Effect of Organic Modification on Multiwalled Carbon Nanotube Dispersions in Highly Concentrated Emulsions.

Highly concentrated water-in-oil emulsions incorporating multiwalled carbon nanotubes (MWCNTs) are prepared. Homogeneous and selective dispersions of MWCNTs throughout the oil phase of the emulsions are investigated. The practical insolubility of carbon nanotubes (CNTs) in aqueous and organic media necessitates the disentanglement of CNT "agglomerates" through the utilization of functionalized CNTs. The design and synthesis of two tetra-alkylated pyrene derivatives, namely, 1,3,6,8-tetra(oct-1-yn-1-yl)pyrene (TOPy) and 1,3,6,8-tetra(dodec-1-yn-1-yl)pyrene (TDPy), for the noncovalent organic modification of MWCNTs are reported. The modifier molecules are designed in such a manner that they facilitate an improved dispersion of individualized MWCNTs in the continuous-oil phase of the highly concentrated emulsion (HCE). Transmission electron microscopic analyses suggest that the alkylated pyrene molecules are adsorbed on the MWCNT surface, and their adsorption eventually results in the debundling of MWCNT agglomerates. Fourier transform infrared, Raman, and fluorescence spectroscopic analyses confirm the π-π interaction between the alkylated pyrene molecules and MWCNTs. The noncovalent modification significantly improves the effective debundling and selective dispersion of MWCNTs in HCEs.

Introduction

Highly concentrated emulsions (HCEs) possess an internal (dispersed) phase volume fraction (ϕ) greater than 0.74, the maximum possible volume fraction of nondeformable monodisperse spheres in hexagonal close packing. In HCEs, ϕ can be as high as 96–99%, and the structure consists of a polydisperse–polyhedral droplet system separated by a thin film of the continuous phase.[1−3] HCEs retain their original structure even at extremely high values of ϕ by virtue of the polydispersity and the deformation of spherical droplets to polyhedral ones.[4,5] Apart from its extensive use in the food industry, cosmetics, paints, oil recovery, and commercial explosives, HCEs are also widely used as reaction media for chemical- and enzyme-catalyzed reactions[6−8] as well as templates for the preparation of low-density materials like poly-high internal phase emulsion (polyHIPE) foams.[9−11] Emulsion explosives that are widely used for commercial blasting in the mining industry are primarily highly concentrated water-in-oil (w/o) emulsions.

Carbon nanotubes (CNTs) are being widely used to improve the thermal, electrical, and mechanical properties of a variety of materials, including ceramics, metals, and polymers. Several studies were reported on HCEs incorporating nanoparticles, including multiwalled carbon nanotubes (MWCNTs), either as Pickering stabilizers or co-stabilizers.[12−16] CNTs with various surface functionalities were incorporated into HCEs to prepare conductive polyHIPE foams with a low electrical percolation threshold.[17−21] A significant level of understanding about the formation of electrically percolating MWCNT networks in HCEs has been gained from these studies on polyHIPE composite foams. Kim et al. have reported a percolation threshold of less than 1 wt % in polyHIPE foams.[18]

In our previous investigation, nanotube-incorporated HCEs were prepared, wherein MWCNTs were added into the continuous phase of w/o HCEs to achieve a MWCNT “network-like” structure within the emulsion.[16] The oil phase of less than 10 wt % of the total emulsion and hence, the MWCNTs network can be achieved with a reasonably low amount of MWCNTs, provided MWCNTs are effectively debundled and selectively dispersed in the oil phase of the emulsion. It is crucial that the MWCNTs should remain in the continuous phase of the emulsion and not diffuse to the oil–water interface, because this may adversely affect the emulsion stability. An efficient MWCNT network within the emulsion could impart additional characteristics or modify the thermal and electrical properties of the emulsion. It is believed that an improvement in the emulsion thermal conductivity could significantly enhance the performance of emulsion explosives.

However, developing a homogeneous dispersion involving “individualized” MWCNTs in the oil blend (which forms the continuous phase of the HCE) remains a difficult task. It was noticed earlier that a fraction of MWCNTs remained in the agglomerated-state in the emulsion matrix.[16] Moreover, the fraction of remaining MWCNT agglomerates increased with the increase in MWCNT concentration in the HCE. CNTs aggregate into bundles or ropes with strong van der Waals interactions during their synthesis, limiting their superior properties.[22] Many of the proposed applications of CNTs have been limited by their poor solubility in aqueous and organic solvents.[23] Hence, it is necessary to debundle the CNT agglomerates to prepare functional CNT-based composites or suspensions.

In view of this, the present investigation focuses on the effective debundling along with the selective dispersion of MWCNTs to localize them in the oil phase of the HCE. Several covalent and noncovalent functionalization strategies were reported to debundle and to disperse CNTs in various dispersion media. Although the covalent functionalization of CNT surfaces can significantly improve the dispersibility of CNTs in various solvents, it alters the sp2 carbon framework and adversely affects the intrinsic physical properties of CNTs.[23] In contrast, the noncovalent modification of CNTs preserves their intrinsic properties to a great extent, and it involves the adsorption of appropriate modifier molecules onto the CNT surfaces via hydrophobic, van der Waals, and/or electrostatic interactions.[24] Conventional surfactants are not recommended for dispersing MWCNTs in the oil phase of HCEs, as they could disturb the structure and the stability of the emulsion. During the preparation of conductive polyHIPE foams through HCEs, it was noticed that the utilization of water-soluble surfactants to disperse CNTs resulted in unstable HCEs.[18,25] This is because the surfactant destabilizes the film between the droplets, which is stabilized by the emulsifier of the HCE.

