ARTICLE pubs.acs.org/Langmuir
SolGel Transition in Dispersions of Layered Double-Hydroxide Nanosheets Vikrant V. Naik and Sukumaran Vasudevan* Department of Inorganic and Physical Chemistry, Indian Institute of Science, Bangalore 560012, India
bS Supporting Information ABSTRACT: Surfactant-intercalated layered double-hydroxide solid MgAl LDHdodecyl sulfate (DDS) undergoes rapid and facile delamination to its ultimate constituent, single sheets of nanometer thickness and micrometer size, in a nonpolar solvent such as toluene to form stable dispersions. The delaminated nanosheets are electrically neutral because the surfactant chains remain tethered to the inorganic layer even on exfoliation. With increasing volume fraction of the solid, the dispersion transforms from a free-flowing sol to a solidlike gel. Here we have investigated the solgel transition in dispersions of the hydrophobically modified MgAl LDHDDS in toluene by rheology, SAXS, and 1H NMR measurements. The rheo-SAXS measurements show that the sharp rise in the viscosity of the dispersion during gel formation is a consequence of a tactoidal microstructure formed by the stacking of the nanosheets with an intersheet separation of 3.92 nm. The origin and nature of the attractive forces that lead to the formation of the tactoidal structure were obtained from 1D and 2D 1H NMR measurements that provided direct evidence of the association of the toluene solvent molecules with the terminal methyl of the tethered DDS surfactant chains. Gel formation is a consequence of the attractive dispersive interactions of toluene molecules with the tails of DDS chains anchored to opposing MgAl LDH sheets. The toluene solvent molecules function as molecular “glue” holding the nanosheets within the tactoidal microstructure together. Our study shows how rheology, SAXS, and NMR measurements complement each other to provide a molecular-level description of the solgel transition in dispersions of a hydrophobically modified layered double hydroxide.
’ INTRODUCTION Gelation occurs in a wide range of dispersed systems where particles interact via attractive forces.1,2 Despite its ubiquity and significance, the molecular mechanisms involved in the formation of a gel are far from understood. One of the most extensively investigated processes is the gelation in dispersions of clay particles in both aqueous and nonaqueous media. The process is of theoretical interest as well as industrial importance because clay gels have wide applications in drilling fluids, paints, ceramic additives, and cosmetic and pharmaceutical formulations.39 Clay suspensions are composed of anisotropic disklike charged particles that interact via an attractive screened Coulomb potential and undergo a transition from a fluidlike sol to a solidlike gel with increasing volume fraction of the clay in the dispersion.10 Various gelation mechanisms and gel microstructures have been proposed for clay dispersions.1115 These include the 3D “houseof-cards” aggregated structure originating from electrostatic attraction between oppositely charged double layers at the edges and faces of the particles and tactoidal structures formed by the regular stacking of the clay particles. Gel formation in the latter is a consequence of electrical double-layer repulsion between clay particles, the so-called Wigner glass.15 Organoclays, a closely related system, are formed by the intercalation of long-chain surfactants in the clay.16 These are widely used to control the r 2011 American Chemical Society
rheology of hydrocarbon fluids and as flow modifiers in a range of applications that include oil-field drilling fluids, paints, and lubricating greases.59 A gel state can be induced in the organoclay dispersion by changing the volume fraction of the solvent and is also facilitated by additives such as water molecules.4,17 The underlying mechanism of the gelling process in the organoclays, however, remains obscure. Recently, we outlined a simple strategy to delaminate layered double-hydroxide (LDH) solids to their ultimate constituent, intact single layers of nanometer thickness and micrometer size in a nonpolar solvent such as toluene.18 The layered double hydroxides are anionic clays derived from the Brucite (Mg(OH)2) structure by the isomorphous substitution of M2+ with M3+ with charge neutrality being preserved by interlamellar exchangeable anions. The delamination procedure involved the ion-exchange intercalation of an ionic surfactant to form a hydrophobic anchored surfactant bilayer in the interlamellar space of the solid. Delamination was effected by simply stirring the surfactantintercalated layered solid in the solvent. The method was rapid but at the same time gentle enough to produce exfoliated nanosheets with Received: July 25, 2011 Revised: September 17, 2011 Published: September 19, 2011 13276
dx.doi.org/10.1021/la202876g | Langmuir 2011, 27, 13276–13283
Langmuir Scheme 1. Delamination of a Surfactant-Intercalated LDH in Toluene to Give Colloidal Dispersions18
ARTICLE
The volume fraction ϕv is defined as18 ϕv ¼ ϕm
regular hexagonal morphology (Scheme 1). A unique feature of the procedure is that the delaminated nanosheets are electrically neutral because the ionic surfactants remain anchored to the sheets. With increasing volume fraction of the LDH solid in the dispersion, stable gels are obtained, similar to that in clay dispersions. Unlike the clays, however, in these dispersions the surfactant-tethered LDH nanosheets do not interact through spatially separated surface charges of opposite sign because they are electrically neutral. Their gelling behavior at low volume fractions and the nature of attractive forces in dispersions of the LDH are therefore rather surprising. Here we describe the solgel transition in dispersions of dodecyl sulfate (DDS)-intercalated Mg0.67Al0.33(OH)2-layered double-hydroxide (MgAl LDHDDS) nanosheets in toluene. We have explored the gelation process and its dependence on the volume fraction by rheological measurements. The dispersions are non-Newtonian fluids exhibiting a shear banding instability at higher shear rates. The microstructure of the dispersions and its changes during the gelling process were characterized by smallangle X-ray scattering (SAXS) measurements. An analysis of the SAXS measurements indicates a tactoidal structure of the dispersion that successfully reproduces the value of the volume fraction at which gelling is observed in the rheology measurements. These measurements, of course, do not provide an understanding of the nature of the attractive forces responsible for the gelling of the dispersion. This is provided by 1H NMR measurements of the dispersions. The nature of the attractive forces in the dispersion was deciphered from the direct evidence of the interaction of the toluene solvent molecules with the surfactant-anchored dispersed LDH sheets obtained from the saturation transfer difference NMR and 2D-NOESY NMR spectra. The NMR results in conjunction with the rheology and SAXS data allow a molecular perspective of the macroscopic phenomena associated with the solgel transition in dispersions of the hydrophobically modified layered double-hydroxide nanosheets. To the best of our knowledge, we believe that this is the first study of the solgel transition in dispersions of layered solids to do so.
