Spontaneous Partition of Carbon Nanotubes in Polymer-modified ...
Supporting Information for
Spontaneous Partition of Carbon Nanotubes in Polymer-modified Aqueous Phases Constantine Y Khripin, Jeffrey A. Fagan, and Ming Zheng* Materials Science and Engineering Division, National Institute of Standards and Technology 100 Bureau Drive, Gaithersburg, MD 20899 *Correspondence to: [email protected]
Preparation of Aqueous Two-Phase (ATP) System for Separation. The two polymer components of the ATP system were dextran (DX, from Leuconostoc mesenteroides, MW 64,000 – 78,000 Da, Sigma-Aldrich), and polyethylene glycol (PEG, 6,000 Da, Alpha Aesar). DX was dissolved in DI H2O to make a stock solution of 20 % wt., PEG was dissolved in DI H2O to make a stock solution of 50 % wt.. Other components were: SWCNT dispersion (in 2 % wt. SC), 10 % wt. SDS, 10 % wt. SC, and DI H2O. Table S1 gives the compositions of experimental systems used in this work. The proportion of each component given in this table can be used to design ATP system of different total volumes for SWCNT separation. Table S1 Compositions of the ATP systems used for SWCNT separation. Modifications for specific experiments were described below. DX 70 kDa 20 % wt.
PEG 6 kDa 50 % wt.
SC 10 % wt.
SDS 10 % wt.
SWCNT in 2 % SC
DI H2O
1 L scale 300 mL 120 mL 90 mL 70 mL 1 mL 420 mL separation 510 µL scale 150 µL 60 µL 40 µL* 35 µL 25 µL* 200 µL* separation * The quantities of these components were adjusted to accommodate SWCNT samples of different volume while keeping the final SC concentration constant.
A typical ATP separation experiment began by mixing in an eppendorf tube the six components listed in Table S1. We observed that the partition outcome was independent of the order by which these components were mixed. The mixture was vortexed for 10 sec and then allowed to settle for 15 min, after which it was centrifuged at 4,000 g for 5 min. This led to a clear formation of two aqueous phases and partition of SWCNTs in the two phases. The length dependence study (Figure S1a) was performed at 19 ºC. First, appropriate amounts of different length fractions from a size-exclusion chromatography (SEC) run were diluted to 225 µL final volume with 2 % SC, so that the final SWCNT mass concentrations for all samples were the same. The relative SWCNT mass concentration was determined according S1
to a method given in reference 1. These samples were added to an ATP system in total volume of 505 µL. Note that in this case no additional SC was required, making the final recipe 150 µL 20 % DX, 60 µL 50 % PEG, 35 µL 10 % SDS, 35 µL DI H2O and 225 µL SWCNT sample in 2 % SC. For the surfactant concentration dependence study (Figure S1b), experiments were carried out at 9 ºC. The amount of SWCNT sample and SDS were reduced to achieve the desired total surfactant concentration while keeping the SDS:SC ratio constant. For the NaSCN addition study (Figure S1c), appropriate volume of 1 M NaSCN was added to the ATP system in place of an equal volume of DI water to achieve the desired concentration. The study was carried out at 23 ºC. At this temperature, all SWCNTs normally partition into the top phase in the absence of NaSCN. Addition of NaSCN gradually pushed SWCNTs from the top phase into the bottom phase, allowing full range of partition to be reached. The temperature dependence study (Figure S1d) was carried out by preparing a series of identical ATP samples of 505 µL. The temperature of water baths was checked with a NIST traceable digital thermometer (Control Co., Friendswood, TX). Each sample was vortexed for 10 sec, incubated at its required temperature for 5 min, then vortexed again and incubated for at least 15 min. After this the two phases were observed to form and the sample was centrifuged at 4,000 g for 5 min to produce a clear phase boundary. The SWCNT concentration dependence study (Figure 2a) was carried out at 9 ºC. The same temperature was used for the large-scale separation (Figure 2b). For the latter, the ATP mixture without SWCNTs was incubated at 9 ºC for 2 hours, then the SWCNT sample was added and the mixture was incubated overnight to allow the phases to establish a well-defined boundary. As a general guidance, SWCNT partition equilibrium in an ATP system can be conveniently tuned by temperature, surfactant concentrations, and NaSCN concentration, as indicated by Figure S1. Thus, if