PFG NMR Investigations of Tetraalkylammonium−Silica Mixtures - The

J. Phys. Chem. C 2010, 114, 8449–8458

8449

PFG NMR Investigations of Tetraalkylammonium-Silica Mixtures Xiang Li and Daniel F. Shantz* Artie McFerrin Department of Chemical Engineering, Texas A&M UniVersity, College Station, Texas 77843-3122 ReceiVed: February 2, 2010; ReVised Manuscript ReceiVed: March 22, 2010

The results from PFG NMR studies of several tetraalkylammonium-silica mixtures are reported. pH data of these mixtures correlate well with previously reported literature. 1H NMR shows line broadening and shifting of resonances for hydrogen atoms on the periphery of the TAA cations as the silica content is increased. Transverse relaxation measurements show a strong correlation with the pH data in that T2 decreases strongly until the silica critical aggregation concentration (cac) is reached at which point the T2 plateaus. These results point to organocation-silica interactions, which were quantified by using pulsed-field gradient NMR. The self-diffusion coefficient of a series of TAA cations in the presence of silica were measured, and a systematic decrease of the diffusion coefficient is observed as the cation/silica ratio decreases. Subsequent analysis gives confidence that the diffusion behavior of the TAA cations can be described as a two-state system, where cations are either free in solution or bound to the silica nanoparticles in the mixture. On the basis of this binding isotherms can be constructed and binding energies can be determined. These values are in the range of -10 to -14 kJ/mol, with tetramethylammonium on the high end and tetrapropylammonium cations on the low end. One conclusion drawn from this work is that the more hydrophilic TMA cations associate more strongly with the silica nanoparticles than TPA. The investigation shows that PFG NMR is a powerful tool for investigating these clear mixtures, and the results are discussed in the context of zeolite nucleation and growth. Introduction The nucleation and growth of high-silica zeolites have been the subject of intense study in the zeolite science community, and remain as unresolved issues in the field. Many of these studies have looked at the nucleation and growth of silicalite-1 (MFI). This has been driven by several factors. Silicalite-1 is the all-silica form of ZSM-5, an industrially important catalyst. Also, silicalite-1 can be rapidly formed at low temperatures from optically transparent mixtures containing only a silica source (often tetraethyl orthosilicate), tetrapropylammonium hydroxide, and water.1-7 Numerous laboratories have observed the presence of silica particles with sizes in the range of 3-5 nm in these mixtures prior to heating. Upon heating silicalite-1 is observed to form over the time scale of several hours.6,8-13 Given the rapid formation of zeolite crystals and the optically transparent nature of the mixture it represents a model system for the investigation of zeolite formation. Thus many techniques such as scattering,5-9,12,14-19 transmission electron microscopy (TEM),1,20-22 NMR,19,23-28 mass spectrometry,29,30 and others have been used to investigate these mixtures. Even though the presence of these colloidal stable nanoparticles is widely agreed upon, their structure, composition, and role in zeolite nucleation remain under debate. Schoeman6 proposed that the nucleation of silicalite-1 is via aggregation of monomeric and oligomeric silica (less than 1-2 nm) and dissolution of these nanoparticles. Hence, the nanoparticles may serve as a source of nutrients during crystal growth. Works by Kirschhock and co-workers conclude that the nanoparticles formed in these mixtures possess many structural similarities to the MFI topology.9,16,19,20,26 The Tsapatsis lab22 has shown that precursor nanoparticles contribute to growth, and their addition to the growing crystal can be rate limiting. However, * To whom correspondence should be addressed. Phone: (979) 845-3492. Fax: (979) 845-6446. E-mail: [email protected].

seeded growth experiments decoupled nucleation from crystal growth and the derived mechanisms cannot elucidate the direct formation of silicalite-1 from dilute clear solutions.3,31 More recently along this theme Davis and co-workers1 proposed a mechanism that the nanoparticles show no silicalite-1 structure yet participate in crystallization by directly adding to the growing crystal during their evolution at room temperature. A series of papers from the Lobo and Vlachos laboratories10,32-37 suggest these particles are initially amorphous, consistent with the work reported from the Tsapatsis lab. The work reported included conductivity, pH, and small-angle scattering studies of silica nanoparticles formed in tetraalkylammonium and alkalimetal silicate solutions. They found that all the cation solutions have a well-defined critical aggregation concentration (cac) at ∼1:1 ratio of [SiO2]/[OH-], below which silica is in the form of dissolved monomers and small oligomers, whereas above which silica forms uniform nanoparticles. The average particle sizes slightly decrease with solution pH and are nearly independent of the size of the alkyl group in TAA. On the basis of small-angle neutron and X-ray scattering measurements they proposed the particles are amorphous hydrated silica and ellipsoidal. Consistent with the works above, Cheng and Shantz showed that formation of these nanoparticles seems to be quite general beyond even TAAOH/silica solutions, as nanoparticles were observed by SAXS in a series of organocation/silica solutions.12 Jorge et al.38 presented a lattice model describing the formation of silica nanoparticles in dilute clear solution. They concluded that nanoparticles are mainly stabilized by electrostatic interactions between the negatively charged silica surface and a layer of organic cations. The work summarized above has contributed significantly to both the understanding of the structural nature of the precursor silica particles and the kinetics of silicalite-1 growth. In contrast, the role of the organocation in this process is still poorly understood. Of central importance is how the TPA cations affect