Pyrene-containing compounds have often been used for the noncovalent functionalization due to the very high affinity of pyrene toward the CNT surface.[26−37] Pyrene derivatives adsorb on the CNT surface via π–π interactions, preventing the formation of bundles and enabling its dispersion in solvents.[38,39] Apart from their use as noncovalent stabilizing units, pyrene derivatives were also used to anchor various systems to CNTs in diverse applications. Furthermore, the pyrene ring offers several possibilities of peripheral group modifications, which can affect the condensed-matter structure.[40] It has been reported that the unsubstituted pyrene forms monoclinic crystals, whereas some of the tetra-substituted pyrene derivatives exhibit stacking and form liquid crystalline columnar phases.[41,42]

Noncovalent functionalization of MWCNTs with pyrene derivatives of long hydrocarbon chains may be an efficient way to achieve a stable and selective dispersion of debundled MWCNTs in HCEs. The pyrene ring will ensure the adsorption of the modifier molecules on the surface of the MWCNT through π–π stacking via coupling of π–π interactions (between the sp2 carbons of the aromatic pyrenyl group and MWCNT), and the hydrocarbon chains would interact and allow the MWCNTs to be localized in the oil blend. Two alkylated pyrene derivatives have been chosen for the surface modification of MWCNTs, namely, 1,3,6,8-tetra(oct-1-yn-1-yl)pyrene (TOPy) and 1,3,6,8-tetra(dodec-1-yn-1-yl)pyrene (TDPy). The two derivatives differ in the side chains that are located on the 1,3,6,8-positions of the pyrene core. TOPy consists of an n-octyl chain; on the other hand, TDPy contains an n-dodecyl group. Two alkynes were chosen to assess the influence of the side chain length on the solubilizing power of the modifier molecule. Side chains could have a profound influence on the molecular arrangement as well as on the interaction with the surrounding environment.

These two pyrene derivatives have been synthesized, and their effectiveness as a noncovalent modifier has been investigated in detail. The effect of the noncovalent modification on MWCNT debundling and the interaction between the modifier and MWCNTs were assessed. Furthermore, the modified MWCNTs were incorporated into w/o HCEs, and the influence of noncovalent modification on the dispersion state of MWCNTs was investigated. A model w/o emulsion system with ϕ greater than 90 wt %, which has rheological characteristics identical to those of emulsion explosive precursors, was chosen for this study.

Results and discussion

Influence of the Noncovalent Modification on the Morphology and the “Agglomerate State” of MWCNTs

The nature of interactions between MWCNTs and the dispersing medium dictates the “dispersibility” of MWCNTs in a medium. In the case of the modified MWCNTs, the interaction between MWCNTs and the modifier molecules is also important. To establish an understanding of the dispersion state and the extent of the debundling of MWCNTs, the investigation of the dispersion of MWCNTs was carried out at the “microscale” and “nanoscale” levels. The average agglomerate size of MWCNTs was determined by microscale investigations, such as optical microscopy and sedimentation observations, whereas the debundling of MWCNT agglomerates was examined on the nanoscale with the help of scanning electron microscopy (SEM) and transmission electron microscopy (TEM) analyses.

Sedimentation behavior of nanotubes in the presence of the modifier molecule is an important parameter to estimate the dispersibility of MWCNTs. The unmodified and modified MWCNTs were dispersed in tetrahydrofuran (THF), and their sedimentation behavior was examined. The dispersion of the unmodified MWCNTs was highly unstable, and sedimentation was observed after a few minutes of ultrasonication. On the other hand, both 1:1 and 1:2 modified MWCNTs were stable for a few days in THF (Figure S1 in the Supporting Information).

Furthermore, dispersion studies of MWCNTs were carried out in the oil blend, where the oil blend composition was the same as that of the continuous phase of the emulsion (Figure S2 in the Supporting Information). This investigation could provide crucial insights into the influence of the noncovalent modification on the dispersion state of MWCNTs in the oil phase of the emulsion. The size of the MWCNT agglomerates that remained in the oil blend–MWCNT dispersion was determined using optical microscopic analysis for the unmodified and modified MWCNTs (Figure S2 in the Supporting Information). The average agglomerate size for the unmodified MWCNTs was observed to be ∼218.2 μm2, whereas for 1:1 (w/w) TOPy- and TDPy-modified MWCNTs, it was ∼109.9 and ∼93.4 μm2, respectively (Figure S3 in the Supporting Information). With the increase in the modifier weight ratio, the agglomerate size was further reduced to ∼75.6 and ∼73.2 μm2, respectively, for the 1:2 (w/w) TOPy- and TDPy-modified MWCNTs. Thus, the dispersion with unmodified MWCNTs exhibited a higher average agglomerate size as compared to all corresponding modified MWCNTs. However, there was no significant variation in the average agglomerate sizes of modified MWCNTs, between these two modifiers at a specific modifier concentration. For the modified MWCNTs, the average size of MWCNT agglomerates reduced with increasing weight ratio of TOPy and TDPy. Similar observations were reported for MWCNTs that were surface-modified using 1-pyrenecarboxaldehyde and sodium/lithium salt of 6-aminohexanoic acid.[43,44]

Pyrene derivatives have been reported to interact with the nanotube surface with a high surface coverage because of the very high affinity of pyrene toward the CNT surface.[26,27,30−34] Pyrene derivatives adsorb onto the surface of CNTs through π–π interactions, preventing the bundle formation and enabling its dispersion in solvents.[38,39] It is also likely that the TOPy and TDPy molecules may form a “layer-like structure”, while covering the surface of MWCNTs, owing to better packing offered by the matching aromatic structure. Lerner et al. demonstrated the stacking of pyrene molecules on a nanotube surface through an increased thickness of the nanotube, as observed in AFM studies.[38] Furthermore, it was demonstrated through molecular simulation that the stability of the complex of pyrene with the nanotube is greater than that between two pyrene molecules.[27] Thus, the pyrene molecule and its derivatives have the tendency to interact with the nanotube surface and form a complex rather than assembling themselves.

