- Email: [email protected]
The interaction between the surfactant and the co-structure directing agent in anionic surfactant-templated mesoporous silicas
Microporous and Mesoporous Materials 163 (2012) 291–299
Contents lists available at SciVerse ScienceDirect
Microporous and Mesoporous Materials journal homepage: www.elsevier.com/locate/micromeso
The interaction between the surfactant and the co-structure directing agent in anionic surfactant-templated mesoporous silicas Marc Florent, Daniella Goldfarb ⇑ Department of Chemical Physics, Weizmann Institute of Science, Rehovot 76100, Israel
a r t i c l e
i n f o
Article history: Received 14 May 2012 Received in revised form 7 June 2012 Accepted 29 June 2012 Available online 10 July 2012 Keywords: Ordered mesoporous materials Co-structure directing agent EPR ESEEM
a b s t r a c t Anionic surfactant-templated mesoporous silicas (AMS) are synthesized with a co-structure directing agent (CSDA) that interacts with both the organic and inorganic components of the system. The AMS materials structure is controlled by pH. We investigated the formation of AMS cubic and hexagonal phases, prepared under the same conditions, except pH, by EPR spectroscopic measurements. We used silica-like and surfactant-like spin probes added to the reaction mixtures in minute amounts. Through the spin probes we resolved the specific interactions of the CSDA (N-trimethoxylsilylpropyl-N,N,N-trimethyl ammonium chloride (TMAPS)) with the surfactant (sodium myristate (C14AS)) and the polymerizing silica. We observed for both phases a fast formation of a mesophase upon addition of the silica precursor (TEOS, tetraethoxysilane) and the CSDA into the surfactant solution, attributed to the strong attraction between the CSDA and the anionic surfactant. This is then followed by a slow condensation of the silica. Electron spin echo envelope modulation (ESEEM) spectra of both spin probes in the as-synthesized materials indicated the presence of two types of CSDA molecules; one interacting with the surfactant and the other with the silica wall. Continuous wave EPR spectra showed different spin probe motilities in the two as-synthesized materials that indicated that the relative populations of the two CSDA types are different in the two phases. We attribute this difference to the pH differences in the reaction mixtures. A soft extraction of the surfactant from the pores did not alter the structure of the final materials, but it abolished the observed molecular level differences between them. The extraction allowed the pending ammonium groups to acquire a high degree of freedom and accessibility to water molecules. Ó 2012 Elsevier Inc. All rights reserved.
1. Introduction Synthesis of ordered mesoporous silicas relies on the selfassembly of a surfactant template and inorganic silica precursors into a mesostructure [1,2]. This self-assembly depends on the intimate interactions between the surfactant and the inorganic oxide source. Accordingly, different synthesis routes were classified according to the type of the interaction: S+I, SI+, S+XI+, SX+I, S0I0, S0H+XI+, and so on, where S, I and X represent the surfactant, the inorganic (silica) precursors, and a counter ion, respectively, with their charge depending on whether they are in a cationic, anionic or neutral form [3,4]. The synthesis of ordered mesoporous materials is based on a sol–gel process, in acidic or basic environment [5]. Therefore, the choice of surfactant and reaction conditions determine the interactions driving the mesostructure formation. For example, in the case of silica materials the pH is crucial, considering that the isoelectric
⇑ Corresponding author. Address: Weizmann Institute of Science, Chemical Physics Department, Rehovot 76100, Israel. Tel.: +972 8 934 2341; fax: +972 8 934 4123. E-mail address: [email protected] (D. Goldfarb). 1387-1811/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.micromeso.2012.06.050
point of silicates species is around 2 [6]. Cationic and neutral nonionic surfactants have been proven very successful in forming diverse mesostructures, in acidic or basic conditions [1,7–10]. However, anionic surfactants have not yielded ordered mesostructure silicas unless a co-structure directing agent (CSDA) was added to the synthesis [11]. This CSDA has an alkoxysilane on one end that can co-condense with the silica precursor, and an ammonium group on the other end that can interact with the negatively charged head group of the surfactant, actually acting as glue. Using this strategy, a variety of structures, including cubic, tetragonal, hexagonal, bicontinuous cubic, and lamellar, as well as chiral structures, were successfully prepared [11–16]. The different mesostructures obtained were explained to some extent by the packing parameter of the surfactant g = v/a0l, where v is the chain volume of the surfactant, a0 is the effective hydrophobic/hydrophilic interfacial area and l is the kinetic surfactant chain length [17,18]. This parameter depends on the structure of the components of the system and predict the curvature of the aggregate, and therefore the final mesostructure. The g parameter has been shown to be sensitive to the surfactant concentration, temperature and the presence of other species in the solution [19,20].
292
M. Florent, D. Goldfarb / Microporous and Mesoporous Materials 163 (2012) 291–299
Fig. 1. A schematic representation of the change of curvature of the surfactant micelles by modifying the ionization degree of the anionic surfactant through a change of pH. The positively charged CSDA is shown as well.
It has been proposed that the packing parameter of the anionic surfactant can be controlled through the degree of ionization of the surfactant, which depends on the pH, as shown on Fig. 1. The degree of ionization of the polar head of the carboxylic fatty acids increases with pH, thereby increasing its effective headgroup area due to repulsion and leading to a lower g value. Indeed, by increasing the pH of the reaction mixture a transition from lamellar to cylindrical and to cage-type mesophases was observed. This, in turn, was correlated with an increase of curvature due to the decrease of g [14]. Other parameters reported to affect the structure of the anionic surfactant-templated silicas are the time delay between the addition of CSDA and the silica source [16], the structures of the CSDA and of the surfactant [13], and the molecular CSDA/surfactant ratio [15]. In addition to the possibility to control the structure of the final phase, the anionic pathway received attention also for its ability to easily provide functionalized mesoporous silicas. The possible removal of the surfactant by extraction, instead of calcination, allows retaining the amino or ammonium group of the CSDA in the pore as it is bound to the silica walls [21]. 13C NMR and elemental analysis of the functionalized material thus obtained showed a homogeneous distribution of amine or ammonium groups. Their loading amount was reported to be limited by the quantity of CSDA that can interact electrostatically with the anionic surfactant [21]. A previous study on the synthesis of AMS, using the acid form of a surfactant and a CSDA with an amino group, showed that the reaction takes different paths, depending on the time between the addition of CSDA and the addition of the silica source, yielding different phases. Following its addition into the solution, the CSDA was reported to quickly interact with the surfactant and to start condensing. An organo-silica layer made of condensed CSDA was shown to first form around the micelles in solution. The inorganic silica species condense later to form an outer inorganic silica shell. As the reaction goes on, the system just shows an increase of the fraction of condensed inorganic silica [22]. Herein we used spin probe EPR methods to explore the interplay between the surfactant, CSDA and the silica in hexagonal (p6mm) and cubic (Fd3m) structures. We used a surfactant in its salt form, sodium myristate (C14AS), and a CSDA with a quaternary
ammonium group, N-trimethoxylsilylpropyl-N,N,N-trimethyl ammonium chloride (TMAPS). In this combination, the ionization degree of the CSDA is not modified by pH while that of the surfactant is. C14AS is a common surfactant whose phase diagram in water is well known [23]. Its 1% wt Krafft point has been determined around 44 °C, above which it dissolves in water to form a clear micellar solution [24]. It is known that at equilibrium a dilute aqueous solution of sodium myristate (C < 20 mM, corresponding to pH < 10) have a protonation fraction of about 1%, but this has a strong influence on the solution and phase behavior [25]. By adding NaOH the protonation can be further decreased, but the solubility of C14AS decreases as well. The addition of other electrolytes, and especially tetraalkyammonium salt affects also significantly the solution behavior of C14AS [24]. Spin probe EPR has been proven successful to characterize pores and surface properties [26–30]. The formation of mesoporous silica materials, synthesized with cationic or neutral surfactant templates, has also been monitored using in situ continuous wave (CW)-EPR [31–36], and more advanced pulse EPR experiments, such as electron spin echo envelope modulation (ESEEM) [37– 39]. In general, surfactant-like nitroxides, were used to probe local changes in the organic phase, and a nitroxide linked to an alkoxysilane was employed to give information on the formation of the silica wall [40]. The CW-EPR spectrum of the spin probes provide information on the dynamics and polarity of the microenvironment of the nitroxide label, while ESEEM experiments give information on its local structure via the hyperfine interaction with nearby magnetic nuclei [41,42]. To investigate the formation mechanism and inner surface properties of as-synthesized cubic (Fd3m) and 2D-hexagonal (p6mm) materials, with emphasis on the CSDA interactions with the other components of the reaction, we used 5-doxyl stearic acid (5-DSA) as a surfactant-like probe, and a silica-like probe, 2,2,6,6tetramethyl-4-[3-(triethoxysilyl)propylamino]-1-piperidinooxy (SL1SiEt) that can co-condense with the silica precursor (see Fig. 2). Whatever the phase, we observed that the addition of the silica mixture (CSDA + silica precursor, (TEOS, tetraethoxysilane)) leads to a fast (2–8 min) formation of a mesophase, whose exact kinetics is below our experimental resolution. We could only observe the
Fig. 2. (a) The nitroxide radical, SL1SiEt, used to probe the silica region and (b) the 5-DSA employed for probing the surfactant part of the system.
M. Florent, D. Goldfarb / Microporous and Mesoporous Materials 163 (2012) 291–299
evolution of the silica condensation, which is slower. Therefore we focused on the properties of the as-synthesized materials. These underlined the strong interaction between the CSDA and the negatively charged surfactant evidenced by clear interactions of the spin label with the 14N nuclei of CSDA. In addition, our results indicate the existence of CSDA interacting with the silica surface. The ratio between these two CSDA populations varies from one phase to the other and this is related to their different reaction pH. This is also reflected in the local mobility of the nitroxide probes in the vicinity of the wall and at the surfactant–CSDA interface. A soft removal of the surfactant generates pores covered by pending ammonium groups, which are highly mobile and accessible to small molecules. 2. Materials and methods 2.1. Materials Chemicals were used as received. Sodium myristate (C14AS) was purchased from Fluka, tetraethoxysilane (TEOS) from Aldrich and the CSDA, N-trimethoxylsilylpropyl-N,N,N-trimethyl ammonium chloride (TMAPS) from Gelest, Inc. The nitroxide spin probe 5-doxyl stearic acid (5-DSA) was purchased from Aldrich and the silica nitroxide spin probe, 2,2,6,6-tetramethyl-4-[3-(triethoxysilyl)propylamino]-1-piperidinooxy (SL1SiEt), was synthesized as described in the literature [40,43]. Nickel(II) ethylene diaminediacetic acid (NiEDDA) was prepared as described in the literature [44]. 2.2. Synthesis Materials were synthesized following a procedure described in the literature [14,21]. The anionic surfactant, C14AS, was first dissolved in water (or deuterated water) at 80 °C. Sodium hydroxide was added when the aimed final material was cubic. Cooling to 60 °C, a mixture of silica precursor, TEOS, and CSDA, TMAPS, was added while stirring. After stirring for ten more minutes, the reaction mixture was aged for two days at 60 °C. The precipitate was then filtered, washed and dried overnight at 80 °C to obtain the as-synthesized material. Two paramagnetic probes, 5-DSA, a surfactant-like, or SL1SiEt, a silica-like nitroxide probe, shown in Fig. 2, were added in minute amount to the reaction mixture to probe the reaction and the final material. The molar composition of the reaction mixture was C14AS/TMAPS/TEOS/H2O/Spain Probe/NaOH = 1/1/7/1389/0.006/x, where x = 0, or 0.1 to obtain the 2D-hexagonal (p6 mm) or the cu bic (Fd3m) materials respectively. 2.3. Extraction of the surfactant Final materials free of surfactant were obtained according to the two following procedures. The as-synthesized silica material was calcined at 600 °C for 6 h to give a surfactant-free mesoporous silica. Alternatively, to get a surfactant-free mesoporous silica functionalized with alkoxy ammonium group, the surfactant was extracted from the as-synthesized material by an ethanolic solution of HCl (11 vol.%), for 24 h at boiling temperature. The material was then filtered, washed with ethanol, and dried overnight at 80 °C.
