Tuning the Surface Ordering of Self-Assembled Ionic Surfactants on

for the lineshape comparison; multiplication factors are shown when applicable (this scaling may also affect the noise and nonresonant background ...
0 downloads 0 Views 1018KB Size
Download PDF
Subscriber access provided by STEPHEN F AUSTIN STATE UNIV

Interface-Rich Materials and Assemblies

Tuning the Surface Ordering of Self-Assembled Ionic Surfactants on Semiconducting Single-Walled Carbon Nanotubes: Concentration, Tube Diameter and Counterions Soha T Algoul, Sanghamitra Sengupta, Thomas Tom Bui, and Luis Velarde Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.8b01813 • Publication Date (Web): 14 Jul 2018 Downloaded from http://pubs.acs.org on July 15, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Tuning the Surface Ordering of Self-Assembled Ionic Surfactants on Semiconducting Single-Walled Carbon Nanotubes: Concentration, Tube Diameter and Counterions Soha T. Algoul†, Sanghamitra Sengupta†, Thomas T. Bui, and Luis Velarde*

AUTHOR ADDRESS: Department of Chemistry, University at Buffalo, State University of New York, Buffalo, NY 14260-3000

†

S.T.A and S. S. contributed equally to this work.

1 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ABSTRACT We report direct spectroscopic measurements of the macromolecular organization of ionic surfactants on the surface of semiconducting single-walled carbon nanotubes (SWCNTs) within solution-processed thin films. By using vibrational sum frequency generation (VSFG) spectroscopy, sensitive measurements of interfacial surfactant ordering were obtained as a function of surfactant concentration for sodium dodecyl sulfate (SDS) encapsulated (6,5) and (7,6) SWCNTs with and without excess electrolytes. Anionic surfactants are known to effectively stabilize SWCNTs. Current models suggest a strong influence of the dispersion conditions on the surfactant interfacial macromolecular organization and self-assembly. Direct experimental probes of such organization using nanotubes of specific chirality are needed in order to validate existing models. We found that as the bulk SDS concentration increases near the surfactant critical micelle concentration (CMC), the interfacial ordering increased approaching the formation of cylindrical-like micelles with the nanotube at the core. At the higher surfactant concentrations measured here, the (6,5) SWCNTs produced more ordered structures relative to those with the (7,6) SWCNTs. The relatively larger diameter (7,6) chiral tubes support enhanced van der Waals (vdW) interactions between the tube carbon surface and the surfactant methylene chain groups that likely increases the density of gauche defects. A new effect arises when the precursor solution is exposed to a small concentration of divalent Ca2+ counterions. We postulate that a salt bridging configuration on such highly curved surfaces decreases the ordering of interfacial surfactant molecules resulting in compact, disordered structures. However, this phenomenon was not observed with excess Na+ ions at the same ionic strength. Instead, a modest increase in surfactant ordering was observed with the excess monovalent electrolyte. These results provide new insights for thin film solution processing of vdW nanomaterials and demonstrate that VSFG is a sensitive probe of surfactant organization on nanostuctuctures.

2 ACS Paragon Plus Environment

Page 2 of 32

Page 3 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

INTRODUCTION The use of surfactants is widespread in modern solution processing of nanomaterials.1 This has created a great interest from both fundamental and applied perspectives in the understanding of non-covalent modification of carbon and nanostructured materials via vdW forces.2-4 For instance, density-gradient separations using different surfactant systems can effectively be used to sort SWCNTs by electronic type, chirality and diameter, opening the door to the fabrication of integrated circuits with impressive performance using solution processing techniques.5 The socalled “purified-and-placed” method has proven to be advantageous compared to chemical vapor deposition, where a major concern is the effective removal of metallic nanotubes which may cause undesirable device characteristics.6-8 In brief, carbon nanotubes (CNTs) can be readily suspended in solution using different surfactants. Semiconducting and metallic nanotubes are then separated via a sonication and centrifugation process, and lastly, the extracted semiconducting SWCNTs are assembled in controlled arrays on a wafer using surface chemistry techniques. This method presents a great potential for scalable manufacturing technologies.9 One of the most extensively studied systems in the field of surfactant-stabilized CNTs is sodium dodecyl sulfate (SDS)–SWCNT dispersions in aqueous solutions.10-11 The structure and dynamics associated with SDS self-assembly around cylindrical SWCNTs nanostructures has attracted substantial attention from experimental12-17 and theoretical groups.18-20 It provides a model system to clearly understand the interfacial structure and phase behavior of ionic surfactants upon changing the physical and chemical factors associated with the nanotube microenvironment, such as addition of salts, co-solvents, temperature, and nanotube diameter. Despite large amounts of data on SDS–SWCNTs dispersions, there are a limited number of studies providing a direct observation of the supramolecular structure of the SDS–SWCNT assemblies in thin films prepared under controlled surface coverage, electrolytes, and nanotube curvatures.