’ EXPERIMENTAL SECTION Materials and Methods. Intercalated Mg0.67Al0.33(OH)2(DDS)0.33, [MgAl LDHDDS] was synthesized and characterized using previously reported procedures (Supporting Information).19 The delamination of MgAl LDHDDS was carried out by stirring known weights of MgAl LDHDDS into 3 mL of AR-grade toluene, followed by sonication for 5 min. Dispersions with volume fractions (ϕv) of 0.0005, 0.005, 0.024, 0.048, 0.091, 0.13, 0.166, 0.2, and 0.23 of MgAl LDHDDS in toluene were prepared. These volume fractions correspond to concentrations of 0.001, 0.01, 0.1, 0.2, 0.3, 0.4, 0.5, and 0.6 g/mL of MgAl LDHDDS in toluene.
F ½1 ϕm LDH þ ϕm Ftoluene
!1
where ϕm is the partial mass of LDHDDS in the dispersion and Ftoluene is the density of toluene. FLDH is the density of LDHDDS and was assumed to be same as that of hydrotalcite (2 g/mL).18 Measurement Techniques. Rheology. Rheological measurements were performed on a Physica Anton Paar 100 rheometer using cone and plate geometry (CP 25-2) with a Peltier temperature controller. Solvent evaporation was minimized by covering the rheology plate with a toluene-soaked sponge in the chamber and by carrying out the measurements at subambient temperature (10 °C). All measurements were made by fixing the gap between the cone and plate at 0.05 mm. The rotational flow curve (zero shear stress) measurements were performed in the shear rate range of 1 105 to 1 102 s1 at intervals of 30 s. SAXS Measurements. The small-angle X-ray scattering experiments were performed on a Bruker NANOSTAR SAXS instrument. Incident X-rays were generated using a rotating copper target (λ = 1.54 Å), and the scattered intensity was collected using a 2D multiwire channel detector. The measurements were carried out under vacuum. The collected SAXS data was corrected for transmission and background and averaged azimuthally as a function of the scattering vector, q. Typically, the SAXS intensity profiles covered the q range of 0.1 to 2.1 nm1. The final intensities of the scattered radiation were obtained as the difference between the observed intensities and that of pure toluene. The SAXS measurements were carried out for volume fractions of 0.0005, 0.005, 0.024, and 0.048. NMR Measurements. The 1H NMR spectra of the dispersions in deuterated D8 toluene (Aldrich 99.99%) were recorded on a Bruker AV500 NMR spectrometer (500 MHz for 1H) with a 1H high-resolution magic angle spinning (HRMAS) probe. The HRMAS spectra were obtained using a 40 μL HRMAS zirconia rotor spun at frequencies of between 2 and 12 kHz, depending on the nature of the experiment. A total of eight transients were collected to obtain a reasonable signal-tonoise ratio. Chemical shifts were calibrated with respect to TMS. The NMR measurements were recorded for dispersions with volume fractions of 0.0005, 0.048, 0.091, 0.13, and 0.166. Saturation transfer difference (STD) and 2D NOESY experiments were performed at a spinning speed of 7 kHz on a dispersion with a volume fraction of 0.091. In the NOESY experiment, the mixing time (tm) was kept at 5 ms. In the STD experiment, the toluene peaks were saturated for 3 s. SAXS Analysis. The X-ray scattering intensity in SAXS depends on a number of factors, viz., the number of scatterers (or colloidal particles), the structure factor, S(q), which depends on the interparticle interaction, and the form factor, F2(q), which depends on the shape of the particles.20 Following the usual practice, the SAXS data was analyzed by assuming that the scattering intensity can be expressed as a product of S(q) and F2(q). A preliminary analysis of the X-ray scattering intensities showed that they had a q2 dependence that is consistent with the form factor of dispersed thin disks.20,21 The scattering intensities were analyzed by modeling the LDH sheets as thin disks with radius R and thickness 2H (Supporting Information). The LDH samples were prepared by precipitation methods and hence their lateral dimensions are likely to show polydispersivity. This was accounted for by convoluting the form factor with a normalized Gaussian distribution with the fwhm, ΔR.21 At lower volume fractions, the dispersions can be thought of as consisting of fully dispersed LDH sheets, but the assumption may not be valid at higher volume fractions where the sheets could stack with an intersheet distance D. The intersheet distance could also exhibit a distribution (σD) and so also 13277
dx.doi.org/10.1021/la202876g |Langmuir 2011, 27, 13276–13283
Langmuir
ARTICLE
Figure 1. (a) Plot of shear stress as a function of shear rate for different volume fractions of MgAl LDHDDS in toluene. (b) Zero-shear viscosity as a function of volume fraction. The gel point from the test tube inversion is indicated. 0the number of LDH sheets, N, in a stack.22,23 The latter may be accounted for by convoluting the structure factor with a Gaussian distribution with width ΔN. The X-ray scattering intensity, I(q), expressed as the product of the form factor and structure factor, can be written as IðqÞ ¼ A ðF2 ðqÞfH, Rg X ΨðRÞfR, ΔgÞ ðSðqÞfN, σD , Dg X ΨðNÞfN, ΔN gÞ þ B
ð1Þ
