Pulsed Field Gradient NMR Investigations of Alkyltripropylammonium

20178

J. Phys. Chem. C 2010, 114, 20178–20188

Pulsed Field Gradient NMR Investigations of Alkyltripropylammonium-Silica Mixtures Alejandra Rivas-Cardona and Daniel F. Shantz* Artie McFerrin Department of Chemical Engineering, Texas A&M UniVersity, 3122 TAMU, College Station, TX 77843-3122, United States ReceiVed: August 30, 2010; ReVised Manuscript ReceiVed: October 13, 2010

Pulsed field gradient nuclear magnetic resonance (PFG NMR) spectroscopy has been used to study the organocation-silica interactions in silica-alkyltripropylammonium solutions. Mixtures containing silica and methyltripropylammonium, ethyltripropylammonium, and tetrapropylammonium cations have been investigated to determine how the organocation structure and hydrophobicity affect organocation-silica interactions. PFG NMR results show that, in the presence of silica, the diffusion coefficients of the tetraalkylammonium organocations decreased compared with those of the cations in water alone. This decrease is attributed to a reduction in the cation mobility as the cations interact with the silica nanoparticles. The diffusion coefficients as a function of the organocation concentration were measured to determine the fraction of organocations bound to the silica through a model that considers the cation in either a free or a bound state. The resulting adsorption isotherms were fit with the Langmuir model, from which the free energy of binding was determined to be between 10 and 13 kJ/mol. This value appeared insensitive to the organocation identity and the silica content of the mixture, although the most hydrophilic cation had the highest binding energy. The results show that the organocation-silica interactions play an important role in zeolite formation and can help to develop a more general approach to making high-silica zeolites from optically transparent mixtures containing silica nanoparticles. Introduction Zeolites are microporous crystalline alumino-silicate materials widely used in catalysis, adsorption, separations, and ionexchange operations.1-3 To date, approximately 150 distinct zeolites have been identified, and this number will likely increase.4 High-silica zeolites (Si/Al > 10), which are usually synthesized in the presence of organocations,5,6 have found widespread use in catalysis7,8 and have been touted for many emerging areas, including thin films, sensors, and ion conductors.9 One of the most exciting ideas in the field of high-silica zeolite synthesis is that, at some level, the pore topology of the zeolite is affected by the molecular structure of the organocation used in the synthesis.10,11 Thus, to design zeolite structures, it is important to understand how (or even if), at a molecular level, the organocation is involved in zeolite nucleation and growth. The nucleation and growth of high-silica zeolites have been extensively studied, but an overarching description of the process is lacking. Several methods, mainly experimental, have been used to understand zeolite formation. In particular, the synthesis of silicalite-1 zeolite from optically transparent solutions of TAA+ (tetraalkylammonium cations), TEOS (tetraethyl orthosilicate), and water has been intensely studied with methods, including small-angle scattering and light scattering,12-31 NMR32-41 spectroscopy, calorimetry,42-47 and electron microscopy.26,48-50 These investigations have shown the existence of small (99.0%) and deuterium oxide (D2O, CIL, 99.9%) were used in the silicalite-1 synthesis. Phosphate pH buffer (Beckman Coulter, pH 7.00 ( 0.01 at 25 °C) and carbonate pH buffer (Beckman Coulter, pH 10.01 ( 0.01 at 25 °C) were used for the calibration of the pH meter. A

20180

J. Phys. Chem. C, Vol. 114, No. 47, 2010

Rivas-Cardona and Shantz

Figure 1. DOSY bipolar pulse pair stimulated echo (Dbppste) sequence.69

conductivity calibration solution (VWR, 718 micro-Mho ( 1 at 25 °C, 0.005 N KCl) was used to calibrate the conductivity meter. Synthesis of Structure-Directing Agent. Methyltripropylammonium iodide (MTPAI) and ethyltripropylammonium iodide (ETPAI) were prepared by reacting iodomethane and iodoethane, respectively, with tripropylamine (N(C3H7)3) by the following procedure: Tripropylamine ( 60 mL, 0.3 mol) was added to 200 mL of butanone in a round-bottom flask. The corresponding iodoalkane (70 mL or 1.1 mol of iodomethane and 75 mL or 0.9 mols of iodoethane) was placed in an addition funnel and then added dropwise to the tripropylamine mixture. The resulting solution was stirred under reflux for 24 h in the absence of light. After this time, the reacting solution was cooled; the solids were recovered by filtration, rinsed with ethyl acetate, and dried at room temperature. The typical yield was greater than 75% for MTPAI and greater than 90% for ETPAI. Each of these alkyltripropylammonium iodide salts was ionexchanged twice to the hydroxide form using a column with the anion-exchange resin. The resulting alkyltripropylammonium hydroxide (RN(C3H7)3+OH-) solution was titrated with hydrochloric acid using m-cresol purple and phenolphthalein as indicators. The exchange efficiency was always greater than 80% and usually greater than 90%. Silicalite-1 Synthesis: Precursor Solutions. Mixtures with a composition of x RN(C3H7)3+OH-:y TEOS:7290 H2O:810 D2O (where x ) 1-18 mol, y ) 0-60 mol, and R ) Me, Et, and Pr) were prepared as follows. The organocation was first diluted with deionized water and deuterium oxide (which was added to provide a lock signal), and the resulting solution was then mixed with TEOS. The organocation-TEOS-water mixture was placed in a screw-cap Teflon container and was aged for 24 h while mixing at room temperature to ensure full hydrolysis of TEOS. Conductivity and pH. The conductivity measurements were performed at 25 °C with an Amber Science (model 2052) conductivity meter and Pt cell (Amber Science Inc., 545 MultiPurpose Cell, 10.03 cm-1 cell constant). The conductivity meter was calibrated at 25 °C using calibration solution. pH measurements were performed with a Beckman Φ340 pH meter and a glass electrode (model 511052) at room temperature. The pH meter was calibrated with pH 7 and pH 10 buffer solutions at 25 °C. Dynamic Light Scattering. DLS measurements were performed using a Brookhaven ZetaPALS instrument. A laser with a wavelength of 660 nm was used as the incident beam. The scattered light was detected at a 90° scattering angle at 25 °C. The measurements were performed three times on each sample to check reproducibility. The light scattering data were analyzed with the BI-DLSW control software, and the non-negative constrained least-squares (NNLS) algorithm was used to deter-