3 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Vibrational Sum Frequency Generation spectroscopy (VSFG) is an optical second order spectroscopic technique.21-25 Under the electric dipole approximation, the optical sum frequency process is forbidden in centrosymmetric media. VSFG is therefore selective to interfacial species due to the inherent symmetry breaking that occurs at the surface plane. This technique has been increasingly gaining recognition for its ability to provide detailed structural analysis of selfassembled molecules at nanoscale surfaces and interfaces.26-31 Not only because its exquisite surface selectivity and sensitivity, but also due to its ability to determine molecular orientation and ordering at the particle surface.32-33 Our group recently reported the sensitivity of VSFG towards the interfacial organization of surfactant molecules around SWCNTs within a film as a function of surfactant concentration.15 The spectroscopic results provided here expand these initial studies and will help clarify existing questions on surfactant self-assembly processes on semiconducting CNT surfaces and the environmental conditions that lead to the formation of irregular structures (which may lead to the unfavorable exposure of the hydrophobic SWCNT surface and surfactant tails to the aqueous environment). While a number of theoretical and experimental studies have attempted to characterize the SDS self-assembly with SWCNT surfaces in the solution as a function of diameter, concentration and electrolyte addition, there is an absence of direct surfaceselective measurements; furthermore, knowledge on how SDS interacts with SWCNTs within a film is still missing. Deposition of SWCNTs into films for integration into electronic devices require them to be purified and dispersed into solutions. The final device properties are known to be influenced by the nanotube chirality and SWCNT dispersion conditions.34 The information provided here will be helpful in deciphering and improving dispersion and separation techniques for vdW nanostructures and associated solution-processing techniques in thin film deposition methods.

EXPERIMENTAL 4 ACS Paragon Plus Environment

Page 4 of 32

Page 5 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Sample preparation. Two widely used semiconducting nanotubes (6,5) and (7,6) enriched SWCNT species (referred by their manufacturer’s numbers as SG65 and SG76, respectively) were used as received from Sigma-Aldrich (CoMoCAT processing). Three stock aqueous dispersions of each species were prepared by mixing 1 mg of the corresponding nanotube powder with 10 ml of 7 mM, 10 mM, and 14 mM of SDS (ultrapure ≥99.0%, Sigma-Aldrich), respectively, in ultrapure deionized water (Millipore water purification system). CNT exfoliation was then carried out by the well-known sonication method. The temperature of the mixture was monitored constantly during sonication and kept at ~20−25 °C. A 5 ml aliquot of the supernatant fluid was extracted as the isolated SWCNT dispersion for thin film deposition on clean silica substrates by dip coating. A representative scanning electron imaging microscopy (SEM) image is provided in Figure S2 of Ref. 15 showing an ultrathin film with strands of the SWCNTs that lay flat on the surface. The glass slides (1 mm thick) were thoroughly cleaned by overnight soaking in freshly prepared NOCHROMIX solution and rinsed with copious amounts of ultrapure water. The slides were further cleaned by sequential sonication in Alconox detergent, acetone and finally in ultrapure water for 10 min each. The substrates were then dried in an oven at 110 ᵒC followed by UV-Ozone treatment for 30 min yielding a clean hydrophilic surface. To study the influence of excess electrolytes, the procedure described above was modified by the addition of crystalline salts to the mixture prior to sonication. To examine the interfacial effects arising from excess counterions on the SDS–SWCNT solution-processed films, we limited the surfactant concentration to 14 mM SDS. Both NaCl and CaCl2·2H2O crystalline powders were obtained from J.T Baker. To avoid contribution from impurities in the powder, the crystals were purified by solution filtration.35 Baking prior to filtration was applied only to NaCl due to the low melting point of the CaCl2·2H2O salt. In this work, we first used a 0.7 mM (aq) concentration for each salt. Such small amount was chosen because it exhibits measurable interactions with the SDS head group while minimizing inter-ionic forces which can play a dominant role at higher counterion concentrations. In addition, it is likely that at higher electrolyte concentrations the nanotube-nanotube head-group repulsion is screened, facilitating fusion of particles.36-37 Indeed, the critical coagulation concentration (CCC), defined as the 5 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