where F2(q; R, H) is the form factor, ψ(R){R, Δ} is the normalized Gaussian distribution in R, S(q){N, σD, D} is the structure factor, and ψ(N){N, ΔN} is the normalized Gaussian distribution in N. A and B are instrument constants, and X indicates convolution. A more detailed description of the SAXS analysis is provided as part of the Supporting Information. A MATLAB code was written to carry out the nonlinear least-squares fit.24
’ RESULTS AND DISCUSSION Rheology. Viscoelastic samples are characterized by their resistance to flow under applied stress. The flow curves and shear stress, σ, as a function of shear rate for different volume fractions of LDH dispersed in toluene, measured using simple rotational flow curve measurements, are shown in Figure 1a. For all volume fractions, the shear stress increases linearly with shear rate at low shear rates characteristic of a Newtonian fluid. However, on increasing the shear rate above ∼104 s1, the flow curves exhibit a region of instability. The point at which the samples begin to show this instability is almost independent of the sample concentration. On further increases in the shear rate (>5 s1), the linear regime is recovered, albeit with a reduced
slope as compared to the slope at low shear rates. The decrease in viscosity (defined as the ratio of shear stress to shear rate) at high shear indicates shear thinning.25 The observed discontinuity in the flow curves is characteristic of shear banding, a phenomenon that occurs as a consequence of the existence of two or more flow regimes with different shear rates under the same shear stress.26 Shear banding is manifested in a measured stressshear-rate flow curve in the form of a zeroslope stress plateau or a discontinuity.27 It should be noted that in the cone and plate geometry used in the present measurements the applied stress is homogeneous and hence the observed shear banding discontinuity is not an artifact of the instrument. The stress plateau/discontinuity seen in Figure 1a indicates that when stress is imposed the shear rate is not stable and evolves either toward a lower or upper bound. In the present measurements, the flow instabilities are absent above a critical shear rate, 5 s1, at which the flow curve recovers a positive slope. Shear banding has been observed for a large number of colloidal suspensions and has been extensively studied for laponite dispersions.2831 The microscopic origin of the phenomenon is, however, poorly understood. The behavior is associated with a dynamic transition involving changes in the microstructure of the dispersion due to coupling between the shear stress and local organization. In the present system, this instability is probably a consequence of the conflicting response to shear resulting in a transition from a jammed disk microstructure to a less-viscous microstructure where the disks are aligned. The two microstructures have a difference of about 4 orders of magnitude in their viscosity (Figure 1a). The zero-shear viscosity as a function of volume fraction, ϕv, is shown in Figure 1b. The zero-shear viscosity was calculated from the slope of the linear region at lower shear rates in Figure 1a. As expected, the dispersions show an increase in viscosity with increasing volume fraction. Dispersions at low volume fractions have low viscosity and are almost fluidlike. On increasing the concentration (ϕv > 0.048), the viscosity increases rapidly, and at a volume fraction of about 0.091, the dispersion ceases to flow, as confirmed by the tube inversion test.18 Two types of microstructures can explain the observed viscosity behavior. The first is the classical house-of-cards structure that consists of sheets arranged in a random edgeto-face orientation that leads to the formation of a gel.12 This randomly arranged microstructure would lead to a high viscosity of the dispersion. On application of shear, the sheets would align themselves parallel to each other, thereby decreasing the friction and hence the viscosity. Another possible structure is a tactoidal arrangement wherein the LDH sheets stack in a parallel platelike arrangement.14 The application of a small shear would be sufficient for the already parallel platelets to slide over each other and show shear-induced thinning. Either microstructure, the house-of-cards” or the tactoidal microstructure, could explain the observed viscosity behavior. However, small-angle X-ray scattering measurements (SAXS) can distinguish the two. Small Angle X-ray Scattering. The shape plots of the scattering intensity I(q) versus q for four concentrations of the MgAl LDHDDS dispersion in toluene are shown in Figure 2. It may be seen that at low volume fractions the scattering intensity decays monotonically with increasing q. At higher volume fractions of the LDH in the dispersion, however, the X-ray scattering plots show a distinctive but broad hump at q = 1.60 nm1. This value of q is independent of the volume 13278
dx.doi.org/10.1021/la202876g |Langmuir 2011, 27, 13276–13283
Langmuir
ARTICLE
Figure 2. Shape plots of the scattering intensity I(q) vs wave vector, q, for different volume fractions of MgAl LDHDDS dispersed in toluene. The red line represents the best fit of eq 1 to the experimental data. The parameters of the fit are provided in Table 1.