293
300 mesh copper grid and air-dried. Digital images were recorded by a Gatan US1000 CCD camera, and fast Fourier transforms (FFT) of the images were calculated by DigitalMicrograph software. Further characterization of the final materials was done by Xray diffraction (XRD) on a Rigaku TTRAX-III diffractometer, using a CuKa radiation (k = 0.154 nm) over the range of 0.3–5° 2h with steps of 0.02° 2h, or by small angle X-ray scattering (SAXS), on a home-built diffractometer equipped with Franks mirror and a linear position-sensitive detector, also using a Ni-filtered CuKa radiation. In both cases quartz capillaries of 1.5 mm diameter were used. 29 Si MAS NMR (magic angle spinning nuclear magnetic resonance) experiment was performed on a Bruker DSX spectrometer operating at 59.6 MHz, equipped with a Bruker 4 mm MAS probe. The experiment was carried out at a spinning frequency of 5 kHz, by employing a spin-echo pulse sequence with a 90° excitation pulse of 4.4 ls, an echo delay of 200 ls, and a repetition delay of 600 s. The chemical shift ppm scale referred to tetramethylsilane (TMS). Proton decoupling was employed during the acquisition, using the two pulse phase modulation (TPPM) scheme with 70 kHz decoupling power and Du = 17°. CW-EPR spectra were recorded on a Bruker Elexsys 500 spectrometer or a modified Varian E12 X-band spectrometer equipped with a temperature controller. Solid materials were introduced into EPR quartz tubes of 3 mm outer diameter. For the experiments of solvent accessibility, the product was suspended in water, or water solutions of NiEDDA, and introduced into a glass capillary of 0.75 mm diameter. EPR spectra were recorded at room temperature, for different microwave power. Measurements during the formation of the AMS were done by sampling the reaction mixture at different times of the synthesis. Here a small amount was transferred into glass capillaries. The spectra were recorded at the reaction temperature, 60 °C. Sampling a multiphase mixture is always a problem in terms of the relative quantity taken of each phase. Unfortunately, attempted in situ synthesis did not produce ordered materials, probably because of the too small reaction vessel required and the associated slow rate of ethanol evaporation. Simulations of the EPR spectra of the wet mesophases at the end of reaction were done using the EasySpin functions chili and garlic, which calculate slow motion and fast motion spectra respectively [45]. Three pulse ESEEM traces were recorded with an X-band pulse Bruker Elexsys 580 spectrometer at 85 K. The ESEEM experiments were done using the three-pulse pulse sequence p/2–s–p/2–T–p/ 2–s–echo, with a four-step phase cycle [46]. The interval s between the first two pulses was calculated to give the maximum deuterium signal, when 2H was the nucleus of interest, or the maximum 14 N signal and minimal proton signal when 14N was the nucleus of interest [47]. ESEEM traces were all treated identically: after phase correction and normalization, the background decay of the normalized data was divided using a biexponential fit, then the data was apodized with a Hamming window, zero filled to 512 points and Fourier transformed (FT) with cross-term averaging [48]. The FTESEEM spectra consist of peaks centered at the Larmor frequency of the nuclei in the vicinity of the nitroxide. The intensity of these peaks reflects the proximity and density of nuclei around the nitroxide.
2.4. Methods
3. Results
The final surfactant free materials were characterized by transmission electron microscopy (TEM) with a Philips CM120 supertwin microscope operated at 120 kV. A dispersion of final material in ethanol was deposited on a carbon/collodion-coated
3.1. As-synthesized cubic and hexagonal AMS Two different phases, hexagonal (p6mm) and cubic (Fd3m), were prepared by varying the pH of the same reaction mixture.
294
M. Florent, D. Goldfarb / Microporous and Mesoporous Materials 163 (2012) 291–299
The final materials were characterized by TEM (transmission electron microscopy) and XRD (X-ray diffraction) (Supporting information Fig. S1). The cell parameters, a = 5.9 nm and a = 21.2 nm for the hexagonal and the cubic materials, respectively, are similar to that reported in the literature [14]. 29 Si MAS NMR spectrum of the as-synthesized cubic AMS (Fig. S2) shows the presence of CSDA molecules condensed with the silica. Furthermore it shows that a large amount of inorganic silica is only partially condensed, with a fair amount of Q3 and an appreciable amount of Q2 silica species, as reported in the literature [49]. The CW-EPR spectra of 5-DSA and SL1SiEt in the dry as-synthesized hexagonal (obtained without NaOH) and cubic materials (obtained with NaOH, x = 0.1, see experimental section) recorded at 150 K, room temperature (RT) and 335 K are shown in Fig. 3. The spectrum recorded at 150 K is identical for both materials and is typical for an immobilized nitroxide spin label and is therefore shown for reference. The spectra of the surfactant-like spin probe, 5-DSA, show that at RT it is rigid on the EPR time scale (rotational correlation time is longer than 1 ls) and becomes slightly more mobile as the temperature increases to 335 K. The SL1SiEt probe, by contrast, is much less rigid. It is in a slow motion regime already at room temperature, and by increasing temperature, the probe becomes much more mobile. Here we observe differences between the hexagonal and cubic phases; in the hexagonal phase; in the hexagonal phase, SL1SiEt is less mobile. We attribute the difference in mobility between the two different probes, 5-DSA and SL1SiEt, to differences in their structure (Fig. 2). In 5-DSA, the nitroxide ring is directly connected to the alkyl chain through a common shared carbon atom. Therefore the nitroxide ring does not experience local motion (bond rotation, etc.) and it can only rotate along with the whole alkyl chain. In the case of SL1SiEt, the nitroxide ring is connected to the polymerizing silica through four single bonds, allowing considerable local motion. To learn more about the local environment of the two spin probes we carried out ESEEM measurements. Fig. 4 shows the
FT-ESEEM of 5-DSA in the dry hexagonal and cubic as-synthesized materials prepared in D2O (the time domain traces are shown in Fig. S3). The FT-ESEEM spectra recorded under conditions optimizing 14N or 2H modulations show peaks at the 14N, 2H, and 1H Larmor frequencies, arising from weakly coupled nuclei. 5-DSA clearly reports the presence of deuterium in its vicinity, which arises from deuterated surfactant (exchangeable OD groups) or D2O. In contrast, the FT-ESEEM spectra of SL1SiEt (Fig. 5) in both phases, acquired under the same experimental conditions, comprise signals of 14N and 1H only. This indicates that SL1SiEt probes a region farther away from the surfactant head groups. Interestingly, both probes reveal clear 14N modulation from the ammonium group nitrogen of the CSDA. To exclude the possibility that the 14N signal observed for SL1SiEt arises from 14N of the probe itself we carried out ESEEM measurements on SL1SiEt in toluene; no 14 N modulations were observed. The 14N modulation cannot arise from the 14N of the nitroxide group because it is strongly coupled and have large hyperfine couplings that lead to nuclear frequencies far away from the Larmor frequency. The presence of the 14N modulation from the co-surfactant nitrogen provides direct experimental evidence for its proximity to the surfactant-like probe, 5-DSA. This confirms the CSDA–surfactant interaction mechanism, where the negatively charged polar heads of the micelle interact with the ammonium group of the CSDA. Moreover, the observation of 14N ESEEM also for SL1SiEt indicates that it is also close to some co-surfactant molecule. However, the fact that the silica probe does not report on 2H nuclei around it, in contrast to 5-DSA, but still experience 14N modulation from CSDA molecules suggests that SL1SiEt reports on CSDA molecules which are not interacting with the anionic head of the surfactant. The intensity of the 14N peak in the SL1SiEt FT-ESEEM spectra is lower than that reported by the 5-DSA probe. This implies that the SL1SiEt probe is farther away from the ammonium group of the CSDA or that there are less interacting CSDA molecules around it. Interestingly, one can notice that the intensity of 14N in the cubic material is about the same as in the hexagonal one, though, in the
Fig. 3. X-band CW-EPR spectra of (a and b) 5-DSA and (c and d) SL1SiEt in the as-synthesized (a and c) hexagonal and (b and d) cubic materials recorded at 150 K, RT and 335 K, as indicated in the figure. In (c), and (d) the CW-EPR spectra of SL1SiEt in the final material obtained after surfactant extraction by HCl/ethanol treatment of the cubic and hexagonal materials at RT are also presented. The asterisk indicates the presence of a highly mobile component in small quantity.
M. Florent, D. Goldfarb / Microporous and Mesoporous Materials 163 (2012) 291–299
295
Fig. 4. FT-ESEEM traces of 5-DSA in the as-synthesized (a and c) hexagonal, and (b and d) cubic materials, recorded with (a and b) s ’ 220 ns, optimized for 2H modulation, and with (c and d) s ’ 480 ns, optimized for 14N modulation and minimization of 1H modulation.