mine the size distributions. Note that, for the samples here, the reagents were filtered, but the reaction mixtures were not, prior to analysis. NMR Spectroscopy. Single-pulse 1H, 1H relaxation (spinlattice and spin-spin), and 1H diffusion NMR measurements were performed on a Varian INOVA spectrometer operating at 500 MHz for 1H, equipped with a 5 mm indirect detection probe and a z-axis PFG coil that provides a maximum gradient strength of 30 G/cm. All the NMR experiments were carried out at 25 °C, using 600 µL of sample and without spinning. Single-pulse 1 H experiments were performed using 128 scans. The 90° pulse length was calibrated for each sample, and the values were between 6.9 and 7.6 µs. The spin-lattice relaxation time, T1, was determined by inversion recovery experiments [π - τ π/2 - acquisition]. The spin-spin relaxation time, T2, was measured using the Carr-Purcell-Melboom-Gill (CPMG) pulse sequence [π/2 - (τ - π - τ)n - acquisition]. For the CPMG measurements, τ ) 4 ms and the maximum number of counters used was 800. To measure the diffusion coefficients, diffusion ordered spectroscopy (DOSY) experiments were carried out with the DOSY bipolar pulse pair stimulated echo (Dbppste) sequence shown in Figure 1. This stimulated echo sequence has a [hsgt-π-hgst] homospoil block at the beginning. The sequence uses pairs of bipolar gradient pulses of duration δ and strength g, followed by a gradient stabilization delay time τ. The first pair of gradient pulses spatially encodes the nuclear spins of the probe molecules; the spins are allowed to evolve during the diffusion delay time ∆ after which the second pair of gradient pulses is applied to decode the nuclear spins. The gradient strength calibration was performed using a 10 mol % D2O in H2O sample with a self-diffusion coefficient of (2.26 ( 0.01) × 10-9 m2/s.68 The first delay d1 was 2 s; τ was 500 µs; ∆ was 50 ms, unless mentioned otherwise; δ was 4 ms; and g was varied from 3 to 19 G/cm using 40 gradient levels that were acquired with 64 scans per level and 4 steadystate transients at the start of each level. The NMR data was processed using Varian’s VNMRJ 2.2C software.70 To extract the diffusion data from the DOSY experiments, the baseline of all the spectra was corrected and the threshold was adjusted such that all signals of interest were above it. The line broadening was set to 1 Hz, and the FIDs were zero-filled. The dosy macro was then used. This macro determines the heights of all signals above the threshold and fits the decay curve for each signal to a Gaussian. A 2D spectrum is constructed with this information, displaying the chemical shift in the first dimension and the diffusion coefficients in the second. Results and Discussion Conductivity and pH (or Hydroxide Concentration). Figure 2 shows the conductivity and hydroxide concentration [OH-] of MTPAOH-TEOS-water and ETPAOH-TEOS-water

PFG NMR of Alkyltripropylammonium-Silica Mixtures

J. Phys. Chem. C, Vol. 114, No. 47, 2010 20181

Figure 2. Conductivity and hydroxide concentration as a function of the total concentration of alkyltripropylammonium hydroxide.