minimum concentration of counterions to induce coagulation of the colloidal particles, were reported as ~96 and ~13 mM NaCl for SG65 and SG76, respectively and ~2.8 and ~0.6 mM CaCl2 for SG65 and SG76, respectively.38-39 To study specific ion effects, we also performed a second experiment where a CaCl2 concentration of 0.23 mM was used in order to keep the same ionic strength (I) of 21 mM obtained with 0.7 mM NaCl in the dispersion as calculated using the Debye-Hückel formula.

Raman Spectroscopy and Nanotube Diameter. Raman spectra of the SDS–SWCNT films (provided in the Supporting Information (SI), Figure S1) prepared from SG65 and SG76 dispersions with 14 mM SDS were obtained with a Renishaw inVia Raman microscope using 514 nm Raman excitation. Each spectrum was the result of four accumulations acquired for 10 seconds each using 20X magnification, a 2400 l/mm grating, and a RenCam CCD detector (Renishaw). Using the observed radial breathing mode (RBM) frequencies at 311.3 cm-1 and 264.5 cm-1 for the SG65 and SG7 films, respectively, we estimate the average tube diameter to be 0.75 and 0.89 nm, respectively (see the SI for the calculations). These values are in excellent agreement with those reported by Oyama et al.40 and close to the average values for the pristine sample provided in the manufacturer specifications (0.78 and 0.83 nm for SG65 and SG76, respectively).

VSFG Spectroscopy. VSFG spectra were obtained using a home-built41-42 broadband spectrometer43 based on a regenerative amplifier (Legend HE+, Coherent, Inc.), seeded by a titanium-sapphire oscillator (Vitara, Coherent, Inc.). The amplifier output has a pulse width of ~35 femtoseconds (fs) centered at ~804 nm with a 1-kHz repetition rate. A portion of this output (~2.3 mJ/pulse) is used to generate mid-infrared (IR) pulses using an optical parametric amplifier (OPA). The signal and idler outputs from the OPA (Topas PRIME, Light Conversion) are fed to a difference frequency generation unit (NDFG, Light Conversion) to generate tuneable broadband mid-IR pulses. A second portion of the amplifier output (~1.7 mJ/pulse) is used to 6 ACS Paragon Plus Environment

Page 6 of 32

Page 7 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

generate narrowband picosecond (ps) pulses (~5 ps) at 400 nm in a second harmonic bandwidth compressor (SHBC, Light Conversion). The 400 nm pulses pump a picosecond OPA to generate tuneable narrowband (~7 cm-1) visible (Vis) pulses. For this work, the IR frequency was held fix at a center frequency of 2900 cm-1 and the visible up-converting Vis wavelength was set at 660 nm to avoid electronic resonances of the nanotubes. The spectra were calibrated using the 2850.15 cm-1 absorption line of a polystyrene film placed momentarily in the IR path while obtaining the broadband VSFG signal of a clean gold film. The angle of incidence for the IR and Vis beams were 70ᵒ and 45ᵒ, respectively, from the surface normal. The IR and Vis laser power was kept at ~4 mW and ~10.0 mW, respectively. The IR beam was focused to a spot size of ~200 µm. The Vis beam was mildly focused to a spot size of 500 µm at the sample where it was overlapped spatially and temporally with the IR beam. The resultant SFG beam was collimated, guided, and focused at the entrance slit of our spectrograph coupled to a liquid nitrogen cooled charge coupled device (CCD) camera (PYL100BRX, Princeton Instruments). A 650 nm short pass filer was used to prevent the incoming beams from reaching the detector. Each spectrum shown here is the average of three replicates on three different spots and on three different films to account for potential spatial heterogeneities and rule out contributions from islands of free surfactants directly in contact with the substrate. Each individual spectrum was acquired for 10 min, background subtracted, and normalized to the non-resonant SFG response of a bare gold film to account for the spectral power distribution of the broadband IR pulse. In order to ensure that the laser beams caused no damage to the film and/or any degradation of the sample quality, we recorded the SFG response at the same surface spot after 30 min of continuous illumination and obtained statistically equivalent intensity and spectral profiles when compared to our original SFG spectrum (Figure S2 in the SI).

