Tunable Porosity in Bridged Organosilicas Using Self-Organizing

Langmuir 2008, 24, 12539-12546

12539

Tunable Porosity in Bridged Organosilicas Using Self-Organizing Precursors Andreas Ide,† Gudrun Scholz,‡ and Arne Thomas*,† Department of Colloid Chemistry, Max Planck Institute of Colloids and Interfaces, Research Campus Golm, D-14424 Potsdam, Germany, and Institute of Chemistry, Humboldt UniVersity of Berlin, D-12489 Berlin, Germany ReceiVed May 2, 2008. ReVised Manuscript ReceiVed July 17, 2008 Functionalized, mesoporous organosilicas with tunable porosity were prepared by a direct and simple approach from rationally designed precursors, combining the function of a network builder and a porogen in one molecule. The precursors are synthesized using a dual hydroboration reaction, fulfilling the criteria of “click-chemistry”, first on an ethylene-bridged organosilica and then on a long-chain alkene. Thus, in the final molecule the boron atom connects the sol-gel precursor (the bridged organosilica) with the porogen (the long-chain alkene). The so-prepared precursors do self-organize when hydrolysis of their inorganic moiety takes place via an aggregation of their organic side chains into hydrophobic domains. The length of the attached chain influences the size of the hydrophobic domain and thus, after a condensation-aminolysis sequence, the finally observed porosity of the organosilicas. Depending on chain length micro- to mesoporous materials with average pore sizes from 1.5 to 4.1 nm (for attached pentene to hexadecene chains) are observed. Furthermore, the boron entity enables the subsequent introduction of various functional groups into the pore walls of the organosilica networks. Amine or hydroxyl functionalities can be easily introduced, dependent on the experimental conditions used during the borane cleavage and extraction step. The accessibility of these functionalities can be proven by a significant metal adsorption onto the functional organosilica walls.

Introduction Mesoporous materials have received great attention in recent years, as they combine small pore sizes (2-50 nm) with high surface areas.1 Since the conventional mesoporous silicas have some restrictions regarding functionality and polarity, there is a broad interest in the incorporation of organic functionalities into these structures.2-4 In this respect, porous bridged organosilanes, as represented by aerogels of polysilsesquioxanes5 or the surfactant-mediated periodic mesoporous organosilicas (PMOs),6-9 are promising materials. Starting from bridged organosilane precursors, templating and sol-gel approaches similar to the preparation of pure mesoporous silicas can be applied to form porous materials with comparable pore sizes and structures. In comparison to porous silicas where the functional groups are incorporated by grafted organic moieties,10,11 bridged organosilicas exhibit some important advantages, mainly the high organic loading possible under preservation of the mechanical stability while even smaller pores are only insignificantly blocked. * To whom correspondence should be addressed. E-mail: Arne.Thomas@ mpikg.mpg.de. † Max Planck Institute of Colloids and Interfaces. ‡ Humboldt University of Berlin.

(1) Davis, M. E. Nature 2002, 417, 813. (2) Sanchez, C.; Soler-Illia, G.; Ribot, F.; Lalot, T.; Mayer, C. R.; Cabuil, V. Chem. Mater. 2001, 13, 3061. (3) Sayari, A.; Hamoudi, S. Chem. Mater. 2001, 13, 3151. (4) Hoffmann, F.; Cornelius, M.; Morell, J.; Froba, M. Angew. Chem., Int. Ed. 2006, 45, 3216. (5) Loy, D. A.; Jamison, G. M.; Baugher, B. M.; Russick, E. M.; Assink, R. A.; Prabakar, S.; Shea, K. J. J. Non-Cryst. Solids 1995, 186, 44. (6) Asefa, T.; Yoshina-Ishii, C.; MacLachlan, M. J.; Ozin, G. A. J. Mater. Chem. 2000, 10, 1751. (7) Yoshina-Ishii, C.; Asefa, T.; Coombs, N.; MacLachlan, M. J.; Ozin, G. A. Chem. Commun. 1999, 2539. (8) Inagaki, S.; Guan, S.; Fukushima, Y.; Ohsuna, T.; Terasaki, O. J. Am. Chem. Soc. 1999, 121, 9611. (9) Melde, B. J.; Holland, B. T.; Blanford, C. F.; Stein, A. Chem. Mater. 1999, 11, 3302. (10) Lim, M. H.; Stein, A. Chem. Mater. 1999, 11, 3285. (11) Walcarius, A.; Etienne, M.; Lebeau, B. Chem. Mater. 2003, 15, 2161.