The morphology of the unmodified and modified MWCNTs, as observed in TEM and SEM analyses, is shown in Figure . The unmodified MWCNTs exhibited a highly agglomerated state, whereas the TOPy- and TDPy-modified MWCNTs showed a less agglomerated state. TEM micrographs (Figure a–e) illustrate the debundling of MWCNTs upon modification. As observed from the high-resolution TEM (HR-TEM) micrographs, which were imaged at 300k magnification, the unmodified MWCNTs show a relatively smooth surface (Figure f), whereas the TEM micrographs of the modified MWCNTs exhibit roughness on the outer wall, which is expected to have originated from the modifier molecules that adhered to the surface of the MWCNTs (Figure g–j). The SEM micrographs of unmodified and modified MWCNTs depict an “entangled” network of MWCNTs. However, it is expected that the “agglomerate-strength” of the MWCNTs may decrease with the modification due to the effective debundling.

Figure 1

TEM and SEM micrographs of the unmodified and modified MWCNTs: (a)–(e) at 50k magnification; (f–j) high-resolution TEM images of unmodified and modified MWCNTs, at 300k magnification; (k–o) SEM images at 100k magnification.

Furthermore, the average diameters of the unmodified and modified MWCNTs were estimated from the HR-TEM images. The increase in the average diameter of MWCNTs suggests the adsorption of the modifier molecules on the MWCNT surface. Average diameters of the unmodified and modified MWCNTs that were estimated from the TEM micrographs are shown in Table . SEM and TEM observations suggest that the TOPy and TDPy molecules are adsorbed on the MWCNT surface to a thickness of ca. 1.5–2 nm, and this subsequently results in debundling of the MWCNT agglomerates. The TOPy and TDPy molecules interact with the MWCNT surface through π–π interactions and these adsorbed molecules weaken the intertube attractive forces resulting in the debundling of MWCNTs.

Table 1

Average Diameter (Determined from HR-TEM Images) of the Unmodified and Modified MWCNTs

sample		Davg (nm)
unmodified MWCNTs		10.2 ± 2
TOPy-modified MWCNTs	1:1 (w/w)	12.4 ± 1
	1:2 (w/w)	12.3 ± 1.7
TDPy-modified MWCNTs	1:1 (w/w)	13.2 ± 2.5
	1:2 (w/w)	14.3 ± 1.4

Interaction between Alkylated Pyrene Modifier Molecules and MWCNTs

The performance of CNT-based nanocomposites and suspensions relies strongly on the interaction of CNTs with their environment, which is in turn influenced by the chemical structure and nature of the dispersion medium and the presence of other modifier molecules. There can be various noncovalent interactions that act between CNTs and molecular, macromolecular, or ionic species that constitute the dispersion medium.[33] The highly aromatic nature of the pyrenyl group leads to a strong π–π interaction with the sidewalls of CNTs.[45] Various techniques, such as fluorescence spectroscopy, Fourier transform infrared (FTIR), Raman spectroscopy, and NMR can be used to probe the π–π interaction between the pyrene derivative molecules and CNTs.[27] A detailed assessment of interactions between the modifier molecules and MWCNTs would bring interesting insights into the understanding of MWCNT dispersions in the oil phase of emulsions that have been investigated here. FTIR spectra, Raman spectra, and fluorescence emission spectra of unmodified and modified MWCNTs were analyzed to assess the interaction between the modifier molecules and MWCNTs.

FTIR Spectroscopic Measurements

FTIR spectra of unmodified, 1:1 TOPy-, and 1:1 TDPy-modified MWCNTs in the full wavelength range is presented in Figure S4 in the Supporting Information. Figure a exhibits the FTIR spectra of TOPy and 1:1 TOPy-modified MWCNTs. As observed from the FTIR spectrum of TOPy, the peaks at 721.6 and 830.8 cm–1 are associated with the C–H wagging vibrations of the pyrenyl group. These two wagging vibration bands appeared in the TOPy-modified MWCNTs as well. However, they showed a downshift to 717.8 and 828.3 cm–1. These downshifts of the wagging vibrational peaks indicate the strong π–π stacking interaction and the possible hydrophobic interactions between the aromatic rings of TOPy and the MWCNT surface.[27,43]Figure b shows the FTIR spectra of the unmodified and 1:1 TOPy-modified MWCNTs. Figure c illustrates the downshift of the wagging vibrational bands in TDPy-modified MWCNTs. The vibrational bands have shown a downshift to 716.3 and 828.5 cm–1, respectively, from 720.6 and 832.5 cm–1 corresponding to TDPy. Figure d shows the FTIR spectra of the unmodified and 1:1 TDPy-modified MWCNTs. Thus, a shift of the wagging vibrational bands to lower wavenumbers suggests the strong π–π interaction between the aromatic rings of the modifier molecules and the MWCNT surface.

Figure 2

FTIR spectra of (a) TOPy and 1:1 (w/w) TOPy-modified MWCNTs; (b) unmodified and 1:1 (w/w) TOPy-modified MWCNTs; (c) TDPy and 1:1 (w/w) TDPy-modified MWCNTs; (d) unmodified and 1:1 (w/w) TDPy-modified MWCNTs.

Raman Spectroscopic Measurements

Figure a shows the Raman spectra of unmodified, TOPy-, and TDPy-modified MWCNTs. The Raman spectrum of the unmodified MWCNTs exhibits two characteristic peaks: a D-band at ∼1344.2 cm–1 and a G-band centered at ∼1581.6 cm–1. Raman scattering involves inelastic scattering of light, and the Raman spectrum depicts the scattered light intensity with respect to the excitation wavelength. The shape and intensity of the D-band correspond to the sp3 hybridized carbon atoms,[46] which is correlated with the defect concentration or the measure of disorders in the C–C bonds and the chemical sidewall functionalization.[47] The G-band is attributed to the tangential C=C bond stretching in an ordered graphitic structure.[48]

Figure 3

Raman spectroscopic investigation of the unmodified and modified MWCNTs: (a) Raman spectra in the full scan range for unmodified, TOPy-, and TDPy-modified MWCNTs; (b) Raman spectra in the 1000–2000 cm–1 range, illustrating the shifts in the D-band and G-band of the modified MWCNTs.