10.1021/jp101006f  2010 American Chemical Society Published on Web 04/08/2010

8450

J. Phys. Chem. C, Vol. 114, No. 18, 2010

the evolution of the nanoparticles and participate in silicalite-1 nucleation and growth. Along this theme previous work from our lab has shown that silicalite-1 is unusual in the sense it is very difficult to form other siliceous zeolites from clear solutions.12 We have also shown that small perturbations to the silicalite-1 synthesis mixture lead to pronounced differences in rates of growth.14,15 Thus the role of the organocation in this process appears significant, and any insights gained into the strength and nature of the cation-nanoparticle interactions should be very helpful in advancing the understanding of zeolite formation and growth. Pulse Field Gradient (PFG) NMR has emerged as a powerful technique to study, noninvasively, the binding of small molecules to macromolecules via changes in the small-molecule self-diffusion value.39-44 PFG NMR can be used to measure the self-diffusivity of molecules in solution via labeling the nuclear spins by their Larmor frequencies in a spatially varying magnetic field. This is possible through the use of gradient pulses to label the spins, and the subsequent motion of the molecule in solution attenuates the ability to refocus the echo signal.45,46 This experiment can be extended to a two-dimensional experiment via methods such as diffusion ordered 2D-NMR spectroscopy (DOSY).47-49 When DOSY is applied to a mixture, the conventional NMR spectrum is displayed in one dimension, while the “diffusion spectrum”, based on molecular diffusion rate, is displayed in the second dimension, facilitating analysis of mixtures that contain overlapping resonances. Diffusion NMR in the current work is coupled with dynamic light scattering (DLS) that is used to determine the size/diffusion coefficient of the silica precursor particles. PFG NMR and light scattering techniques have been used in combination previously to study self-assembly and polymer-surfactant interaction.50-53 In the current work we report 1H PFG NMR studies of tetraalkylammonium cations in the presence of silica nanoparticles. These results are complemented as necessary with dynamic light scattering. The current work demonstrates the validity of this method for measuring cation-nanoparticle interactions as binding isotherms. Clear differences can be observed for different organocations and a detailed comparative study between tetramethylammonium cations and tetrapropylammonium cations is presented. The implications of these findings in terms of zeolite nucleation and growth are discussed. Experimental Section Materials. All compounds used in the current work were used as received. Tetramethylammonium hydroxide (TMAOH, 25% w/w), tetraethylammonium hydroxide (TEAOH, 35% w/w), tetrapropylammonium hydroxide (TPAOH, 40% w/w), tetrabutylammonium hydroxide (TBAOH, 40% w/w), and tetramethylammonium bromide (TMABr, 98%) were purchased from Alfa Aesar. Tetrapropylammonium bromide (TPABr, 98%) was purchased from Aldrich. Tetraethyl orthosilicate (TEOS, >99%) was purchased from Fluka. Sodium hydroxide (NaOH, 99%) was purchased from BDH Chemicals. Ethanol (CH3CH2OH, >99.5%) was purchased from Acros Organic. Deuterium oxide (D2O, 99.96% D, Cambridge Isotopes) was used as received. Sample Preparation. Two different preparation methods were used to formulate the mixtures investigated. For one set of experiments it was desired to prepare samples of varying silica content at fixed TAAOH concentration for determination of the cac. In this case solutions of molar composition 9TAAOH: xSiO2:9500H2O:4xEtOH (x varied from 0 to 120) were synthesized in two steps. First, the desired tetraalkylammonium hydroxide was added to deionized water and the mixture was

Li and Shantz TABLE 1: Mixture Compositions of Samples Investigated C0

9NaOH:0TEOS:9500H2O

C1 C2 C3 C4

9NaOH:5TEOS:9500H2O:20EtOH 9NaOH:20TEOS:9500H2O:80EtOH 9NaOH:40TEOS:9500H2O:160EtOH 9NaOH:80TEOS:9500H2O:320EtOH

stirred for approximately 1 h. Then tetraethyl orthosilicate was added to the mixture, and the resulting solution was stirred for a minimum of 12 h to ensure the complete hydrolysis of TEOS. The solutions used for PFG NMR measurements were prepared following the method above, but instead with deuterated water. To make a comparison with hydrolyzed TEOS solutions, mixtures with the equivalent amount of water and ethanol were prepared by first diluting TPAOH in deuterated water and then adding the desired amount of ethanol to yield a final composition of 9TPAOH:4xEtOH:9500D2O (x ) 20-120). Samples with fixed silica content wherein the TAA content was varied were also made to measure the cation self-diffusion coefficient as a function of TAA content. In this case solutions of molar composition 9NaOH:yTAABr:xSiO2:9500D2O:4xEtOH (x ) 0, 5, 20, 40, 80; for a given value of x, y was varied between 0.25 and 36) were prepared as follows. NaOH was dissolved in deuterated water. After complete dissolution of the sodium hydroxide, TEOS was added to give a mixture composition of 9NaOH:xSiO2:9500D2O:4xEtOH. These solutions were stirred for at least 12 h, and are hereafter denoted as C0-C4 based on the silica content as noted in Table 1. After full hydrolysis of TEOS, TPABr or TMABr was added to the above solutions, and the resulting mixtures are denoted as yTAA/C0, C1, C3, C4 (TAA ) TMABr, TPABr; y ) 0.25-36). The compositions were chosen as they are representative of those used in silica precursor particle studies3,32,33 and silicalite-1 growth experiments.34 Sodium hydroxide was used to set the solution initial pH when varying the TAA cation concentration, and thus control the nanoparticle formation. All samples used for NMR measurements were allowed to equilibrate for a minimum of 24 h before measuring. Solution pH values were determined with a Fisher Scientific AB15/15+ pH meter and an Accument glass body Ag/AgCl reference electrode (Pittsburgh, PA, USA). The pH meter was calibrated with standardized pH 7 and 10 buffer solutions (Ricca Chemical Co.). NMR Measurements. All NMR experiments were performed on a Varian INOVA 500 MHz spectrometer equipped with a 5 mm broadband indirect detection probe and a z gradient coil (up to 32 G/cm). The temperature was regulated at 25 °C for all experiments and the temperature calibration was performed with methanol. Chemical shifts are reported relative to an internal reference of DSS at 0 ppm (sodium 2,2-dimethyl-2silapentane-5-sulfonate, Cambridge Isotopes). Approximately 500 µL of mixture was added to the 5-mm tube (Wilmad Labglass). The samples were allowed to thermally equilibrate for at least 15 min prior to analysis. 1 H PFG NMR experiments were carried out with static samples, using the bipolar pulse pair stimulated-echo pulse sequence (BPPSTE).46,54 The 90° pulse length was typically 7 to 9 µs. The gradient strength (g) was varied from 1 G/cm to 25, 30, or 32 G/cm. The bipolar pulse gradient duration (δ) was 2 ms (half-sine shape, 1 ms duration of individual pulses), the gradient recovery delay (τ) was 300 µs, and the diffusion period (∆) was varied from 200 ms to 1s in order to attenuate the signal intensity to approximately 10% of the value obtained at 1 G/cm. At least 16 transients were collected for each

PFG NMR of Tetraalkylammonium Silica Mixtures

J. Phys. Chem. C, Vol. 114, No. 18, 2010 8451

increment step with a relaxation delay of 4 s. The field gradient strength was calibrated by measuring the value of self-diffusivity (Ds) of 10 mol % 1H2O in D2O (99.96% D atom, Cambridge Isotopes), Ds ) (1.92 ( 0.06) × 10-10 m2/s at 25 °C.55 For the BPP-STE pulse sequence the resonance intensity (I) is related to the self-diffusion coefficient, Ds, by