Fig. 5. FT-ESEEM traces of SL1SiEt in the as-synthesized (a and c) hexagonal, and (b and d) cubic material, recorded with (a and b) s ’ 220 ns, optimized for 2H modulation, and with (c and d) s ’ 480 ns, optimized for 14N modulation and minimization of 1H modulation. FT-ESEEM traces after surfactant extraction by HCl/ethanol in (e) the hexagonal and (f) cubic materials.
synthesis of the latter, the lower degree of ionization of the surfactant should lead to a lower number of CSDA molecules interacting with it (see Fig. 1) and therefore to a lower intensity of the 14N peak. 3.2. AMS materials after template removal Mesoporous materials can be used in various applications only after the pores are freed from the surfactant. The classical method
to remove the template is to calcine the solid [50]. Another method is to extract the surfactant with an acid ethanolic solution [21,51,52]. In the case of the AMS materials this soft method is particularly interesting because it yields a functional material with the alkyl ammonium groups pending in the pores [21]. The as-synthesized cubic and hexagonal materials were subjected to such a treatment and the final structure of the phases did not change. The materials synthesized with 5-DSA had no EPR signal after this treatment showing that the surfactant-like spin probe is washed
296
M. Florent, D. Goldfarb / Microporous and Mesoporous Materials 163 (2012) 291–299
out along with the anionic surfactant. In contrast, the signal of SL1SiEt persists also after the extraction of the surfactant, as shown in Fig. 3. This indicates that this probe is attached to the silica wall, confirming that its siloxane group polymerized with TEOS. Yet, the signal to noise ratio of the spectrum decreased after the surfactant extraction, most probably because part of the nitroxides were transformed into hydroxylamines by the acid treatment [53]. The lineshape of SL1SiEt in the final surfactant free hexagonal and cubic phases shows the presence of two components: a majority species that is considerably more mobile than SL1SiEt before solvent extraction, and a minor more rigid species. Another very mobile species in a negligible quantity can also be seen and is indicated by ⁄ on Fig. 3. The amount of the latter is a little higher in the cubic phase than in the hexagonal phase, but beside that both phases present similar spectra after extraction. The higher mobility of the probe arises from the free space inside the pores obtained after the surfactant removal, allowing the pending probes to fully extend and rotate. The minor rigid species is assigned to polymerized SL1SiEt trapped inside the wall. In order to differentiate between SL1SiEt in the wall and pending into the pores we carried out accessibility measurements. The accessibility of the silica spin label to a water soluble paramagnetic relaxation agent, NiEDDA, was measured by CW-EPR. Here we expect to see broadening upon the interaction with the paramagnetic NiEDDA. The spectra of the final hexagonal material in different solutions of paramagnetic NiEDDA are shown on Fig. 6. It appears that before the surfactant extraction the SL1SiEt actually exists in three different environments: a rigid one, a slow motion one and a small quantity of a fast motion one. The free component disappears with the addition of NiEDDA, showing a high accessibility to water. This very free component may arise from some probes fixed at the external surface of the material and therefore they are highly accessible to the solution. The two others types do not seem to be affected by the introduction of NiEDDA, implying that the NiEDDA complex cannot enter the pores because they are occupied by the surfactants. After the HCl treatment, in the water suspension, the probe becomes very free to move, due to the removal of the surfactant, and the lineshape broadens with the addition of NiEDDA, showing that the probes are now more accessible to water. ESEEM measurements were carried out on the hexagonal and cubic materials with SL1SiEt after HCl treatment. In both materials the 14N FT-ESEEM signal, displayed in Fig. 5, shows a slight increase compared to the situation before the extraction, indicating a subtle increase of CSDA density in the vicinity of the SL1SiEt. After freeing the pores from the surfactant both the silica probe and the co-surfactant are pending into the pores and can freely move, bringing them closer to each other.
3.3. Measurements during synthesis True in situ CW-EPR measurements could not be carried out, as the synthesis performed in the small vessel required for EPR measurements did not yield ordered mesoporous materials. Therefore, samples for EPR measurements were drawn from the heterogeneous reaction mixture into EPR capillaries at different time points along the reaction. A major inconvenience of this method is that the relative quantity taken of each phase (solution or precipitate) every time, is hard to control. CW-EPR spectra of the silica probe SL1SiEt recorded during the synthesis did not undergo any change with time. The spectrum consists of three very narrow lines all along the reaction, demonstrating that the forming silica gel stays very fluid during the entire synthesis. In contrast, the spectrum of the surfactant-like spin probe 5-DSA undergoes significant changes as shown in Fig. 7. The changes are characteristic of a population of two species; the relative amount of one increases with time while that of the other decreases. At time zero, corresponding to the addition of the CSDA + TEOS mixture into the surfactant solution, the spectrum comprises a mobile species only. New features, marked with squares, appear already after 8 min. As the time progresses the intensities of these features, corresponding to a more rigid species, increase and those of the mobile species decrease. This behavior is
Fig. 7. CW-EPR spectra of 5-DSA in the reaction mixture during the synthesis of the cubic material recorded at different times t of the reaction. From top to bottom, t = [0;2;8;15;25;40;60;90;120;180;240;300:460;1210;1330] min where t = 0 min is the time of addition of the TEOS and CSDA mixture. s, Fast motion component; h, slow motion component. Below, the spectrum of the dry as-synthesized material is shown for comparison.
Fig. 6. CW-EPR spectra of dispersions of the hexagonal material in water, and water solutions of NiEDDA, (a) before (as-synthesized material) and (b) after HCl treatment (final material). s, Isotropic component; h, rigid component; , slow motion component. Inset: expanded view of the central lines.
M. Florent, D. Goldfarb / Microporous and Mesoporous Materials 163 (2012) 291–299
Fig. 8. CW-EPR spectra (black line) of 5-DSA in the synthetic gel, obtained after centrifugation of the reaction mixture at the end of the syntheses of the hexagonal and cubic phase, with their simulation (red line). The asterisks indicate the features of a free nitroxide in very small quantity. The simulations were done combining a slow motion component and a free one that holds for 1.5% of the total. Parameters used in the simulations were g = [2.0214 2.0196 2.0153], A = [13.8 13.6 103] MHz and sc = 0.4 ns for the free component. The same parameters were used for the slow motion one, and the parameters given in Table 1. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
typical or a A?B transformation. The appearance of the new signal, assigned to the formed mesophase, is associated with the formation of the precipitate. In general, the same behavior was observed for both hexagonal and cubic materials. The spectra of the precipitate (mesophase), which was separated by centrifugation of the final reaction mixture, without drying, are shown in Fig. 8. The spectra are typical of a slow motion regime with rotational motion restricted to some orientations due to an orienting potential. A small part, marked with *, corresponds to remaining 5-DSA in solution that give a fast motion component, but holds only for about 1.5 % of the whole nitroxides. As explained above, the rotational restriction can be observed because of the structure of 5-DSA that allows the nitroxide ring to rotate only with the whole surfactant chain. Simulations of these spectra gave the parameters listed in Table 1. The spectrum of the wet as-synthesized materials of the two phases have very similar properties, except for a small difference in the correlation time, with the hexagonal phase exhibiting slightly higher mobility.
4. Discussion Using EPR spectroscopy of two spins probes, a surfactant-like, 5-DSA, and a silica-like, SL1SiEt, we explored the formation mechanism of a cubic and hexagonal AMS material where the reaction mixture differ only in pH. In the following discussion we first address the features that were common to both and then illuminate minor differences that can be related to the pH difference. 5-DSA was found to be a better reporter for the synthesis. It revealed a relatively fast (2–8 min) two-step transformation from a micellar-like spectrum to a mesophase-like spectrum for both structures. What we actually observed is the rate of polymerization
Table 1 Hyperfine splitting Azz, rotational correlation time sc and k2,0 of the slow component of 5-DSA in the synthetic gel of the hexagonal and of the cubic materials. k2,0 is the coefficient of the orienting potential, which reflects the constraints upon the nitroxide that restrict its range of orientations.
Hexagonal Cubic
Azz [MHz]
sc [ns]
k2,0
101.3 ± 1.5 100.3 ± 1.5
2.1 ± 0.3 2.7 ± 0.3
1.7 ± 0.1 1.8 ± 0.1
297
that leads to precipitation, and this is slow. In contrast, the spectrum of SL1SiEt was invariant throughout the reaction, revealing in addition to the flexibility of the probe itself also the fluidity of the forming silica gel silica. Drying entails rigidity to the silica wall as manifested in the considerable reduced motion of SL1SiEt. The addition of water, after drying restores partially the mobility of the as-synthesized materials. The same behavior is observed in the material after surfactant extraction. This suggests that the silica wall is not fully polymerized, and thus can get partially rehydrated. This behavior was observed earlier in the synthesis of cationic surfactant templated silica MCM-41 [40], and is supported by our 29Si MAS NMR results showing the presence of a fair amount of Q2 and Q3 silica species. SL1SiEt in the surfactant-extracted materials exhibit a high mobility because the pores have been freed from the surfactant which restricted their motion in the assynthesized materials. A similar behavior is expected for the CSDA. Furthermore the wall surface that may have been rich in SiO groups due to the basic character of the reaction [54], became fully protonated after the acidic treatment, affecting the possible interaction between the wall and the CSDA. This, in turn makes the region of the SL1SiEt less crowded. Direct experimental evidence for the interaction of the negatively charged template and the positively charged co-structure directing agent (CSDA) came from the appearance of 14N signals in the FT-ESEEM spectra of 5-DSA in both phases. The interaction between the nitroxide spin label and the 14N nuclei is dipolar and depends on the inverse of the cube of the interspin distance, and is therefore short range (<8 nm), particularly for low c nuclei such as 14N. Therefore, the appearance of significant 14N modulations indicates close proximity of the polar head of the surfactant and the ammonium of the CSDA. This confirms the proposed surfactant–CSDA interaction mechanism [14]. Similar to 5-DSA, the silica probe exhibited a 14N signal in the FT-ESEEM spectra in both phases, again without a noticeable difference between the hexagonal and cubic phases. These results provide evidence that SL1SiEt indeed copolymerized with the TEOS and the CSDA, and that it is pending into the pore as the CSDA. The functionalized pores are well accessible to molecules after the surfactant was removed. An intriguing, unexpected observation was the absence of 2H nuclei in the vicinity of SL1SiEt in both the as-synthesized cubic and hexagonal materials prepared in D2O. This is in marked contrast to the 5-DSA results. This suggests that spin label in SL1SiEt is situated close to the surface and that this surface is not hydroxylated, but mostly constituted of SiO. This is due to the high pH of the reaction, silanol having a isoelectric point pI 2, which leads to a surface rich in SiO [6,54]. Alternatively, SL1SiEt could be trapped in the silica wall. Indeed the EPR spectrum of SL1SiEt after solvent extraction contains a fraction of rigid probes, but the majority is highly mobile, indicating that most of it is in the pores. The absence of 2H near SL1SiEt implies that the 5-DSA and SL1SiEt probe different regions of the pores. 5-DSA probes the surfactant–CSDA interface, while SL1SiEt probes the silica wall region. Accordingly, the CSDA molecules sensed by each probes belong to different populations, CSDAs interacting with the surfactant, probed by 5-DSA, and the other, CSDAw, bending towards the wall, probed by SL1SiEt, as shown schematically on Fig. 9. The most significant experimental difference we observed between the two phases is in the degree of mobility of the spin probes. 5-DSA is slightly more mobile in the final wet gel of the hexagonal phase reaction while SL1SiEt is more mobile in the assynthesized cubic phase. We attribute these differences to the pH difference as follows: The lower pH in the hexagonal phase reaction yields less negatively charged surfactants molecules that interact directly with the positively charged CSDA, namely less CSDAs. This allows some additional motional freedom to the surfactant in the hexagonal phase (see Fig. 1).
298
M. Florent, D. Goldfarb / Microporous and Mesoporous Materials 163 (2012) 291–299
surfactants exist at higher pH, and so more CSDA can interact with them. A soft extraction method by an acid ethanolic solvent preserved the final phase structure, yet it removed the molecular differences between the two phases. It freed the pores of surfactant molecules presents in the as-synthesized material, and protonate the surface allowing the pending ammonium groups a high degree of freedom and accessibility to other molecules. Acknowledgement We thank Dr. Ronit Popovitz-Biro for the TEM measurements, Dr. Tali Amitay-Rosen for the NMR measurements and Dr. Veronica Frydman for the synthesis of the spin probes. This research was made possible in part by the historic generosity of the Harold Perlman Family. D.G. holds the Erich Klieger Professorial Chair in Chemical Physics. Appendix A. Supplementary data Fig. 9. A schematic representation of a pore with CSDAs interacting with negatively charged surfactants and the CSDAs interacting with the silica wall.