Raman spectra of unmodified and modified MWCNTs in the range of 1000–2000 cm–1 are shown in Figure b to clearly illustrate the shifts in the characteristic peaks of the modified MWCNTs. The Raman spectra of the modified MWCNTs exhibit an upward shift in the G-band: ∼1589.5 and ∼1591.4 cm–1 for 1:1 and 1:2 (w/w) TOPy-modified MWCNTs, respectively; ∼1590.5 and ∼1591.7 cm–1 for 1:1 and 1:2 (w/w) TDPy-modified MWCNTs, respectively. Furthermore, the D-band of the modified MWCNTs also exhibited an upward shift, but relatively lower when compared to that of the G-band: ∼1346.5 and ∼1348.5 for 1:2 (w/w) TOPy- and TDPy-modified MWCNTs, respectively.

The D-band and G-band peaks originate due to the “resonance phenomenon”, and the peak position could be associated with the electronic and vibrational densities of state.[43] The adsorption of the alkylated pyrene molecules on MWCNTs may alter the vibrational densities of state and could eventually affect the vibrational modes of MWCNTs. Furthermore, the tangential displacement G-mode is also sensitive to the charge that is exchanged between the CNTs and the adsorbed molecule.[49] Hence, the shift in the G-band to a higher wavenumber for the modified MWCNTs could be due to the π–π interaction between the adsorbed modifier molecules and MWCNTs. A similar shift to higher wavelengths in the Raman spectroscopic peaks of pyrene-modified MWCNTs has been reported earlier.[27] Furthermore, a shift to a higher wavenumber can also arise due to weaker intertube interactions, suggesting the “de-agglomeration” of MWCNTs in the presence of the modifier molecules.

The unmodified MWCNTs exhibit a ID/IG value of ∼1.22, whereas the 1:1 and 1:2 (w/w) TOPy-modified MWCNTs show ∼0.82 and ∼0.98, respectively. For the TDPy-modified MWCNTs, the ID/IG values are 0.95 and 0.99, respectively, for 1:1 and 1:2 (w/w) modifications. An increase in the ID/IG ratio indicates a higher defect fraction in the ordered graphitic structure.[43,50] Furthermore, a reduced ID/IG ratio suggests an improved debundling of MWCNT agglomerates and an enhanced resonance process after the exfoliation of MWCNTs.[51,52] The persistent decrease in the ID/IG ratio of modified MWCNTs is due to the enhanced debundling of MWCNTs in the presence of alkylated pyrene molecules.

UV–Visible and Fluorescence Spectroscopy Measurements

The UV–visible absorption spectra of unmodified, 1:2 (w/w) TOPy-, and TDPy-modified MWCNT dispersions in THF along with the absorption spectra of TOPy and TDPy free molecules in THF are shown in Figure . The unmodified MWCNTs in THF exhibited an absorption maximum at ∼253 nm. The absorption spectra of both modified MWCNTs were almost identical to those of the free modifier molecules in THF. This observation implies that the absorption spectrum is primarily governed by the respective modifier molecules in the corresponding bulk solution.

Figure 4

UV–visible absorption spectra of dilute solutions of TOPy and TDPy, and dispersions of unmodified, TOPy-, and TDPy-modified MWCNTs in THF.

Fluorescence spectra of the pyrene monomers in the solution phase show significant vibronic bands, and the pyrene derivatives are often used as fluorescence probes due to their high fluorescence quantum yield.[27,53] The strong fluorescence from the TOPy and TDPy molecules is a marker to demonstrate the interaction of these molecules with MWCNTs. The fluorescence spectra of TOPy- and TDPy-modified MWCNT dispersions in THF were recorded using solutions that possessed identical absorbance at the Soret band maximum (426 nm). Figure exhibits the fluorescence emission spectra of all materials in THF. Upon excitation, the fluorescence spectra of the modified MWCNTs show superimposable profiles with respect to the free TOPy and TDPy molecules in THF. Although there is no change in the spectroscopic feature for both modified MWCNTs, a significant quenching in the fluorescence emission intensity of the modified MWCNTs is observed when compared to the emission intensity of the modifier molecules alone in THF. However, there is no significant spectral shifting in the presence of MWCNTs. Fluorescence quenching by CNTs through energy transfer has been reported in several publications.[27,32,54−59] The fluorescence quenching is attributed to the energy transfer from the high-energy pyrene singlet which is at the excited state to the low-energy manifold of CNTs.[60] The residual fluorescence might be derived from the free, nonimmobilized modifier molecules in the solution.

Figure 5

Fluorescence emission spectra of dilute solutions of TOPy and TDPy, and dispersions of unmodified, TOPy-, and TDPy-modified MWCNTs in THF. All samples, except for the dispersions of unmodified MWCNTs, possessed a matching absorption at 426 nm. This is a qualitative trend of the fluorescence emission intensities and cannot be assumed to be quantitative for comparisons (λex = 426 nm, excitation and emission slit width is 2.5 nm).

To further investigate the interaction between the modifier molecules and MWCNTs at the molecular level, a set of fluorescence spectroscopy experiments were run at varied concentrations of MWCNTs in the modifier solution in THF. Upon excitation, the fluorescence spectra of MWCNTs again showed superimposable profiles with respect to the free TOPy and TDPy molecules in THF. The addition of MWCNTs did not give rise to a new spectroscopic feature; however, there was substantial quenching in the fluorescence emission intensity in the presence of MWCNTs when compared to the emission intensity of the modifier molecules alone in THF (Figure S5 in the Supporting Information).

The fluorescence measurements strongly suggest a mapping of the modifier molecules onto the surface of the MWCNTs through π–π stacking. Thus, the residual fluorescence quenching may presumably be due to π–π interactions. Since there is no indication of charge transfer between the modifier molecule and MWCNTs, the fluorescence quenching could have resulted from an efficient energy transfer from the pyrene moiety in TOPy and TDPy to the MWCNTs through vibrational coupling. However, it is also reported that the observed reduction in the fluorescence intensity cannot be assumed to be quantitative. Fluorescence quenching could also result from other factors such as light partitioning between the modifier molecule and the MWCNTs (hence, a fraction of the photons excites the CNT scaffold) and scattering of the emitted light due to the presence of MWCNTs in the dispersion.[58,61]

Figure illustrates the schematic of the noncovalent modification of MWCNTs using the alkylated pyrene derivatives, which leads to the debundling of MWCNT agglomerates. Details of the interaction between the alkylated pyrene modifier molecules and MWCNTs at the molecular level are indefinite at the current stage. However, the spectroscopic results suggest that the physical adsorption of the modifier molecules on the sidewall of the MWCNTs can be attributed to π–π interactions and van der Waals interactions.