However, most of the so far reported bridged polysilsesquioxanes, exhibiting a significant porosity after condensation, consist of small, stiff, but nonfunctional organic bridging groups, e.g., methylene, ethylene, ethenylene, or phenylene linkers, as these organic moieties are stable enough to preserve the high surface energies accompanied with the introduction of meso- or microporosities into such networks. Furthermore, the homogeneous distribution of these organic groups in the silica framework does not necessarily make them be located at the pore walls and accessible for further modification. Usually a significant fraction of the moieties is buried throughout condensation in the pore walls. For example, it was demonstrated that ethylene bridges in a PMO material are not fully accessible for hydroboration12 or bromination9 reactions. Interestingly, a crystal-like arrangement of organic groups in phenylene silicas apparently yields a better accessibility, and an amination of the phenylene groups was achieved with, at least, close to 28% conversion.13 As a further functionalization of bridged organosilanes is hard to achieve, the introduction of valuable functionalities such as amine groups into mesoporous silicas was most often realized via a postsynthetic grafting process, for example, using aminopropyltriethoxysilane (APTS) reactants.11,14-17 For bridged organosilanes the examples are so far restricted to long organic amine bridges18,19 or cyclam moieties,20,21 which, however, cannot be seen as a supporting element of the porous network. Thus, usually a high amount of a pure silica precursor has to be admixed during synthesis to (12) Asefa, T.; Kruk, M.; MacLachlan, M. J.; Coombs, N.; Grondey, H.; Jaroniec, M.; Ozin, G. A. J. Am. Chem. Soc. 2001, 123, 8520. (13) Ohashi, M.; Kapoor, M. P.; Inagaki, S. Chem. Commun. 2008, 841. (14) Yoshitake, H.; Yokoi, T.; Tatsumi, T. Chem. Mater. 2002, 14, 4603. (15) Yokoi, T.; Yoshitake, H.; Tatsumi, T. J. Mater. Chem. 2004, 14, 951. (16) Vrancken, K. C.; Vandervoort, P.; Possemiers, K.; Vansant, E. F. J. Colloid Interface Sci. 1995, 174, 86. (17) Jun, Y. S.; Huh, Y. S.; Park, H. S.; Thomas, A.; Jeon, S. J.; Lee, E. Z.; Won, H. J.; Hong, W. H.; Lee, S. Y.; Hong, Y. K. J. Phys. Chem. C 2007, 111, 13076. (18) Wahab, M. A.; Imae, I.; Kawakami, Y.; Ha, C. S. Chem. Mater. 2005, 17, 2165. (19) Wahab, M. A.; Kim, I.; Ha, C. S. J. Solid State Chem. 2004, 177, 3439.

10.1021/la801374u CCC: $40.75  2008 American Chemical Society Published on Web 09/30/2008

