 1 O2 /H2O + λ)) ⎟ exp⎜⎜ ⎟ 4kBTλ 4πλkBT ⎠ ⎝

(M5) where kB is the Boltzmann constant, T is absolute temperature (K), EO2 /H2O is the electrochemical potential of O2/H2O couple (pHdependent, units of eV). The DOS of the oxidizing species (Dox) can be calculated from H


Article

ACS Nano Dox = CoxWox(λ , E , pH)

002) and a Singapore NMRC Ministry of Health Industry Alignment Fund Category 2 (Project Number: NMRC/ MOHIAFCAT2/003/2014, Application No.: MOHIAFCAT2005). Wang acknowledges the support of Nanyang Technological University through a Ph.D. Research Scholarship. We also acknowledge helpful discussions with Y. X. Thong, M. Tam, and S. Conner.

(M6)

We can relate d[h+] to dEF as follows: [e−] =

∞

∫−∞ ρ(E)·f (E , EF) dE

∂[e−] ∂[h+] ∂ =− = ∂E F ∂E F ∂E F ∞

=

∫−∞

(M7)

∞

∫−∞ f (E , EF)·ρ(E) dE

∂f (E , E F) · ρ(E) dE ∂E F

REFERENCES (M8)

(1) Franklin, A. D. Electronics the Road to Carbon Nanotube Transistors. Nature 2013, 498, 443−444. (2) Franklin, A. D.; Luisier, M.; Han, S. J.; Tulevski, G.; Breslin, C. M.; Gignac, L.; Lundstrom, M. S.; Haensch, W. Sub-10 nm Carbon Nanotube Transistor. Nano Lett. 2012, 12, 758−762. (3) Bindl, D. J.; Wu, M.-Y.; Prehn, F. C.; Arnold, M. S. Efficiently Harvesting Excitons from Electronic Type-Controlled Semiconducting Carbon Nanotube Films. Nano Lett. 2011, 11, 455−460. (4) Arnold, M. S.; Zimmerman, J. D.; Renshaw, C. K.; Xu, X.; Lunt, R. R.; Austin, C. M.; Forrest, S. R. Broad Spectral Response Using Carbon Nanotube/Organic Semiconductor/C60 Photodetectors. Nano Lett. 2009, 9, 3354−3358. (5) Jain, R. M.; Howden, R.; Tvrdy, K.; Shimizu, S.; Hilmer, A. J.; McNicholas, T. P.; Gleason, K. K.; Strano, M. S. Polymer-Free nearInfrared Photovoltaics with Single Chirality (6,5) Semiconducting Carbon Nanotube Active Layers. Adv. Mater. 2012, 24, 4436−4439. (6) Roberts, M. E.; LeMieux, M. C.; Bao, Z. Sorted and Aligned Single-Walled Carbon Nanotube Networks for Transistor-Based Aqueous Chemical Sensors. ACS Nano 2009, 3, 3287−3293. (7) Hammock, M. L.; Chortos, A.; Tee, B. C. K.; Tok, J. B. H.; Bao, Z. 25th Anniversary Article: The Evolution of Electronic Skin (ESkin): A Brief History, Design Considerations, and Recent Progress. Adv. Mater. 2013, 25, 5997−6037. (8) Takahashi, T.; Takei, K.; Gillies, A. G.; Fearing, R. S.; Javey, A. Carbon Nanotube Active-Matrix Backplanes for Conformal Electronics and Sensors. Nano Lett. 2011, 11, 5408−5413. (9) Javey, A.; Guo, J.; Wang, Q.; Lundstrom, M.; Dai, H. Ballistic Carbon Nanotube Field-Effect Transistors. Nature 2003, 424, 654− 657. (10) Avouris, P.; Freitag, M.; Perebeinos, V. Carbon-Nanotube Photonics and Optoelectronics. Nat. Photonics 2008, 2, 341−350. (11) Cao, Q.; Han, S. J. Single-Walled Carbon Nanotubes for HighPerformance Electronics. Nanoscale 2013, 5, 8852−8863. (12) Coleman, J. N.; Khan, U.; Blau, W. J.; Gun’ko, Y. K. Small but Strong: A Review of the Mechanical Properties of Carbon Nanotube− Polymer Composites. Carbon 2006, 44, 1624−1652. (13) Tour, J. M. Materials Chemistry: Seeds of Selective Nanotube Growth. Nature 2014, 512, 30−31. (14) Hersam, M. C. Progress Towards Monodisperse Single-Walled Carbon Nanotubes. Nat. Nanotechnol. 2008, 3, 387−394. (15) Zhang, H. L.; Wu, B.; Hu, W. P.; Liu, Y. Q. Separation and/or Selective Enrichment of Single-Walled Carbon Nanotubes Based on Their Electronic Properties. Chem. Soc. Rev. 2011, 40, 1324−1336. (16) Collins, P. G.; Arnold, M. S.; Avouris, P. Engineering Carbon Nanotubes and Nanotube Circuits Using Electrical Breakdown. Science 2001, 292, 706−709. (17) Jin, S. H.; Dunham, S. N.; Song, J.; Xie, X.; Kim, J.-H.; Lu, C.; Islam, A.; Du, F.; Kim, J.; Felts, J.; et al. Using Nanoscale Thermocapillary Flows to Create Arrays of Purely Semiconducting Single-Walled Carbon Nanotubes. Nat. Nanotechnol. 2013, 8, 347− 355. (18) Xie, X.; Jin, S. H.; Wahab, M. A.; Islam, A. E.; Zhang, C.; Du, F.; Seabron, E.; Lu, T.; Dunham, S. N.; Cheong, H. I.; et al. Microwave Purification of Large-Area Horizontally Aligned Arrays of SingleWalled Carbon Nanotubes. Nat. Commun. 2014, 5, 5332. (19) Liu, H.; Feng, Y.; Tanaka, T.; Urabe, Y.; Kataura, H. DiameterSelective Metal/Semiconductor Separation of Single-Wall Carbon Nanotubes by Agarose Gel. J. Phys. Chem. C 2010, 114, 9270−9276.