Figure 6

Schematic representation of physical adsorption of alkylated pyrene derivatives on the surface of MWCNTs leading to the debundling of MWCNT agglomerates (the dimensions of MWCNTs and modifier molecules are not in scale).

Modified MWCNT-Incorporated HCEs

To facilitate the selective dispersion and localization, MWCNTs were originally dispersed in the oil blend through ultrasonication. The oil blend comprises a mixture of hydrocarbon oils and an emulsifier. The oil blend–MWCNT dispersion was used as the oil phase for the preparation of the nanotube-incorporated HCE and gradually incorporated into the salt solution to prepare the emulsion. First, a stable and homogeneous dispersion of the individualized MWCNTs in the oil blend is necessary to effectively incorporate MWCNTs into the emulsion matrix. Even though the modified MWCNTs could provide a much improved dispersion of debundled MWCNTs in the oil blend, some amount of MWCNTs still remained in the agglomerated state. This could be due to the very high concentration of MWCNTs in the oil blend. The amount of modified MWCNTs in the dispersion ranged between 2.24 and 17.9 mg/mL and was considerably higher when compared to the other high-concentration CNT dispersions reported in any organic solvent (∼2 mg/mL).[62] Even an extended ultrasonication also failed to cause any considerable improvement in the dispersion state of MWCNTs. The remaining MWCNT agglomerates were more noticeable in dispersions that had a higher loading of MWCNTs. The MWCNT–oil blend dispersions of unmodified and modified MWCNTs with varying concentrations of MWCNTs were used for preparing the nanotube-incorporated HCEs.

Emulsion Microstructure

Laser scanning confocal microscopy was used to obtain differential interference contrast (DIC) images and fluorescence images of emulsions. Figure illustrates the fluorescence images and overlapped DIC-fluorescence images for 0.5NT and 1NT emulsions incorporated with modified MWCNTs when excited with a laser at 488 nm. Emulsions incorporated with modified MWCNTs exhibit fluorescence, which arises from the adsorbed alkylated pyrene molecules excited by the laser (Figure a–h). As observed from micrographs, the fluorescence phenomenon is from the thin layer around the dispersed droplets, which suggests that a major fraction of the MWCNTs is localized and confined to the continuous phase of HCEs (Figure a–h). It should also be noted that there is a possibility that some fraction of the adsorbed modifier molecules from the modified MWCNTs gets dissolved in the oil blend. This might also contribute to the total fluorescence from the continuous phase.

Figure 7

Laser scanning confocal micrographs of emulsions incorporated with modified MWCNTs, which show fluorescence images (right) and overlapped DIC-fluorescence images (left) for 0.5NT and 1NT emulsions: (a–d) emulsions incorporated with 1:1 and 1:2 (w/w) TOPy-modified MWCNTs; (e–h) emulsions incorporated with 1:1 and 1:2 (w/w) TDPy-modified MWCNTs.

The modified MWCNTs allow better dispersion of the individualized MWCNTs, as well as allow them to be localized in the continuous phase of the emulsion so that the MWCNT network can be achieved within the HCE. Laser scanning confocal microscopic analyses qualitatively suggest that the majority of MWCNTs remain selectively dispersed in the continuous phase of the emulsion; however, some fraction of MWCNTs remained in the agglomerated state in the emulsion matrix. Furthermore, it needs to be investigated whether the amount of MWCNTs dispersed in the HCE is sufficient to alter any emulsion characteristics by forming an efficient three-dimensional network within the HCE. On the other hand, it is not possible to incorporate a very high amount of MWCNTs into the oil phase as any concentration beyond ∼5 wt % in the oil phase hardly produces stable emulsions or cannot accommodate the same, very high, aqueous phase volume fraction, even when some emulsions are formed.

Dispersion State of the MWCNTs in the Emulsion: Microscale and Nanoscale Analyses

Average Agglomerate Size of the MWCNTs in the Emulsion: Influence of Noncovalent Modification

Microscopic analyses were carried out to assess the role of the noncovalent modifiers in effectively debundling the agglomerates of MWCNTs to achieve an improved dispersion in the emulsion. The size of the MWCNT agglomerates that remained in the emulsion matrix was determined using confocal microscopic analysis (Figure a–l). It can be seen that the remaining MWCNT agglomerates are more prominent in emulsions with unmodified MWCNTs (Figure a–d), whereas a substantial improvement is observed in the case of emulsions with modified MWCNTs (Figure e–l).

Figure 8

Laser scanning confocal micrographs that show the remaining MWCNT agglomerates in the 0.25NT, 0.5NT, 1NT, and 2NT emulsions incorporated with unmodified and modified MWCNTs: (a–d) emulsions incorporated with unmodified MWCNTs; (e–h) emulsions incorporated with 1:2 (w/w) TOPy-modified MWCNTs; and (i–l) emulsions incorporated with 1:2 (w/w) TDPy-modified MWCNTs.

Around 25–30 micrographs were analyzed for each emulsion system, and the data numerically processed to assess the enhancement of MWCNT dispersion in the corresponding emulsion. Figure demonstrates the variation in average agglomerate size (Davg) and the percentage area of agglomerates in the unmodified and 1:2 (w/w) modified nanotube-incorporated emulsions with different concentrations of MWCNTs. For the emulsion having 0.25 wt % MWCNTs, Davg is 9.8 μm for the unmodified MWCNTs, whereas for the TOPy- and TDPy-modified MWCNTs, it is 6.8 and 6.5 μm, respectively. For the 2 wt % nanotube-incorporated emulsions, Davg is 12.4 μm for the unmodified MWCNTs and 10.6 and 10.9 μm, respectively, for the TOPy- and TDPy-modified MWCNTs. Furthermore, the trend in the calculated area ratios of the agglomerates is consistent with that of the agglomerate size. For 0.25NT and 2NT emulsions incorporated with unmodified MWCNTs, the area ratio is 0.55 and 8.8%, respectively. In contrast, the corresponding area ratios are 0.18 and 5.5% for the emulsions incorporated with TOPy-modified MWCNTs and 0.21 and 5.1% for the emulsions incorporated with TDPy-modified MWCNTs.