SWCNT population is initially found only in the top PEG phase, one can push the equilibrium towards having more tubes in the bottom DX phase by lowering temperature, by adding more SC or NaSCN, or by a combination of these measures. Conversely, if majority SWCNTs are initially in the bottom DX phase, one can shift the equilibrium towards having more tubes in the top PEG phase by increasing temperature, or by adding more SDS. Spectroscopic Analysis. The relative abundance of metallic and semiconducting SWCNTs in the PEG and DX phase was determined by UV–vis–NIR absorption measurement. The chirality distribution of CoMoCAT SWCNTs in the two phases was determined by fluorescence measurement. UV–vis–NIR measurements were carried out on a Varian Cary 5000 spectrophotometer using a 10 mm path length microcuvette. Samples were diluted into 1 wt. % SDC by at least 2x for those from the PEG phase and 3x for those from the DX phase. Blank ATP samples without SWCNTs were prepared to obtain accurate baseline measurements. The partition coefficient of metallic and semiconducting SWCNTs was calculated from the absorbance at the M11 and S22 excitonic transition peaks. The off-peak contribution of metallic S2
and semiconducting nanotubes was approximated from the spectra of the purest metal and semiconductor fractions obtained in this study. Fluorescence measurement was performed on a JY-Horiba Nanolog-3 spectrofluorimeter with a liquid nitrogen cooled InGaAs detector, with 10 nm excitation and 5 nm emission slits. Excitation wavelengths were 580 nm for the (6,4), 6,5) and (8,4) chiralities, 650 nm for the (7,5) and (7,6) chiralities, and 670 nm for the (8,3) chirality. The partition coefficient was calculated by using fluorescence intensity as a measure of concentration for each SWCNT chirality. ATP extracted SWCNT samples were diluted by 10 x into 1 wt. % SDC for measurement. We determined that residual polymer had no effect on the fluorescence signal.
Preparation of Length-Fractionated SWCNT Samples. SWCNTs synthesized using arcdischarge method were obtained from Hanwha Nanotech (average diameter 1.4 nm, grade ASP100F, Incheon, Korea). CoMoCAT SWCNTs were graciously provided by SouthWest NanoTechnologies (average diameter 0.8 nm, grade SG65i, Norman, OK). SWCNTs were first dispersed using 1 % wt. sodium deoxycholate (SDC) as follows. Fifteen mg of nanotube powder was mixed with 15 mL of the surfactant solution in a rosette sonication cell and sonicated for 1 hr in an ice bath at 20 W using a ¼ ” probe and a VCX130 sonicator (all purchased from Sonics and Materials, Inc., Newtown, CT). To remove bundles and nanotubes with poor optical qualities, we used a rate-zonal centrifugation method described in reference 2. In two centrifuge tubes (Beckman-Coulter Optiseal #362183), 7.2 ml of SWCNT solution was layered over 29 ml of 10 % iodixanol/1 % SDC solution and centrifuged at 5236 Rad/s in a VTi 50 rotor (avg. force ≈ 206 000 * g) for 300 minutes. The top nanotube band (≈ 7.5 mL occurring in the middle of the centrifuge tube) was collected from both centrifuge tubes and combined. The solution was diluted 10 x with 1 % SDC, and concentrated to a final volume of 5 mL using a pressure cell (Amicon 8003) with a 100 kDa cutoff filter (regenerated cellulose YM100), all purchased from Millipore, Billerica, MA. This step has the additional effect of reducing the iodixanol concentration by 10-fold. Length fractionation by SEC was carried out according to the method described in reference 3. Preparation of Non-Length-Fractionated SWCNTs. The above procedure was modified as follows to produce SWCNT samples which could be used directly in ATP separation without length sorting. The dispersion was carried out in 2 wt. % SC with 0.33 g / L to 4 g / L nanotube mass concentration. In the case of 10 g / L, 3 wt. % SC was used since it was found to perform better. The dispersion was either carried out at the 15 mL scale as described above, or in a 1 mL centrifuge tube, using a smaller 2 mm probe and 7 W sonication power. In the former case, the dispersion was centrifuged in a fixed angle rotor for 2 hours at 20,000 g. In the latter case, the dispersion was divided into 100 uL aliquots and centrifuged for 75 min at 17,000 g. AFM Analysis. SWCNT length distribution was measured by AFM. AFM sample preparation was done according to a procedure described in reference 3.