12540 Langmuir, Vol. 24, No. 21, 2008

maintain the high surface areas of the resulting materials. Nevertheless these networks were able to hold and bind a high amount of transition metal ions. Another approach toward tunable surface functionalities was recently reported, involving the surfactant-mediated condensation of a bromophenyl-bis(silane) precursor. This system enables a simple postcondensation functionalization by bromine substitution. A further substitution of the bromine yielded carboxylic acid, vinyl groups, or phosphonic esters incorporated in the pore walls.22 Recently, we presented a new concept for the introduction of functional groups into the pore walls of porous bridged organosilicas, where the position of the organic functionality is controlled to be exclusively along the pore interface.23 This was achieved by a hydroboration reaction on an ethylene-bridged silica precursor and the subsequent fixation of a sterically demanding group to this functional monomer. During condensation this precursor self-organized comparable to amphiphilic molecules by a microphase separation of the organic moiety into hydrophobic domains. Subsequent aminolysis yielded a mesoporous organosilica with amine groups placed exclusively at the pore surface. This approach differed significantly from other strategies, where organic bridging groups of the precursors were used as the template for porosity. There, the organic bridging group is removed, e.g., via calcination, creating a porosity,24-27 which is, however, accompanied with a structural weakening of the organosilica framework. In contrast, in our approach just appending parts of the organic moieties were selectively cleaved maintaining the bridging organic group, while introducing a viable functionality on the pore wall, only. With the use of enantioselective hydroboration, it was furthermore possible to extend this approach toward chiral, porous, amino-functionalized silicas28 or even to PMOs with chiral building blocks.29 The topological design of mesoporous silica materials, that is, their pore architecture and pore size, are still major issues in areas such as catalysis, adsorption, separation, and host-guest chemistry.30 Successful methods of controlling pore size include the use of surfactants of different chain lengths31-34 or in different concentrations35 as templates but also postsynthesis treatment36,37 or the use of swelling agents38-40 during synthesis. (20) Corriu, R. J. P.; Mehdi, A.; Reye, C.; Thieuleux, C. Chem. Commun. 2002, 1382. (21) Corriu, R. J. P.; Mehdi, A.; Reye, C.; Thieuleux, C. Chem. Commun. 2003, 1564. (22) Kuschel, A.; Polarz, S. AdV. Funct. Mater. 2008, 18, 1272. (23) Voss, R.; Thomas, A.; Antonietti, M.; Ozin, G. A. J. Mater. Chem. 2005, 15, 4010. (24) Boury, B.; Chevalier, P.; Corriu, R. J. P.; Delord, P.; Moreau, J. J. E.; Chiman, M. W. Chem. Mater. 1999, 11, 281. (25) Boury, B.; Corriu, R. J. P. AdV. Mater. 2000, 12, 989. (26) Boury, B.; Corriu, R. J. P.; Le Strat, V. Chem. Mater. 1999, 11, 2796. (27) Shea, K. J.; Loy, D. A.; Webster, O. J. Am. Chem. Soc. 1992, 114, 6700. (28) Ide, A.; Voss, R.; Scholz, G.; Ozin, G. A.; Antonietti, M.; Thomas, A. Chem. Mater. 2007, 19, 2649. (29) Polarz, S.; Kuschel, A. AdV. Mater. 2006, 18, 1206. (30) Park, S. S.; Ha, C. S. Chem. Rec. 2006, 6, 32. (31) Goltner, C. G.; Henke, S.; Weissenberger, M. C.; Antonietti, M. Angew. Chem., Int. Ed. 1998, 37, 613. (32) Zhao, D. Y.; Feng, J. L.; Huo, Q. S.; Melosh, N.; Fredrickson, G. H.; Chmelka, B. F.; Stucky, G. D. Science 1998, 279, 548. (33) Muth, O.; Schellbach, C.; Froba, M. Chem. Commun. 2001, 2032. (34) Matos, J. R.; Kruk, M.; Mercuri, L. P.; Jaroniec, M.; Asefa, T.; Coombs, N.; Ozin, G. A.; Kamiyama, T.; Terasaki, O. Chem. Mater. 2002, 14, 1903. (35) Thomas, A.; Schlaad, H.; Smarsly, B.; Antonietti, M. Langmuir 2003, 19, 4455. (36) Nishiyama, N.; Tanaka, S.; Egashira, Y.; Oku, Y.; Ueyama, K. Chem. Mater. 2003, 15, 1006. (37) Grosso, D.; Balkenende, A. R.; Albouy, P. A.; Ayral, A.; Amenitsch, H.; Babonneau, F. Chem. Mater. 2001, 13, 1848. (38) Sayari, A.; Kruk, M.; Jaroniec, M.; Moudrakovski, I. L. AdV. Mater. 1998, 10, 1376.

Ide et al.

Herein we show how a small and facile variation of the molecular structure of the precursor affects the porosity of the resulting materials. With the use of alkenes with different chain lengths in the second hydroboration step, precursors with various lengths of the hydrophobic alkyl moiety can be synthesized, subsequently condensing into porous functional materials with different pore sizes. Although the concept of altering the chain length of alkyl silane precursors to vary the dimension of the resulting aggregates and subsequent pore structures has been reported, these systems were mainly based on monoalkylated silanes and could therefore not enable an access to functionalized materials.41,42 In contrast, the “lizard” template presented by Zhang et al. enables the chemospecific cleavage of the alkyl tail under preservation of the functional group,43 yielding, however, dense silica structures with the functional groups grafted on the pore wall. In contrast, in the here shown approach the structural factor determining the resulting pore size is introduced in the bridging group of the precursor yielding functionalized bridged organosilanes with tunable pore sizes.