I δ τ ) exp -Dsγ2g2δ2 ∆ - Io 3 2

[

(

)]

(1)

where the quantity γδg is often referred to as K. From a suitable plot of the intensity versus K2 one can determine the self-diffusivity. The 1D PFG NMR spectra can be further transformed into a 2D DOSY spectrum with chemical shifts on one axis and Gaussian distribution of self-diffusivity on the other. Details of the DOSY processing algorithm are well documented.46,47,56,57 In this work, DOSY processing was used to resolve the overlapping resonances and was performed with the Varian VNMR, VnmrJ operating system. The sweep width of the attenuated stack spectra was manually adjusted to approximately 3-4 kHz for data storage purposes. In the 1H dimension, the free induction decays (FIDs) were zero filled to 32 767 data points and processed with a decaying exponential apodization function equivalent to 0.2 Hz line broadening. A total of 128 complex increments were used in the diffusion dimension (plot of peaks corresponding to their diffusion coefficient values). Peak heights were utilized as signal intensity (I) in the analyses. Spin-spin relaxation (T2) measurements were performed with the Carr-Purcell-Meiboom-Gill (CPMG) pulse sequence, [90°x-τ-(180°y-2τ)n-Acq], which utilizes a 180° pulse train to attenuate signals from relaxing species.42,58,59 The T2 delay (τ) is set to 1 ms and a half-echo was recorded every 4nτ. The relaxation delay was 20 s and the number of scans per sample was 4 or 16. When the experimental conditions were such that the signal-to-noise was poor (e.g., samples with small TAA cations concentration, long diffusion delay time (∆), long pulse train length (4nτ)), the spectra were carefully examined to include only clearly identifiable peaks. Dynamic Light Scattering (DLS) Experiments. The DLS experiments were performed with a BIC ZetaPALS with a BI9000AT correlator. The wavelength of the incident laser beam (λ) was 660 nm and the detector angle (θ) was 90°. To eliminate any dust, the tested samples were filtered by using a 0.2 µm PES syringe filter (Corning Co.) prior to loading into the cuvette (VWR). For each sample, three measurements were performed and the elapsed time was 5 min to ensure good signal-to-noise. The sampling and analysis were carried out in the self-beating mode. The delay time increased from 2 µs to 20 ms and the measurement temperature was 25 °C. The intensity autocorrelation functions were analyzed with the non-negative constrained least-squares method (NNLS).60,61 NNLS fitting yields a particle size (d)/translational diffusion (Dt) distribution for polydisperse solutions from the following equation and the largest population is set to 100%:

Dt )

1 τq2

(2)

where q is the scattering vector and τ is the relaxation time. The refractive index (n) of the solution was taken to be 1.33. Given the level of dilution in these mixtures (approximately 1.3 vol % silica in the C4 mixture) particle-particle interactions can be reasonably neglected. Thus to a reasonable first approximation the translational diffusion coefficient measured is approximately equal to the self-diffusion coefficient (Ds), which

Figure 1. Hydroxide concentration as a function of the total silica content in solution. The dashed line denotes the approximate value of the critical aggregation concentration (cac).

can be related to the particle hydrodynamic radius (Rh) via the Stokes-Einstein62 equation

Ds )

kT 6πηRh

(3)

where k is the Boltzmann constant, T is the temperature in Kelvin, and η is the viscosity of the solution, which is taken to be that of deuterated water (1.097 cP at 298 K).63 Results and Discussion Figure 1 shows the pH values for several TAAOH/silica mixtures as a function of total silica concentration. In general the results shown are in good agreement with previous work by Fedeyko and co-workers.32,33 When the silica concentration is below the critical aggregation concentration (cac) the solution pH decreases sharply upon increasing the silica concentration. After the cac point, there is only a very small decrease in the pH with increasing silica concentration. Previous work in the literature has described this behavior in detail23,32,38;the result in Figure 1 is included here as baseline information, and also for comparison to the NMR relaxation and diffusion measurements described below. 1 H NMR spectra for 0.25TMA/C0-C4 and 0.5TPA/C0-C4 mixtures are shown in Figure 2. The ethanol resonances due to TEOS hydrolysis are at chemical shift values of 3.63 and 1.17 ppm and neither the line positions nor line widths change appreciably with increasing silica content. However, the chemical shift of TMABr, originally at 3.17 ppm in 0.25TMA/C0 sample, i.e., in the absence of silica, moves progressively downfield as the silica content is increased, to a maximum shift of 3.24 ppm in the 0.25TMA/C4 sample. The TMA cation resonances also broaden with increasing silica content. A simple interpretation of these results is that there are interactions between the TMA cations and the silica in solution. The resonances assigned to TPABr in the C0 solution in the absence of silica are at 3.11, 1.68, and 0.92 ppm. The lines gradually lose their fine structure (i.e., J coupling) and their line widths increase monotonically upon increasing silica concentration. Additionally, the resonance of the methyl group of TPA shifts approximately 40 Hz downfield to approximately

8452

J. Phys. Chem. C, Vol. 114, No. 18, 2010

Li and Shantz

Figure 2. (A) 1H NMR spectra of 0.25TMA/C0-C4 (a-e) and (B) 1H NMR spectra of 0.5TPA/C0-C3 (a-d) and (e) 1.0TPA/C4.

Figure 3. Spin-spin relaxation time T2 versus silica content for (a) TMA-silica and (b) TPA-silica mixtures. Note that in the C4 mixture for 1.0 TPA, not 0.5 equiv of TPABr were used. And for the sensitivity limit, T2 values of the methyl group of TPA cation are reported here. The lines are shown as a guide to the eye.