Diffusion Ordered Spectroscopy (DOSY) NMR. Our DOSY NMR spectra were taken by using an Inova 500 MHz system (Varian) using a pulse field gradient (PFG) technique and processed using the Vnmrj software. All NMR spectra were recorded at room temperature (see 7 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

the SI for experimental details). We performed a pre-saturation experiment to suppress the large water signal obscuring the SDS signal by using a preset pulse sequence in the Vnmrj software prior to recording the DOSY spectra.

RESULTS AND DISCUSSION Effect of tube diameter and surfactant concentration on interfacial SDS ordering. We previously showed that the degree of interfacial ordering for perdeuterated SDS (or d-SDS) surfactants adsorbed on SG76 within a solid thin film depended directly on the surfactant concentration in the precursor dispersion.15 This degree of ordering for the SWCNT-bound SDS molecules was probed by VSFG spectroscopy in the C−D stretching region. The highly ordered surface structures observed at higher d-SDS concentrations were attributed to cylindrical-like micelle assemblies having the SWCNT at the core. As the d-SDS concentration in the solution decreased, the interfacial order was found to decrease as well. This loss of ordering at lower surfactant concentrations was associated to a more random (or wrapping) adsorption of surfactants on the nanotube surfaces.15 Here, we present new studies with hydrogenated SDS (or h-SDS) with both SG65 and SG76 SWCNTs. The chiral enrichments for the two SWCNTs show a similar roll-up angle (∼27°) with near armchair configuration, a major difference being their average tube diameter (0.75 and 0.89 nm for SG65 and SG76, respectively, as obtained from our Raman measurements).44-45 Since adsorption density and organization of surface SDS molecules have been proposed to depend on the nanotube curvature, it is important to determine if such difference in tube diameter has measurable effects on the VSFG spectra, providing therefore a direct observation of diameter-dependent ordering under the same microenvironments.18-19 The VSFG experiments were first carried out in two different polarization combinations, namely ssp and ppp, where “s” stands for light with electric field vector perpendicular to the plane of incidence and “p” for a field vector parallel to the incidence plane. The indexing order corresponding to the SFG, Vis and IR beams, respectively. In the CH stretching region, the ssp 8 ACS Paragon Plus Environment

Page 8 of 32

Page 9 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

spectrum is known to be more sensitive to symmetric vibrations, whereas ppp more responsive to asymmetric contributions.46-47 Figure 1 displays the average ssp and ppp VSFG spectra for films deposited from respective aqueous dispersions of SG65 and SG76 prepared with 14 mM SDS. All spectra were fitted to a sum of complex Lorentzians according to the following equation:48-51  ∝

  

      



     Γ

 1

where is the effective susceptibility for a particular polarization, Aq is the amplitude of a 

specific vibrational mode, ωq is the resonance frequency, 2Γq is the Lorentzian full-width at half

maximum (FWHM) of the peak centered at ωq, and ANR corresponds to the non-resonant

contribution with phase angle " with respect to the vibrationally resonant part.

9 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 1: VSFG response in the CH stretching region for ssp (red) and ppp (blue) polarization combinations for films deposited from 14mM SDS and (a) (6,5) SWCNT and (b) (7,6) SWCNT dispersions. The dots represent the experimental data points and the solid lines are the Lorentzian fits to Eq. 1.