Table 1. Structural Parameters Obtained from an Analysis of the SAXS Data volume fraction 0.0005
0.005
2R (Å)
360 ( 2.5
360 ( 2.5
Δ (Å)
40 ( 10
40 ( 10
5 ( 0.5
5 ( 0.5
0.024
0.048
400 ( 3.5
410 ( 4.0
60 ( 10
60 ( 10
5 ( 0.5
5 ( 0.5
N
2
3
ΔN Dmax (Å)
2.2 ( 0.4 39.2
2.7 ( 0.3 39.2
σD
0.01 ( 0.0005
0.01 ( 0.0005
2H (Å)
fraction of MgAl LDHDDS in toluene. The corresponding value of the peak in real space is 39.2 Å. This value is much higher than the 27.6 Å interlayer spacing in crystalline MgAl LDHDDS.19 For all volume fractions, the scattering intensity I(q) roughly follows a q2 power-law decay at small q. This variation is consistent with the form factor, F2(q), for randomly oriented thin disks.20,21 It may thus be concluded that the dilute dispersions of MgAl LDHDDS consist of fully delaminated inorganic sheets whereas the presence of the broad hump at q = 1.60 nm1 (d = 39.2 Å) indicates the stacking of the sheets. The results suggest that the gel retains some form of tactoidal ordering of the MgAl LDHDDS sheets. The SAXS patterns were analyzed following the procedure described in the Experimental Section. The nonlinear
least-squares fit of eq 1 to the experimental data is shown in red in Figure 2, and the parameters that gave the best fit are tabulated in Table 1. Although the nonlinear algorithm has a number of variables (Table 1), the fact that the thickness of the sheet (2H) for all volume fractions is in close agreement with the thickness of an MgAl LDH layer32 (4.8 Å) gives confidence in the uniqueness of the fit parameters in Table 1 and also confirms that the MgAl LDHDDS solid is completely delaminated into isolated single sheets in the dispersion. The delaminated sheets have the DDS bilayer anchored to them, but because of the low scattering cross section of the organic moiety, only the X-ray scattering from the inorganic MgAl LDH would be observed. The delaminated sheets have a mean lateral dimension of ∼400 Å. At higher volume fractions, the sheets stack with an intersheeet separation of 39.2 Å. The number of sheets, N, in a stack appears to increase with the volume fraction, but as shown in Table 1, there is a fairly large distribution in this number. To summarize, the SAXS results indicate that at low volume fractions MgAl LDHDDS exists as isolated single sheets in the dispersion but with increasing volume fraction a tactoidal microstructure, formed by the stacking of the delaminated nanosheets with an intersheet distance of 39.2 Å, evolves, leading to the formation of the gel. The SAXS results can be related to the rheology measurements by considering the effective hydrodynamic volume, Veff, of a tactoid given by33 Vef f ¼ 13279
4Rt 3 3
ð2Þ dx.doi.org/10.1021/la202876g |Langmuir 2011, 27, 13276–13283
Langmuir
ARTICLE
Table 2. Assignment of the 1H NMR Resonances in the HRMAS Spectra of MgAl LDHDDS Dispersed in Toluene (Ov = 0.091) chemical shift (δ)/ppm
Figure 3. 1H NMR spectra of (a) MgAl LDHDDS dispersed in D8 toluene with a volume fraction of ϕv = 0.0005 (red) and (b) NaDDS in D8 toluene (black). (c) HRMAS spectra of MgAl LDHDDS in D8 toluene in the gel state with a volume fraction of ϕv = 0.091 (blue) and (d) MgAL LDHDDS in CDCl3 in the gel state at a volume fraction of 0.166.
where Rt is the radius of the tactoid. This volume is much larger than its actual displaced volume, V, V ¼ πRt 2 ht
ð3Þ
where ht is the thickness of the tactoid. A transition from a sol to a gel may be expected when the tactoidal spheres “touch” and fill space. This would occur at a volume fraction of ϕv*34 ϕ ¼ ϕMRJ
V 3ht ¼ ϕMRJ Vef f 4Rt
ð4Þ
where ϕMRJ (= 0.30) is the volume fraction associated with random close packing or the “maximally random jammed” configuration of spheres.35 For a tactoidal microstructure formed by the stacking of three MgAl LDHDDS sheets (N = 3), Rt = 200 Å and ht = 93 Å, the critical volume fraction ϕ* is 0.099. This value may be compared to the value of the volume fraction ϕv = 0.091 at which the dispersion ceases to flow (Figure 1b). The agreement in the two measurements is reasonable considering the fact that the calculation did not consider the variation in the aspect ratios and the polydispersity in the lateral dimensions of the tactoids. The rheology and SAXS measurements provide a consistent picture of the microstructure of the MgAl LDHDDS dispersion that explains the observed solgel transition with increasing volume fraction. It does not, however, provide an explanation as to why the tactoidal structure is formed with the sheets stacked with an intersheet distance of 39.2 Å, a value that is much larger than that in crystalline MgAl LDHDDS (27.6 Å). Clay suspensions are known to form space-filling elastic “gels” at dilute volume fractions with either a tactoidal or house-of-cards microstructure.1115 However, here the microstructure arises because of the presence of charges on the surfaces of the platelets that allows them to interact via long-range electrostatic interactions. However, the MgAl LDHDDS sheets dispersed in toluene are electrically neutral and therefore not expected to interact through spatially separated opposite surface charges as in the clays. The origin of the attractive interactions between the MgAl LDHDDS nanosheets dispersed in toluene remains an open question. In an earlier molecular dynamics study, we had suggested that the delamination of crystalline MgAl LDHDDS
assignment
0.33
LDH water
0.83
ω-CH3 of DDS chain
1.01
?