1.0 ppm in the C4 mixture with the highest silica content, whereas no chemical shift change of the ethylene groups was detected. The chemical shift change observed for the methyl groups of the TPA is similar to that observed for TMA. Again, a simple explanation for the line shifts and broadening observed is that there are interactions between the silica in solution and the organocations. The relative intensity ratios of the TPA and ethanol resonances are summarized in the Supporting Information, and indicate nearly all the TPA is observable. To investigate this further, transverse relaxation measurements (T2) were performed on the same mixtures shown in Figure 2, which are shown in Figure 3. The most interesting finding is that these plots show a very similar behavior to the pH data shown in Figure 1. Namely there is a systematic decrease in the value of T2 until the silica cac is reached, at which point the value of T2 plateaus. The results in Figure 3 are consistent with the idea that the silica-organocation interactions are the origin of the chemical shift changes and line broadening observed in Figure 2. The fact that the value of the relaxation time is essentially constant after the cac seems to imply that increasing the amount of

nanoparticles does not lead to changes in the cation-silica interaction. In other words, upon crossing the cac adding more silica leads to more nanoparticles until gelation occurs. On the basis of the results above the addition of silica to these mixtures leads to a perturbation of the environment of the organocation as manifested by the movement of certain resonances and the line broadening observed, which is reflected in the decreasing values of T2. To study this in more detail PFG NMR was used to investigate these mixtures. Figure 4a shows the measured organocation diffusion coefficient for TAAOH/ silica mixtures as a function of the silica content in the mixture. For all organocations it is clearly observed that the self-diffusion coefficient decreases with increasing silica content. The values of the self-diffusion coefficient at the lowest silica content are quite similar to those of the value of the TAAOH in water. Table 2 summarizes the self-diffusion coefficients of the TAAOH in water at the same concentration as the results shown in Figure 4. As shown in Figure 4b, the observed diffusion coefficient (Dobs) of TPAOH in the two kinds of solutions decreases linearly with addition of more TEOS and ethanol. However, the value of Dobs in each TEOS hydrolyzed solution

PFG NMR of Tetraalkylammonium Silica Mixtures

J. Phys. Chem. C, Vol. 114, No. 18, 2010 8453

Figure 4. (a) Self-diffusion coefficient for a series of TAAOH-silica mixtures with constant TAAOH content but varying silica content. (b) Self-diffusion coefficient of TPAOH in ethanol/water (solid circles) and TEOS/water (open circles) mixtures. The dashed line denotes the approximate value of the critical aggregation concentration (cac).

is smaller than that in the corresponding water/ethanol mixture, and the discrepancy between the two values becomes larger with increasing silica contents. A simple interpretation of the data in Figure 4 is that the cation-silica interactions lead to a reduction in the observed diffusion coefficient. Our attempts to describe this in more detail have focused on a two-state model, wherein the organocation can be described as either “free” in solution or “bound” to the nanoparticle. This simple physical model is chosen for several reasons. First, small molecule binding in macromolecular systems has been successfully described with this approach. The main points of this methodology have been explained elsewhere.40,42,43,64,65 Second, this appears to be a reasonable physical model for physical adsorption of the organocation on the nanoparticle surface and is consistent with existing models of these systems which report these to be organocation-silica core-shell materials. Given the potentially confounding effects of chemical exchange rate on diffusion experiments,66 we performed PFG NMR experiments on a series of TAA-silica mixtures with various TAA concentration for diffusion times (∆) ranging from 0.05 to 1.0 s (∆ < T1) (Supporting Information). These experiments show that the diffusion coefficients measured as a function of ∆ are essentially independent of diffusion (no attenuation above 5-10%). This and the observation of single exponential intensity decay behavior are both consistent with the cation being in the fast exchange limit. Thus, an averaged diffusion coefficient between the value of free and bound cations is observed in practice and given by

Dobs ) fbDb + ffDf

(4)

where the subscripts b and f denote the bound and free species, respectively. The concentration of bound cation can then be determined by

[TAA+]bound ) fb[TAA+]total

(5)

DLS was employed to estimate the diffusion coefficient of the bound state, as it should to a first approximation be equal to the diffusion coefficient of the silica nanoparticle. These experiments should provide a more quantitative analysis of the binding equilibrium between the organocations and silica

TABLE 2: Self-Diffusion Coefficients (Ds)a of TAAOH in Water Solution Determined by 1H PFG NMR [TAAOH] (mol/kg)

Ds (TMAOH)

Ds (TEAOH)

Ds (TPAOH)

0.047

9.54 ( 0.07

6.64 ( 0.03

4.94 ( 0.01

a

Diffusion coefficients are expressed in units of 10 measured at 298 K.

Ds (TBAOH) 3.90 ( 0.02 -10

m2/s and

nanoparticles. The remainder of the work will focus on TMA-silica and TPA-silica solutions. This is motivated by TPA cations being widely used as an SDA for the synthesis of silicalite-1 from clear solutions, and TMA is a much more hydrophilic cation, incapable of making zeolites under these conditions (i.e., pure silica mixtures, low-temperature heating). Figure 5 shows the diffusion coefficient measured versus the organocation concentration for a variety of TMA-silica and TPA-silica mixtures, i.e. TAA/C0-C4 (mixture compositions are shown in Table 1). Three key observations can be made. First, measurements of TMA/C0 and TPA/C0 water solutions show that the self-diffusion coefficient is nearly constant over the range of cation concentrations investigated. This allows us to conclude that TAA aggregation under these conditions is not significant. Second, for C1 solutions only with dissolved silica species (the silica concentration is below the cac point), the observed diffusion coefficients of TMA and TPA are essentially constant with addition of cations and only show a minor variation from the values in the C0 water mixture. Third, for mixtures with silica concentrations above the cac, i.e., C2-C4 solutions, the observed diffusion coefficients (Dobs) increase with increasing organocation. Particularly, at high cation concentrations the measured diffusion coefficients plateau near the value for the cation in water (i.e., no silica) in all mixtures. It is also clear that the diffusion coefficients are the lowest in the dilute cation limit, i.e., when total [TAA] is less than 5 mM. As an example, Dobs for TPABr in the 0.75TPA/C4 solution (1.00 × 10-10 m2/s) reduces to 20% of its value at a similar concentration of C0 solution in the absence of silica (4.69 × 10-10 m2/s). As another example there is a 10-fold decrease in Dobs of TMABr in going from 9.59 × 10-10 m2/s in the 0.25TMA/C0 water solution to 0.62 × 10-10 m2/s in the 0.25TMA/C4 mixture. However, the

8454

J. Phys. Chem. C, Vol. 114, No. 18, 2010

Li and Shantz

Figure 5. Observed diffusion coefficient of (a) TMA and (b) TPA as a function of cation concentration for several mixtures, which are indicated in the legend.