For films of both SG65 and SG76 SDS–SWCNT hybrids, the spectral features in Figure 1 are assigned as follows:46-47, 52 prominent in the ssp polarization are the CH2 symmetric stretch (d+) at ~2850 cm-1, symmetric CH3 stretch (r+) ~2875 cm-1, unresolved CH2 asymmetric stretch (d−) and Fermi Resonance (FR) modes (dFR) in the 2910-2930 cm−1 region, and a Fermi Resonance (FR) enhanced CH3 bend overtone (rFR) ~2934 cm-1. The asymmetric CH3 stretch (r−) at ~2960 cm-1 is the most intense mode in the ppp spectra. The presence of CH2 stretching bands (better resolved in the ssp spectra) indicates the existence of gauche defects in the long hydrocarbon chains of the adsorbed SDS molecules on the SWCNT surfaces.15, 24, 47, 52-53 In brief, this is understood as follows: due to the requirement for a non-centrosymmetric environment in SFG spectroscopy, contributions from CH2 (methylene) stretching modes will be negligible if all methylene groups in the alkyl chain are in the all-trans centrosymmetric configuration. In such a highly ordered scenario, only the SFG contribution from the terminal CH3 (methyl) stretching modes would be observed in this region. In addition, SFG signals from ordered structures are amplified by in-phase coherent superposition of the individual molecular responses, resulting in a more intense CH3 response. When gauche defects are present in the alkyl chains (essentially in the form of kinks), the inversion symmetry in the long-chain is lost and therefore vibrational SFG signals from the CH2 groups near the gauche conformation sites contribute to the spectrum with SFG intensity related to the density of such defects. The increased disorder is also manifested by a reduction in the SFG intensity of the CH3 bands due to multiple conformations. Based on this, the intensity (or amplitude) ratio between CH3 and CH2 vibrational modes has been used as a direct tool to measure the degree of ordering in long alkyl hydrocarbon chains at interfaces. The symmetric stretches corresponding to such vibrational modes in the ssp spectrum 10 ACS Paragon Plus Environment

Page 10 of 32

Page 11 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

are typically used to determine the order ratio54-55 of such alkyl chains since asymmetric modes are not well-resolved.52, 56 Therefore, from here on, we limit the analysis to the ssp spectra. To rule out organic impurities perturbing our quantitative analysis, we performed VSFG experiments on all bare clean glass slides and found no SFG response from the substrates under identical acquisition conditions (Figure S3 in the SI). The SFG response of a surfactant-free film spun-coated from the nanotubes dissolved in DMF (shown previously in Fig. 2 of Ref. 15) has only a negligible contribution from residual DMF solvent. Previously, we reported that while there was a negligible change in the spectral lineshape (therefore no significant change in the order ratio) when only d-SDS molecules were deposited on the silica substrates at three different concentrations (6mM, 10mM, and 14mM). However, a clear trend could be observed for the spectra for the SG76 SDS–SWCNT films at such concentrations, where larger order ratios (higher ordering) were observed for the films prepared with higher d-SDS concentrations.15 In Figure 2, we show the VSFG spectra obtained for films of two different diameter SWCNTs (SG65 and SG76), exfoliated using three different h-SDS concentrations. The surfactant concentrations were guided by the critical micellar concentration (CMC) of SDS in water (~ 8.0 mM). We selected three different concentrations of SDS: 7 mM (slightly below the CMC), 10 mM and 14 mM (both moderately above CMC ). The prescence of CNTs is known to delay the transition of free SDS to micelles giving an adjusted or apparent CMC, which is larger than that of the pure SDS. The adsorption and saturation of surfactant on the nanotubes being implicated in this shift in the CMC through the overall mass balance (i.e. reduced free SDS in solution).14

11 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 32

Figure 2: Comparative plots of the VSFG spectra (ssp polarization) obtained from films made with (a) (6,5) SWCNT and (b) (7,6) SWCNT dispersions using 7mM (green), 10mM (blue), and 14mM (purple) SDS. The dots represent the data points and the solid lines represents the Lorentzian fits to Eq. 1. Each spectrum is scaled to the 2875 cm-1 r+ mode for lineshape comparison, multiplication factors are shown when applicable (this scaling may also affect the noise and non-resonant background level). The error bars are the 95% confidence intervals calculated from the intensity associated with experimental replicates. We fitted the VSFG spectra in Figure 2 using Eq. 1 and estimated the macroscopic

susceptibility  for the qth mode as  ∝  /Γ obtained directly from the fitting 

parameters.32-33,

57

Then ratio between the VSFG transition strength  of methyl (r+) to

methylene (d+) symmetric stretches is the order ratio,58 formally defined as  r% / d% . Because 



 is proportional to the number density of contributing surface groups, the amplitude ratio for 12 ACS Paragon Plus Environment