Experimental Section Materials. Bis(triethoxysilyl)ethene (BTSE) was synthesized by alkene metathesis of triethoxysilylethene.44 Triethoxysilylethene was purchased from Gelest. Hydroxylamine-O-sulfonic acid, borane-dimethyl sulfide complex (2 M in THF), tetraethoxysilane (TEOS), all 1-alkenes, and CuCl2 for adsorption tests were obtained from Aldrich. Solvents for reactions under moisture exclusion were prepared according standard procedures. All further commercial chemicals were used without additional purification. Preparation of the Long Chain Boron Silica (LCBS-Cx) Precursor. The reaction was carried out under N2 atmosphere. In a typical synthesis 2.13 g (6.04 mmol) of BTSE was dissolved in 25 mL of dry THF at 0 °C, and then 3.02 mL (6.04 mmol) of borane-dimethyl sulfide complex was added. The mixture was stirred for 19 h and allowed to warm up to room temperature (RT). The appropriate 1-alkene (6.04 mmol) was then added at RT, and the solution was stirred for another 4 h, followed by the addition of 0.53 mL (9.06 mmol) of dry ethanol. The residual solvent was removed under vacuum yielding the LCBS–Cx precursors. 1H NMR. LCBS-C . Colorless liquid: 3.82 (m, 2 H, BOCH), 5 3.80 (t, J ) 7.0 Hz, 6 H, SiOCH), 3.79 (t, J ) 6.3 Hz, 6 H, SiOCH), 1.53 (m, 2 H, CHalliphat), 1.24 (m, 6 H, CHalliphat), 1.18 (m, 21 H, CHalliphat), 0.87 (m, 6 H, CHalliphat). LCBS-C10. Colorless liquid: 3.81 (m, 2 H, BOCH), 3.80 (t, J ) 7.1 Hz, 6 H, SiOCH), 3.79 (t, J ) 6.3 Hz, 6 H, SiOCH), 1.54 (m, 2 H, CHalliphat), 1.18 (m, 37 H, CHalliphat), 0.86 (m, 6 H, CHalliphat). LCBS-C12. Colorless liquid: 3.82 (m, 2 H, BOCH), 3.80 (t, J ) 7.1 Hz, 6 H, SiOCH), 3.79 (t, J ) 8.6 Hz, 6 H, SiOCH)), 1.54 (m, 2 H, CHalliphat), 1.23 (m, 41 H, CHalliphat), 0.86 (m, 6 H, CHalliphat). LCBS-C16. Colorless viscous liquid: 3.82 (m, 2 H, BOCH), 3.80 (t, J ) 7.1 Hz, 6 H, SiOCH), 3.79 (t, J ) 6.7 Hz, 6 H, SiOCH)), 1.55 (m, 2 H, CHalliphat), 1.23 (m, 49 H, CHalliphat), 0.86 (m, 6 H, CHalliphat). LCBS-C20. White solid: 3.82 (m, 2 H, BOCH), 3.80 (t, J ) 7.0 Hz, 6 H, SiOCH), 3.79 (t, J ) 6.8 Hz, 6 H, SiOCH)), 1.55 (m, 1 H, CHalliphat), 1.23 (m, 58 H, CHalliphat), 0.86 (m, 6 H, CHalliphat). Condensation of Boron-Containing Precursor. All organosilicas were synthesized using the following molar ratios: LCBS/TEOS/ EtOH/HCl/H2O ) 1:4:9.5:2.77:15.4. In a typical synthesis 0.54 mmol of the LCBS precursor and 450 mg (2.16 mmol) of TEOS were (39) Lind, A.; Spliethoff, B.; Linden, M. Chem. Mater. 2003, 15, 813. (40) Blin, J. L.; Su, B. L. Langmuir 2002, 18, 5303. (41) Shimojima, A.; Liu, Z.; Ohsuna, T.; Terasaki, O.; Kuroda, K. J. Am. Chem. Soc. 2005, 127, 14108. (42) Shimojima, A.; Kuroda, K. Chem. Rec. 2006, 6, 53. (43) Zhang, Q. M.; Ariga, K.; Okabe, A.; Aida, T. J. Am. Chem. Soc. 2004, 126, 988. (44) Marciniec, B.; Maciejewski, H.; Gulinski, J.; Rzejak, L. J. Organomet. Chem. 1989, 362, 273.