which implies dE F = −

d[h+] ∞ ∂f (E , E F) ·ρ(E) −∞ ∂E F

∫

dE

(M9)

which is valid for both metallic and semiconducting species. We solve equations M3, M4, and M9 in parallel by setting d[h+]met to 10−6 holes per carbon-atom for each step of met-SWNT electron transfer and computing the corresponding d[h+]semi, dEF_met, and dEF_semi. To obtain the total concentration of holes, we can integrate eq M3 as follows:

[h+]semi =

∫0

[h+] for r → 0

rsemi d[h+]met rmet

(M10)

In eq M10, the concentration of holes is zero at the intrinsic Fermi level and we assume the electron transfer rate approaches zero when the SWNT Fermi level reaches the energy level where the Wox value is equal to 10% of the Wox_max (the maximum point in the Dox curve in Figure 1). With this assumption, the [h+]semi at pH = 7 was calculated to be practically zero which indicates essentially no electron transfer from semi-SWNTs to the O2/H2O couple in this condition. Figure 2D shows the results of our numerical integrations to estimate the pH dependence of the concentrations of holes generated in met- and semi-SWNTs.

ASSOCIATED CONTENT S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsnano.5b05795. Additional notes and equations, Raman spectra, XPS spectra, SWNT length distribution, yield calculation, photography of the sorting process, DOS of met- and semi-SWNTs with different diameters and redox potential of O2/H2O couple, purity of semi-SWNTs as a function of pH and the sorting results of HiPco SWNTs (PDF)

AUTHOR INFORMATION Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. Notes

The authors declare the following competing financial interest(s): Chan-Park, Wang and Tan have recently started a company dealing with sorted semi-SWNTs exploiting this technology discovery.

ACKNOWLEDGMENTS This work was supported by a Competitive Research Program grant from the Singapore National Research Foundation (NRFCRP2-2007-02) and also partially funded and supported by a Singapore Ministry of Education Tier 3 grant (MOE2013-T3-1I