1.21
CH2 chain
1.41 2.01
? CH3 of toluene
3.25
α-CH2 of the DDS chain
3.944.38
LDH protons
6.827.08
toluene ring
in toluene occurs because of the disruption of the dispersive interactions between chains anchored on opposing LDH sheets by the toluene solvent molecules and also the possibility of toluenemediated stacking of the delaminated MgAl LDHDDS.18 A molecular understanding of the nature of the interactions of the toluene solvent molecules and the dispersed MgAl LDHDDS sheets is clearly a prerequisite to a microscopic explanation of solgel formation in this system. We have attempted to do so by 1 H NMR measurements of dispersions of MgAl LDHDDS in deutrated toluene at different volume fractions. NMR Spectroscopy. The proton NMR spectrum of a dilute dispersion (ϕv = 0.0005) of MgAl LDHDDS in D8 toluene is shown in Figure 3. This composition corresponds to the first point in the zero-shear viscosity versus volume fraction plot (Figure 1b). The NMR spectrum shows features belonging to the anchored DDS chain and trace amounts of protonated toluene present as an impurity in the deuterated solvent. For comparison, the spectrum of the sodium salt of DDS in toluene is also shown (Figure 3). It may be seen that the positions of the resonances in the two spectra are comparable. The NMR spectrum shows three distinct peaks for the DDS chain: ω or terminal CH3, a broad and not entirely fully resolved resonances for the CH2 protons corresponding to the methylene units of the DDS chain and the α protons that are attached to the headgroup (Table 2). At higher volume fractions, the 1H spectra are broadened and poorly resolved as a consequence of dipolar interactions that become significant at these concentrations and because the absence of mobility in these viscous gel samples leaves these interactions only partially averaged (Supporting Information).36 Better-resolved spectra with narrower lines are obtained by spinning the samples at high speeds (210 kHz) at the magic angle, 54.74°. The high-resolution magic angle spinning (HRMAS) spectrum of a dispersion with a volume fraction of 0.091 at a spinning speed of 2 kHz is shown in Figure 3. A comparison of the HRMAS spectrum of the dispersion with a volume fraction of 0.091 with that of the dilute dispersion shows the presence of two additional features at 1.01 and 1.41 ppm (shaded peaks in Figure 3). The assignments of the 1H NMR peak positions are given in Table 2. The HRMAS spectra were also recorded for different volume fractions of MgAl LDHDDS in the dispersion (Supporting Information). With increasing concentration, the intensities of the two new features at 1.01 and 1.41 ppm grow at the expense of the original DDS resonances. Above a volume fraction of 0.13, the original peaks are completely absent and only the new resonances are seen. These spectra suggest that with increasing 13280
dx.doi.org/10.1021/la202876g |Langmuir 2011, 27, 13276–13283
Langmuir
Figure 4. Saturation transfer difference spectrum on irradiation of the toluene ring resonance for dispersions of MgAl LDHDDS in D8 toluene in the gel state at a volume fraction of 0.091. The numbers indicate enhancement factors for the resonances at 1.01 and 1.41 ppm. The spectra were recorded under HRMAS at a spinning speed of 7 kHz.
volume fraction there is a continuous transformation of the DDS chain characterized by the resonances at 0.83 and 1.21 ppm to that characterized by the resonances at 1.01 and 1.41 ppm. Before attempting to interpret the origin of these features, it is important to rule out the possibility that these features are not artifacts of the HRMAS experiment. It is well known that spurious features can appear in the HRMAS measurements. These can arise, for example, from the heating of the sample at high spinning speeds. In gels, there is also the possibility that the microstructure is disrupted because of mechanical perturbation. This possibility may be ruled out by the fact that spectra are reproducible over a number of repeated recordings. To eliminate the possibility that these new resonances are not an artifact of spinning, the HRMAS spectra were recorded as a function of spinning speed (Supporting Information). Although the intensity of the original DDS peaks remains roughly the same, the intensity of new resonances rises sharply and the peaks narrow down as the spinning speed is increased from 2 to 10 kHz. There is, however, no change in the positions of any of the resonances with spinning speed. The spectra are reversible and reproducible with the changing of spinning speeds, which clearly suggests that the microstructure of the gels is not disrupted by vortex formation inside the HRMAS rotor. Spinlattice relaxation (T1) measurements showed no change in T1 values with spinning speeds, clearly ruling out any rise in temperature during the HRMAS experiment. The new peaks at 1.01 and 1.41 ppm are genuine features associated with the formation of the gel. The fact that the crystalline MgAl LDHDDS is easily delaminated in toluene whereas the corresponding nitrate-intercalated MgAl LDHNO3 is not suggests a close association of toluene solvent with the anchored DDS chains.18 It may therefore be anticipated that the aromatic ring current of the toluene molecules could cause partial shielding of the methyl and methylene parts of the DDS chains that are in close proximity. To verify this possibility, the HRMAS spectra were recorded for gels formed by the dispersions of MgAl LDHDDS in CDCl3. MgAl LDHDDS delaminates in chloroform in a manner similar to that in toluene and forms gels at higher concentrations of the dispersed MgAl LDHDDS nanosheets. However, the spectra of the chloroform dispersions, even at the highest volume
ARTICLE
Figure 5. NOESY spectrum of MgAl LDHDDS in D8 toluene in the gel state with a volume fraction of 0.091. The cross peak between the toluene ring and the terminal CH3 of the tethered surfactant chain is indicated. The spectra were recorded under HRMAS at a spinning speed of 7 kHz.