observed diffusion value decreases slightly from that of the corresponding water mixture in the high TAA concentration limit. For example, Dobs of 36TMA/C4 is 6.13 × 10-10 m2/s and that of 36TPA/C4 is 3.12 × 10-10 m2/s. Clearly, the reduction of the observed self-diffusion coefficient in C2-C4 silica solutions compared to its value in C0 water solution varies with the ratio of [TAA]total/[SiO2]total, i.e., bound cation fraction of TAA silica mixture. The results could be attributed to changes of the solution viscosity with varying silica content. However, we performed viscosity measurements of C1-C4 mixtures at 25 °C (Supporting Information) and the viscosity variation of the silica mixtures is negligible compared with the observed diffusion coefficient change of cations in their dilute limit. The significant reduction of Dobs is thus mainly due to the binding effect. In the limit where the added [TAA] , [SiO2], i.e., fb is very close to 1, a dramatic decrease of the Dobs is observed at low cation concentrations (1-2 mM) for both TMA and TPA, shown in Figure 5. One might anticipate that the bound state dominates and that the observed diffusion coefficient would approximate the value of Db. Thus, an averaged diffusion value biased most toward Db is obtained for the 0.25TMA/C4 sample with a minimum [TAA]total/[SiO2]total ratio, i.e, maximum bound cation. By contrast in the limit of excess cation one would expect the free states to dominate, i.e., fb decreases, and the diffusion coefficient trends to the value observed in the absence of silica. This is in fact observed for 36TAA/C2-C4 (which have the most TAA). Additionally, the observed diffusion coefficient systematically increases and then ultimately plateaus. The trends appear at least qualitatively consistent with monolayer-type adsorption, which would be consistent with the core-shell model proposed previously by others, and we believe justifies our description of this system as a two-state problem.32,34 Two important parameters needed in the two-site model are the diffusion coefficients of the free and the bound state. To estimate the diffusion coefficient of the free state, mixtures were prepared wherein the TEOS was replaced with 4 equiv of ethanol (i.e., yTAABr/4xEtOH/9500H2O) to generate mixtures identical with those in Figure 5, except that they contain no silica. PFG NMR results of those mixtures show a modest (and non-negligible) decrease in the self-diffusivity of the cations, but which is not nearly as significant as compared to the samples

TABLE 3: Comparison of Self-Diffusion Coefficients Determined by NMR and DLS Experiments for Silicate Solutions DLSa [SiO2]total b solution (mol/kg) Ds d (nm) population (%) C2 C3 C4

0.105 0.210 0.421

0.68 0.63 0.63

5.77 6.32 6.26

62 100 100

NMRa Dobsc

fb

2.83 ( 0.04 0.73 0.68 ( 0.04 0.99 0.62 ( 0.01 0.99

a All diffusion coefficients are expressed in units of 10-10 m2/s and measured at 298 K. b Ds of silica nanoparticles with a hydrodynamic diameter of ∼6 nm. c Observed diffusion coefficient (Dobs) of 0.25TMA/C2-C4 samples.

in the presence of TEOS. Thus these values were used as the free diffusion coefficient for C2-C4 solutions. For the bound state, if one assumes that the silica nanoparticle bound-TAA cations diffuse at the same rate as the nanoparticles the diffusion coefficient can be estimated by DLS. There are two main reasons to use Dt rather than Dobs extrapolated to fb ≈ 1 obtained by NMR. First, Dobs measured by PFG NMR is weighted toward smaller, more rapidly diffusing molecules as their signals attenuate faster. Therefore, Db derived from the NMR method could potentially be biased, and larger than the true value of bound TAA cations, especially for samples at low nanoparticle concentration, i.e., C2 solutions. DLS also provides direct detection of the particles. Table 3 shows the results obtained from the DLS and the autocorrelation functions and corresponding particle size distributions are included in the Supporting Information. The DLS results indicate that the bulk of the particles are in the 3-5 nm size range and a small amount of silica aggregates is observed, consistent with previous work by others.3,5,11,67,68 Table 3 also shows a comparison between Dt of the particles measured by DLS and Dobs of TMA silica mixtures with fb ≈ 1 measured by NMR. It is noteworthy that the agreement between NMR and DLS improves as one goes to mixtures of increasing total silica content, which should have higher fractions of bound cations. With the results above in hand it is now possible to determine the fraction of bound cation for the various solutions. Figure 6 shows the fraction of TAA bound as a function of the total cation concentration. The lowest concentration mixtures were chosen based on the detection limit of the NMR measurements. The

PFG NMR of Tetraalkylammonium Silica Mixtures

J. Phys. Chem. C, Vol. 114, No. 18, 2010 8455

Figure 6. Fraction of bound (fb) (a) TMA and (b) TPA as a function of cation concentration for several TAA-silica mixtures, which are indicated in the legend.

Figure 7. Binding isotherms for (a) TMA-silica mixtures and (b) TPA-silica mixtures, which are indicated in the legend.

high concentration mixtures were chosen based on the observed diffusion coefficient becoming insensitive to changes in concentration. In Figure 6, the bound fraction of both TMA and TPA decreases rapidly upon addition of more cations at first and gradually converges to a constant value in the high cation concentration region. The data in Figure 6 can be replotted as the concentration of bound cation versus total solution concentration (i.e., binding isotherms, Figure 7). The behavior observed in Figure 7 is consistent with classic monolayer (Langmuir) behavior. The strongest cation binding occurs at low concentrations of TAABr and is characterized by a sharp rise in the amount of TAA adsorbed. Following this sharp increase in the amount of TAA bound the value plateaus as the total amount of TAA further increased above ∼31.5 mM/kg for TMA and 47 mM/ kg for TPA. Also noteworthy is that the maximum amount of TAA bound appears to scale with the silica concentration, implying the amount bound scales with the number of nanoparticles. If one is willing to assume that the nanoparticle surface is uniform and TAA cations adsorbed onto the surface form a monolayer coverage the Langmuir model can be used to describe

the behavior observed in Figure 7. This is consistent with previous literature, which indicates that TAA cations form aggregates in water at molarities of approximately 1 mol/L,69 which is well above the concentrations used in our studies (1.45 mM/L-0.21 mol/L). For the dilute solutions of this study, the general equation for a Langmuir-type adsorption is similar to that for gas adsorption, expressed as70,71

Cf Cf 1 ) + m m Xb KadXb Xb

(6)

where Xb is the molar ratio of bound [TAA]/[SiO2]nanoparticle (the concentration of silica in the form of nanoparticles) and Xbm is the ratio at adsorption saturation. Kad is the adsorption equilibrium constant, and Cf is the concentration of free TAA cations in solution at equilibrium. From a plot of (Cf)/(Xb) versus Cf one can obtain Xbm from the slope and Kad from intercept. The adsorption free energy ∆Goad can then be calculated72,73

∆G°ad ) -RT ln Kad

(7)

To fully analyze the data and calculate total adsorbed amount at saturation [TAA]bm (mol/kg), we need an estimate of

8456

J. Phys. Chem. C, Vol. 114, No. 18, 2010

Li and Shantz

Figure 8. Data (circles) and fit (solid line) for TMA cations/silica solutions as described by the Langmuir isotherm model for (a)TMA/C2, (b)TMA/ C3, and (c) TMA/C4.