Article

ACS Nano (20) Zheng, M.; Jagota, A.; Semke, E. D.; Diner, B. A.; McLean, R. S.; Lustig, S. R.; Richardson, R. E.; Tassi, N. G. DNA-Assisted Dispersion and Separation of Carbon Nanotubes. Nat. Mater. 2003, 2, 338−342. (21) Tulevski, G. S.; Franklin, A. D.; Afzali, A. High Purity Isolation and Quantification of Semiconducting Carbon Nanotubes via Column Chromatography. ACS Nano 2013, 7, 2971−2976. (22) Arnold, M. S.; Green, A. A.; Hulvat, J. F.; Stupp, S. I.; Hersam, M. C. Sorting Carbon Nanotubes by Electronic Structure Using Density Differentiation. Nat. Nanotechnol. 2006, 1, 60−65. (23) Antaris, A. L.; Seo, J. W. T.; Green, A. A.; Hersam, M. C. Sorting Single-Walled Carbon Nanotubes by Electronic Type Using Nonionic, Biocompatible Block Copolymers. ACS Nano 2010, 4, 4725−4732. (24) Khripin, C. Y.; Fagan, J. A.; Zheng, M. Spontaneous Partition of Carbon Nanotubes in Polymer-Modified Aqueous Phases. J. Am. Chem. Soc. 2013, 135, 6822−6825. (25) Wang, H. L.; Mei, J. G.; Liu, P.; Schmidt, K.; Jimenez-Oses, G.; Osuna, S.; Fang, L.; Tassone, C. J.; Zoombelt, A. P.; Sokolov, A. N.; et al. Scalable and Selective Dispersion of Semiconducting ArcDischarged Carbon Nanotubes by Dithiafulvalene/Thiophene Copolymers for Thin Film Transistors. ACS Nano 2013, 7, 2659−2668. (26) 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. Scalable and Effective Enrichment of Semiconducting Single-Walled Carbon Nanotubes by a Dual Selective Naphthalene-Based Azo Dispersant. J. Am. Chem. Soc. 2013, 135, 5569−5581. (27) Mistry, K. S.; Larsen, B. A.; Blackburn, J. L. High-Yield Dispersions of Large-Diameter Semiconducting Single-Walled Carbon Nanotubes with Tunable Narrow Chirality Distributions. ACS Nano 2013, 7, 2231−2239. (28) Tanaka, T.; Jin, H.; Miyata, Y.; Kataura, H. High-Yield Separation of Metallic and Semiconducting Single-Wall Carbon Nanotubes by Agarose Gel Electrophoresis. Appl. Phys. Express 2008, 1, 114001. (29) Mesgari, S.; Poon, Y. F.; Yan, L. Y.; Chen, Y.; Loo, L. S.; Thong, Y. X.; Chan-Park, M. B. High Selectivity cum Yield Gel Electrophoresis Separation of Single-Walled Carbon Nanotubes Using a Chemically Selective Polymer Dispersant. J. Phys. Chem. C 2012, 116, 10266− 10273. (30) Tu, X.; Manohar, S.; Jagota, A.; Zheng, M. DNA Sequence Motifs for Structure-Specific Recognition and Separation of Carbon Nanotubes. Nature 2009, 460, 250−253. (31) Liu, H.; Nishide, D.; Tanaka, T.; Kataura, H. Large-Scale SingleChirality Separation of Single-Wall Carbon Nanotubes by Simple Gel Chromatography. Nat. Commun. 2011, 2, 309. (32) Ghosh, S.; Bachilo, S. M.; Weisman, R. B. Advanced Sorting of Single-Walled Carbon Nanotubes by Nonlinear Density-Gradient Ultracentrifugation. Nat. Nanotechnol. 2010, 5, 443−450. (33) Hirano, A.; Tanaka, T.; Urabe, Y.; Kataura, H. pH- and SoluteDependent Adsorption of Single-Wall Carbon Nanotubes onto Hydrogels: Mechanistic Insights into the Metal/Semiconductor Separation. ACS Nano 2013, 7, 10285−10295. (34) Gui, H.; Streit, J. K.; Fagan, J. A.; Hight Walker, A. R.; Zhou, C.; Zheng, M. Redox Sorting of Carbon Nanotubes. Nano Lett. 2015, 15, 1642−1646. (35) Antaris, A. L.; Seo, J.-W. T.; Brock, R. E.; Herriman, J. E.; Born, M. J.; Green, A. A.; Hersam, M. C. Probing and Tailoring pHDependent Interactions between Block Copolymers and Single-Walled Carbon Nanotubes for Density Gradient Sorting. J. Phys. Chem. C 2012, 116, 20103−20108. (36) Flavel, B. S.; Kappes, M. M.; Krupke, R.; Hennrich, F. Separation of Single-Walled Carbon Nanotubes by 1-DodecanolMediated Size-Exclusion Chromatography. ACS Nano 2013, 7, 3557− 3564. (37) Strano, M. S.; Huffman, C. B.; Moore, V. C.; O’Connell, M. J.; Haroz, E. H.; Hubbard, J.; Miller, M.; Rialon, K.; Kittrell, C.; Ramesh, S. Reversible, Band-Gap-Selective Protonation of Single-Walled Carbon Nanotubes in Solution. J. Phys. Chem. B 2003, 107, 6979− 6985.