systems as a function of the total concentration of alkyltripropylammonium hydroxide. In general, the values of conductivity and hydroxide concentration are very similar for both organocations. This is expected as MTPA+ and ETPA+ are very similar in structure and have the same net charge. It can be noticed that conductivity and hydroxide concentration drop with the addition of TEOS and that, in the presence of silica, the conductivity and hydroxide concentration increase slowly with the total concentration of alkyltripropylammonium hydroxide. These results are in agreement with previous results presented by Fedeyko20 for the region above the critical aggregation concentration (cac) where uniform silica nanoparticles are observed in equilibrium with oligomeric species. The observed behavior was expected because, in this work, the amount of silica added to the samples with the lowest silica content (samples with y ) 20 mol of TEOS ) 0.13 mol/kg) is above the cac value (0.092 mol/kg) reported by Fedeyko20 for a similar sample with a composition of 18 TPAOH:9500 H2O. NMR Spectroscopy. Figure 3 shows the one-pulse 1H NMR spectra of MTPAOH-TEOS-water and ETPAOH-TEOSwater systems. In the absence of silica, the signal intensities of MTPA+ (Figure 3a) increase monotonically as the MTPAOH concentration increases. In these mixtures, no changes in either the chemical shift position or the line width are observed. However, in the presence of silica (Figure 3b,c), the MTPA+ resonances show broadening and a considerable shift of some NMR resonances, the details of which are summarized in Table 1. Figure 3b shows broad MTPA+ signals at low cation:silica ratios, which become narrower as the cation:silica ratio increases. Moreover, the triplet resonance corresponding to the methyl moiety (A) moves upfield as the cation:silica ratio increases. The ethanol resonances (G and H) coming from the hydrolysis of TEOS do not show broadening or systematic

changes in the chemical shift. That the methyl group chemical shift changes, but the other groups do not, is consistent with organic-particle interactions. The electronic environment of the methyl groups is perturbed the most by the particles, and this is reflected in the change in chemical shift values. Figure 3c shows that further addition of silica (y ) 60) produces signal broadening in such a way that the observation of MTPA+ peaks at low concentrations (x e 1.5) becomes difficult. This figure also shows an upfield shift of the methyl peak (A) with increasing cation:silica ratio. A similar trend is observed in the mixtures with ETPAOH. In the absence of silica, the chemical shift and line width of the ETPA+ signals (Figure 3d) do not change with the ETPAOH concentration. However, upon addition of silica (Figure 3e), the ETPA+ signals show line broadening as the cation:silica ratio decreases. The cation methyl resonances (A and F) shift upfield (see Table 1) as the ETPAOH concentration increases. Further addition of silica (Figure 3f) shows stronger line broadening, causing signal loss at x e 3. In this case, an upfield shift of both methyl resonances (A and F) is observed. The upfield shift of the methyl peak (A) of ETPA+ is slightly and consistently greater than the upfield shift of the analogous methyl peak of MTPA+. The change in chemical shift of the second methyl peak (F) from the lowest to the highest ETPAOH concentration cannot be determined due to low intensity and broadening of the signals. Changes in line shape (i.e., broadening) and chemical shift can be attributed to cation-silica interactions. Signal broadening can be related to a reduction in the mobility of the cations (or cation segments) experiencing an association with the silica particles. Thus, a possible cause for broadening of the resonances is the decrease in cation mobility. In contrast, upfield changes in chemical shift can be linked to a greater shielding likely due to weaker cation-silica interactions and, therefore, more mobile cations.

20182



Figure 3. 1H NMR spectra for mixtures with a composition of x MTPAOH:y TEOS:7290 H2O:810 D2O, where (a) y ) 0, (b) y ) 20, and (c) y ) 60, and for mixtures with a composition of x ETPAOH:y TEOS:7290 H2O:810 D2O, where (d) y ) 0, (e) y ) 20, and (f) y ) 60.

TABLE 1: Chemical Shift Changes of the Cation Methyl Resonances of Figure 2a composition

shifted peaks

δ (x ) 1), ppm

δ (x ) 18), ppm

∆δ, ppm (∆ν, Hz)

x MTPAOH:7290 H2O:810 D2O x MTPAOH:20 TEOS:7290 H2O:810 D2O

none A

0.87

0.84

x MTPAOH:60 TEOS:7290 H2O:810 D2O

A

0.87

0.84

0.03 (15) 0.03 (15)

x ETPAOH:7290 H2O:810 D2O x ETPAOH:20 TEOS:7290 H2O:810 D2O

none A

0.88

0.84

F

1.18

1.15

A

0.88

0.84

F

1.16a

1.14

x ETPAOH:60 TEOS:7290 H2O:810 D2O

a

Chemical shift value at x ) 9.

0.04 (20) 0.03 (15) 0.04 (20) 0.02 (10)



Figure 4. (a) Spin-lattice relaxation (T1) time for mixtures with a composition of x MTPAOH:y TEOS:7290 H2O:810 D2O. (b) Spin-spin relaxation (T2) time for mixtures with a composition of x MTPAOH:y TEOS:7290 H2O:810 D2O. (c) Spin-lattice relaxation (T1) time for mixtures with a composition of x ETPAOH:y TEOS:7290 H2O:810 D2O. (d) Spin-spin relaxation (T2) time for mixtures with a composition of x ETPAOH:y TEOS:7290 H2O:810 D2O.