Figure 9

Variation in the (a) average agglomerate size and (b) area ratio of the MWCNT agglomerates that remained in the emulsion matrix for the emulsions incorporated with unmodified and modified MWCNTs.

The microscale analysis suggests that the noncovalent surface modification has effectively improved the dispersion state of the MWCNTs in the emulsion. This can be envisaged as a result of the effective debundling of the MWCNTs in the presence of the noncovalent modifier molecules. Moreover, the rapid re-aggregation of the dispersed MWCNTs might have been hindered by the adsorbed modifier molecules on the MWCNT surface. It is also expected that the noncovalent modification facilitates the localization of the individualized MWCNTs predominantly in the oil phase mediated via the specific interaction between the alkyl chain of the organic modifier and the oil phase. At room temperature, the dispersed aqueous droplets are made of a supersaturated and supercooled salt solution. Hence, it is crucial that the dispersed MWCNTs stay in the continuous-oil phase of the emulsion and not migrate to the oil–water interface, as this may adversely affect the stability of the emulsion.

Dispersion State of MWCNTs in HCEs: Nanoscale Microstructural Analysis Using Cryo-FEG-SEM

Cryo-FEG-SEM analysis was performed to assess the dispersion state of MWCNTs at the nanoscale level. The dispersion state of various MWCNTs has been analyzed by considering the 1NT emulsion samples as the representative emulsions. The emulsion samples were frozen at −190 °C, and the corresponding SEM images were taken from the fractured surfaces. The morphologies of the 1NT emulsion incorporated with unmodified MWCNTs (Figure a,b), the corresponding emulsion incorporated with 1:2 (w/w) TOPy-modified MWCNTs (Figure c,d), and the corresponding emulsion incorporated with 1:2 (w/w) TDPy-modified MWCNTs (Figure e,f) are shown in the SEM micrographs (Figure a–f). The morphology, as observed from the Cryo-FEG-SEM micrographs, suggests that it is quite difficult to differentiate between the dispersion states of different MWCNTs in the corresponding emulsions. The dispersed MWCNTs in the emulsion matrix can be clearly observed in the micrographs of fractured surfaces. However, the selective dispersion and localization of MWCNTs in the oil phase of HCEs cannot be explicitly established from these micrographs. Hence, the Cryo-FEG-SEM observation of the fractured surface morphology alone is inadequate to elucidate the dispersion states of the unmodified and modified MWCNTs in the respective emulsions.

Figure 10

Cryo-FEG-SEM images of 1NT emulsion incorporated with unmodified and modified MWCNTs at high and low magnifications: (a, b) 1NT emulsion incorporated with unmodified MWCNTs; (c, d) 1NT emulsion incorporated with 1:2 (w/w) TOPy-modified MWCNTs; (e, f) 1NT emulsion incorporated with 1:2 (w/w) TDPy-modified MWCNTs.

Conclusions

Two tetra-alkylated pyrene derivatives that can act as effective noncovalent surface modifiers for MWCNTs were proposed and synthesized. The tetra-alkylated pyrene molecules were intended to produce a better dispersion of the individualized MWCNTs, and allow them to be localized in the continuous-oil phase of the highly concentrated w/o emulsions. These modifier molecules interact efficiently with the MWCNT surface through π–π interactions. The adsorption of the modifier molecules onto the surface of the MWCNTs resulted in the effective debundling of MWCNT agglomerates due to the weakened force of attraction between the nanotubes. Both the G-band shift in Raman spectroscopic analysis and the downshift of wagging vibrational bands in FTIR analysis, which were exhibited by the modified MWCNTs, along with the fluorescence quenching of the alkylated pyrene molecules in the presence of the MWCNTs all confirm the π–π interaction between the tetra-alkylated pyrene molecules and MWCNTs.

The modified MWCNTs were then incorporated into HCEs, and the influence of the noncovalent surface modification on the MWCNT dispersion in emulsions was investigated. The fluorescence localization from the thin layer around the dispersed droplets, in emulsions incorporated with modified MWCNTs, suggested that the major fraction of the incorporated MWCNTs are localized and confined to the continuous-oil phase of HCEs. However, the selective dispersion and localization of MWCNTs in the oil phase of HCEs was not unambiguously concluded from the Cryo-FEG-SEM micrographs. Developing an efficient protocol to investigate the state of dispersion of MWCNTs in these emulsions is important and extremely challenging. Moreover, a comparatively minor fraction of MWCNTs remained in the agglomerated state in the emulsion, even after the noncovalent modification. However, there was a substantial reduction in the average size and the area ratio of the MWCNT agglomerates that remained in the emulsion matrix in these emulsions when compared to the corresponding emulsions composed of unmodified MWCNTs. However, there was no significant difference in the average agglomerate sizes of modified MWCNTs between the two modifiers. The improved dispersion state of MWCNTs in the emulsion can be envisaged as a result of the effective debundling in the presence of the noncovalent modifier molecules and the impeded rapid re-aggregation of the dispersed MWCNTs by the adsorbed modifier molecules.

Experimental Section

Materials

Synthesis of Alkylated Pyrenes

1,3,6,8-Tetrabromopyrene, 1-octyne, 1-dodecyne, bis(triphenylphosphine)palladium(II) dichloride (Pd(PPh3)2Cl2) and copper(I) iodide (CuI) were procured from Sigma-Aldrich. The solvents, dichloromethane (DCM) and petroleum benzine, were procured from Merck Ltd., and anhydrous tetrahydrofuran (THF) and diisopropylamine were obtained from Sigma-Aldrich.