(38) Zheng, M.; Diner, B. A. Solution Redox Chemistry of Carbon Nanotubes. J. Am. Chem. Soc. 2004, 126, 15490−15494. (39) Bard, A. J.; Faulkner, L. R. Electrochemical Methods: Fundamentals and Applications, 2nd ed.; John Wiley & Sons: New York, 2000; pp 125−130. (40) Hu, H.; Bhowmik, P.; Zhao, B.; Hamon, M. A.; Itkis, M. E.; Haddon, R. C. Determination of the Acidic Sites of Purified SingleWalled Carbon Nanotubes by Acid-Base Titration. Chem. Phys. Lett. 2001, 345, 25−28. (41) Vichchulada, P.; Cauble, M. A.; Abdi, E. A.; Obi, E. I.; Zhang, Q.; Lay, M. D. Sonication Power for Length Control of Single-Walled Carbon Nanotubes in Aqueous Suspensions Used for 2-Dimensional Network Formation. J. Phys. Chem. C 2010, 114, 12490−12495. (42) Chakrapani, V.; Angus, J. C.; Anderson, A. B.; Wolter, S. D.; Stoner, B. R.; Sumanasekera, G. U. Charge Transfer Equilibria between Diamond and an Aqueous Oxygen Electrochemical Redox Couple. Science 2007, 318, 1424−1430. (43) Collins, P. G.; Bradley, K.; Ishigami, M.; Zettl, A. Extreme Oxygen Sensitivity of Electronic Properties of Carbon Nanotubes. Science 2000, 287, 1801−1804. (44) Chakrapani, V.; Pendyala, C.; Kash, K.; Anderson, A. B.; Sunkara, M. K.; Angus, J. C. Electrochemical Pinning of the Fermi Level: Mediation of Photoluminescence from Gallium Nitride and Zinc Oxide. J. Am. Chem. Soc. 2008, 130, 12944−12952. (45) Levesque, P. L.; Sabri, S. S.; Aguirre, C. M.; Guillemette, J.; Siaj, M.; Desjardins, P.; Szkopek, T.; Martel, R. Probing Charge Transfer at Surfaces Using Graphene Transistors. Nano Lett. 2011, 11, 132−137. (46) Ignaczak, A.; Schmickler, W.; Bartenschlager, S. Electrochemical Reduction of the O2 Molecule to the Radical Ion − A Theoretical Approach. J. Electroanal. Chem. 2006, 586, 297−307. (47) Heller, I.; Kong, J.; Williams, K. A.; Dekker, C.; Lemay, S. G. Electrochemistry at Single-Walled Carbon Nanotubes: The Role of Band Structure and Quantum Capacitance. J. Am. Chem. Soc. 2006, 128, 7353−7359. (48) Blanch, A. J.; Lenehan, C. E.; Quinton, J. S. Optimizing Surfactant Concentrations for Dispersion of Single-Walled Carbon Nanotubes in Aqueous Solution. J. Phys. Chem. B 2010, 114, 9805− 9811. (49) Moonoosawmy, K. R.; Kruse, P. Cause and Consequence of Carbon Nanotube Doping in Water and Aqueous Media. J. Am. Chem. Soc. 2010, 132, 1572−1577. (50) Kim, K. K.; Bae, J. J.; Park, H. K.; Kim, S. M.; Geng, H.-Z.; Park, K. A.; Shin, H.-J.; Yoon, S.-M.; Benayad, A.; Choi, J.-Y.; et al. Fermi Level Engineering of Single-Walled Carbon Nanotubes by AuCl3 Doping. J. Am. Chem. Soc. 2008, 130, 12757−12761. (51) Petit, T.; Girard, H. A.; Trouve, A.; Batonneau-Gener, I.; Bergonzo, P.; Arnault, J.-C. Surface Transfer Doping Can Mediate Both Colloidal Stability and Self-Assembly of Nanodiamonds. Nanoscale 2013, 5, 8958−8962. (52) Suksrichavalit, T.; Yoshimatsu, K.; Prachayasittikul, V.; Bulow, L.; Ye, L. ″Clickable″ Affinity Ligands for Effective Separation of Glycoproteins. J. Chromatogr. A 2010, 1217, 3635−3641. (53) Thong, Y. X.; Poon, Y. F.; Chen, T. Y.; Li, L. J.; Chan-Park, M. B. Direct Intermolecular Force Measurements between Functional Groups and Individual Metallic or Semiconducting Single-Walled Carbon Nanotubes. Small 2014, 10, 750−757. (54) LeMieux, M. C.; Roberts, M.; Barman, S.; Jin, Y. W.; Kim, J. M.; Bao, Z. N. Self-Sorted, Aligned Nanotube Networks for Thin-Film Transistors. Science 2008, 321, 101−104. (55) Jagota, A.; Diner, B.; Boussaad, S.; Zheng, M. Carbon Nanotube−Biomolecule Interactions: Applications in Carbon Nanotube Separation and Biosensing. In Applied Physics of Carbon Nanotubes: Fundamentals of Theory, Optics and Transport Devices; Springer: Berlin, 2005; pp 258−261. (56) Moshammer, K.; Hennrich, F.; Kappes, M. M. Selective Suspension in Aqueous Sodium Dodecyl Sulfate According to Electronic Structure Type Allows Simple Separation of Metallic from Semiconducting Single-Walled Carbon Nanotubes. Nano Res. 2009, 2, 599−606. J