The cation relaxation rates (T1 and T2) are also important given that these both are an indicator of molecular motion and affect the diffusion NMR measurements. Therefore, the relationship between the relaxation times and the total concentration of alkyltripropylammonium hydroxide was investigated. Figure 4 shows the spin-lattice (T1) and spin-spin relaxation (T2) times of MTPAOH-TEOS-water and ETPAOH-TEOSwater systems for the methyl group D of the MTPA+ and of the methyl group F of the ETPA+, respectively. The error bars represent the error between the experimental data and the fitting to its respective relaxation function. Figure 4a shows that the spin-lattice relaxation times do not vary much for MTPAOHwater mixtures over the concentration range investigated but decrease with the addition of silica, particularly, at low MTPAOH concentrations. Figure 4b shows that the spin-spin relaxation time of MTPA+ in water does not vary strongly over the concentration range investigated, but a drop in the T2 values occurs if silica is added to the mixtures. Furthermore, in the presence of silica, T2 becomes dependent on the MTPAOH concentration and its value drops as the MTPAOH decreases. Similar trends are observed for ETPAOH-TEOS-water mixtures. The spin-lattice (Figure 4c) and spin-spin relaxation (Figure 4d) times are practically independent of the ETPAOH concentration for ETPAOH-water systems. However, the relaxation times decrease in the presence of silica; in particular, the spin-spin relaxation time decreases significantly. Relaxation times could not be determined at low concentrations of alkyltripropylammonium hydroxide (either MTPAOH or ETPAOH) and high concentrations of silica because of the low intensity and broadening of the signals. This can be particularly

observed in Figure 4d, where the spin-spin relaxation time can only be determined at high ETPAOH concentrations. This behavior is consistent with the previous discussion that, at low concentrations of alkyltripropylammonium hydroxide and high concentrations of silica, the cation-silica interactions cause a reduction in the mobility of the cations and, therefore, line broadening, which is associated in this figure with short T2 times. On the other hand, the fact that T2 is shorter than T1 can be attributed to chemical exchange because these mixtures have relatively low viscosities (i.e., similar to the viscosity of water). PFG NMR was used to further quantify the organocation mobility. Figure 5 shows the diffusion NMR results for a series of MTPAOH-TEOS-water, ETPAOH-TEOS-water, and TPAOH-TEOS-water mixtures. The error bars represent the line width of the signal in the diffusion dimension determined by the estimated error of the diffusion coefficient obtained from the fitting process.70 In the absence of silica, the diffusion coefficients of the cations appear insensitive to alkyltripropylammonium hydroxide concentration, and the values are shown in Table 2. The self-diffusion coefficients presented in this table were calculated as the average of the coefficients of all data points shown in Figure 5 for cation-water mixtures, and the error represents its standard deviation. By contrast, the addition of silica leads to a decrease in the observed organocation diffusion coefficients. The lowest values of diffusion coefficients were detected at the highest silica content and were (0.254 ( 0.027) × 10-9 m2/s at 18.7 mM MTPAOH, (0.262 ( 0.072) × 10-9 m2/s at 18.7 mM ETPAOH, and (0.276 ( 0.042) × 10-9 m2/s at 18.7 mM TPAOH. Diffusion coefficients could not be

20184



Figure 5. Diffusion coefficients as a function of the total concentration of alkyltripropylammonium hydroxide for mixtures with compositions of (a) x MTPAOH:y TEOS:7290 H2O:810 D2O, (b) x ETPAOH:y TEOS:7290 H2O:810 D2O, and (c) x TPAOH:y TEOS:7290 H2O:810 D2O.

TABLE 2: Self-Diffusion Coefficients of the Cations in Water at 25 °C cation

Df, 10-9 m2/s

MTPA (methyltripropylammonium) ETPA+ (ethyltripropylammonium) TPA+ (tetrapropylammonium)

0.719 ( 0.006 0.694 ( 0.011 0.648 ( 0.011a

+

a The diffusion coefficient of TPA+ in dilute aqueous solution at 25 °C reported in the literature is 0.623 × 10-9 m2/s.71

determined at lower alkyltripropylammonium hydroxide concentrations because of the low signal intensity as discussed above. Also noteworthy is that, at the highest cation contents, the diffusion coefficient approaches the value of the cation in aqueous solution (i.e., in the absence of silica), shown in Table 2. One simple way to explain the results in Figure 5 is that, in

the presence of silica particles, the cation samples two distinct states, a free state and a bound state. Therefore, the organocation experiences a reduction in its mobility that is manifested as a decrease in the diffusion coefficient. The relaxation and diffusion data presented above indicate that the silica perturbs the mobility of the organocation because the relaxation times and the diffusion coefficients show significant changes as compared with the values of the organocation in water. A simple way to model the cation-silica interactions is by using the two-site model previously shown in Scheme 1, which considers that the cation can exist free in aqueous solution or bound to the silica nanoparticles. As described in the Introduction, the decay constants in eq 2 for molecules exchanging between both sites can be simplified in the fast exchange limit (eq 3). In this limit, the exchange is fast compared with

Figure 6. Signal decays obtained from the diffusion experiments as a function of the diffusion time for mixtures with compositions of (a) 3 MTPAOH:20 TEOS:7290 H2O:810 D2O and (b) 3 ETPAOH:20 TEOS:7290 H2O:810 D2O. The decays correspond to the methyl signal highlighted in the molecular structure of the cation.