The postulation of the existence of two distinct populations of CSDA also accounts for the observed difference in the mobility of SL1SiEt in the as-synthesized cubic and hexagonal phases. If less CSDA interact with the surfactant, more will bend toward the wall (CSDAw), where they may be attracted by the presence of SiO groups, producing a more crowded environment at the surface of silica wall. This in turn will be manifested by a lower mobility of the SL1SiEt probe. Similarly, increasing the pH to prepare the cubic materials will increase the number of ionized surfactants, thus shifting the equilibrium CSDAw ¢ CSDAs , towards CSDAs. This will decrease the mobility at the surfactant–CSDA interface, and increase the mobility close to the wall. While difference in the envisioned relative populations of CSDAs and CSDAw could be picked up by the probe mobilities, it was too small to be resolved by the ESEEM experiments. In the case of 5-DSA the faster motion in the hexagonal phases was resolved only in the wet gel because in the as-synthesized material the motion is too close to the rigid limit and is therefore not sensitive to small changes. The differences in mobility of SL1SiEt observed in as-synthesized hexagonal and cubic materials are abolished after solvent extraction as now there are no surfactant molecules that were the source of the differences. Our experimental results are not sufficient for resolving the significance of the observed two populations of the CSDA molecules, and particularly their ratio, in controlling the final AMS phase structure. This would require new experiments aiming at detecting two types of CSDA molecules in the reaction mixture, exploring the time evolution of their ratio and its pH dependence.
5. Conclusion CW EPR and ESEEM measurements gave experimental evidence for the Coulomb surfactant–CSDA interaction that drives the formation of the AMS materials. This interaction leads to a fast assembly of a mesophase followed by a slow precipitation process. Throughout this process the forming silica remains rather fluidic and becomes rigid only upon drying. The EPR results were interpreted in terms of the presence of two populations of CSDA molecules; one interacting with the anionic surfactant and the other with the silica wall, which is mostly ionized. The major observed molecular level differences between the two as-synthesized phases were attributed to different relative population of these two types of CSDA. This, in turn, was correlated with the pH; more ionized
XRD patterns and TEM micrographs of the final materials, 29Si MAS NMR spectrum of the as-synthesized cubic material and ESEEM traces of 5-DSA in the as-synthesized materials. Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.micromeso. 2012.06.050. References [1] J.S. Beck, J.C. Vartuli, W.J. Roth, M.E. Leonowicz, C.T. Kresge, K.D. Schmitt, C.T.W. Chu, D.H. Olson, E.W. Sheppard, S.B. Mccullen, J.B. Higgins, J.L. Schlenker, J. Am. Chem. Soc. 114 (1992) 10834. [2] C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.C. Vartuli, J.S. Beck, Nature 359 (1992) 710. [3] Q.S. Huo, D.I. Margolese, U. Ciesla, P.Y. Feng, T.E. Gier, P. Sieger, R. Leon, P.M. Petroff, F. Schüth, G.D. Stucky, Nature 368 (1994) 317. [4] Y. Wan, D.Y. Zhao, Chem. Rev. 107 (2007) 2821. [5] L.L. Hench, J.K. West, Chem. Rev. 90 (1990) 33. [6] R.K. Iler, The Chemistry of Silica: Solubility, Polymerization, Colloid and Surface Properties and Biochemistry of Silica, Wiley, 1979. [7] F. Kleitz, S.H. Choi, R. Ryoo, Chem. Commun. (2003) 2136. [8] F. Kleitz, D.N. Liu, G.M. Anilkumar, I.S. Park, L.A. Solovyov, A.N. Shmakov, R. Ryoo, J. Phys. Chem. B 107 (2003) 14296. [9] S.D. Shen, A.E. Garcia-Bennett, Z. Liu, Q.Y. Lu, Y.F. Shi, Y. Yan, C.Z. Yu, W.C. Liu, Y. Cai, O. Terasaki, D.Y. Zhao, J. Am. Chem. Soc. 127 (2005) 6780. [10] D.Y. Zhao, Q.S. Huo, J.L. Feng, B.F. Chmelka, G.D. Stucky, J. Am. Chem. Soc. 120 (1998) 6024. [11] S. Che, A.E. Garcia-Bennett, T. Yokoi, K. Sakamoto, H. Kunieda, O. Terasaki, T. Tatsumi, Nat. Mater. 2 (2003) 801. [12] S. Che, Z. Liu, T. Ohsuna, K. Sakamoto, O. Terasaki, T. Tatsumi, Nature 429 (2004) 281. [13] C. Gao, Y. Sakamoto, O. Terasaki, K. Sakamoto, S. Che, J. Mater. Chem. 17 (2007) 3591. [14] C.B. Gao, H.B. Qiu, W. Zeng, Y. Sakamoto, O. Terasaki, K. Sakamoto, Q. Chen, S.A. Che, Chem. Mater. 18 (2006) 3904. [15] C.B. Gao, Y. Sakamoto, O. Terasaki, S.A. Che, Chem. Eur. J. 14 (2008) 11423. [16] A.E. Garcia-Bennett, N. Kupferschmidt, Y. Sakamoto, S. Che, O. Terasaki, Angew. Chem. Int. Ed. 44 (2005) 5317. [17] J.N. Israelachvili, D.J. Mitchell, B.W. Ninham, J. Chem. Soc. Farad. T2 (72) (1976) 1525. [18] G.J.D. Soler-illia, C. Sanchez, B. Lebeau, J. Patarin, Chem. Rev. 102 (2002) 4093. [19] Q.S. Huo, R. Leon, P.M. Petroff, G.D. Stucky, Science 268 (1995) 1324. [20] S.H. Tolbert, C.C. Landry, G.D. Stucky, B.F. Chmelka, P. Norby, J.C. Hanson, A. Monnier, Chem. Mater. 13 (2001) 2247. [21] H.Q. Zheng, C.B. Gao, S.N. Che, Microporous Mesoporous Mater. 116 (2008) 299. [22] A.E. Garcia-Bennett, U. Brohede, R.P. Hodgkins, N. Hedin, M. Strømme, Langmuir 23 (2007) 9875. [23] C. Madelmont, R. Perron, Colloid Polym. Sci. 254 (1976) 581. [24] B. Lin, A.V. McCormick, H.T. Davis, R. Strey, J. Colloid Interface Sci. 291 (2005) 543. [25] X.Y. Wen, E.I. Franses, J. Colloid Interface Sci. 231 (2000) 42. [26] R.G. Herman, J.H. Lunsford, H. Beyer, P.A. Jacobs, J.B. Uytterhoeven, J. Phys. Chem. 79 (1975) 2388. [27] A. Moscatelli, A. Galarneau, F. Di Renzo, M.F. Ottaviani, J. Phys. Chem. B 108 (2004) 18580. [28] M.F. Ottaviani, R. Mazzeo, N.J. Turro, X.G. Lei, Microporous Mesoporous Mater. 139 (2011) 179.
M. Florent, D. Goldfarb / Microporous and Mesoporous Materials 163 (2012) 291–299 [29] B. Jee, K. Koch, L. Moschkowitz, D. Himsl, M. Hartman, A. Pöppl, J. Phys. Chem. Lett. 2 (2011) 357. [30] L. Kevan, M. Hartmann, in: B.C. Gilbert, N.M. Atherton, M.J. Davies (Eds.), Electron Paramagnetic Resonance, The Royal Society of, Chemistry, 1998. [31] J. Zhang, D. Goldfarb, Microporous Mesoporous Mater. 48 (2001) 143. [32] J.Y. Zhang, Z. Luz, D. Goldfarb, J. Phys. Chem. B 101 (1997) 7087. [33] J.Y. Zhang, H. Zimmermann, Z. Luz, D. Goldfarb, Mesoporous Mol. Sieves 1998 117 (1998) 535. [34] H. Caldararu, A. Caragheorgheopol, F. Savonea, D.J. Macquarrie, B.C. Gilbert, J. Phys. Chem. B 107 (2003) 6032. [35] M.F. Ottaviani, A. Moscatelli, D. Desplantier-Giscard, F. Di Renzo, P.J. Kooyman, B. Alonso, A. Galarneau, J. Phys. Chem. B 108 (2004) 12123. [36] A. Galarneau, F. Di Renzo, F. Fajula, L. Mollo, B. Fubini, M.F. Ottaviani, J. Colloid Interface Sci. 201 (1998) 105. [37] J.Y. Zhang, P.J. Carl, H. Zimmermann, D. Goldfarb, J. Phys. Chem. B 106 (2002) 5382. [38] S. Ruthstein, V. Frydman, S. Kababya, M. Landau, D. Goldfarb, J. Phys. Chem. B 107 (2003) 1739. [39] S. Ruthstein, V. Frydman, D. Goldfarb, J. Phys. Chem. B 108 (2004) 9016. [40] D. Baute, V. Frydman, H. Zimmermann, S. Kababya, D. Goldfarb, J. Phys. Chem. B 109 (2005) 7807.
299
[41] D. Goldfarb, in: M. Brustolon, E. Giamello (Eds.), Electron Paramagnetic Resonance – A Practitioner’s Toolkit, Wiley, 2009. [42] S. Ruthstein, D. Goldfarb, Electron Paramagnetic Resonance, vol. 21, The Royal Society of Chemistry, 2008. 184p. [43] R. Ciriminna, J. Blum, D. Avnir, M. Pagliaro, Chem. Commun. (2000) 1441. [44] K. Joon Oh, C. Altenbach, R.J. Collier, W.L. Hubbell, in: O. Holst (Ed.), Bacterial toxins: Methods and Protocols, vol. 145, Humana Press, 2000, p. 147. [45] S. Stoll, A. Schweiger, J. Magn. Reson. 178 (2006) 42. [46] J.M. Fauth, A. Schweiger, L. Braunschweiler, J. Forrer, R.R. Ernst, J. Magn. Reson. 66 (1986) 74. [47] L. Kevan, R.N. Schwartz, Time domain electron spin resonance, John Wiley & Sons, New York, 1979. 279p. [48] S. Van Doorslaer, G.A. Sierra, A. Schweiger, J. Magn. Reson. 136 (1999) 152. [49] R. Yokoi, H. Yoshitake, T. Tatsumi, Pts a - C 154 (2004) 519. [50] F. Kleitz, W. Schmidt, F. Schüth, Microporous Mesoporous Mater. 65 (2003) 1. [51] Z.L. Hua, J.L. Shi, L. Wang, W.H. Zhang, J. Non-Cryst, Solids 292 (2001) 177. [52] C.B. Gao, S.A. Che, Adv. Funct. Mater. 20 (2010) 2750. [53] E.G. Rozantsev, V.D. Sholle, Russ. Chem. Rev. 40 (1971) 233. [54] N. Baccile, G. Laurent, C. Bonhomme, P. Innocenzi, F. Babonneau, Chem. Mater. 19 (2007) 1343.