Article

ACS Nano (57) Thendie, B.; Miyata, Y.; Kitaura, R.; Miyauchi, Y.; Matsuda, K.; Shinohara, H. Rapid Single-Stage Separation of Micrometer-Long and High-Purity Semiconducting Carbon Nanotubes by Gel Filtration. Appl. Phys. Express 2013, 6, 065101. (58) Edmonds, M. T.; Pakes, C. I.; Mammadov, S.; Zhang, W.; Tadich, A.; Ristein, J.; Ley, L. Work Function, Band Bending, and Electron Affinity in Surface Conducting (100) Diamond. Phys. Status Solidi A 2011, 208, 2062−2066. (59) Dettlaff-Weglikowska, U.; Skákalová, V.; Graupner, R.; Jhang, S. H.; Kim, B. H.; Lee, H. J.; Ley, L.; Park, Y. W.; Berber, S.; Tománek, D.; et al. Effect of SOCl2 Treatment on Electrical and Mechanical Properties of Single-Wall Carbon Nanotube Networks. J. Am. Chem. Soc. 2005, 127, 5125−5131. (60) Petit, T.; Pfluger, M.; Tolksdorf, D.; Xiao, J.; Aziz, E. F. Valence Holes Observed in Nanodiamonds Dispersed in Water. Nanoscale 2015, 7, 2987−2991. (61) Tarábek, J.; Kavan, L.; Dunsch, L.; Kalbac, M. Chemical States of Electrochemically Doped Single Wall Carbon Nanotubes as Probed by Raman Spectroelectrochemistry and ex situ X-Ray Photoelectron Spectroscopy. J. Phys. Chem. C 2008, 112, 13856−13861. (62) Ballutaud, D.; Simon, N.; Girard, H.; Rzepka, E.; Bouchet-Fabre, B. Photoelectron Spectroscopy of Hydrogen at the Polycrystalline Diamond Surface. Diamond Relat. Mater. 2006, 15, 716−719. (63) Qi, D.; Chen, W.; Gao, X.; Wang, L.; Chen, S.; Loh, K. P.; Wee, A. T. S. Surface Transfer Doping of Diamond (100) by TetrafluoroTetracyanoquinodimethane. J. Am. Chem. Soc. 2007, 129, 8084−8085. (64) Williams, O. A.; Hees, J.; Dieker, C.; Jäger, W.; Kirste, L.; Nebel, C. E. Size-Dependent Reactivity of Diamond Nanoparticles. ACS Nano 2010, 4, 4824−4830. (65) Akai, Y.; Saito, S. Electronic Structure, Energetics and Geometric Structure of Carbon Nanotubes: A Density-Functional Study. Phys. E 2005, 29, 555−559.

K


Selective Surface Charge Sign Reversal on Metallic Carbon

Recommend Documents