Page 13 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

the methylene and methyl stretch modes is therefore an effective way to quantify the fraction of gauche defects in the surface alkyl chains. Other approaches include the use of the fractional peak area for the methylene modes out of all the CH stretches to estimate the degree of disorder.59 Our fitting results show a direct correlation between the order ratio and bulk SDS concentration (see Table S1 and Figure 3). In agreement with our perdeuterated d-SDS results reported previously, when the SDS concentration is increased, the value of the order ratio increases for the SDS-SWCNT adducts. This is indicative of a micelle-like organization (standing-up surfactants) at higher concentrations. This trend is observed for films deposited from either SG65 or SG76 SWCNT dispersions. However, this data provides new insights into the effect of tube diameter. While probing a larger selection of nanotube diameters is desirable, we were limited by commercially available semiconducting SWCNTs at the time of the experiments. At higher SDS concentrations, our SFG results consistently showed distinct order ratios for the two types of nanotubes, indicating some clear differences in the extent of ordering for the SDS molecules around the two nanotube surfaces. Interestingly, at 14mM and 10mM SDS, the order ratio is consistently higher for the smaller diameter nanotube SG65 (diameter ~ 0.75 nm) compared to the larger diameter SG76 (diameter ~ 0.89 nm). This is evident in Figure 3. At 7 mM SDS, the order ratio becomes comparable in the two nanotube types -both films, however, are considered highly disordered at this concentration, consistent with a wrapping-like configuration.

13 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 3: Bulk SDS concentration dependencies of the order ratio (see Table S1) for SDSSWCNT films using (6,5) SWCNT (blue circles) and (7,6) SWCNT (red squares) dispersions. The shaded area is a guide to the eyes. SDS molecules bare an amphiphilic nature since they are comprised of a hydrophobic tail (12 carbon saturated alkyl chain) and a hydrophilic head (–OSO3−). Attractive vdW interactions dictate their adsorption and self-assembly on the nanotube surface. This normally results in the negatively charged sulfonate groups pointing away from the surface, assisting in the hydration process and in the formation of stable dispersions. In previous SFG analyses of ligand ordering on nanoparticle surfaces,27-28,

33

the surface molecules bind to the particle through strong

chemical interactions supported by a well-defined anchoring group, such as thiol groups or phosphonic acids. The alkyl chains are therefore in the “standing mode” geometry and the degree of ordering is dictated mainly by lateral vdW interactions and the available conical volume to the hydrocarbon chains to explore minimum free energy conformations. In contrast, the SDS– SWCNT system encounters competing vdW surface interactions between (i) neighboring longchain groups and (ii) those of the chain hydrocarbon groups with the carbon walls of the nanotube (which can alter ligand adsorption motifs). This raises important questions on the effects of curvature on the macromolecular organization and packing density in these cylindrical entities with competing vdW interactions. Molecular dynamics (MD) calculations predict that when the curvature of a CNT is small (large diameter), vdW interactions show no strong preference for an SDS molecule to be either 14 ACS Paragon Plus Environment

Page 14 of 32

Page 15 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

wrapped around or standing on the CNT surface, resulting in random or fan-like17 (for larger tubes) adsorption. In contrast, the simulations have shown that SDS molecules are more prone to the standing mode when the curvature of the CNT is large (diameter in the order of the chain length),18-19 such as in the nanotubes in this study. Due to the high surface curvature, the terminal CH3 groups of the surfactant arrange close to the nanotube surface, with most of the heads orienting nearly perpendicular to the CNT axis and extending into the outer aqueous phase. Surfactants in this configuration self-assemble into a cylinder-like micelle with the SWCNT located at the core. However, in the case of low surfactant packing density, there may be sufficient free space on the nanotube surface to accommodate the surfactant molecules in a wrapping configuration.18, 60 Consequently, surfactant molecules under these conditions tend to form less dense structures with no definite structured organization, even in small diameter nanotubes. This is the case observed here for both SWCNTs at 7 mM SDS. At higher SDS concentrations, our measurements support the MD predictions that a smaller nanotube diameter favors more ordered structures.18 We attribute this to both a reduction in available conical volume at the surface and reduced space on the tube sp2 carbon surface that limits the number of chain methylene groups directly interacting with the CNT surface. While we only compare two sizes in this work (due to the commercial availability of purified singlechirality semiconducting SWCNTs), our results are in sharp contrast with the previous SFG reports on spherical nanoparticles,27-28, 33 where increased chain order was observed for larger nanoparticles. These particles, however, had no competing interactions between the inorganic surface and the hydrocarbon groups in the alkyl chains. The combination of competitive vdW interactions within the surfactants themselves and with the carbon nanotube surface, the surfactant headgroup point away from the SWNT and the cylindrical geometries may provide significant differences that can explain the deviation from the previous SFG results on spherical nanoparticles. The ssp spectra of the SG65 films are ~1.2 times more intense than those for the SG76 films, this being consistent with either a more ordered geometry and/or a higher surface 15 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