Tunable Porosity in Bridged Organosilicas dissolved in 300 µL (5.14 mmol) of ethanol, and then 150 µL of hydrochloric acid (pH ) 2.0) was added. The solution was stirred in a closed vial for 10 min under heating to approximately 50 °C to ensure complete homogenization. The silica solution was then aged for 2 days at 60 °C to give the monolithic solids. Amine Functionalization. The crude silica was mechanically crushed, and a quantity of silica powder corresponding to 0.42 mmol of boron atoms was dispersed in 5.0 mL of diglyme. To this mixture 0.21 g (1.86 mmol) of hydroxylamine-O-sulfonic acid was added, and the mixture was stirred at 60 °C for 4 h. After the suspension was cooled down to RT the solvent was removed by centrifugation. The obtained solid was stirred in a mixture of 20 mL of HCl (pH ) 2) and 20 mL of ethanol overnight and then extracted for 12 h with ethanol and THF, respectively. Additional purification of the organosilica was carried out by washing with ammonia (1 N) and then repeated centrifugation and washing with THF. All solvents were removed under vacuum at 60 °C yielding a colorless solid. Hydroxyl Functionalization. An amount of 0.25 g of the mechanically crushed organosilica was dispersed in a solution of 1.0 mL of hydrogen peroxide (30%) and 30 µL of sodium hydroxide (0.1 M), and the mixture was stirred at RT for 5 h. The reaction mixture was neutralized with 0.5 mL of hydrochloric acid (pH ) 2), and 5 mL of ethanol was additionally added. The liquid phase was centrifuged off, and the solid was repeatedly washed with water (twice), ethanol (twice), and THF (twice). All solvents were removed under vacuum at 60 °C yielding a colorless solid. Methods. Transmission electron microscopy (TEM) images were taken using a Zeiss EM 912Ω operated at an acceleration voltage of 120 kV. Samples were ground in a ball mill and dispersed in acetone. One droplet of the suspension was applied to a 400 mesh carbon-coated cooper grid and left to dry in air. Nitrogen adsorption data were obtained with a Quantachrome Autosorb-1 at liquid nitrogen temperature. Small-angle X-ray scattering (SAXS) measurements were done using a Bruker XRD D8-advance. Cu atomic absorption spectroscopy (AAS) measurements were done with a Perkin-Elmer 1100B Atom Adsorption spectrometer at 324.8 nm. 1H, 11B, and 29Si NMR spectra were recorded on a Bruker DPX 400 spectrometer in CDCl3 solution. The 11B NMR spectrum was background-corrected, and 29Si spectrum was recorded under utilization of an hmbc pulse program. Calibration was conducted using CDCl3 signals for 1H (δ ) 7.24 ppm), BF3 · OEt2 for 11B (δ ) 0 ppm), and tetramethylsilane for 29Si (δ ) 0 ppm). 11B, 1H, and 1H-29Si cross polarization magic-angle spinning (CP MAS) NMR spectra were recorded on a Bruker AVANCE 400 spectrometer (Larmor frequencies: ν29Si ) 79.5 MHz, ν11B ) 128.4 MHz, ν1H ) 400.1 MHz) using a 4 mm MAS probe (Bruker Biospin) and applying a spinning speed of 10 kHz. The contact time for the 1H-29Si CP MAS experiments was determined as 4 ms. The spectra were recorded with a spectrum width of 20 kHz and a recycle delay of 5 s. 11B MAS NMR (I ) 3/ ) spectra were recorded with an excitation pulse duration of 1 µs 2 and referenced with respect to the chemical shift of 11B in BF3 · OEt2. The recycle delay was chosen as 1 s, and the accumulation number was 1800. The contact times of the 1H-13C CP MAS NMR experiments were optimized as 1 ms for the boron containing precursor. 1H MAS studies were made with a π/2 pulse length of 2.7 µs and a recycle delay of 10 s. Existent background signals of 1H could be completely suppressed with the application of a phase-cycled depth pulse sequence according to Cory and Ritchey.45 Values of the isotropic chemical shifts of 1H and 29Si are given with respect to TMS. Metal Sorption Experiments. For the Cu adsorption experiments 50 mg of the appropriate mechanically crushed silica was dispersed in a solution of 0.1 M CuCl2 in 0.1 M sodium acetate. The solution was stirred for 1 h. The resulting mixture was centrifuged and then washed four times with in total 4 mL of bidest water. The Cu-loaded samples were dried at 60 °C for 16 h. For any silica sample, 40 mg of the sample was extracted with 3.0 mL of hydrochloric acid. After (45) Cory, D. G.; Ritchey, W. M. J. Magn. Reson. 1988, 80, 128.