S3
Figure S1 Factors controlling partition coefficient K for the arc-discharge CNTs. Plotted are ln K for metallic (blue dots) and semiconducting (pink dots) tubes as a function of a. nanotube length; b. total surfactant concentration at a constant ratio of 9:7 for [SC]:[SDS]; c. chaotropic salt NaSCN concentration; and d. temperature. Other conditions are the same as those given in Figure 1.
S4
Normalized count (a. u.)
a
As-dispersed Semi-sorted
10
b
100 SWCNT length
c
As-dispersed
500 nm
1000
Semi-sorted
500 nm
Figure S2 Length histograms and representative AFM images of arc-discharge SWCNTs. a. Length histograms of 368 SWCNTs measured for an “as-dispersed” sample and 204 SWCNTs for a semiconductor-enriched (“semi-sorted”) sample obtained after 9 rounds of ATP extraction. b,c Representative AFM image of the “as-dispersed” and “semi-sorted” sample, respectively.
S5
Disclaimer. Certain commercial equipment, instruments, or materials are identified in this paper in order to specify the experimental procedure adequately. Such identification is not intended to imply recommendation or endorsement by the national institute of standards and technology, nor is it intended to imply that the materials or equipment identified are necessarily the best available for the purpose. Unless noted otherwise, all reagents were obtained from standard sources.
References (1) (2) (3)
Khripin, C.; Tu, X.; Howarter, J. A.; Fagan, J. A.; Zheng, M. Anal. Chem. 2012, 84, 8733–8739. Fagan, J. A.; Huh, J. Y.; Simpson, J. R.; Blackburn, J. L.; Holt, J. M.; Larsen, B. A.; Hight Walker, A. R. ACS Nano 2011, 5, 3943-3953. Khripin, C. Y.; Tu, X.; Heddleston, J. M.; Silvera-Batista, C.; Hight Walker, A. R.; Fagan, J.; Zheng, M. Anal. Chem. 2013, 85, 1382-1388.
S6
Spontaneous Partition of Carbon Nanotubes in Polymer-modified Aqueous Phases Constantine Y Khripin, Jeffrey A. Fagan, and Ming Zheng* Materials Science and Engineering Division, National Institute of Standards and Technology 100 Bureau Drive, Gaithersburg, MD 20899 *Correspondence to: [email protected]
Preparation of Aqueous Two-Phase (ATP) System for Separation. The two polymer components of the ATP system were dextran (DX, from Leuconostoc mesenteroides, MW 64,000 – 78,000 Da, Sigma-Aldrich), and polyethylene glycol (PEG, 6,000 Da, Alpha Aesar). DX was dissolved in DI H2O to make a stock solution of 20 % wt., PEG was dissolved in DI H2O to make a stock solution of 50 % wt.. Other components were: SWCNT dispersion (in 2 % wt. SC), 10 % wt. SDS, 10 % wt. SC, and DI H2O. Table S1 gives the compositions of experimental systems used in this work. The proportion of each component given in this table can be used to design ATP system of different total volumes for SWCNT separation. Table S1 Compositions of the ATP systems used for SWCNT separation. Modifications for specific experiments were described below. DX 70 kDa 20 % wt.
PEG 6 kDa 50 % wt.
SC 10 % wt.
SDS 10 % wt.
SWCNT in 2 % SC
DI H2O
1 L scale 300 mL 120 mL 90 mL 70 mL 1 mL 420 mL separation 510 µL scale 150 µL 60 µL 40 µL* 35 µL 25 µL* 200 µL* separation * The quantities of these components were adjusted to accommodate SWCNT samples of different volume while keeping the final SC concentration constant.