Emulsion Preparation

All of the constituents of the oil phase of the emulsion, such as the methylated canola oil, Exxol D130 (paraffinic oil), and emulsifier (polyisobutylene succinic anhydride (PIBSA)–diethanolamine derivative surfactant) were supplied by Orica Australia Pty. Ltd. The emulsifier is based on alkanolamine derivatives of PIBSA, which is prepared by reacting PIBSA and diethanolamine in 1:1 M ratio. Further details of the PIBSA-derivative emulsifier can be found elsewhere.[16] The constituents of the aqueous phase, such as ammonium sulfate, ammonium chloride, and other trace additives, were procured from Sigma-Aldrich. Deionized water was used to prepare the aqueous phase, and the trace additives were added to adjust the pH of the salt solution. MWCNTs were procured from Nanocyl S.A., Belgium (grade: NC 3100; Davg = 9.5 nm, L = 1.5 μm and purity greater than 95% as per manufacturer).

Synthesis

General Methods

Thin-layer chromatography was performed using silica gel plates (Merck) that were precoated with a fluorescent indicator with visualization by ultraviolet light. Column chromatography was performed on silica gel (60–120 mesh). 1H and 13C NMR spectra were recorded using a Bruker Avance-400 spectrometer (1H: 400 MHz and 13C: 100 MHz). The chemical shifts are recorded in ppm using the solvent as an internal standard: chloroform-d at 7.24 and 77.23 ppm for 1H and 13C NMR, respectively. High-resolution mass spectra (HRMS) were recorded using a Q-TOF analyzer in the positive ion electrospray ionization (ESI) mode.

TOPy has previously been synthesized.[40,63] The synthesis protocol of TDPy was adapted from that of TOPy. Tetrabromopyrene was functionalized with 1-octyne and 1-dodecyne under Sonogashira conditions to obtain TOPy and TDPy, respectively. Synthesis protocols for TOPy and TDPy are detailed below:

1,3,6,8-Tetrabromopyrene (2 g, 3.86 mmol), CuI (42 mg, 0.22 mmol), and (Pd(PPh3)2Cl2) (142 mg, 0.202 mmol) were added to a degassed solution of anhydrous THF (40 mL) and diisopropylamine (40 mL) in a round-bottomed flask under N2. The mixture was again degassed and an excess of 1-octyne (4 mL, 27.11 mmol, 7 equiv) was added. The mixture was degassed once again and stirred at 80 °C under N2 for 24 h. The reaction mixture was then allowed to cool to room temperature and, subsequently, filtered over celite and washed with 250 mL of DCM. The solvent was then removed, and the residue was dried under vacuum. The crude product was purified by column chromatography with a solvent that comprised 90% of petroleum benzine and 10% of DCM. Furthermore, precipitation from DCM–diethyl ether yielded a bright yellow powder of TOPy (1.36 g, 55.6% yield) (Scheme ).

Scheme 1

1,3,6,8-Tetra(oct-1-yn-1-yl)pyrene (TOPy)

Melting point: 95.6 °C. 1H NMR (400 MHz, CDCl3): δ 8.57 (s, 4H); 8.17 (s, 2H); 2.66 (t, 8H); 1.79 (m, 8H); 1.61 (m, 8H); 1.43 (m, 16H); 0.97 (m, 12H) ppm (Figure S6 in the Supporting Information). 13C NMR (100 MHz, CDCl3): δ 133.6, 131.4, 126.3, 124.1, 119.3, 97.0, 79.0, 31.4, 28.9, 28.8, 22.7, 19.9, 14.1 ppm (Figure S7 in the Supporting Information). HRMS (ESI): calcd for C48H58Na [(M + Na)]+ 657.4431; found, 657.4435 (Δm +0004 and error +0.6 ppm).

1,3,6,8-Tetrabromopyrene (2 g, 3.86 mmol), CuI (58 mg, 0.30 mmol), and Pd(PPh3)2Cl2 (164 mg, 0.23 mmol) were added to a degassed solution of anhydrous THF (40 mL) and diisopropylamine (40 mL) in a round-bottomed flask under N2. The mixture was again degassed and an excess of 1-dodecyne (3 mL, 14.03 mmol, 7 equiv) was added. The mixture was degassed once again and stirred at 85 °C under N2 for 24 h. The resulting mixture was filtered over celite and washed with 250 mL of DCM. The solvent was then removed, and the residue was dried under vacuum. The crude product was purified by column chromatography with a solvent that comprised 90% of petroleum benzine and 10% of DCM. Precipitation from DCM–diethyl ether yielded a bright yellow powder of TDPy (684 mg, 40% yield) (Scheme ).

Scheme 2

1,3,6,8-Tetra(dodec-1-yn-1-yl)pyrene (TDPy)

Melting point: 81.8 °C. 1H NMR (400 MHz, CDCl3): δ 8.57 (s, 4H); 8.17 (s, 2H); 2.65 (t, 8H); 1.78 (m, 8H); 1.6 (m, 8H); 1.3 (m, 48H); 0.90 (m, 12H) ppm (Figure S8 in the Supporting Information). 13C NMR (100 MHz, CDCl3): δ 133.62, 131.38, 126.35, 124.08, 119.28, 96.96, 78.99, 31.92, 29.62, 29.58, 29.35, 29.2, 29.08, 28.94, 22.69, 19.92, 14.1 ppm (Figure S9 in the Supporting Information). HRMS (ESI): calcd for C64H90Na [(M + Na)]+ 881.6935; found, 881.6931 (Δm +0004 and error +0.4 ppm). FTIR, UV–visible, and fluorescence spectra of TOPy and TDPy are provided in Figures S10 and S11 in the Supporting Information.