Langmuir, Vol. 24, No. 21, 2008 12541 Scheme 1. Synthesis of the LCBS-Cx Precursors

1.5 h of shaking the solid material was centrifuged off and washed four times with bidest water. The combined aqueous phases were diluted to a total volume of 25 mL and then directly measured using AAS.

Results and Discussion Micro- and mesoporous amine-functionalized organosilicas were synthesized from precursors prepared via a doublehydroboration reaction on first an ethylene-bridged organosilica precursor and second a terminal alkene (Scheme 1). By choosing the right experimental conditions, the hydroborane species reacts first with the ethylene-bridged siloxane and second with the respective terminal alkene. Quenching the reaction with ethanol yielded products denoted as LCBS-Cx, where x gives the number of carbon in the n-alkyl chain. The synthesis and structure of the precursors are shown in Scheme 1. It should be noted that hydroboration of alkenes with borane (“BH3”) principally can yield in mono-, di-, and trisubstituted boranes and that these products are usually in equilibrium, dependent on the initial molar ratio of the reactants.46 A similar reaction pathway would rather form a complex reaction mixture than single products as shown in Scheme 1. However, the products of hydroboration reactions are crucially determined by the bulkiness of the alkene substituents. Thus, given an excess of the alkene, terminal alkenes tend to form trisubstituted alkylboranes, whereas trans- and cis-alkenes generally form not more than the disubstituted products. Bulky side groups furthermore decrease the rate of substitution. Thus, it was shown that the addition of boronhydride on an excess of 1-methyl-2-tert-butylethylene just yields the monosubstituted compound.47 Consequently, for alkenes with two triethoxysilane groups in a vicinal position, just monosubstituted products as shown in Scheme 1 (46) Brown, H. C.; Kramer, G. W.; Levy, A. B.; Midland, M. M. Organic Synthesis Via Boranes; Wiley-Interscience: New York, 1975. (47) Schwier, J. R.; Brown, H. C. J. Org. Chem. 1993, 58, 1546.

12542 Langmuir, Vol. 24, No. 21, 2008

Ide et al.

Figure 1. (a) 1H NMR of LCBS-C10 (for comparison the NMR spectra of the reactants bis(triethoxysilyl)ethene (BTSE) and 1-decene are also shown); (b) FT-IR spectra of LCBS-C10 (compared with the spectrum of 1-decene).

Figure 2. MAS NMR results of pure condensed LCBS-C10: (a) 1H–29Si CP, (b) 11B, and (c) 1H–13C, exemplary for all LCBS organosilica materials.

can be expected. Indeed control experiments using an excess of BTSE over the borane (2:1 molar ratio) left half of the alkene groups unreacted, pointing to monosubstitution of the borane, exclusively. Even though a second addition of an alkene with bulky substituents is hindered, the remaining active hydroborane sites can still react with terminal alkenes, exhibiting a much lower sterical hindrance. This enables the addition of long-chain alkenes to the preformed boron-organosilica precursor (step 2 in Scheme 1). Afterward the reactions sphere of the small boron atom is geometrically saturated; thus, no further addition of alkene groups can be observed, and even a large excess of terminal alkenes in this reaction step yielded no disubstitution of terminal alkenes but the products shown in Scheme 1. Thus, all precursors exhibit the same structural features except for the length of the alkyl moiety. The formation of the LCBS precursor can be monitored using 1H NMR and FT-IR measurements (Figure 1). NMR measurements confirm the successful hydroboration of the present alkene groups. Thus, the ethylene protons of the reactants have vanished in the spectra of the products, proving the complete conversion after addition of the borane. 11B and 29Si NMR show the incorporation of a single boron species (31 ppm) and the incorporation of two silicon nuclei (-50 and -44 ppm) (Supporting Information Figure S1). These results are supported by FT-IR measurements, showing the appearance of C-H, C-B, O-B deformation modes between 1250-1460 cm-1. As expected, no signals above 3000 cm-1 are observed for LCBS precursors, which would be indicative for CdC-H