Figure 9. Data (circles) and fit (solid line) for TPA cations/silica solutions as described by the Langmuir isotherm model for (a) TPA/C2, (b) TPA/C3, and (c) TPA/C4.

TABLE 4: Langmuir Constants and Derived Thermodynamic Parameters Obtained for Adsorbing TAA+ on Silica Nanoparticles in Dilute Clear Solutions TMAb

TPAc

solution

[SiO2]nanoparticlea (mol/kg)

Xbm

[TMA]m b (mol/kg)

Kad (kg/mol)

∆G°ad (kg/mol)

Xbm

[TPA]m b (mol/kg)

Kad (kg/mol)

∆G°ad (kg/mol)

C2 C3 C4

0.063 0.169 0.379

0.072 0.069 0.064

0.0045 0.0117 0.0243

202.27 295.62 284.35

–13.15 –14.09 –14.00

0.062 0.055 0.040

0.0039 0.0092 0.0152

69.81 95.51 148.64

–10.52 –11.29 –12.39

a

Calculated with the value of [SiO2]cac (0.042 mol/kg) obtained from the pH measurement of sodium silicate solutions, using eqs 8 and 9. c m For TMA/C2-C4, the average uncertainty in Xm b and [TMA]b is 3.33%, and that in Kad and ∆G°ad is 25.41%. For TPA/C2-C4, the average + m uncertainty in Xm b and [TMA ]b is 2.96%, and that in Kad and ∆G°ad is 17.31%

b

[SiO2]nanoparticle. On the basis of previous work we assume the dominant silica species are the dissolved monomers and nanoparticles.23,24,74,75 Using the work of Rimer and co-workers34 we can described the composition of silica species as:

[SiO2]tot ) [SiO2]monomer + [SiO2]nanoparticles

(8)

Also from these previous works we can estimate the monomer concentration32-34 from the cac point

[SiO2]cac ) 24.147[OH-] + 0.016

(9)

Figure 8 shows the results of these fits for the various TMA-silica solutions and Figure 9 shows the fits for the TPA-silica solutions. R2 values of all fits are between 0.969 (TMA/C2) and 0.998 (TPA/C4), with P values less than 0.0001, showing that the experimental results fit the Langmuir adsorption model well. The deviations from the regression fits consistently appear at the lowest solution concentrations of TMA and TPA. A simple interpretation of this result is that in the dilute solution

limit nonhomogeneity of the surface can be detected, or alternatively that the first cations adsorbed bind at the strongest binding sites. Table 4 displays the Langmuir constants and derived thermodynamic parameters for TMA and TPA in silicate solutions with different silica content. All ∆G°ad values in Table 4 vary between -10 and -14 kJ/ mol, indicating a nonspecific adsorption of silica particles to TAA cations. Comparing the ∆G°ad and Kad results shown in Table 4, TMA displays larger values than TPA with approximately the same silica nanoparticle concentration, suggesting a stronger binding strength between TMA and silica particles. Another notable finding is that more TMA adsorbs on the nanoparticles than TPA for monolayer coverage. Considering the difference in surface charge density of organic cations, TMA may have stronger electrostatic interactions with the nanoparticles than TPA. The other notable difference between the two cations is that TPA is more hydrophobic; however, work below, we believe, points to the differences

PFG NMR of Tetraalkylammonium Silica Mixtures

J. Phys. Chem. C, Vol. 114, No. 18, 2010 8457

Figure 10. Schematic view of TAA cations adsorption on silica nanoparticles in solution.

TABLE 5: Monolayer Coverage of TAA Cations Per Particle of Different Silica Mixtures TMA cations solution

[SiO2]total (mol/kg)

no. of TAA/ particlea

SAb (nm2)

C2 C3 C4

0.105 0.210 0.421

26 25 23

1.64 1.71 1.84

TPAcations no. of TAA/ particlea 22 20 14

a Number of TAA cations adsorbed per particle. surface area occupied by a TAA cation.

SAb (nm2) 1.90 2.14 2.95

b

The particle

observed being due primarily to electrostatic effects. In addition, compared to TMA cations (∼3.4 Å), the larger ionic radius of TPA cations (∼4.5 Å)76 could hinder its diffusion through the double-layer region to the silica particles surface and results in lower adsorption loading. Another, potentially simpler explanation is that since the TMA is smaller more of it can be accommodated on a given unit surface than TPA. The adsorption process of TAA cations on silica particles is schematically summarized in Figure 10. The number of TAA cations adsorbed per particle for monolayer coverage can be estimated based on data in Table 4 and values reported by Fedeyko and Rimer et al.33,36 The nanoparticle surface area is estimated to be 42 nm2 with a particle core containing 356 silicon atoms. The calculated results are shown in Table 5. More TMA cations adsorb per particle than TPA, consistent with the smaller molecular size of TMA. The particle surface area (SA) covered by a TMA cation and a TPA cation are larger than the approximate area occupied by them, as reported by Claesson et al. of 0.54 nm2 for TMA and 0.77 nm2 for a TPA, respectively.77 Hence, the bound TMA and TPA cations are not likely closely packed on the particle surface. Also the number of cations per particle for C4 shows a slight decrease compared with that of C2 and C3; however, the change is not as significant as the increase of particle concentration from C2 to C4. This might be due to the larger SiO2 unit per particle for C4 than the constant value assumed. Also the simple calculation does not take into account the real surface properties of the particles due to the lack of that knowledge. Conclusions The interactions between TAA cations and silica nanoparticles in a series of clear solutions have been investigated by PFG NMR. The results confirm that binding of TAA cations to silica is a general phenomenon in these mixtures. The binding isotherm of TMA cations and TPA cations in silica mixtures demonstrated a nonlinear behavior and can be well described by the Langmuir adsorption model. From this model, we estimated the monolayer coverage of TAA cations on silica nanoparticles, as well as adsorption equilibrium constants (Kad) and adsorption free energy (∆G°ad) for different TAA-silica mixtures. The results indicate that TMA has a stronger adsorption strength and larger adsorption loading on the silica particles than TPA cations. A sharp decrease of T2 values of organocations in the presence of silica implies that rotational motion of