Figure 7. Amount of bound cations as a function of the total concentration of alkyltripropylammonium hydroxide for mixtures with compositions of (a) x MTPAOH:y TEOS:7290 H2O:810 D2O, (b) x ETPAOH:y TEOS:7290 H2O:810 D2O, and (c) x TPAOH:y TEOS:7290 H2O:810 D2O. Open symbols denote analysis if all particles are 200 nm in diameter; filled symbols denote analysis if all particles are 5 nm in diameter.

the relaxation and diffusion scale. One way to ascertain if the cation exchange is in the fast limit is by observing the signal decays of the relaxation and diffusion experiments. If monoexponential decay of the signals in the relaxation and diffusion experiments is observed and the observed diffusion coefficient (Dobs) is independent of the diffusion time (∆), then exchange (τex) is very fast compared with the relaxation and diffusion scales.64 The relaxation data in Figure 4 is well described by monoexponential decays. In Figure 6, we show the signal decays from the diffusion experiments as a function of the diffusion time (∆). Figure 6a shows the signal decays for a mixture with a composition of 3 MTPAOH:20 TEOS:7290 H2O:810 D2O, whereas Figure 6b shows the signal decays for a mixture of ETPAOH with an analogous composition (3 ETPAOH:20 TEOS:7290 H2O:810 D2O). In this figure, it can be observed that the signal decays monoexponentially. In addition, the slope of the normalized intensity does not change with the diffusion time; and therefore, the diffusion coefficient (which can be calculated from the slope) is independent of the diffusion time. On the basis of the results in Figure 6, it can be concluded that the systems are in the fast exchange limit. Thus, the observed diffusion coefficient (Dobs) is given by eq 4 as a weighted average between the free and the bound states

Dobs ) xfDf + xbDb

(4)

where f denotes the free state and b denotes the bound state. The diffusion coefficient for the former (Df) is taken as the selfdiffusion coefficient of the cation determined from organo-

cation-water mixtures, and these values are summarized in Table 2. The diffusion coefficient for the latter (Db) is taken as the diffusion coefficient of the silica particle and can be estimated from dynamic light scattering, where the bound diffusion coefficient (Db) is obtained from the particle size using the Stokes-Einstein equation

Db )

kBT 6πηR

(5)

where kB is the Boltzmann’s constant, T is the absolute temperature, η is the viscosity, and R is the hydrodynamic radius of the spherical particle. DLS analysis of the mixtures yielded several observations. First, in all mixtures, there was a bimodal distribution of small (5 nm or less) particles and large (100 nm or larger) particles. The relative amount of small particles increased with increasing OH content (i.e., increasing organocation content). Example NNLS size distributions are shown in the Supporting Information for samples with a composition of 3 RN(C3H7)3+OH-:60 TEOS:7290 H2O:810 D2O. It is important to point out that the distributions in these histograms are weighted by the intensity, not the particle number density. Given that the scattering intensity scales nonlinearly with particle size, the plots overestimate the number density of large particles. Although this can potentially complicate the analysis of the PFG NMR data, a few important points can be made. First, it is unlikely that the cations bound to the large aggregates contribute to the diffusion NMR spectra. This is due to the fact that their rotational diffusion rate will be slow on the NMR time scale.

20186



Figure 8. Lineweaver-Burk plots and fits for (a) x MTPAOH:y TEOS:7290 H2O:810 D2O, (b) x ETPAOH:y TEOS:7290 H2O:810 D2O, and (c) x TPAOH:y TEOS:7290 H2O:810 D2O.

In other words, if the particles do not isotropically reorient rapidly on the NMR time scale, narrow, liquid-like resonances will not be observed for cations bound to the aggregates. Second, the influence of the errors of the input parameters (Dobs, Df, and Db) on the bound fraction of cations (xb) calculated with eq 4 was analyzed by varying the input parameters within their error, and the analysis showed that the bound fraction is considerably sensitive to the errors of Dobs and Df, but not very sensitive to the error of Db. On the basis of these two observations, it seems unlikely that cations bound to the large aggregates contribute significantly to the diffusion NMR results. It is also important to note that no sedimentation of silica is observed from any of these samples. Measurements performed on mixtures after several days are identical to “fresh” samples. With this information in hand, it is then possible to calculate the fractions of free and bound cations, and from that, one can determine the amount of cations bound (Γ) as a function of the total concentration of alkyltripropylammonium hydroxide. This information is presented in Figure 7. The plots show the amount of bound cations bound for two cases, the first being if all the nanoparticles are 5 nm in size (filled symbols), the second being if the nanoparticles are all 200 nm in size (open symbols). Interestingly, one can observe that the fraction of cations bound is very similar for both scenarios. One observes a fairly linear increase in the amount of cations bound at low cation contents, and then the convergence toward a plateau can be observed at high concentrations of alkyltripropylammonium hydroxide. It is likely due to the saturation of the silica nanoparticle surface. It can also be noticed that the calculated amount of the cation

bound to 5 nm silica particles is very similar (a difference of less than 15%) to that of the cation bound to 200 nm silica aggregates. From the data in Figure 7, it is possible to determine both the binding energies and the cation surface concentration at saturation. One of the many ways to analyze the data is through the use of the Lineweaver-Burk linearization of the Langmuir isotherm