vibrational modes. However, it should be noted that also in the BTSE precursor the CH double-bond vibrational modes are very weak; thus, this finding can just be used for monitoring the hydroboration of the terminal alkene chains. Since this synthesis includes just quantitative reaction steps and the LCBS-Cx precursors are formed as single products (fulfilling the criteria of “click-chemistry”48), as can been seen from the depicted spectra, further purification steps can be conceptually avoided. Porous organosilicas were prepared from the LCBS-Cx precursors using a bulk condensation route. Even though porous organosilicas were observed using the pure LCBS precursors as well, it turned out that admixing TEOS stabilizes the resulting framework. A molar ratio TEOS/LCBS 4:1 was found to yield organosilicas with the highest surface areas and porosities; thus, this ratio was used throughout the following experiments. LCBS and TEOS were dissolved in ethanol, and hydrochloric acid was added. After complete homogenization of the mixture, ethanol was removed by evaporation, yielding transparent monolithic gels. A closer insight into the local chemical structure of the condensates is provided by MAS NMR spectroscopy, exemplarily shown for a pure organosilica and one with admixed TEOS derived from LCBS-C10 (Figure 2 and Supporting Information Figure S2). 29Si resonances appear at -65 and -100 ppm, which are assigned to T-type and Q-type species, respectively. The presence of Q sites in the spectra shows that some Si-C bond cleavage has to be taken into account throughout the condensation. (48) Kolb, H. C.; Finn, M. G.; Sharpless, K. B. Angew. Chem., Int. Ed. 2001, 40, 2004.

Tunable Porosity in Bridged Organosilicas

An exact assignment of the peaks to T1, T2, and T3 is difficult due to an expected overlap of the chemically different silicon nuclei located at the ethylene bridge. Additionally, the 1H-29Si CP MAS NMR spectra for the applied 1:4 molar ratio of precursor to TEOS is given in the Supporting Information. 11B MAS NMR for the pure LCBS-C10 proves the successful condensation under preservation of the boron species. The NMR spectra indicate the appearance of different boron species, probably caused by slightly varied chemical environments yielding different coordination of the boron groups in the solid state. Thus boron atoms connected to silane precursors with different condensations states (T1, T2, T3) could exhibit a shift in the 11B NMR peak. 13C NMR underlines the incorporation of the aliphatic species into the silica matrix and furthermore the successful hydroboration of the alkenes by the absence of signals attributable to double bonds. Signals at 14 and 7 ppm (broad) can be attributed to SiCH2 and SiCHB carbons, respectively. The sharp peaks between 10 and 40 ppm can be attributed to the grafted alkyl chains, whereas the peaks at 60 and 64 ppm can be assigned to ethoxy groups either covalently bound on the boron center or as residual ethoxy groups on the silanes. Splitting of the carbon-boron bonds in the organosilica network was carried out by stirring the materials in hydroxylamine-O-sulfonic acid/diglyme solutions, leading to amine functionalization of the organic bridges. Thus, bridged organosilicas with amine functionalities directly located at the bridging ethyl group were obtained. (Scheme 2). While aminolysis of the LCBS-Cx silicas yielded the corresponding amine-functionalized organosilica (LCBS-Cx-NH2), an oxidative hydrolysis yielded the corresponding hydroxyl-functionalized organosilica (LCBS-Cx-OH) used for control experiments. The successful aminolysis and extraction step of LCBS-Cx can be monitored using FT-IR spectroscopy (Figure 3). The CH3, CH2, and CH stretching vibrations at 2840-2960 cm-1 nearly disappeared in the functionalized samples, proving the successful cleavage and removal of the pendant alkyl chains. Also the B-C, B-O deformations vibrations at 1320-1470 cm-1 vanished completely in the functionalized samples. As aminolysis yields almost complete removal of the hydrocarbons, the remaining CH groups cause only very weak signals. Vibrational modes for amines are often rather weak in amine-functionalized organosilicas, which was shown by a comparison with an APTSfunctionalized mesoporous silica, and thus often not suitable for characterization. A small peak at 1635 cm-1 can be identified in the product, assignable to NH2 deformation modes. The typically existing NH stretching modes at 3100 cm-1 are presumably buried under the broad vibration band of adsorbed water and therefore not suitable for characterization. Nitrogen sorption measurements of the resulting organosilicas revealed the significant influence of the alkyl moiety on the finally observed porosity. All organosilicas (except for LCBS-C5-NH2) exhibit comparable high surface areas of 700-800 m2/g (Table 1). As the LCBS/TEOS molar ratio was held constant throughout the experiments, a higher overall organic loading for organosilicas with increasing alkyl chain length is assumed, which reflects the increasing porosity in the series C10-C16. On the other hand the surface area stays relatively constant, proving that the bridged