TAA is significantly reduced mainly due to binding. 1H NMR spectral line broadening and the frequency shift of TAA cations with increasing silica content are consistent with the change of T2. Taken together, the findings in this study have important implications for understanding the organic-inorganic interaction and further, the role of TAA-silica particles in zeolite nucleation. Ongoing work is investigating how these properties change in the presence of electrolyte and heating and will be reported elsewhere. Acknowledgment. The authors acknowledge financial support from the National Science Foundation (CHE-0646052). The authors also acknowledge Mr. Steve Silber and Dr. K. P. Sarathy of the NMR facility in the Department of Chemistry for generous technical and organizational support. Supporting Information Available: Integrated intensities of ethanol and TPA in selected mixtures, T1 and T2 data for several TMA and TPA mixtures, measured diffusion coefficient as a function of the diffusion time (∆) for several mixtures, measured diffusion coefficients for TMA and TPA solutions as a function of organocation concentration, measured diffusion coefficients for TMA and TPA in water/cation/ethanol mixtures, and selected autocorrelation functions (DLS) for several mixtures. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Davis, T. M.; Drews, T. O.; Ramanan, H.; He, C.; Dong, J.; Schnablegger, H.; Katsoulakis, M. A.; Kokkoli, E.; McCormick, A. V.; Penn, R. L.; Tsapatsis, M. Nat. Mater. 2006, 5, 400–408. (2) Mintova, S.; Olson, N. H.; Senker, J.; Bein, T. Tetrahedron Lett. 1994, 35, 1003–1006. (3) Nikolakis, V.; Kokkoli, E.; Tirrell, M.; Tsapatsis, M.; Vlachos, D. G. Chem. Mater. 2000, 12, 845–853. (4) Persson, A. E.; Schoeman, B. J.; Sterte, J.; Ottesstedt, J. E. Zeolites 1994, 14, 557–567. (5) Schoeman, B. J. Microporous Mater. 1997, 9, 267–271. (6) Schoeman, B. J. Microporous Mesoporous Mater. 1998, 22, 9–22. (7) Schoeman, B. J.; Regev, O. Zeolites 1996, 17, 447–456. (8) de Moor, P. P. E. A.; Beelen, T. P. M.; van Santen, R. A. J. Phys. Chem. B 1999, 103, 1639–1650. (9) Ravishankar, R.; Kirschhock, C. E. A.; Knops-Gerrits, P. P.; Feijen, E. J. P.; Grobet, P. J.; Vanoppen, P.; De Schryver, F. C.; Miehe, G.; Fuess, H.; Schoeman, B. J.; Jacobs, P. A.; Martens, J. A. J. Phys. Chem. B 1999, 103, 4960–4964. (10) Kragten, D. D.; Fedeyko, J. M.; Sawant, K. R.; Rimer, J. D.; Vlachos, D. G.; Lobo, R. F.; Tsapatsis, M. J. Phys. Chem. B 2003, 107, 10006–10016. (11) Yang, S.; Navrotsky, A.; Wesolowski, D.; Pople, J. Chem. Mater. 2004, 16, 210–219. (12) Cheng, C.-H.; Shantz, D. F. J. Phys. Chem. B 2005, 109, 7266– 7274. (13) Hould, N. D.; Lobo, R. F. Chem. Mater. 2008, 20, 5807–5815. (14) Cheng, C.-H.; Shantz, D. F. J. Phys. Chem. B 2005, 109, 13912– 13920. (15) Cheng, C.-H.; Shantz, D. F. J. Phys. Chem. B 2005, 109, 19116– 19125. (16) Kirschhock, C. E. A.; Ravishankar, R.; Van Looveren, L.; Jacobs, P. A.; Martens, J. A. J. Phys. Chem. B 1999, 103, 4972–4978. (17) Schoeman, B. J.; Sterte, J.; Otterstedt, J. E. Zeolites 1994, 14, 568– 575. (18) Watson, J. N.; Iton, L. E.; Keir, R. I.; Thomas, J. C.; Dowling, T. L.; White, J. W. J. Phys. Chem. B 1997, 101, 10094–10104. (19) Aerts, A.; Follens, L. R. A.; Haouas, M.; Caremans, T. P.; Delsuc, M. A.; Loppinet, B.; Vermant, J.; Goderis, B.; Taulelle, F.; Martens, J. A.; Kirschhock, C. E. A. Chem. Mater. 2007, 19, 3448–3454. (20) Kirschhock, C. E. A.; Buschmann, V.; Kremer, S.; Ravishankar, R.; Houssin, C. J. Y.; Mojet, B. L.; van Santen, R. A.; Grobet, P. J.; Jacobs, P. A.; Martens, J. A. Angew. Chem., Int. Ed. 2001, 40, 2637–2640. (21) Ramanan, H.; Kokkoli, E.; Tsapatsis, M. Angew. Chem., Int. Ed. 2004, 43, 4558–4561. (22) Kumar, S.; Wang, Z.; Penn, R. L.; Tsapatsis, M. J. Am. Chem. Soc. 2008, 130, 17284–17286.