1 1 1 + ) Γ Γmax ΓmaxKadC where Γ is the amount of bound cations, Γmax is the maximum amount of bound cations, Kad is the adsorption constant, and C is the equilibrium concentration of cations in solution. The Lineweaver-Burk fits of the isotherms are shown in Figure 8. These plots show the inverse of the amount of cations absorbed to 5 nm particles as a function of the inverse of the equilibrium concentration. The values of the adsorption constant (K) and maximum amount of bound cations (Γmax) obtained from the LineweaverBurk fit, as well as the change in the Gibbs energy, are summarized in Table 3. In this table, it can be observed that the adsorption constant decreases as the amount of TEOS increases in the mixtures. In general, for a given TEOS concentration, the adsorption constants of MTPA+ are larger than the adsorption constants of ETPA+, which are larger than those of TPA+. The exception to this is the mixture with x TPA: 20 TEOS that also has the lowest determination coefficient (R2)

PFG NMR of Alkyltripropylammonium-Silica Mixtures TABLE 3: Langmuir Constants and Adsorption Energy for Samples with a Composition of x RN(C3H7)3+OH-:y TEOS:7290 H2O:810 D2O cation MTPA+ y ) 20 y ) 60 ETPA+ y ) 20 y ) 60 TPA+ y ) 20 y ) 60

Kad, mol/L

Γmax, mol

Γ*maxa

R2

σb

∆Gad, kJ/mol

182.9 70.5

1.81 8.15

0.218 0.981

0.972 0.994

0.091 0.011

-12.91 -10.55

78.1 65.9

2.42 7.77

0.291 0.935

0.980 0.990

0.105 0.015

-10.80 -10.38

101.1 51.4

1.38 7.94

0.166 0.955

0.939 0.978

0.219 0.025

-11.44 -9.77

a Maximum amount of adsorbed cations (Γmax) per amount of silica nanoparticles. The amount of silica nanoparticles was calculated with a value of [SiO2]cac ) 0.052 mol/kg (or ycac ) 8.31 mol) reported by Fedeyko20 for a sample with a composition of 9 TPAOH:9500 H2O. b Residual standard deviation defined as σ ) (Q/n - p)1/2, where Q is the objective function, n is the number of observations, and p is the number of parameters in the model.

and the highest residual standard deviation (σ). In general, the determination coefficient and the residual standard deviation show that the experimental data fit well to the Langmuir isotherm. However, deviations of the model might occur because the Lineweaver-Burk linear regression, which is the double reciprocal of the Langmuir equation, might be susceptible to data errors at low concentrations. We believe that this is likely due to the fact that the Langmuir model assumes that all adsorption sites are uniform. Particularly at low cation concentrations, this might not be the case as one would anticipate that cations bind at the most favorable binding sites first. The MTPA cation displays larger values of the adsorption constant and Gibbs energy compared with the values of ETPA and TPA cations for analogous concentrations of TEOS. Therefore, the strongest cation-silica binding strength is associated with the MTPA cation. One simple explanation for this result is that this molecule has a small C/N ratio and thus is the most hydrophilic and also has the highest charge density. The values presented in Table 3 are used to calculate the number of cations adsorbed per silica particle that was assumed to have a core containing 356 silicon atoms, as reported by Vlachos.51 Although many assumptions go into this calculation that may not be precisely correct (uniformity of surface coverage, etc.), most of the numbers trend to approximately the same value range of approximately 55 cations per particle. The low silica TPA mixture appears much lower; as mentioned above, the fitting results appeared to be the least statistically robust. The low ETPA content mixture appears to give a much higher value. The origin of this is unclear currently. One reasonable conclusion from Table 4 is that the cation monolayer coverage of silica nanoparticles seems to be fairly independent of cation identity and silica content. The former makes sense as the size of the three organocations is very comparable. The latter is consistent with the existing model that, above the cac, addition of silica leads mainly to more particles, not to a dramatic change in the size of the particles. It is also important to point out that these numbers are different from those that we have reported previously for TPA-silica mixtures.57 The differences are due to the preparation method. In our previous work, the nanoparticles were formed using NaOH, followed by titration of TPABr into the mixture, resulting in a high dilution of the TPA and the presence of an excess of sodium cations. In the current work, the nanoparticle formation is controlled by the addition of organocations and there is no sodium present.

J. Phys. Chem. C, Vol. 114, No. 47, 2010 20187 TABLE 4: Cation Monolayer Coverage of Silica Nanoparticles in Mixtures with a Composition of x RN(C3H7)3+OH-:y TEOS:7290 H2O:810 D2O cation