Surface Modification of MWCNTs Using Alkylated Pyrene Molecules

Both TOPy and TDPy demonstrated very high solubility in common organic solvents, such as hexane, toluene, chloroform, DCM, and THF. MWCNTs were solubilized in THF, and subsequently, the TOPy- and TDPy-functionalized MWCNTs were prepared. MWCNT/modifier ratios of 1:1 and 1:2 (w/w) were employed for surface modification of MWCNTs. MWCNTs were initially dispersed in THF by ultrasonication (Ultrasonic Processor 750W, Sonics & Materials Inc.) for 10 min. The modifier solution in THF was then slowly added to the dispersion of MWCNTs followed by ultrasonication for another 30 min. Subsequently, the resultant dispersion of MWCNTs–modifier molecules in THF was continuously stirred at room temperature to evaporate the solvent and obtain the dry powder of modified MWCNTs. The resultant mixture of MWCNTs–modifier molecules was then vacuum-dried in an oven at 40 °C for 24 h to remove the traces of THF.

Preparation of MWCNT-Incorporated HCEs

A highly concentrated w/o emulsion system with ϕ = 93.5 wt % was chosen for the current study. The dispersed phase of the emulsion comprises a supersaturated inorganic salt solution, in which water constitutes 55 wt % and the remaining materials are ammonium sulfate and ammonium chloride salts along with some trace additives. The mixture of a PIBSA-derivative emulsifier in the blend of canola oil and Exxol D130 oil constitutes the continuous phase. A detailed description of the emulsion composition can be found elsewhere.[16]

A high-torque Caframo BDC 1850 mixer along with a high-shear Jiffy impeller (Jiffy Mixer LM, 32 mm diameter) was used to prepare the HCE samples. MWCNTs were initially dispersed in the oil blend, and the resulting oil blend–MWCNT dispersion was gradually incorporated into the aqueous salt solution to prepare the nanotube-incorporated HCEs. The concentration of MWCNTs in the oil blend was varied from 0.25 to 2 wt %, which corresponds to 0.016–0.13 wt % of the total emulsion. The aqueous phase composition and the aqueous-to-oil phase ratio were kept unaltered for all HCE samples. The detailed procedure for the preparation of the nanotube-incorporated HCEs could be found elsewhere.[16,64] It is to be noted that HCEs under study are stabilized by a relatively large amount of emulsifier, and the MWCNTs dispersed in the oil blend do not show any surfactant-like behavior. To obtain the emulsions with identical droplet size and polydispersity, emulsions with varying MWCNT concentrations were subjected to different mixing times. The targeted average droplet size of 5 μm was reached with 6–15 min of refining, depending on the level of MWCNT loading. The compositions of different emulsion samples along with their sample codes are presented in Table .

Table 2

Sample Codes and Their Composition for the MWCNT-Incorporated Emulsions

	concentration of MWCNTs (wt %)
sample code	in oil phase	in total emulsion	types of MWCNTs used
neat	0	0	unmodified, 1:1 TOPy-modified, 1:2 TOPy-modified, 1:1 TDPy-modified, 1:2 TDPy-modified
0.25NT	0.25	0.0163
0.5NT	0.5	0.0325
1NT	1	0.065
2NT	2	0.13

Characterization Techniques

UV–Visible Spectroscopic Analysis

UV–visible spectroscopic measurements were performed on dispersions of pristine and modified MWCNTs in THF using a Shimadzu Varian Cary 100 Bio UV–vis spectrometer. Around 1 mg of unmodified/modified MWCNTs was ultrasonicated in 20 mL of THF with a bath-type sonicator (PCi Analytics, India, frequency 20 kHz) for 20 min.

Fluorescence Spectroscopy

Fluorescent spectra were obtained using a Varian Cary Eclipse fluorescence spectrometer. Measurements were carried out on the solutions of noncovalent modifier molecules and dispersions of unmodified and modified MWCNTs in THF. The excitation wavelength was set at 426 nm.

Transmission Electron Microscopic (TEM) Analysis

TEM micrographs of the unmodified and modified MWCNTs were captured using a JEOL JEM-2100 F (Japan) field-emission electron microscope. To prepare samples for TEM investigation, around 1 mg of unmodified/modified MWCNTs was ultrasonicated in 20 mL of THF with a bath-type sonicator (PCi Analytics, 20 Hz) for 20 min. After ultrasonication, the supernatant was carefully pipetted, and a drop of the dispersion was placed on the TEM grid and it was dried under vacuum.

Scanning Electron Microscopic (SEM) Analysis

SEM analyses on unmodified and modified MWCNTs were performed using FEG-SEM (JSM-7600F, Japan) at an accelerating voltage of 10 kV. A MWCNT dispersion in THF was drop-cast on a silicon wafer, after which the wafer was vacuum-dried. The wafer surface was gold-sputtered prior to imaging.

Raman Spectroscopic Analysis

Raman spectroscopic investigations of unmodified and modified MWCNTs were carried out on a HR 800 micro-Raman spectrometer (HORIBA Jobin Yvon, France). The Raman spectra were obtained in the scanning range of 1000–3000 cm–1 with an incident laser excitation wavelength of 532 nm.

Fourier Transform Infrared Spectroscopic (FTIR) Analysis

Fourier transform infrared spectroscopic (FTIR) analysis of the unmodified and modified MWCNTs was carried out using a Vertex 80 FTIR spectrometer (Bruker, Germany). The FTIR spectra were recorded over the scanning range of 400–4000 cm–1 at room temperature.

Laser Scanning Confocal Microscopic Analysis

Laser scanning confocal microscopic analyses were carried out on an Olympus IX 81 confocal laser scanning microscope. Laser scanning confocal micrographs were captured using a thin layer of the emulsion that was retained between glass slides. The sample preparation for the confocal microscopic investigation could be found elsewhere.[64] The droplet size and droplet-size distribution of the emulsion samples were estimated from the confocal micrographs using the ImageJ software.

Cryo-FEG-SEM Analysis

The dispersion state of MWCNTs in HCEs was examined using a Cryo-FEG-SEM facility: JSM-7600F FEG-SEM with a PP3000T cryo-preparation system, Japan. The Cryo-FEG-SEM micrographs were captured from the fractured surfaces under 5 kV acceleration voltage. The sample preparation for the Cryo-FEG-SEM analysis can be found elsewhere.[64]