A typical ATP separation experiment began by mixing in an eppendorf tube the six components listed in Table S1. We observed that the partition outcome was independent of the order by which these components were mixed. The mixture was vortexed for 10 sec and then allowed to settle for 15 min, after which it was centrifuged at 4,000 g for 5 min. This led to a clear formation of two aqueous phases and partition of SWCNTs in the two phases. The length dependence study (Figure S1a) was performed at 19 ºC. First, appropriate amounts of different length fractions from a size-exclusion chromatography (SEC) run were diluted to 225 µL final volume with 2 % SC, so that the final SWCNT mass concentrations for all samples were the same. The relative SWCNT mass concentration was determined according S1
to a method given in reference 1. These samples were added to an ATP system in total volume of 505 µL. Note that in this case no additional SC was required, making the final recipe 150 µL 20 % DX, 60 µL 50 % PEG, 35 µL 10 % SDS, 35 µL DI H2O and 225 µL SWCNT sample in 2 % SC. For the surfactant concentration dependence study (Figure S1b), experiments were carried out at 9 ºC. The amount of SWCNT sample and SDS were reduced to achieve the desired total surfactant concentration while keeping the SDS:SC ratio constant. For the NaSCN addition study (Figure S1c), appropriate volume of 1 M NaSCN was added to the ATP system in place of an equal volume of DI water to achieve the desired concentration. The study was carried out at 23 ºC. At this temperature, all SWCNTs normally partition into the top phase in the absence of NaSCN. Addition of NaSCN gradually pushed SWCNTs from the top phase into the bottom phase, allowing full range of partition to be reached. The temperature dependence study (Figure S1d) was carried out by preparing a series of identical ATP samples of 505 µL. The temperature of water baths was checked with a NIST traceable digital thermometer (Control Co., Friendswood, TX). Each sample was vortexed for 10 sec, incubated at its required temperature for 5 min, then vortexed again and incubated for at least 15 min. After this the two phases were observed to form and the sample was centrifuged at 4,000 g for 5 min to produce a clear phase boundary. The SWCNT concentration dependence study (Figure 2a) was carried out at 9 ºC. The same temperature was used for the large-scale separation (Figure 2b). For the latter, the ATP mixture without SWCNTs was incubated at 9 ºC for 2 hours, then the SWCNT sample was added and the mixture was incubated overnight to allow the phases to establish a well-defined boundary. As a general guidance, SWCNT partition equilibrium in an ATP system can be conveniently tuned by temperature, surfactant concentrations, and NaSCN concentration, as indicated by Figure S1. Thus, if SWCNT population is initially found only in the top PEG phase, one can push the equilibrium towards having more tubes in the bottom DX phase by lowering temperature, by adding more SC or NaSCN, or by a combination of these measures. Conversely, if majority SWCNTs are initially in the bottom DX phase, one can shift the equilibrium towards having more tubes in the top PEG phase by increasing temperature, or by adding more SDS. Spectroscopic Analysis. The relative abundance of metallic and semiconducting SWCNTs in the PEG and DX phase was determined by UV–vis–NIR absorption measurement. The chirality distribution of CoMoCAT SWCNTs in the two phases was determined by fluorescence measurement. UV–vis–NIR measurements were carried out on a Varian Cary 5000 spectrophotometer using a 10 mm path length microcuvette. Samples were diluted into 1 wt. % SDC by at least 2x for those from the PEG phase and 3x for those from the DX phase. Blank ATP samples without SWCNTs were prepared to obtain accurate baseline measurements. The partition coefficient of metallic and semiconducting SWCNTs was calculated from the absorbance at the M11 and S22 excitonic transition peaks. The off-peak contribution of metallic S2
and semiconducting nanotubes was approximated from the spectra of the purest metal and semiconductor fractions obtained in this study. Fluorescence measurement was performed on a JY-Horiba Nanolog-3 spectrofluorimeter with a liquid nitrogen cooled InGaAs detector, with 10 nm excitation and 5 nm emission slits. Excitation wavelengths were 580 nm for the (6,4), 6,5) and (8,4) chiralities, 650 nm for the (7,5) and (7,6) chiralities, and 670 nm for the (8,3) chirality. The partition coefficient was calculated by using fluorescence intensity as a measure of concentration for each SWCNT chirality. ATP extracted SWCNT samples were diluted by 10 x into 1 wt. % SDC for measurement. We determined that residual polymer had no effect on the fluorescence signal.