Contents lists available at SciVerse ScienceDirect
Microporous and Mesoporous Materials journal homepage: www.elsevier.com/locate/micromeso
The interaction between the surfactant and the co-structure directing agent in anionic surfactant-templated mesoporous silicas Marc Florent, Daniella Goldfarb ⇑ Department of Chemical Physics, Weizmann Institute of Science, Rehovot 76100, Israel
a r t i c l e
i n f o
Article history: Received 14 May 2012 Received in revised form 7 June 2012 Accepted 29 June 2012 Available online 10 July 2012 Keywords: Ordered mesoporous materials Co-structure directing agent EPR ESEEM
a b s t r a c t Anionic surfactant-templated mesoporous silicas (AMS) are synthesized with a co-structure directing agent (CSDA) that interacts with both the organic and inorganic components of the system. The AMS materials structure is controlled by pH. We investigated the formation of AMS cubic and hexagonal phases, prepared under the same conditions, except pH, by EPR spectroscopic measurements. We used silica-like and surfactant-like spin probes added to the reaction mixtures in minute amounts. Through the spin probes we resolved the specific interactions of the CSDA (N-trimethoxylsilylpropyl-N,N,N-trimethyl ammonium chloride (TMAPS)) with the surfactant (sodium myristate (C14AS)) and the polymerizing silica. We observed for both phases a fast formation of a mesophase upon addition of the silica precursor (TEOS, tetraethoxysilane) and the CSDA into the surfactant solution, attributed to the strong attraction between the CSDA and the anionic surfactant. This is then followed by a slow condensation of the silica. Electron spin echo envelope modulation (ESEEM) spectra of both spin probes in the as-synthesized materials indicated the presence of two types of CSDA molecules; one interacting with the surfactant and the other with the silica wall. Continuous wave EPR spectra showed different spin probe motilities in the two as-synthesized materials that indicated that the relative populations of the two CSDA types are different in the two phases. We attribute this difference to the pH differences in the reaction mixtures. A soft extraction of the surfactant from the pores did not alter the structure of the final materials, but it abolished the observed molecular level differences between them. The extraction allowed the pending ammonium groups to acquire a high degree of freedom and accessibility to water molecules. Ó 2012 Elsevier Inc. All rights reserved.
1. Introduction Synthesis of ordered mesoporous silicas relies on the selfassembly of a surfactant template and inorganic silica precursors into a mesostructure [1,2]. This self-assembly depends on the intimate interactions between the surfactant and the inorganic oxide source. Accordingly, different synthesis routes were classified according to the type of the interaction: S+I, SI+, S+XI+, SX+I, S0I0, S0H+XI+, and so on, where S, I and X represent the surfactant, the inorganic (silica) precursors, and a counter ion, respectively, with their charge depending on whether they are in a cationic, anionic or neutral form [3,4]. The synthesis of ordered mesoporous materials is based on a sol–gel process, in acidic or basic environment [5]. Therefore, the choice of surfactant and reaction conditions determine the interactions driving the mesostructure formation. For example, in the case of silica materials the pH is crucial, considering that the isoelectric
⇑ Corresponding author. Address: Weizmann Institute of Science, Chemical Physics Department, Rehovot 76100, Israel. Tel.: +972 8 934 2341; fax: +972 8 934 4123. E-mail address: [email protected] (D. Goldfarb). 1387-1811/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.micromeso.2012.06.050
point of silicates species is around 2 [6]. Cationic and neutral nonionic surfactants have been proven very successful in forming diverse mesostructures, in acidic or basic conditions [1,7–10]. However, anionic surfactants have not yielded ordered mesostructure silicas unless a co-structure directing agent (CSDA) was added to the synthesis [11]. This CSDA has an alkoxysilane on one end that can co-condense with the silica precursor, and an ammonium group on the other end that can interact with the negatively charged head group of the surfactant, actually acting as glue. Using this strategy, a variety of structures, including cubic, tetragonal, hexagonal, bicontinuous cubic, and lamellar, as well as chiral structures, were successfully prepared [11–16]. The different mesostructures obtained were explained to some extent by the packing parameter of the surfactant g = v/a0l, where v is the chain volume of the surfactant, a0 is the effective hydrophobic/hydrophilic interfacial area and l is the kinetic surfactant chain length [17,18]. This parameter depends on the structure of the components of the system and predict the curvature of the aggregate, and therefore the final mesostructure. The g parameter has been shown to be sensitive to the surfactant concentration, temperature and the presence of other species in the solution [19,20].
292
M. Florent, D. Goldfarb / Microporous and Mesoporous Materials 163 (2012) 291–299
Fig. 1. A schematic representation of the change of curvature of the surfactant micelles by modifying the ionization degree of the anionic surfactant through a change of pH. The positively charged CSDA is shown as well.
It has been proposed that the packing parameter of the anionic surfactant can be controlled through the degree of ionization of the surfactant, which depends on the pH, as shown on Fig. 1. The degree of ionization of the polar head of the carboxylic fatty acids increases with pH, thereby increasing its effective headgroup area due to repulsion and leading to a lower g value. Indeed, by increasing the pH of the reaction mixture a transition from lamellar to cylindrical and to cage-type mesophases was observed. This, in turn, was correlated with an increase of curvature due to the decrease of g [14]. Other parameters reported to affect the structure of the anionic surfactant-templated silicas are the time delay between the addition of CSDA and the silica source [16], the structures of the CSDA and of the surfactant [13], and the molecular CSDA/surfactant ratio [15]. In addition to the possibility to control the structure of the final phase, the anionic pathway received attention also for its ability to easily provide functionalized mesoporous silicas. The possible removal of the surfactant by extraction, instead of calcination, allows retaining the amino or ammonium group of the CSDA in the pore as it is bound to the silica walls [21]. 13C NMR and elemental analysis of the functionalized material thus obtained showed a homogeneous distribution of amine or ammonium groups. Their loading amount was reported to be limited by the quantity of CSDA that can interact electrostatically with the anionic surfactant [21]. A previous study on the synthesis of AMS, using the acid form of a surfactant and a CSDA with an amino group, showed that the reaction takes different paths, depending on the time between the addition of CSDA and the addition of the silica source, yielding different phases. Following its addition into the solution, the CSDA was reported to quickly interact with the surfactant and to start condensing. An organo-silica layer made of condensed CSDA was shown to first form around the micelles in solution. The inorganic silica species condense later to form an outer inorganic silica shell. As the reaction goes on, the system just shows an increase of the fraction of condensed inorganic silica [22]. Herein we used spin probe EPR methods to explore the interplay between the surfactant, CSDA and the silica in hexagonal (p6mm) and cubic (Fd3m) structures. We used a surfactant in its salt form, sodium myristate (C14AS), and a CSDA with a quaternary
ammonium group, N-trimethoxylsilylpropyl-N,N,N-trimethyl ammonium chloride (TMAPS). In this combination, the ionization degree of the CSDA is not modified by pH while that of the surfactant is. C14AS is a common surfactant whose phase diagram in water is well known [23]. Its 1% wt Krafft point has been determined around 44 °C, above which it dissolves in water to form a clear micellar solution [24]. It is known that at equilibrium a dilute aqueous solution of sodium myristate (C < 20 mM, corresponding to pH < 10) have a protonation fraction of about 1%, but this has a strong influence on the solution and phase behavior [25]. By adding NaOH the protonation can be further decreased, but the solubility of C14AS decreases as well. The addition of other electrolytes, and especially tetraalkyammonium salt affects also significantly the solution behavior of C14AS [24]. Spin probe EPR has been proven successful to characterize pores and surface properties [26–30]. The formation of mesoporous silica materials, synthesized with cationic or neutral surfactant templates, has also been monitored using in situ continuous wave (CW)-EPR [31–36], and more advanced pulse EPR experiments, such as electron spin echo envelope modulation (ESEEM) [37– 39]. In general, surfactant-like nitroxides, were used to probe local changes in the organic phase, and a nitroxide linked to an alkoxysilane was employed to give information on the formation of the silica wall [40]. The CW-EPR spectrum of the spin probes provide information on the dynamics and polarity of the microenvironment of the nitroxide label, while ESEEM experiments give information on its local structure via the hyperfine interaction with nearby magnetic nuclei [41,42]. To investigate the formation mechanism and inner surface properties of as-synthesized cubic (Fd3m) and 2D-hexagonal (p6mm) materials, with emphasis on the CSDA interactions with the other components of the reaction, we used 5-doxyl stearic acid (5-DSA) as a surfactant-like probe, and a silica-like probe, 2,2,6,6tetramethyl-4-[3-(triethoxysilyl)propylamino]-1-piperidinooxy (SL1SiEt) that can co-condense with the silica precursor (see Fig. 2). Whatever the phase, we observed that the addition of the silica mixture (CSDA + silica precursor, (TEOS, tetraethoxysilane)) leads to a fast (2–8 min) formation of a mesophase, whose exact kinetics is below our experimental resolution. We could only observe the
Fig. 2. (a) The nitroxide radical, SL1SiEt, used to probe the silica region and (b) the 5-DSA employed for probing the surfactant part of the system.
M. Florent, D. Goldfarb / Microporous and Mesoporous Materials 163 (2012) 291–299
evolution of the silica condensation, which is slower. Therefore we focused on the properties of the as-synthesized materials. These underlined the strong interaction between the CSDA and the negatively charged surfactant evidenced by clear interactions of the spin label with the 14N nuclei of CSDA. In addition, our results indicate the existence of CSDA interacting with the silica surface. The ratio between these two CSDA populations varies from one phase to the other and this is related to their different reaction pH. This is also reflected in the local mobility of the nitroxide probes in the vicinity of the wall and at the surfactant–CSDA interface. A soft removal of the surfactant generates pores covered by pending ammonium groups, which are highly mobile and accessible to small molecules. 2. Materials and methods 2.1. Materials Chemicals were used as received. Sodium myristate (C14AS) was purchased from Fluka, tetraethoxysilane (TEOS) from Aldrich and the CSDA, N-trimethoxylsilylpropyl-N,N,N-trimethyl ammonium chloride (TMAPS) from Gelest, Inc. The nitroxide spin probe 5-doxyl stearic acid (5-DSA) was purchased from Aldrich and the silica nitroxide spin probe, 2,2,6,6-tetramethyl-4-[3-(triethoxysilyl)propylamino]-1-piperidinooxy (SL1SiEt), was synthesized as described in the literature [40,43]. Nickel(II) ethylene diaminediacetic acid (NiEDDA) was prepared as described in the literature [44]. 2.2. Synthesis Materials were synthesized following a procedure described in the literature [14,21]. The anionic surfactant, C14AS, was first dissolved in water (or deuterated water) at 80 °C. Sodium hydroxide was added when the aimed final material was cubic. Cooling to 60 °C, a mixture of silica precursor, TEOS, and CSDA, TMAPS, was added while stirring. After stirring for ten more minutes, the reaction mixture was aged for two days at 60 °C. The precipitate was then filtered, washed and dried overnight at 80 °C to obtain the as-synthesized material. Two paramagnetic probes, 5-DSA, a surfactant-like, or SL1SiEt, a silica-like nitroxide probe, shown in Fig. 2, were added in minute amount to the reaction mixture to probe the reaction and the final material. The molar composition of the reaction mixture was C14AS/TMAPS/TEOS/H2O/Spain Probe/NaOH = 1/1/7/1389/0.006/x, where x = 0, or 0.1 to obtain the 2D-hexagonal (p6 mm) or the cu bic (Fd3m) materials respectively. 2.3. Extraction of the surfactant Final materials free of surfactant were obtained according to the two following procedures. The as-synthesized silica material was calcined at 600 °C for 6 h to give a surfactant-free mesoporous silica. Alternatively, to get a surfactant-free mesoporous silica functionalized with alkoxy ammonium group, the surfactant was extracted from the as-synthesized material by an ethanolic solution of HCl (11 vol.%), for 24 h at boiling temperature. The material was then filtered, washed with ethanol, and dried overnight at 80 °C.