Langmuir, Vol. 24, No. 21, 2008 12543

silanol groups are indeed forming the major part of the pore walls. For LCBS-C5 and LCBS-C20 a different aggregation behavior of the precursor is observed. Thus for LCBS-C5 a very low hydrophilic-hydrophobic contrast of the silanol headgroups and the short alkyl chain can be assumed, leading to the partial incorporation of pentyl chains into the framework, inaccessible for further functionalization and extraction. In contrast, for LCBS-C20 the length of the alkyl moiety is too large to support the formation of well-defined hydrophobic domains surrounded by the inorganic network. Thus, a more ill-defined material is observed exhibiting a rather broad pore size distribution. Indeed, the hydrophobicity of this precursor increases to such a degree that solutions in ethanol are already slightly turbid. Pentyl or eicosyl chains therefore can be seen as the borderline cases for the here shown approach. Nitrogen sorption isotherms revealed the shift of the average pore size with increasing chain length (Figure 4). Whereas LCBS-C5-NH2 exhibits a type I isotherm typical for microporous materials, LCBS-C10-NH2 and LCBS-C12-NH2 show isotherms described for so-called “supermicroporous” materials49,50 or materials with small mesopores. The nonlocal density functional theory (NLDFT) pore size distributions of these two organosilicas reveal an average pore size of 2.5 and 2.7 nm, respectively. For LCBS-C16-NH2 and LCBS-C20-NH2 type IV isotherms are observed, with a welldefined hysteresis in the case of C16, revealing an average pore size of 4.1 nm with a defined pore size maximum of 5.0 nm from NLDFT pore size distributions. Although the condensation of the organosilica precursor does not yield materials with a narrow pore size distribution, there is obviously a tendency to form larger pores when LCBS-Cx precursors with increasing alkyl chains are used. Thus, after hydrolysis the precursors behave like surfactants and form micelles and lyotropic phases, where the size of the hydrophobic part naturally gets larger with increasing chain length of the hydrophobic tail. After aminolysis the hydrophobic domains are reflected in the porosity of the resulting organosilicas. TEM measurements on the materials indicate the presence of a disordered, wormlike pore structure (Figure 5). Also from powder X-ray diffraction (PXRD) measurements no periodicity of the pores could be detected. That shows that even though the alkyl attachments are able to form hydrophobic aggregates and subsequently lyotropic phases these phases are not long-range

Scheme 2. Condensation and Aminolysis to LCBS-Cx-NH2

Figure 3. FT-IR spectra of LCBS-C16 before (solid line) and after (dashed line) amine functionalization.

12544 Langmuir, Vol. 24, No. 21, 2008

Ide et al.

Figure 4. Nitrogen sorption isotherms and NLDFT pore size distribution of functionalized LCBS-Cx-NH2 organosilicas. Table 1. BET Surface Area and Pore Size Characterization of LCBS-Cx-NH2 Organosilicas entry

asurf (BET) [m2/g]a

dP (NLDFT) [nm]b

VP,micro (NLDFT) [nm]c

VP,meso (NLDFT) [cm3/g]d

VP,tot (NLDFT) [cm3/g]e

LCBS-C5 LCBS-C10 LCBS-C12 LCBS-C16 LCBS-C20

372 732 758 799 699

1.43 2.49 2.74 4.06 3.87

0.142 0.086 0.066 0.019 0.059

0.033 0.285 0.347 0.611 0.504

0.175 0.371 0.413 0.630 0.563

a Specific surface area. b Average pore diameter (D ) 4Vp/(Stotal - Sext)). c Micropore volume of pores