8458

J. Phys. Chem. C, Vol. 114, No. 18, 2010

(23) Kinrade, S. D.; Knight, C. T. G.; Pole, D. L.; Syvitski, R. T. Inorg. Chem. 1998, 37, 4272–4277. (24) Knight, C. T. G.; Balec, R. J.; Kinrade, S. D. Angew. Chem., Int. Ed. 2007, 46, 8148–8152. (25) Knight, C. T. G.; Syvitski, R. T.; Kinrade, S. D. Stud. Surf. Sci. Catal. 1995, 97, 483–488. (26) Kirschhock, C. E. A.; Ravishankar, R.; Verspeurt, F.; Grobet, P. J.; Jacobs, P. A.; Martens, J. A. J. Phys. Chem. B 1999, 103, 4965–4971. (27) Cheng, C.-H.; Shantz, D. F. J. Phys. Chem B 2006, 110, 313–318. (28) Fyfe, C. A.; Darton, R. J.; Schneider, C.; Scheffler, F. J. Phys. Chem. C 2008, 112, 80–88. (29) Pelster, S. A.; Schrader, W.; Schuth, F. J. Am. Chem. Soc. 2006, 128, 4310–4317. (30) Pelster, S. A.; Weimann, B.; Schaack, B. B.; Schrader, W.; Schuth, F. Angew. Chem., Int. Ed. 2007, 46, 6674–6677. (31) Nikolakis, V.; Tsapatsis, M.; Vlachos, D. G. Langmuir 2003, 19, 4619–4626. (32) Fedeyko, J. M.; Rimer, J. D.; Lobo, R. F.; Vlachos, D. G. J. Phys. Chem. B 2004, 108, 12271–12275. (33) Fedeyko, J. M.; Vlachos, D. G.; Lobo, R. F. Langmuir 2005, 21, 5197–5206. (34) Rimer, J. D.; Vlachos, D. G.; Lobo, R. F. J. Phys. Chem. B 2005, 109, 12762–12771. (35) Rimer, J. D.; Fedeyko, J.; Vlachos, D. G.; Lobo, R. F. Chem.sEur. J. 2006, 12, 2926–2934. (36) Rimer, J. D.; Lobo, R. F.; Vlachos, D. G. Langmuir 2005, 21, 8960– 8971. (37) Rimer, J. D.; Trofymluk, O.; Navrotsky, A.; Lobo, R. F.; Vlachos, D. G. Chem. Mater. 2007, 19, 4189–4197. (38) Jorge, M.; Auerbach, S. M.; Monson, P. A. J. Am. Chem. Soc. 2005, 127, 14388–14400. (39) Liu, M.; Nicholson, J. K.; Lindon, J. C. Anal. Commun. 1997, 34, 225–228. (40) Fielding, L. Tetrahedron 2000, 56, 6151–6170. (41) Lin, M.; Shapiro, M. J.; Wareing, J. R. J. Am. Chem. Soc. 1997, 119, 5249–5250. (42) Hajduk, P. J.; Olejniczak, E. T.; Fesik, S. W. J. Am. Chem. Soc. 1997, 119, 12257–12261. (43) Altieri, A. S.; Hinton, D. P.; Byrd, R. A. J. Am. Chem. Soc. 1995, 117, 7566–7567. (44) Jayawickrama, D. A.; Larive, C. K. Anal. Chem. 1999, 71, 2117– 2122. (45) Price, W. S. Concepts Magn. Reson. 1997, 9, 299–336. (46) Johnson, C. S. J. Prog. Nucl. Magn. Reson. Spectrosc. 1999, 34, 203–256. (47) Morris, K. F.; Johnson, C. S. J. J. Am. Chem. Soc. 1992, 114, 3139– 3141. (48) Morris, K. F.; Stilbs, P.; Johnson, C. S. J. Anal. Chem. 1994, 66, 211–215. (49) Morris, K. F.; Johnson, C. S. J. J. Am. Chem. Soc. 1993, 115, 4291– 4299. (50) Wassenius, H.; Lofroth, J.; Nyden, M. Starch/Starke 2006, 58, 66– 81.

Li and Shantz (51) Wong, A.; Ida, R.; Spindler, L.; Wu, G. J. Am. Chem. Soc. 2005, 127, 6990–6998. (52) Galantini, L.; Giampaolo, S. M.; Mannina, L.; Pavel, N. V.; Viel, S. J. Phys. Chem. B 2004, 108, 4799–4805. (53) Le Bon, C.; Nicolai, T.; Kuil, M. E.; Hollander, J. G. J. Phys. Chem. B 1999, 103, 10294–10299. (54) Lucas, L. H.; Otto, W. H.; Larive, C. K. J. Magn. Reson. 2002, 156, 138–145. (55) Mills, R. J. Phys. Chem. 1973, 77, 685–688. (56) Morris, K. F.; Cutak, B. J.; Dixon, A. M.; Larive, C. K. Anal. Chem. 1999, 71, 5315–5321. (57) Waldeck, A. R.; Kuchel, P. W.; Lennon, A. J.; Chapman, B. E. Prog. Nucl. Magn. Reson. Spectrosc. 1997, 30, 39–68. (58) Soderman, O.; Stilbs, P.; Price, W. S. Concepts Magn. Reson. A 2004, 23A, 121–135. (59) Flood, C.; Cosgrove, T.; Espidel, Y.; Welfare, E.; Howell, I.; Revell, P. Langmuir 2008, 24, 7875–7880. (60) Brown, W. Dynamic light scattering: the method and some applications; Clarendon Press: Oxford, UK, 1993. (61) Dahneke, B. E. Measurement of suspended particles by quasi-elastic light scattering; Wiley: New York, 1983. (62) Berne, B. J.; Pecora, R. Dynamic Light Scattering with Applications to Chemistry Biology, and Physics; Wiley: New York, 1976. (63) Lapham, J.; Rife, J. P.; Moore, P. B.; Crothers, D. M. J. Biomol. NMR 1997, 10, 255–262. (64) Chen, A.; Wu, D.; Johnson, C. S. J. J. Phys. Chem. 1995, 99, 828– 834. (65) Lin, M. F.; Larive, C. K. Anal. Biochem. 1995, 229, 214–220. (66) Choudhury, R. P.; Scho¨nhoff, M. J. Chem. Phys. 2007, 127, 234702. (67) Mintova, S.; Olson, N.; Senker, J.; Bein, T. Tetrahedron Lett. 1994, 35, 1003–1006. (68) Yang, S.; Navrotsky, A.; Wesolowski, D. J.; Pople, J. A. Chem. Mater. 2004, 16, 210–219. (69) Kinrade, S. D.; Knight, C. T. G.; Pole, D. L.; Syvitski, R. T. Inorg. Chem. 1998, 37, 4278–4283. (70) Thomas, W. J.; Crittenden, B. D. Adsorption technology and design; Butterworth-Heinemann: Boston, MA, 1998. (71) Choucair, A.; Eisenberg, A. J. Am. Chem. Soc. 2003, 125, 11993– 12000. (72) Wang, C.; Groenzin, H.; Shultz, M. J. J. Am. Chem. Soc. 2005, 127, 9736–9744. (73) Adamson, A. W.; Gast, A. P. Physical chemistry of surfaces; Wiley: New York, 1990. (74) Kinrade, S. D.; Schach, A. S.; Hamilton, R. J.; Knight, C. T. G. Chem. Commun. 2001, 1564–1565. (75) Knight, C. T. G.; Kinrade, S. D. J. Phys. Chem. B 2002, 106, 3329– 3332. (76) Zana, R.; Yeager, E. J. Phys. Chem. 1967, 71, 4241–4244. (77) Claesson, P.; Horn, R. G.; Pashley, R. M. J. Colloid Interface Sci. 1984, 100, 250–263.

JP101006F