293
300 mesh copper grid and air-dried. Digital images were recorded by a Gatan US1000 CCD camera, and fast Fourier transforms (FFT) of the images were calculated by DigitalMicrograph software. Further characterization of the final materials was done by Xray diffraction (XRD) on a Rigaku TTRAX-III diffractometer, using a CuKa radiation (k = 0.154 nm) over the range of 0.3–5° 2h with steps of 0.02° 2h, or by small angle X-ray scattering (SAXS), on a home-built diffractometer equipped with Franks mirror and a linear position-sensitive detector, also using a Ni-filtered CuKa radiation. In both cases quartz capillaries of 1.5 mm diameter were used. 29 Si MAS NMR (magic angle spinning nuclear magnetic resonance) experiment was performed on a Bruker DSX spectrometer operating at 59.6 MHz, equipped with a Bruker 4 mm MAS probe. The experiment was carried out at a spinning frequency of 5 kHz, by employing a spin-echo pulse sequence with a 90° excitation pulse of 4.4 ls, an echo delay of 200 ls, and a repetition delay of 600 s. The chemical shift ppm scale referred to tetramethylsilane (TMS). Proton decoupling was employed during the acquisition, using the two pulse phase modulation (TPPM) scheme with 70 kHz decoupling power and Du = 17°. CW-EPR spectra were recorded on a Bruker Elexsys 500 spectrometer or a modified Varian E12 X-band spectrometer equipped with a temperature controller. Solid materials were introduced into EPR quartz tubes of 3 mm outer diameter. For the experiments of solvent accessibility, the product was suspended in water, or water solutions of NiEDDA, and introduced into a glass capillary of 0.75 mm diameter. EPR spectra were recorded at room temperature, for different microwave power. Measurements during the formation of the AMS were done by sampling the reaction mixture at different times of the synthesis. Here a small amount was transferred into glass capillaries. The spectra were recorded at the reaction temperature, 60 °C. Sampling a multiphase mixture is always a problem in terms of the relative quantity taken of each phase. Unfortunately, attempted in situ synthesis did not produce ordered materials, probably because of the too small reaction vessel required and the associated slow rate of ethanol evaporation. Simulations of the EPR spectra of the wet mesophases at the end of reaction were done using the EasySpin functions chili and garlic, which calculate slow motion and fast motion spectra respectively [45]. Three pulse ESEEM traces were recorded with an X-band pulse Bruker Elexsys 580 spectrometer at 85 K. The ESEEM experiments were done using the three-pulse pulse sequence p/2–s–p/2–T–p/ 2–s–echo, with a four-step phase cycle [46]. The interval s between the first two pulses was calculated to give the maximum deuterium signal, when 2H was the nucleus of interest, or the maximum 14 N signal and minimal proton signal when 14N was the nucleus of interest [47]. ESEEM traces were all treated identically: after phase correction and normalization, the background decay of the normalized data was divided using a biexponential fit, then the data was apodized with a Hamming window, zero filled to 512 points and Fourier transformed (FT) with cross-term averaging [48]. The FTESEEM spectra consist of peaks centered at the Larmor frequency of the nuclei in the vicinity of the nitroxide. The intensity of these peaks reflects the proximity and density of nuclei around the nitroxide.
2.4. Methods
3. Results
The final surfactant free materials were characterized by transmission electron microscopy (TEM) with a Philips CM120 supertwin microscope operated at 120 kV. A dispersion of final material in ethanol was deposited on a carbon/collodion-coated
3.1. As-synthesized cubic and hexagonal AMS Two different phases, hexagonal (p6mm) and cubic (Fd3m), were prepared by varying the pH of the same reaction mixture.
294
M. Florent, D. Goldfarb / Microporous and Mesoporous Materials 163 (2012) 291–299
The final materials were characterized by TEM (transmission electron microscopy) and XRD (X-ray diffraction) (Supporting information Fig. S1). The cell parameters, a = 5.9 nm and a = 21.2 nm for the hexagonal and the cubic materials, respectively, are similar to that reported in the literature [14]. 29 Si MAS NMR spectrum of the as-synthesized cubic AMS (Fig. S2) shows the presence of CSDA molecules condensed with the silica. Furthermore it shows that a large amount of inorganic silica is only partially condensed, with a fair amount of Q3 and an appreciable amount of Q2 silica species, as reported in the literature [49]. The CW-EPR spectra of 5-DSA and SL1SiEt in the dry as-synthesized hexagonal (obtained without NaOH) and cubic materials (obtained with NaOH, x = 0.1, see experimental section) recorded at 150 K, room temperature (RT) and 335 K are shown in Fig. 3. The spectrum recorded at 150 K is identical for both materials and is typical for an immobilized nitroxide spin label and is therefore shown for reference. The spectra of the surfactant-like spin probe, 5-DSA, show that at RT it is rigid on the EPR time scale (rotational correlation time is longer than 1 ls) and becomes slightly more mobile as the temperature increases to 335 K. The SL1SiEt probe, by contrast, is much less rigid. It is in a slow motion regime already at room temperature, and by increasing temperature, the probe becomes much more mobile. Here we observe differences between the hexagonal and cubic phases; in the hexagonal phase; in the hexagonal phase, SL1SiEt is less mobile. We attribute the difference in mobility between the two different probes, 5-DSA and SL1SiEt, to differences in their structure (Fig. 2). In 5-DSA, the nitroxide ring is directly connected to the alkyl chain through a common shared carbon atom. Therefore the nitroxide ring does not experience local motion (bond rotation, etc.) and it can only rotate along with the whole alkyl chain. In the case of SL1SiEt, the nitroxide ring is connected to the polymerizing silica through four single bonds, allowing considerable local motion. To learn more about the local environment of the two spin probes we carried out ESEEM measurements. Fig. 4 shows the
FT-ESEEM of 5-DSA in the dry hexagonal and cubic as-synthesized materials prepared in D2O (the time domain traces are shown in Fig. S3). The FT-ESEEM spectra recorded under conditions optimizing 14N or 2H modulations show peaks at the 14N, 2H, and 1H Larmor frequencies, arising from weakly coupled nuclei. 5-DSA clearly reports the presence of deuterium in its vicinity, which arises from deuterated surfactant (exchangeable OD groups) or D2O. In contrast, the FT-ESEEM spectra of SL1SiEt (Fig. 5) in both phases, acquired under the same experimental conditions, comprise signals of 14N and 1H only. This indicates that SL1SiEt probes a region farther away from the surfactant head groups. Interestingly, both probes reveal clear 14N modulation from the ammonium group nitrogen of the CSDA. To exclude the possibility that the 14N signal observed for SL1SiEt arises from 14N of the probe itself we carried out ESEEM measurements on SL1SiEt in toluene; no 14 N modulations were observed. The 14N modulation cannot arise from the 14N of the nitroxide group because it is strongly coupled and have large hyperfine couplings that lead to nuclear frequencies far away from the Larmor frequency. The presence of the 14N modulation from the co-surfactant nitrogen provides direct experimental evidence for its proximity to the surfactant-like probe, 5-DSA. This confirms the CSDA–surfactant interaction mechanism, where the negatively charged polar heads of the micelle interact with the ammonium group of the CSDA. Moreover, the observation of 14N ESEEM also for SL1SiEt indicates that it is also close to some co-surfactant molecule. However, the fact that the silica probe does not report on 2H nuclei around it, in contrast to 5-DSA, but still experience 14N modulation from CSDA molecules suggests that SL1SiEt reports on CSDA molecules which are not interacting with the anionic head of the surfactant. The intensity of the 14N peak in the SL1SiEt FT-ESEEM spectra is lower than that reported by the 5-DSA probe. This implies that the SL1SiEt probe is farther away from the ammonium group of the CSDA or that there are less interacting CSDA molecules around it. Interestingly, one can notice that the intensity of 14N in the cubic material is about the same as in the hexagonal one, though, in the
Fig. 3. X-band CW-EPR spectra of (a and b) 5-DSA and (c and d) SL1SiEt in the as-synthesized (a and c) hexagonal and (b and d) cubic materials recorded at 150 K, RT and 335 K, as indicated in the figure. In (c), and (d) the CW-EPR spectra of SL1SiEt in the final material obtained after surfactant extraction by HCl/ethanol treatment of the cubic and hexagonal materials at RT are also presented. The asterisk indicates the presence of a highly mobile component in small quantity.
M. Florent, D. Goldfarb / Microporous and Mesoporous Materials 163 (2012) 291–299
295
Fig. 4. FT-ESEEM traces of 5-DSA in the as-synthesized (a and c) hexagonal, and (b and d) cubic materials, recorded with (a and b) s ’ 220 ns, optimized for 2H modulation, and with (c and d) s ’ 480 ns, optimized for 14N modulation and minimization of 1H modulation.
