Surfactant Chemistry and Polymer Choice Affect Single-Wall Carbon Nanotube Extraction Conditions in Aqueous Two-Polymer Phase Extraction

Abstract

Quantitative determination of the effects of surfactant chemistry and polymer chain length on the concentration conditions necessary to yield extraction of specific single-wall carbon nanotube (SWNCT) species in an aqueous two-polymer phase extraction (ATPE) separation are reported. In particular, the effects of polyethylene glycol (PEG) chain length, surfactant ratios, and systematic structural variations of alkyl surfactants and bile salts on the surfactant ratios necessary for extraction were investigated using a recently reported fluorescence-based method. Alkyl surfactant tail length was observed to strongly affect the amount of surfactant necessary to cause PEG-phase extraction of nanotube species in ATPE, while variation in the anionic sulfate/sulfonate head group chemistry has less impact on the concentration necessary for extraction. Substitution of different bile salts results in different surfactant packings on the SWCNTs, with substitution greatly affecting the alkyl surfactant concentrations required for (n,m) extraction. Finally, distinct alkyl-to-bile surfactant ratios were found to extract specific (n,m) SWCNTs across the whole effective window of absolute concentrations, supporting the hypothesized competitive adsorption mechanism model of SWCNT sorting. Altogether, these results provide valuable insights into the underlying mechanisms behind ATPE-based SWCNT separations, towards further development and optimization of the ATPE method for SWCNT chirality and handedness sorting.

Keywords: SWCNT; Single-wall carbon nanotube; aqueous two-phase extraction; separation.

Figure 1.

A) Schematic showing the molecular structures and points of variation for the alkyl sulfate/sulfonate co-surfactants. The primary variation investigated is the effect of the number of repeat units, n, in the alkyl tail. B) Schematic of the bile salts used in this contribution. The circles and boxes highlight the chemical moieties that are different or present/absent across the shown set. The most commonly competed ATPE surfactants are SDS and DOC.

Figure 2.

A) Absorbance spectra of the (6,5)/(8,3) rich (purple) and (7,6)/(7,5) rich (blue) parent dispersions as measured in 10.0 g/L DOC. The (8,3) is a small shoulder on the (6,5) lowest energy optical transition (E₁₁) at ≈ 985 nm, but is visible in fluorescence measurements due to its much larger fluorescence cross-section to the 637 nm laser excitation. B) Fluorescence emission for dilute aliquots of the (6,5)/(8,3) rich (purple) and (7,6)/(7,5) rich (blue) parent dispersions at 637 nm excitation.

Figure 3.

Fluorescence intensities of the +(7,5) SWCNT (637 nm excitation) as functions of SDS concentration for different molecular mass PEG polymers at a constant concentration of 35 g/L (3.5 %), with a constant DOC concentration of 0.5 g/L (0.05 %) at 20.0 °C. The effect of PEG addition to the PCCC is significant but is only weakly dependent on its molecular mass. Curves are double finite-width Heaviside fits to the data. PCCCs from Heaviside fitting are listed in Table S1.

Figure 4.

A) Fluorescence intensities of the +(7,5) SWCNT (637 nm excitation) as functions of SDS concentration for different concentrations of DOC in the dispersion with a constant 6 kDa PEG concentration of 35 g/L at 20.0 °C. Curves are double finite-width Heaviside fits to the data. Values for the fluorescence transition concentrations from Heaviside fitting are listed in Table S2. The amount of SDS necessary to change from a DOC-dominated interface to an SDS dominated interface increases with the concentration of DOC in solution. B) PCCCs for the +(7,5), and both enantiomers of the (7,6), (8,3), and (6,5) SWCNTs as a function of DOC concentration show a consistent linear trend (slope), reflecting that the PCCC occurs at an SDS:DOC mass ratio ≈ 15:1 for the +(7,5) SWCNT, and on average 15.6:1 for the (7,6) SWCNTs, 22.6:1 for the (8,3) SWCNTs, and 25.4:1 for the (6,5) SWCNTs species (separate values for both enantiomers are provided in Table S2).

Figure 5.

Fluorescence intensities of the +(7,5) SWCNT (637 nm excitation) as functions of AcS variant concentration at a constant 6 kDa PEG concentration of 35 g/L, constant DOC concentration of 0.5 g/L, and T = 20.0 °C. In A, the concentration of alkyl surfactant necessary to reach the PCCC increases significantly with decreasing AcS tail length, but with similar functionality for tail lengths of 9 to 12 carbons. In B, the unique behavior of SDBS (blue) is shown in comparison with its structurally similar analogues (STS, StS, SDS), which are replotted from A. In both plots the curves are double finite-width Heaviside fits to the data. PCCCs from Heaviside fitting are listed in Table S3.

Figure 6.

Alkyl tail length variant AcS concentration vs. fluorescence emission wavelength (637 nm excitation) 2D contour plots for the (7,6)/(7,5)-rich sample, where the z-scale is the scaled fluorescence intensity, at a constant 6 kDa PEG concentration of 35 g/L, with a constant DOC concentration of 0.5 g/L at 20.0 °C. The labeled dashed lines represent the wavelengths of peak emission [≈ 1035 nm for (7,5), ≈ 1131 nm for (7,6)] for each (n,m) species at 0.5 g/L DOC, 35 g/L 6 kDa PEG. Differences are observed in the peak emission wavelengths at AcS concentrations above the PCCC (intensity transition) that depend on the choice of AcS.

Figure 7.

Fluorescence intensities of the +(7,5) SWCNT (637 nm excitation) as functions of SDS concentration for different bile salts at a constant concentration of 0.5 g/L, with a constant 6 kDa PEG concentration of 35 g/L at 20.0 °C. Curves are double finite-width Heaviside fits to the data. PCCCs from Heaviside fitting are listed in Table S5.

Figure 8.

SDS concentration vs. fluorescence emission wavelength 2D contour plots for the two SWCNT samples, where the z-scale is the scaled fluorescence intensity, for different bile salts at a constant concentration of 0.5 g/L with a constant 6 kDa PEG concentration of 35 g/L at 20.0 °C. The (7,6)/(7,5)-rich sample emission under 637 nm excitation is shown at the top and the (6,5)/(8,3)-rich sample emission under 671 nm excitation is shown at the bottom. The dashed lines represent the wavelengths of peak emission [≈ 1035 nm for (7,5), ≈ 1131 nm for (7,6), ≈ 963 nm for (8,3), and ≈ 984 nm for (6,5)] for each (n,m) species at 0.5 g/L DOC, 35 g/L 6 kDa PEG. Note the initial emission wavelength shifts due to different bile salt coverage, with additional wavelength shifting during the transition to SDS coverage for SC and CDOC.