no. of cations/particle

+

MTPA y ) 20 y ) 60 ETPA+ y ) 20 y ) 60 TPA+ y ) 20 y ) 60

55 56 74 54 42 55

Conclusions Diffusion NMR results of solutions containing TPA mimics and silica are reported. The diffusion NMR results indicate that the cation adsorption free energies are quite comparable to that of TPA for the two TPA mimics investigated. This finding is quite interesting as previous work from our lab has shown that these same TPA mimics lead to dramatically different (and slower) rates of silicalite-1 growth as compared with TPA.23 In particular, MTPA cations were shown to lead to growth rates 50% slower than those of TPA. Thus, the current work shows that small changes in the binding energy to the silica nanoparticles will lead to large changes in the stability of these precursor particles and thus zeolite formation. This has, in fact, been observed in previous work from our lab.72 The current work shows that PFG is uniquely suited to determine the silicateorganocation interactions in a noninvasive way. How these findings translate to in situ studies of zeolite formation are ongoing 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: DLS data for several of the mixtures studied in the current work. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Barrer, R. M. Zeolites and Clay Minerals as Sorbents and Molecular SieVes; Academic Press: London, 1978. (2) Barrer, R. M. Zeolites 1981, 1, 130–140. (3) Breck, D. W. Zeolite Molecular SieVes: Structure, Chemistry and Use; John Wiley: New York, 1974. (4) Foster, M. D.; Simperler, A.; Bell, R. G.; Friedrichs, O. D.; Paz, F. A. A.; Klinowski, J. Nat. Mater. 2004, 3, 234–238. (5) Cundy, C. S.; Cox, P. A. Chem. ReV. 2003, 103, 663–702. (6) Corma, A.; Davis, M. E. ChemPhysChem 2004, 5, 304–313. (7) Saxton, R. J. Top. Catal. 1999, 9, 43–57. (8) Corma, A. Chem. ReV. 1997, 97, 2373–2419. (9) Schueth, F.; Schmidt, W. AdV. Mater. 2002, 14, 629–638. (10) Lobo, R. F.; Zones, S. I.; Davis, M. E. J. Inclusion Phenom. Mol. Recognit. Chem. 1995, 21, 47–78. (11) Davis, M. E.; Lobo, R. F. Chem. Mater. 1992, 4, 756–768. (12) de Moor, P.-P. E. A.; Beelen, T. P. M.; Komanschek, B. U.; Beck, L. W.; Wagner, P.; Davis, M. E.; van Santen, R. A. Chem.sEur. J. 1999, 7, 2083–2088. (13) de Moor, P.-P. E. A.; Beelen, T. P. M.; Komanschek, B. U.; Diat, O.; van Santen, R. A. J. Phys. Chem. B 1997, 101, 11077–11086. (14) de Moor, P.-P. E. A.; Beelen, T. P. M.; van Santen, R. A. J. Phys. Chem. B 1999, 103, 1639–1650. (15) de Moor, P.-P. E. A.; Beelen, T. P. M.; van Santen, R. A.; Beck, L. W.; Davis, M. E. J. Phys. Chem. B 2000, 104, 7600–7611. (16) de Moor, P.-P. E. A.; Beelen, T. P. M.; van Santen, R. A.; Tsuji, K.; Davis, M. E. Chem. Mater. 1999, 11, 36–43.

20188


(17) Iton, L. E.; Trouw, F.; Brun, T. O.; Epperson, J. E.; White, J. W.; Henderson, S. J. Langmuir 1992, 8, 1045–1048. (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) Fedeyko, J.; Vlachos, D. G.; Lobo, R. F. Langmuir 2005, 21, 5197– 5206. (20) Fedeyko, J. M.; Rimer, J. D.; Lobo, R. F.; Vlachos, D. G. J. Phys. Chem. B 2004, 108, 12271–12275. (21) Rimer, J. D.; Fedeyko, J.; Vlachos, D. G.; Lobo, R. F. Chem.sEur. J. 2006, 12, 2926–2934. (22) Rimer, J. D.; Vlachos, D. G.; Lobo, R. F. J. Phys. Chem. B 2005, 109, 12762–12771. (23) Cheng, C.-H.; Shantz, D. F. J. Phys. Chem. B 2005, 109, 13912– 13920. (24) Cheng, C.-H.; Shantz, D. F. J. Phys. Chem. B 2005, 109, 19116– 19125. (25) Cheng, C.-H.; Shantz, D. F. J. Phys. Chem. B 2005, 109, 7266– 7274. (26) 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. (27) Schoeman, B. J. Zeolites 1997, 18, 97–105. (28) Schoeman, B. J. Microporous Mater. 1997, 9, 267–271. (29) Schoeman, B. J. Microporous Mesoporous Mater. 1998, 22, 9–22. (30) Schoeman, B. J.; Babouchkina, E.; Mintova, S.; Valtchev, V. P.; Sterte, J. J. Porous Mater. 2001, 8, 13–22. (31) Schoeman, B. J.; Regev, O. Zeolites 1996, 17, 447–456. (32) Kinrade, S. D.; Donovan, J. C. H.; Schach, A. S.; Knight, C. T. G. J. Chem. Soc., Dalton Trans. 2002, 1250–1252. (33) Kinrade, S. D.; Knight, C. T. G.; Pole, D. L.; Syvitski, R. T. Inorg. Chem. 1998, 37, 4272–4277. (34) Knight, C. T. G. Zeolites 1990, 10, 140–144. (35) Knight, C. T. G.; Syvitski, R. T.; Kinrade, S. D. Stud. Surf. Sci. Catal. 1995, 97, 483–488. (36) Knight, C. T. G.; Wang, J. P.; Kinrade, S. D. Phys. Chem. Chem. Phys. 2006, 8, 3099–3103. (37) Cheng, C.-H.; Shantz, D. F. J. Phys. Chem. B 2005, 110, 313– 318. (38) Fyfe, C. A.; Darton, R. J.; Schneider, C.; Scheffler, F. J. Phys. Chem. C 2008, 112, 80–88. (39) 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. (40) Haouas, M.; Taulelle, F. J. Phys. Chem. B 2006, 110, 3007–3014. (41) Houssin, C. J. Y.; Kirschhock, C. E. A.; Magusin, P.; Mojet, B. L.; Grobet, P. J.; Jacobs, P. A.; Martens, J. A.; Santen, R. A. Phys. Chem. Chem. Phys. 2003, 5, 3518–3524. (42) Navrotsky, A. Curr. Opin. Colloid Interface Sci. 2005, 10, 195– 202. (43) Navrotsky, A.; Trofymluk, O.; Levchenko, A. A. Chem. ReV. 2009, 109, 3885–3902.