Fig. 5. FT-ESEEM traces of SL1SiEt in the as-synthesized (a and c) hexagonal, and (b and d) cubic material, recorded with (a and b) s ’ 220 ns, optimized for 2H modulation, and with (c and d) s ’ 480 ns, optimized for 14N modulation and minimization of 1H modulation. FT-ESEEM traces after surfactant extraction by HCl/ethanol in (e) the hexagonal and (f) cubic materials.
synthesis of the latter, the lower degree of ionization of the surfactant should lead to a lower number of CSDA molecules interacting with it (see Fig. 1) and therefore to a lower intensity of the 14N peak. 3.2. AMS materials after template removal Mesoporous materials can be used in various applications only after the pores are freed from the surfactant. The classical method
to remove the template is to calcine the solid [50]. Another method is to extract the surfactant with an acid ethanolic solution [21,51,52]. In the case of the AMS materials this soft method is particularly interesting because it yields a functional material with the alkyl ammonium groups pending in the pores [21]. The as-synthesized cubic and hexagonal materials were subjected to such a treatment and the final structure of the phases did not change. The materials synthesized with 5-DSA had no EPR signal after this treatment showing that the surfactant-like spin probe is washed
296
M. Florent, D. Goldfarb / Microporous and Mesoporous Materials 163 (2012) 291–299
out along with the anionic surfactant. In contrast, the signal of SL1SiEt persists also after the extraction of the surfactant, as shown in Fig. 3. This indicates that this probe is attached to the silica wall, confirming that its siloxane group polymerized with TEOS. Yet, the signal to noise ratio of the spectrum decreased after the surfactant extraction, most probably because part of the nitroxides were transformed into hydroxylamines by the acid treatment [53]. The lineshape of SL1SiEt in the final surfactant free hexagonal and cubic phases shows the presence of two components: a majority species that is considerably more mobile than SL1SiEt before solvent extraction, and a minor more rigid species. Another very mobile species in a negligible quantity can also be seen and is indicated by ⁄ on Fig. 3. The amount of the latter is a little higher in the cubic phase than in the hexagonal phase, but beside that both phases present similar spectra after extraction. The higher mobility of the probe arises from the free space inside the pores obtained after the surfactant removal, allowing the pending probes to fully extend and rotate. The minor rigid species is assigned to polymerized SL1SiEt trapped inside the wall. In order to differentiate between SL1SiEt in the wall and pending into the pores we carried out accessibility measurements. The accessibility of the silica spin label to a water soluble paramagnetic relaxation agent, NiEDDA, was measured by CW-EPR. Here we expect to see broadening upon the interaction with the paramagnetic NiEDDA. The spectra of the final hexagonal material in different solutions of paramagnetic NiEDDA are shown on Fig. 6. It appears that before the surfactant extraction the SL1SiEt actually exists in three different environments: a rigid one, a slow motion one and a small quantity of a fast motion one. The free component disappears with the addition of NiEDDA, showing a high accessibility to water. This very free component may arise from some probes fixed at the external surface of the material and therefore they are highly accessible to the solution. The two others types do not seem to be affected by the introduction of NiEDDA, implying that the NiEDDA complex cannot enter the pores because they are occupied by the surfactants. After the HCl treatment, in the water suspension, the probe becomes very free to move, due to the removal of the surfactant, and the lineshape broadens with the addition of NiEDDA, showing that the probes are now more accessible to water. ESEEM measurements were carried out on the hexagonal and cubic materials with SL1SiEt after HCl treatment. In both materials the 14N FT-ESEEM signal, displayed in Fig. 5, shows a slight increase compared to the situation before the extraction, indicating a subtle increase of CSDA density in the vicinity of the SL1SiEt. After freeing the pores from the surfactant both the silica probe and the co-surfactant are pending into the pores and can freely move, bringing them closer to each other.
3.3. Measurements during synthesis True in situ CW-EPR measurements could not be carried out, as the synthesis performed in the small vessel required for EPR measurements did not yield ordered mesoporous materials. Therefore, samples for EPR measurements were drawn from the heterogeneous reaction mixture into EPR capillaries at different time points along the reaction. A major inconvenience of this method is that the relative quantity taken of each phase (solution or precipitate) every time, is hard to control. CW-EPR spectra of the silica probe SL1SiEt recorded during the synthesis did not undergo any change with time. The spectrum consists of three very narrow lines all along the reaction, demonstrating that the forming silica gel stays very fluid during the entire synthesis. In contrast, the spectrum of the surfactant-like spin probe 5-DSA undergoes significant changes as shown in Fig. 7. The changes are characteristic of a population of two species; the relative amount of one increases with time while that of the other decreases. At time zero, corresponding to the addition of the CSDA + TEOS mixture into the surfactant solution, the spectrum comprises a mobile species only. New features, marked with squares, appear already after 8 min. As the time progresses the intensities of these features, corresponding to a more rigid species, increase and those of the mobile species decrease. This behavior is
Fig. 7. CW-EPR spectra of 5-DSA in the reaction mixture during the synthesis of the cubic material recorded at different times t of the reaction. From top to bottom, t = [0;2;8;15;25;40;60;90;120;180;240;300:460;1210;1330] min where t = 0 min is the time of addition of the TEOS and CSDA mixture. s, Fast motion component; h, slow motion component. Below, the spectrum of the dry as-synthesized material is shown for comparison.
Fig. 6. CW-EPR spectra of dispersions of the hexagonal material in water, and water solutions of NiEDDA, (a) before (as-synthesized material) and (b) after HCl treatment (final material). s, Isotropic component; h, rigid component; , slow motion component. Inset: expanded view of the central lines.
M. Florent, D. Goldfarb / Microporous and Mesoporous Materials 163 (2012) 291–299
Fig. 8. CW-EPR spectra (black line) of 5-DSA in the synthetic gel, obtained after centrifugation of the reaction mixture at the end of the syntheses of the hexagonal and cubic phase, with their simulation (red line). The asterisks indicate the features of a free nitroxide in very small quantity. The simulations were done combining a slow motion component and a free one that holds for 1.5% of the total. Parameters used in the simulations were g = [2.0214 2.0196 2.0153], A = [13.8 13.6 103] MHz and sc = 0.4 ns for the free component. The same parameters were used for the slow motion one, and the parameters given in Table 1. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
typical or a A?B transformation. The appearance of the new signal, assigned to the formed mesophase, is associated with the formation of the precipitate. In general, the same behavior was observed for both hexagonal and cubic materials. The spectra of the precipitate (mesophase), which was separated by centrifugation of the final reaction mixture, without drying, are shown in Fig. 8. The spectra are typical of a slow motion regime with rotational motion restricted to some orientations due to an orienting potential. A small part, marked with *, corresponds to remaining 5-DSA in solution that give a fast motion component, but holds only for about 1.5 % of the whole nitroxides. As explained above, the rotational restriction can be observed because of the structure of 5-DSA that allows the nitroxide ring to rotate only with the whole surfactant chain. Simulations of these spectra gave the parameters listed in Table 1. The spectrum of the wet as-synthesized materials of the two phases have very similar properties, except for a small difference in the correlation time, with the hexagonal phase exhibiting slightly higher mobility.
4. Discussion Using EPR spectroscopy of two spins probes, a surfactant-like, 5-DSA, and a silica-like, SL1SiEt, we explored the formation mechanism of a cubic and hexagonal AMS material where the reaction mixture differ only in pH. In the following discussion we first address the features that were common to both and then illuminate minor differences that can be related to the pH difference. 5-DSA was found to be a better reporter for the synthesis. It revealed a relatively fast (2–8 min) two-step transformation from a micellar-like spectrum to a mesophase-like spectrum for both structures. What we actually observed is the rate of polymerization
Table 1 Hyperfine splitting Azz, rotational correlation time sc and k2,0 of the slow component of 5-DSA in the synthetic gel of the hexagonal and of the cubic materials. k2,0 is the coefficient of the orienting potential, which reflects the constraints upon the nitroxide that restrict its range of orientations.
Hexagonal Cubic
Azz [MHz]
sc [ns]
k2,0
101.3 ± 1.5 100.3 ± 1.5
2.1 ± 0.3 2.7 ± 0.3
1.7 ± 0.1 1.8 ± 0.1
297
that leads to precipitation, and this is slow. In contrast, the spectrum of SL1SiEt was invariant throughout the reaction, revealing in addition to the flexibility of the probe itself also the fluidity of the forming silica gel silica. Drying entails rigidity to the silica wall as manifested in the considerable reduced motion of SL1SiEt. The addition of water, after drying restores partially the mobility of the as-synthesized materials. The same behavior is observed in the material after surfactant extraction. This suggests that the silica wall is not fully polymerized, and thus can get partially rehydrated. This behavior was observed earlier in the synthesis of cationic surfactant templated silica MCM-41 [40], and is supported by our 29Si MAS NMR results showing the presence of a fair amount of Q2 and Q3 silica species. SL1SiEt in the surfactant-extracted materials exhibit a high mobility because the pores have been freed from the surfactant which restricted their motion in the assynthesized materials. A similar behavior is expected for the CSDA. Furthermore the wall surface that may have been rich in SiO groups due to the basic character of the reaction [54], became fully protonated after the acidic treatment, affecting the possible interaction between the wall and the CSDA. This, in turn makes the region of the SL1SiEt less crowded. Direct experimental evidence for the interaction of the negatively charged template and the positively charged co-structure directing agent (CSDA) came from the appearance of 14N signals in the FT-ESEEM spectra of 5-DSA in both phases. The interaction between the nitroxide spin label and the 14N nuclei is dipolar and depends on the inverse of the cube of the interspin distance, and is therefore short range (<8 nm), particularly for low c nuclei such as 14N. Therefore, the appearance of significant 14N modulations indicates close proximity of the polar head of the surfactant and the ammonium of the CSDA. This confirms the proposed surfactant–CSDA interaction mechanism [14]. Similar to 5-DSA, the silica probe exhibited a 14N signal in the FT-ESEEM spectra in both phases, again without a noticeable difference between the hexagonal and cubic phases. These results provide evidence that SL1SiEt indeed copolymerized with the TEOS and the CSDA, and that it is pending into the pore as the CSDA. The functionalized pores are well accessible to molecules after the surfactant was removed. An intriguing, unexpected observation was the absence of 2H nuclei in the vicinity of SL1SiEt in both the as-synthesized cubic and hexagonal materials prepared in D2O. This is in marked contrast to the 5-DSA results. This suggests that spin label in SL1SiEt is situated close to the surface and that this surface is not hydroxylated, but mostly constituted of SiO. This is due to the high pH of the reaction, silanol having a isoelectric point pI 2, which leads to a surface rich in SiO [6,54]. Alternatively, SL1SiEt could be trapped in the silica wall. Indeed the EPR spectrum of SL1SiEt after solvent extraction contains a fraction of rigid probes, but the majority is highly mobile, indicating that most of it is in the pores. The absence of 2H near SL1SiEt implies that the 5-DSA and SL1SiEt probe different regions of the pores. 5-DSA probes the surfactant–CSDA interface, while SL1SiEt probes the silica wall region. Accordingly, the CSDA molecules sensed by each probes belong to different populations, CSDAs interacting with the surfactant, probed by 5-DSA, and the other, CSDAw, bending towards the wall, probed by SL1SiEt, as shown schematically on Fig. 9. The most significant experimental difference we observed between the two phases is in the degree of mobility of the spin probes. 5-DSA is slightly more mobile in the final wet gel of the hexagonal phase reaction while SL1SiEt is more mobile in the assynthesized cubic phase. We attribute these differences to the pH difference as follows: The lower pH in the hexagonal phase reaction yields less negatively charged surfactants molecules that interact directly with the positively charged CSDA, namely less CSDAs. This allows some additional motional freedom to the surfactant in the hexagonal phase (see Fig. 1).
298
M. Florent, D. Goldfarb / Microporous and Mesoporous Materials 163 (2012) 291–299
surfactants exist at higher pH, and so more CSDA can interact with them. A soft extraction method by an acid ethanolic solvent preserved the final phase structure, yet it removed the molecular differences between the two phases. It freed the pores of surfactant molecules presents in the as-synthesized material, and protonate the surface allowing the pending ammonium groups a high degree of freedom and accessibility to other molecules. Acknowledgement We thank Dr. Ronit Popovitz-Biro for the TEM measurements, Dr. Tali Amitay-Rosen for the NMR measurements and Dr. Veronica Frydman for the synthesis of the spin probes. This research was made possible in part by the historic generosity of the Harold Perlman Family. D.G. holds the Erich Klieger Professorial Chair in Chemical Physics. Appendix A. Supplementary data Fig. 9. A schematic representation of a pore with CSDAs interacting with negatively charged surfactants and the CSDAs interacting with the silica wall.