fractions, do not show any additional features; only the resonances at 0.83 and 1.33 ppm corresponding to the chain methylene protons and the ω-CH3 group, respectively, are observed (spectrum d in Figure 3). It may thus be concluded that the additional features seen at 1.01 and 1.41 ppm in the toluene dispersions originate from the partial shielding of the methylene and methyl protons of the surfactant chain. This explanation would require that the toluene molecules responsible for the shielding have restricted orientational and translational mobilities; otherwise, the shielding would be averaged out and no new resonances would be observed. Saturation Transfer Difference and 2D NOESY. An NMR experiment that can directly establish the close association of the toluene molecules and the anchored DDS surfactant chains is the saturation transfer difference (STD) experiment.37 In the STD experiment, a particular peak is selectively saturated by continuous radiation and the changes in the intensity of other peaks in the spectra are recorded as a difference spectrum between the original and irradiated spectra. The resonances that are in spatial proximity and coupled via dipolar interactions would show a higher enhancement as compared to those that are physically removed. In STD experiments on the dispersions, the CH3 protons (2.01 ppm) and the ring protons (6.827.08 ppm) of toluene were selectively irradiated and the transfer of magnetization to the peaks of DDS chains were recorded. STD experiments were performed on dispersions in the gel state with a volume fraction of 0.091 at a spinning speed of 7 kHz. The selected peaks were irradiated for 3 s, followed by the proton π/2 pulse. On irradiation of the toluene CH3, no appreciable changes in the intensity for any of the peaks were observed. When the toluene ring was irradiated, however, the new resonances at 1.01 and 1.41 ppm showed a high enhancement as seen in the difference spectra in Figure 4. The enhancement factors, η = (I (I0/ I0)) where I is the intensity of the resonance on saturation and I0 is the intensity in the absence of irradiation, are 0.068 and 0.082, respectively. The original DDS resonances did not show up in the difference spectrum. The STD experiment confirms that the new resonances seen in the HRMAS spectra of the concentrated dispersions are 13281
dx.doi.org/10.1021/la202876g |Langmuir 2011, 27, 13276–13283
Langmuir indeed a consequence of the association of DDS chains with the toluene solvent molecules. This association leads to a shift in the resonances of the protons as compared to that of chains that are not associated. The resonance at 1.01 ppm is assigned to the ωCH3 of chains associated with toluene molecules, and the resonance at 1.41 ppm is assigned to the protons of the methylene chain. The difference in the enhancement factor η for the two peaks suggests that the primary association of the toluene molecules is with ω-CH3.The fact that these are downfield shifted with respect to the original DDS peaks indicates that toluene molecules and the chain are oriented such that the methylene and methyl protons experience a diamagnetic shielding.38 This would arise only if the surfactant chains are aligned with the face of the ring and not the edge. Evidence for the association of the toluene solvent molecules with the anchored surfactant DDS chains may also be obtained from 2D NOESY experiments.39 The NOESY spectra of a MgAl LDHDDS dispersion in D8 toluene in the gel state with a volume fraction of 0.091 is shown in Figure 5. The spectra were recorded with a mixing time tm of 5 ms. It should be noted that the mixing times are usually on the order of 1 ms for solid samples and 100 ms for solutions. The spectra were recorded under HRMAS at a spinning speed of 7 kHz. The NOESY spectrum shows a cross peak between the resonance at 1.01 ppm that had been assigned to the terminal ω-CH3 of the surfactant chains associated with toluene molecules and the aromatic toluene ring. No other cross peaks, either from the original DDS resonances or from the resonances at 1.41 ppm and the toluene molecules, are observed. The cross peak indicates a close spatial proximity between the ω-CH3 and toluene molecules. It may be recalled that in the 1D saturation transfer difference (STD) experiment the ω-CH3 protons showed a much higher enhancement than did the CH2 methylene protons on irradiation of the toluene ring peak. The NOESY cross peak is seen on only one side of the diagonal peaks. This is because the exchange rates of the forward and backward processes favor the associated product.40 toluene þ DDS S toluene-DDS In summary, the NMR measurements provide direct evidence of the association of the toluene solvent molecules with the anchored DDS chains of the dispersed MgAl LDH DDS nanosheets. The evidence is the presence of two new downfieldshifted peaks at 1.01 and 1.41 ppm that correspond to those segments of the anchored DDS chains that interact directly with the toluene ring. These new resonances that are downfield shifted with respect to the original DDS resonances are assigned to the ω-CH3 and methylene protons of the DDS chain, with the aromatic ring current of the toluene solvent being responsible for the diamagnetic downfield shift. The 2D NOESY spectrum provides direct evidence of the association of toluene molecules with the ends of the tethered surfactant chain, primarily the terminal methyl groups. With increasing volume fraction of MgAl LDH DDS in the dispersion, the concentration of the chains that show a downfield shift increases with respect to the peaks at the original resonance of the DDS chains and at the concentration where for the dispersion gels only the downfieldshifted resonances are observed. The mechanism of gelling of MgAl LDH DDS in toluene is, therefore, clearly correlated with the association of the terminal ω-CH3 of the anchored DDS chains with the toluene solvent molecules. The associated toluene molecules have reduced translational and orientational
ARTICLE
mobilities that in turn are responsible for the drop in viscosity seen in the macroscopic rheology measurements of the dispersion (Figure 1b).