Rivas-Cardona and Shantz (44) Rimer, J. D.; Trofymluk, O.; Navrotsky, A.; Lobo, R. F.; Vlachos, D. G. Chem. Mater. 2007, 19, 4189–4197. (45) Yang, S.; Navrotsky, A. Chem. Mater. 2002, 14, 2803–2811. (46) Yang, S.; Navrotsky, A. Chem. Mater. 2004, 16, 3682–3687. (47) Yang, S.; Navrotsky, A.; Wesolowski, D. J.; Pople, J. A. Chem. Mater. 2004, 16, 210–219. (48) Kumar, S.; Davis, T. M.; Ramanan, H.; Penn, R. L.; Tsapatsis, M. J. Phys. Chem. B 2007, 111, 3398–3403. (49) Kumar, S.; Wang, Z.; Penn, R. L.; Tsapatsis, M. J. Am. Chem. Soc. 2008, 130, 17284–17286. (50) Ramanan, H.; Kokkoli, E.; Tsapatsis, M. Angew. Chem., Int. Ed. 2004, 43, 4558–4561. (51) Rimer, J. D.; Lobo, R. F.; Vlachos, D. G. Langmuir 2005, 21, 8960– 8971. (52) Cheng, C.-H.; Shantz, D. F. Curr. Opin. Colloid Interface Sci. 2005, 10, 188–194. (53) Nikolakis, V.; Tsapatsis, M.; Vlachos, D. G. Langmuir 2003, 19, 4619–4626. (54) Cejka, J.; Bekkum, H. v.; Corma, A.; Schu¨th, F. Introduction to Zeolite Science and Practice, 3rd ed.; Elsevier: Amsterdam, 2007. (55) Steen, E. v.; Callanan, L. H.; Claeys, M. Recent AdVances in the Science and Thecnology of Zeolites and Related Materials, 1st ed.; Elsevier: Amsterdam, 2004. (56) Nikolakis, V.; Kokkoli, E.; Tirrell, M.; Tsapatsis, M.; Vlachos, D. G. Chem. Mater. 2000, 12, 845–853. (57) Li, X.; Shantz, D. F. J. Phys. Chem. C 2010, 114, 8449–8458. (58) Li, X.; Shantz, D. F. J. Phys. Chem. C 2010, 114, 14561–14570. (59) Petry, D. P.; Haouas, M.; Wong, S. C. C.; Aerts, A.; Kirschhock, C. E. A.; Martens, J. A.; Gaskell, S. J.; Anderson, M. W.; Taulelle, F. J. Phys. Chem. C 2009, 113, 20827–20836. (60) Johnson, C. S. Prog. Nucl. Magn. Reson. Spectrosc. 1999, 34, 203– 256. (61) Price, W. S. Concepts Magn. Reson. 1997, 9, 299–336. (62) Berger, S.; Braun, S. 200 and More NMR Experiments, 1st ed.; Wiley-VCH: Weinheim, Germany, 2004. (63) Cohen, Y.; Avram, L.; Frish, L. Angew. Chem., Int. Ed. 2005, 44, 520–554. (64) Choudhury, R. P.; Schonhoff, M. J. Chem. Phys. 2007, 127, 234702. (65) Ka¨rger, J. Ann. Phys. 1971, 482, 107–109. (66) Ka¨rger, J. AdV. Colloid Interface Sci. 1985, 23, 129–148. (67) Woessner, D. E. Concepts Magn. Reson. 1996, 8, 397–421. (68) Holz, M.; Weingartner, H. J. Magn. Reson. 1991, 92, 115–125. (69) Lucas, L. H.; Yan, J.; Larive, C. K.; Zartler, E. R.; Shapiro, M. J. Anal. Chem. 2003, 75, 627–634. (70) Varian, Inc. Inova and MercuryPlus NMR Systems with VnmrJ 2.2D software, Pub. No. 01-999378-00, Rev. A 0808. (71) CRC Handbook of Chemistry and Physics, 90th ed.; Lide, D. R., Haynes, W. M., Eds.; CRC Press: Boca Raton, FL, 2009-2010. (72) Li, X.; Shantz, D. F. Langmuir 2010. Manuscript under review.

JP1082483

Pulsed Field Gradient NMR Investigations of Alkyltripropylammonium

Recommend Documents