Preparation of Length-Fractionated SWCNT Samples. SWCNTs synthesized using arcdischarge method were obtained from Hanwha Nanotech (average diameter 1.4 nm, grade ASP100F, Incheon, Korea). CoMoCAT SWCNTs were graciously provided by SouthWest NanoTechnologies (average diameter 0.8 nm, grade SG65i, Norman, OK). SWCNTs were first dispersed using 1 % wt. sodium deoxycholate (SDC) as follows. Fifteen mg of nanotube powder was mixed with 15 mL of the surfactant solution in a rosette sonication cell and sonicated for 1 hr in an ice bath at 20 W using a ¼ ” probe and a VCX130 sonicator (all purchased from Sonics and Materials, Inc., Newtown, CT). To remove bundles and nanotubes with poor optical qualities, we used a rate-zonal centrifugation method described in reference 2. In two centrifuge tubes (Beckman-Coulter Optiseal #362183), 7.2 ml of SWCNT solution was layered over 29 ml of 10 % iodixanol/1 % SDC solution and centrifuged at 5236 Rad/s in a VTi 50 rotor (avg. force ≈ 206 000 * g) for 300 minutes. The top nanotube band (≈ 7.5 mL occurring in the middle of the centrifuge tube) was collected from both centrifuge tubes and combined. The solution was diluted 10 x with 1 % SDC, and concentrated to a final volume of 5 mL using a pressure cell (Amicon 8003) with a 100 kDa cutoff filter (regenerated cellulose YM100), all purchased from Millipore, Billerica, MA. This step has the additional effect of reducing the iodixanol concentration by 10-fold. Length fractionation by SEC was carried out according to the method described in reference 3. Preparation of Non-Length-Fractionated SWCNTs. The above procedure was modified as follows to produce SWCNT samples which could be used directly in ATP separation without length sorting. The dispersion was carried out in 2 wt. % SC with 0.33 g / L to 4 g / L nanotube mass concentration. In the case of 10 g / L, 3 wt. % SC was used since it was found to perform better. The dispersion was either carried out at the 15 mL scale as described above, or in a 1 mL centrifuge tube, using a smaller 2 mm probe and 7 W sonication power. In the former case, the dispersion was centrifuged in a fixed angle rotor for 2 hours at 20,000 g. In the latter case, the dispersion was divided into 100 uL aliquots and centrifuged for 75 min at 17,000 g. AFM Analysis. SWCNT length distribution was measured by AFM. AFM sample preparation was done according to a procedure described in reference 3.
S3
Figure S1 Factors controlling partition coefficient K for the arc-discharge CNTs. Plotted are ln K for metallic (blue dots) and semiconducting (pink dots) tubes as a function of a. nanotube length; b. total surfactant concentration at a constant ratio of 9:7 for [SC]:[SDS]; c. chaotropic salt NaSCN concentration; and d. temperature. Other conditions are the same as those given in Figure 1.
S4
Normalized count (a. u.)
a
As-dispersed Semi-sorted
10
b
100 SWCNT length
c
As-dispersed
500 nm
1000
Semi-sorted
500 nm
Figure S2 Length histograms and representative AFM images of arc-discharge SWCNTs. a. Length histograms of 368 SWCNTs measured for an “as-dispersed” sample and 204 SWCNTs for a semiconductor-enriched (“semi-sorted”) sample obtained after 9 rounds of ATP extraction. b,c Representative AFM image of the “as-dispersed” and “semi-sorted” sample, respectively.
S5
Disclaimer. Certain commercial equipment, instruments, or materials are identified in this paper in order to specify the experimental procedure adequately. Such identification is not intended to imply recommendation or endorsement by the national institute of standards and technology, nor is it intended to imply that the materials or equipment identified are necessarily the best available for the purpose. Unless noted otherwise, all reagents were obtained from standard sources.
References (1) (2) (3)
Khripin, C.; Tu, X.; Howarter, J. A.; Fagan, J. A.; Zheng, M. Anal. Chem. 2012, 84, 8733–8739. Fagan, J. A.; Huh, J. Y.; Simpson, J. R.; Blackburn, J. L.; Holt, J. M.; Larsen, B. A.; Hight Walker, A. R. ACS Nano 2011, 5, 3943-3953. Khripin, C. Y.; Tu, X.; Heddleston, J. M.; Silvera-Batista, C.; Hight Walker, A. R.; Fagan, J.; Zheng, M. Anal. Chem. 2013, 85, 1382-1388.
S6