adsorption density. In the following section, we make use of diffusion ordered spectroscopy (DOSY) NMR to probe the micellar structure of aqueous SDS and the perturbations to such micellar structures in the presence of semiconducting SWCNTs. This can potentially reveal further details regarding differences in surfactant packing densities between the two types of colloidal nanotubes.14, 61

DOSY NMR Measurements and Adsorption Density. Measurements of diffusivity may provide additional insights into the relation between surfactant adsorption and SWCNT diameter on the precursor solution. In order to determine the SDS diffusion coefficients, we used DOSY NMR spectroscopy. We measured the diffusivity values for 14mM SDS-only and SG65 and SG76 SWCNTs dispersions in 14mM SDS (see Figure S4). All measurements were done in triplicate. The obtained diffusion coefficients (D) of SDS-only, SDS-(7,6) SWCNT and SDS(6,5) SWCNT dispersions were (3.51±0.04)×10-10 m2/s, (3.68±0.06)×10-10 m2/s and (3.82 ±0.04)×10-10 m2/s, respectively at 298 K. Our experimental DOSY results are summarized in two important observations: i) The D value of 14 mM SDS is smaller compared to both the SDS–SWCNT assemblies. ii) The D value with SG76 dispersions is smaller than those with SG65. Zeigler and co-workers recently studied the SDS concentration dependence of SDS diffusivity by Pulsed-field gradient (PFG) NMR in the presence of CNTs.14 Within solution, SDS tends co-exist as free SDS monomers and micellar structure. The observed SDS diffusion coefficient, Dobs, is the weighted average of the free monomer, Df, and micelle-bound, Dm, values: '()*  + '  +, ', 2 where pf and pm are, respectively, the fraction of surfactant ions in monomeric form or aggregated into micelles, respectively. In this two-site model the surfactant must either be in monomeric form free solution or associated within a micelle, so that we can write pf + pm = 1. Since the diffusivity of SDS micelles, Dm = (6.1 ± 0.9) × 10−11 m2/s, and free monomers, Df = 16 ACS Paragon Plus Environment

Page 16 of 32

Page 17 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

(4.7 ± 0.08) × 10−10 m2/s, have been obtained from the literature,62-63 the fraction of free SDS can be calculated for the concentrations used here with Eq. 2 and the two-state fraction relationship. The fractions are found to vary depending on the bulk SDS concentration, with pf decreasing from 0.7 to 0.3 at concentrations 10 mM to 20 mM, respectively, well above the SDS CMC. Upon addition of SWCNTs to the SDS solution, the amount of SDS left in the solution for micelle formation decreases due to the SDS adsorption to the CNT surfaces. This results in an increase in the free SDS fraction in solution (pf) by a shift in the equilibrium discussed above and consequently in higher Dobs values (see Eq. 2) as shown in Table 1. With this model, the Dobs differences between SG76 and SG65 systems can provide a measurement of the number of strongly adsorbed surfactants on the SWCNT surfaces. The concentration of SDS adsorbed on 012 the SWCNT surfaces (./ ) can be estimated by:14 012 ./



2(345 ./

012 7  '()* '()* 6 012 8 3 '()*  ',

2(345 Where the total concentration of SDS ./  14 mM, the SDS diffusivity in the solution

7 012 without nanotubes at the same total SDS concentration is '()* , and '()* is the observed

diffusivity with the nanotubes in the dispersion. Therefore, the concentration of SDS adsorbed on 012 the SWCNT surface (./ ) was calculated to be 0.8 ± 0.4 mM for SG76 and 1.4 ± 0.4 mM for

SG65. The DOSY NMR results are summarized in Table 1.

Diffusivity (D) (m2/s)

(mM) Estimated :;=:>? ;