’ CONCLUSION Surfactant-intercalated layered double-hydroxide solid MgAl LDHDDS undergoes rapid and facile delamination to its ultimate constituent, single sheets of nanometer thickness and micrometer size, in a nonpolar solvent such as toluene. With increasing volume fraction of the solid, the dispersion transforms from a free-flowing sol to a solidlike gel. The viscosity rises sharply above a volume fraction of 0.048, and the gel point, as determined by test tube inversion test, occurs at a volume fraction of 0.091. Here we have investigated the solgel transition in dispersions of the hydrophobically modified MgAl LDHDDS nanosheets in toluene by rheology, SAXS, and 1H NMR measurements. These techniques provide different insights into the gelling process, allowing for a consistent and holistic perspective of the solgel transition in these dispersions. The flow behavior of the dispersions characterized by rheology measurements shows Newtonian behavior at low shear rates with the viscosity increasing with the volume fraction. At shear rates higher than 104 s1, however, the flow curves exhibit a discontinuity that is characteristic of shear banding, a phenomenon signifying the existence of two or more flow regimes with different shear rates under the same shear stress. SAXS measurements on the dispersions of MgAl LDHDDS in toluene showed that for dilute dispersions the observed scattering could be modeled as arising from isolated disks of thickness 5 Å and diameter 400 Å. The thickness of the disk is consistent with the thickness of the MgAl LDH sheet, 4.8 Å (the SAXS measurements see only the inorganic part of the MgAl LDHDDS sheet). With increasing volume fraction of the solid in the dispersion, the X-ray scattering shows the evolution of a peak at 1.60 nm1. The intensity of the peak increases with concentration, but its position remains constant. The SAXS data was successfully modeled, assuming that with increasing concentration individual disks stack to form a tactoidal structure with an intersheet separation of 39.2 Å. With increasing volume fraction of the solid in the dispersion, the number of disks (sheets) in the tactoidal stacks increases. It is this tactoidal microstructure that is responsible for gel formation, and the percolation threshold or “jamming” calculated from the dimensions obtained from SAXS agrees well with the volume fraction for gelling as obtained from rheology measurements. The breakup of the tactoidal microstructure under shear is responsible for shear banding and the subsequent drop in viscosity. 1 H HRMAS NMR spectra were recorded for different volume fractions of MgAl LDHDDS in D8 toluene. NMR measurements provide direct evidence of the association of the toluene solvent molecules with the anchored DDS chains of the dispersed MgAl LDHDDS sheets. Two new peaks at 1.01 and 1.41 ppm that are downfield shifted with respect to the original DDS resonances are observed. These new resonances are assigned to the ω-CH3 and methylene protons of the DDS chain that interact directly with the toluene ring, with the aromatic ring current of the toluene solvent being responsible for the diamagnetic downfield shift. The 2D NOESY spectrum provides direct evidence of the association of toluene molecules with the ends of the tethered surfactant chain, primarily the terminal methyl groups. With increasing volume fraction of MgAl LDHDDS in the 13282
dx.doi.org/10.1021/la202876g |Langmuir 2011, 27, 13276–13283
Langmuir dispersion, the concentration of the chains that show a downfield shift increases with respect to the peaks at the original resonance of the DDS chains and at the concentration where the dispersion gels only the downfield-shifted resonances are observed. The mechanism of gelling of MgAl LDHDDS in toluene is therefore clearly correlated with the association of the terminal ω-CH3 of the anchored DDS chains with the toluene solvent molecules. The associated toluene molecules have reduced translational and orientational mobilities that in turn are responsible for the drop in viscosity seen in the macroscopic rheology measurements on the dispersion. We stated earlier that gel formation in dispersions of surfactant-anchored LDH nanosheets is rather surprising because of the absence of any obvious evidence of attractive forces between the dispersed sheets. The MgAl LDHDDS sheets dispersed in toluene are electrically neutral and are therefore not expected to interact through spatially separated opposite surface charges as in the clays. The clue to the origin and nature of the attractive forces in this system is the direct evidence of the association between the toluene molecules and the terminal methyl of the DDS chains anchored to the LDH nanosheets provided by the NMR measurements. Gel formation can now be understood as a consequence of the attractive dispersive interactions of toluene molecules with the tails of DDS chains anchored to opposing LDH sheets. The toluene solvent molecules function as a molecular glue cementing the opposing surfactant-anchored LDH sheets and giving rise to the observed tactoidal microstructure. Such a role for the toluene molecules had indeed been suggested by MD simulations of the delamination process, but it is the present measurements that have provided direct evidence of the molecular interactions between the two phases of the dispersion. In conclusion, the present studies have shown how rheology, SAXS, and NMR measurements complement each other in providing a comprehensive molecular-level description of the solgel transition in dispersions of hydrophobically modified layered double-hydroxide nanosheets.
’ ASSOCIATED CONTENT
bS
Supporting Information. Synthesis and characterization of surfactant-intercalated MgAl LDHDDS. Analysis of SAXS data. Dipolar broadened spectrum of MgAl LDHDDS in D8 toluene and the corresponding HRMAS spectrum. 1H HRMAS spectra of dispersions of MgAl LDHDDS in D8 toluene at different volume fractions. Effect of spinning speed on the HRMAS of MgAl LDHDDS in D8 toluene. This material is available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION Corresponding Author
*E-mail:
[email protected]. Tel: +91-80-2293-2661. Fax: +91-80-2360-1552/0683.
’ ACKNOWLEDGMENT We thank Dr. S. Jayanthi and Prof. K. V. Ramanathan for help with the NMR measurements and Prof. Siddhartha P. Sarma for suggesting the STD experiment and for helping us to acquire the NOESY data.