The postulation of the existence of two distinct populations of CSDA also accounts for the observed difference in the mobility of SL1SiEt in the as-synthesized cubic and hexagonal phases. If less CSDA interact with the surfactant, more will bend toward the wall (CSDAw), where they may be attracted by the presence of SiO groups, producing a more crowded environment at the surface of silica wall. This in turn will be manifested by a lower mobility of the SL1SiEt probe. Similarly, increasing the pH to prepare the cubic materials will increase the number of ionized surfactants, thus shifting the equilibrium CSDAw ¢ CSDAs , towards CSDAs. This will decrease the mobility at the surfactant–CSDA interface, and increase the mobility close to the wall. While difference in the envisioned relative populations of CSDAs and CSDAw could be picked up by the probe mobilities, it was too small to be resolved by the ESEEM experiments. In the case of 5-DSA the faster motion in the hexagonal phases was resolved only in the wet gel because in the as-synthesized material the motion is too close to the rigid limit and is therefore not sensitive to small changes. The differences in mobility of SL1SiEt observed in as-synthesized hexagonal and cubic materials are abolished after solvent extraction as now there are no surfactant molecules that were the source of the differences. Our experimental results are not sufficient for resolving the significance of the observed two populations of the CSDA molecules, and particularly their ratio, in controlling the final AMS phase structure. This would require new experiments aiming at detecting two types of CSDA molecules in the reaction mixture, exploring the time evolution of their ratio and its pH dependence.
5. Conclusion CW EPR and ESEEM measurements gave experimental evidence for the Coulomb surfactant–CSDA interaction that drives the formation of the AMS materials. This interaction leads to a fast assembly of a mesophase followed by a slow precipitation process. Throughout this process the forming silica remains rather fluidic and becomes rigid only upon drying. The EPR results were interpreted in terms of the presence of two populations of CSDA molecules; one interacting with the anionic surfactant and the other with the silica wall, which is mostly ionized. The major observed molecular level differences between the two as-synthesized phases were attributed to different relative population of these two types of CSDA. This, in turn, was correlated with the pH; more ionized
XRD patterns and TEM micrographs of the final materials, 29Si MAS NMR spectrum of the as-synthesized cubic material and ESEEM traces of 5-DSA in the as-synthesized materials. Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.micromeso. 2012.06.050. References [1] J.S. Beck, J.C. Vartuli, W.J. Roth, M.E. Leonowicz, C.T. Kresge, K.D. Schmitt, C.T.W. Chu, D.H. Olson, E.W. Sheppard, S.B. Mccullen, J.B. Higgins, J.L. Schlenker, J. Am. Chem. Soc. 114 (1992) 10834. [2] C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.C. Vartuli, J.S. Beck, Nature 359 (1992) 710. [3] Q.S. Huo, D.I. Margolese, U. Ciesla, P.Y. Feng, T.E. Gier, P. Sieger, R. Leon, P.M. Petroff, F. Schüth, G.D. Stucky, Nature 368 (1994) 317. [4] Y. Wan, D.Y. Zhao, Chem. Rev. 107 (2007) 2821. [5] L.L. Hench, J.K. West, Chem. Rev. 90 (1990) 33. [6] R.K. Iler, The Chemistry of Silica: Solubility, Polymerization, Colloid and Surface Properties and Biochemistry of Silica, Wiley, 1979. [7] F. Kleitz, S.H. Choi, R. Ryoo, Chem. Commun. (2003) 2136. [8] F. Kleitz, D.N. Liu, G.M. Anilkumar, I.S. Park, L.A. Solovyov, A.N. Shmakov, R. Ryoo, J. Phys. Chem. B 107 (2003) 14296. [9] S.D. Shen, A.E. Garcia-Bennett, Z. Liu, Q.Y. Lu, Y.F. Shi, Y. Yan, C.Z. Yu, W.C. Liu, Y. Cai, O. Terasaki, D.Y. Zhao, J. Am. Chem. Soc. 127 (2005) 6780. [10] D.Y. Zhao, Q.S. Huo, J.L. Feng, B.F. Chmelka, G.D. Stucky, J. Am. Chem. Soc. 120 (1998) 6024. [11] S. Che, A.E. Garcia-Bennett, T. Yokoi, K. Sakamoto, H. Kunieda, O. Terasaki, T. Tatsumi, Nat. Mater. 2 (2003) 801. [12] S. Che, Z. Liu, T. Ohsuna, K. Sakamoto, O. Terasaki, T. Tatsumi, Nature 429 (2004) 281. [13] C. Gao, Y. Sakamoto, O. Terasaki, K. Sakamoto, S. Che, J. Mater. Chem. 17 (2007) 3591. [14] C.B. Gao, H.B. Qiu, W. Zeng, Y. Sakamoto, O. Terasaki, K. Sakamoto, Q. Chen, S.A. Che, Chem. Mater. 18 (2006) 3904. [15] C.B. Gao, Y. Sakamoto, O. Terasaki, S.A. Che, Chem. Eur. J. 14 (2008) 11423. [16] A.E. Garcia-Bennett, N. Kupferschmidt, Y. Sakamoto, S. Che, O. Terasaki, Angew. Chem. Int. Ed. 44 (2005) 5317. [17] J.N. Israelachvili, D.J. Mitchell, B.W. Ninham, J. Chem. Soc. Farad. T2 (72) (1976) 1525. [18] G.J.D. Soler-illia, C. Sanchez, B. Lebeau, J. Patarin, Chem. Rev. 102 (2002) 4093. [19] Q.S. Huo, R. Leon, P.M. Petroff, G.D. Stucky, Science 268 (1995) 1324. [20] S.H. Tolbert, C.C. Landry, G.D. Stucky, B.F. Chmelka, P. Norby, J.C. Hanson, A. Monnier, Chem. Mater. 13 (2001) 2247. [21] H.Q. Zheng, C.B. Gao, S.N. Che, Microporous Mesoporous Mater. 116 (2008) 299. [22] A.E. Garcia-Bennett, U. Brohede, R.P. Hodgkins, N. Hedin, M. Strømme, Langmuir 23 (2007) 9875. [23] C. Madelmont, R. Perron, Colloid Polym. Sci. 254 (1976) 581. [24] B. Lin, A.V. McCormick, H.T. Davis, R. Strey, J. Colloid Interface Sci. 291 (2005) 543. [25] X.Y. Wen, E.I. Franses, J. Colloid Interface Sci. 231 (2000) 42. [26] R.G. Herman, J.H. Lunsford, H. Beyer, P.A. Jacobs, J.B. Uytterhoeven, J. Phys. Chem. 79 (1975) 2388. [27] A. Moscatelli, A. Galarneau, F. Di Renzo, M.F. Ottaviani, J. Phys. Chem. B 108 (2004) 18580. [28] M.F. Ottaviani, R. Mazzeo, N.J. Turro, X.G. Lei, Microporous Mesoporous Mater. 139 (2011) 179.
M. Florent, D. Goldfarb / Microporous and Mesoporous Materials 163 (2012) 291–299 [29] B. Jee, K. Koch, L. Moschkowitz, D. Himsl, M. Hartman, A. Pöppl, J. Phys. Chem. Lett. 2 (2011) 357. [30] L. Kevan, M. Hartmann, in: B.C. Gilbert, N.M. Atherton, M.J. Davies (Eds.), Electron Paramagnetic Resonance, The Royal Society of, Chemistry, 1998. [31] J. Zhang, D. Goldfarb, Microporous Mesoporous Mater. 48 (2001) 143. [32] J.Y. Zhang, Z. Luz, D. Goldfarb, J. Phys. Chem. B 101 (1997) 7087. [33] J.Y. Zhang, H. Zimmermann, Z. Luz, D. Goldfarb, Mesoporous Mol. Sieves 1998 117 (1998) 535. [34] H. Caldararu, A. Caragheorgheopol, F. Savonea, D.J. Macquarrie, B.C. Gilbert, J. Phys. Chem. B 107 (2003) 6032. [35] M.F. Ottaviani, A. Moscatelli, D. Desplantier-Giscard, F. Di Renzo, P.J. Kooyman, B. Alonso, A. Galarneau, J. Phys. Chem. B 108 (2004) 12123. [36] A. Galarneau, F. Di Renzo, F. Fajula, L. Mollo, B. Fubini, M.F. Ottaviani, J. Colloid Interface Sci. 201 (1998) 105. [37] J.Y. Zhang, P.J. Carl, H. Zimmermann, D. Goldfarb, J. Phys. Chem. B 106 (2002) 5382. [38] S. Ruthstein, V. Frydman, S. Kababya, M. Landau, D. Goldfarb, J. Phys. Chem. B 107 (2003) 1739. [39] S. Ruthstein, V. Frydman, D. Goldfarb, J. Phys. Chem. B 108 (2004) 9016. [40] D. Baute, V. Frydman, H. Zimmermann, S. Kababya, D. Goldfarb, J. Phys. Chem. B 109 (2005) 7807.
299
[41] D. Goldfarb, in: M. Brustolon, E. Giamello (Eds.), Electron Paramagnetic Resonance – A Practitioner’s Toolkit, Wiley, 2009. [42] S. Ruthstein, D. Goldfarb, Electron Paramagnetic Resonance, vol. 21, The Royal Society of Chemistry, 2008. 184p. [43] R. Ciriminna, J. Blum, D. Avnir, M. Pagliaro, Chem. Commun. (2000) 1441. [44] K. Joon Oh, C. Altenbach, R.J. Collier, W.L. Hubbell, in: O. Holst (Ed.), Bacterial toxins: Methods and Protocols, vol. 145, Humana Press, 2000, p. 147. [45] S. Stoll, A. Schweiger, J. Magn. Reson. 178 (2006) 42. [46] J.M. Fauth, A. Schweiger, L. Braunschweiler, J. Forrer, R.R. Ernst, J. Magn. Reson. 66 (1986) 74. [47] L. Kevan, R.N. Schwartz, Time domain electron spin resonance, John Wiley & Sons, New York, 1979. 279p. [48] S. Van Doorslaer, G.A. Sierra, A. Schweiger, J. Magn. Reson. 136 (1999) 152. [49] R. Yokoi, H. Yoshitake, T. Tatsumi, Pts a - C 154 (2004) 519. [50] F. Kleitz, W. Schmidt, F. Schüth, Microporous Mesoporous Mater. 65 (2003) 1. [51] Z.L. Hua, J.L. Shi, L. Wang, W.H. Zhang, J. Non-Cryst, Solids 292 (2001) 177. [52] C.B. Gao, S.A. Che, Adv. Funct. Mater. 20 (2010) 2750. [53] E.G. Rozantsev, V.D. Sholle, Russ. Chem. Rev. 40 (1971) 233. [54] N. Baccile, G. Laurent, C. Bonhomme, P. Innocenzi, F. Babonneau, Chem. Mater. 19 (2007) 1343.