ARTICLE
’ REFERENCES (1) Lu, P. J.; Zaccarelli, E.; Ciulla, F.; Schofield, A. B.; Sciortino, F.; Weitz, D. A. Nature 2008, 453, 499–503. (2) Zaccarelli, E. J. Phys.: Condens. Matter 2007, 19, 323101–323151. (3) Hauser, E. A. Chem. Rev. 1945, 37, 287–321. (4) Swartzen-Allen, S. L.; Matijevic, E. Chem. Rev. 1974, 74, 385– 400. (5) van Olphen, H. An Introduction to Clay Colloid Chemistry, 2nd ed.; John Wiley and Sons: New York, 1977. (6) Weaver, C. E.; Pollard, L. D. The Chemistry of Clay Minerals; Elsevier: Amsterdam, 1973. (7) Neumann, B. S. Rheol. Acta 1965, 4, 250–255. (8) Street, N. J. Aust. J. Chem. 1956, 9, 467–479. (9) Heath, D.; Tadros, Th. F. J. Colloid Interface Sci. 1983, 93, 307–319. (10) Mourad, M. C. D.; Byelov, D. V.; Petukhov, A. V.; de Winter, D. A. M.; Verkleij, A. J.; Lekkerkerker, H. N. W. J. Phys. Chem. B 2009, 113, 11604–11613. (11) Mourchid, A.; Delville, A.; Lambard, J.; LeColier, E.; Levitz, P. Langmuir 1995, 11, 1942–1950. (12) van Olphen, H. Discuss. Faraday. Soc. 1951, 11, 82–84. (13) Rand, B.; Pekenc, E.; Goodwin, J. W.; Smith, R. W. J. Chem. Soc., Faraday Trans. 1 1980, 76, 225–235. (14) Norrish, K. Discuss. Faraday Soc. 1954, 18, 120–134. (15) Ruzicka, B.; Zaccarelli, E .; Zulian, L.; Angelini, R.; Sztucki, M.; Moussad, A.; Narayan, T.; Sciortino, F. Nat. Mater. 2011, 10, 56–60. (16) Whittingham, M. S.; Jacobson, A. Intercalation Chemistry; Academic Press: New York, 1982. (17) Leach, E. S. H.; Hopkinson, A.; Franklin, K.; van Duijneveldt, J. S. Langmuir 2005, 21, 3821–3830. (18) Naik, V. V.; Ramesh, T. N.; Vasudevan, S. J. Phys. Chem. Lett. 2011, 2, 1193–1198. (19) Naik, V. V.; Chalasani, R.; Vasudevan, S. Langmuir 2011, 27, 2308–2316. (20) Guinier, A.; Fournet, G. Small Angle Scattering of X-rays; Wiley: New York, 1955. (21) Kroon, M.; Vos, W. L.; Wegdam, G. H. Phys. Rev. E 1998, 57, 1962–1970. (22) Glatter, O., Kratky, O. Small Angle X-ray Scattering; Academic Press: London, 1982. (23) Richter, D.; Schneiders, D.; Monkenbusch, M.; Willner, L.; Fetters, L. J.; Huang, J. S.; Lin, M.; Mortensen, K.; Farago, B. Macromolecules 1997, 30, 1053–1068. (24) Matlab 7, The Mathworks, Inc.: Natick, MA, 2004. (25) Barnes, H. A.; Hutton, J. E.; Walters, K. An Introduction to Rheology; Elsevier Science Publishers B. V.: Amsterdam, 1993. (26) Callaghan, P. T. Rheol. Acta 2008, 47, 243–255. (27) Rehage, H.; Hoffmann, H. Mol. Phys. 1991, 74, 933–973. (28) Vermant, J. Curr. Opin. Colloid Interface Sci. 2001, 6, 489–495. (29) Olmsted, P. D. Rheol. Acta 2008, 47, 283–300. (30) Schall, P.; van Hecke, M. Annu. Rev. Fluid Mech. 2010, 42, 67–88. (31) Ianni, F.; Di Leonardo, R.; Gentilini, S.; Ruocco, G. Phys. Rev. E 2008, 77, 031406-1–031406-6. (32) Drits, V. A.; Bookin, A. S. Layered Double Hydroxides: Present and Future; Rives, V., Ed.; Nova Science: New York, 2001; p 41. (33) King, H. E., Jr.; Milner, S. T.; Lin, M. Y.; Singh, J. P.; Mason, T. G. Phys. Rev. E 2007, 75, 21403-1–21403-20. (34) Dimon, P.; Sinha, S. K.; Weitz, D. A.; Safinya, C. R.; Smith, G. S.; Varady, W. A.; Lindsay, H. M. Phys. Rev. Lett. 1986, 57, 595–598. (35) Isichenko, M. B. Rev. Mod. Phys. 1992, 64, 961–1043. (36) Mehring, M. High Resolution NMR Spectroscopy in Solids; Springer-Verlag: Berlin, 1976. (37) Mayer, M.; Meyer, B. Angew. Chem. 1999, 38, 1784–1788. (38) Richards, R. E. Proc. R. Soc. London, Ser. A 1960, 255, 72–78. (39) Ernst, R. E.; Bodenhausen, G.; Wokaun, A. Principles of Nuclear Magnetic Resonance in One and Two Dimensions; International Series of Monographs on Chemistry 14; Clarendon Press: Oxford, U.K., 1987. (40) Lee, D.; Lee, W. J. Korean Magn. Reson. 1998, 2, 33–40. 13283
dx.doi.org/10.1021/la202876g |Langmuir 2011, 27, 13276–13283