Use of a Mesoporous Material for Organic Synthesis

Tomasz Witula and Krister Holmberg*

Department of Materials and Surface Chemistry, Chalmers University of Technology,SE-412 96 Goteborg, Sweden

Received November 9, 2004. In Final Form: February 23, 2005

A common problem in synthetic organic chemistry is attaining proper contact between lipophilic organiccompounds and inorganic salts. Various strategies, for example, phase transfer catalysis (Starks, C. M.;Liotta, C. L.; Halpern, M. Phase Transfer Catalysis: Fundamentals, Applications and Industrial Perspectives;Chapman & Hall: New York, 1994) or use of a microheterogeneous medium such as a microemulsion(Hager, M.; Currie, F.; Holmberg, K. Organic Reactions in Microemulsions. In Colloid Chemistry II;Antonietti, M., Ed.; Topics in Current Chemistry 227; Springer-Verlag: Heidelberg, 2003; p 53) have beenworked out to tackle the issue. Here, we report that mesoporous solid materials made from surfactantself-assembly can be used as medium for such reactions. The material is made from silica, and the poresize is large, relatively uniform, and can be controlled with a high degree of precision by the choice ofsurfactant that is being used as template (Palmqvist, A. E. C. Curr. Opin. Colloid Interface Sci. 2003, 8,145). The pores are hydrophilic and are filled with an aqueous solution containing the inorganic salt. Theporous material is dispersed in the lipophilic organic substrate, that is, 4-tert-butylbenzyl bromide, or ina hydrocarbon solution of this substrate. The reaction occurs at the hydrophilic/lipophilic interface, and,because the interface is large, the reaction is fast. A considerable advantage with this new reaction mediumis that the workup procedure is extremely facile. After the reaction is completed, the solid is simplyremoved by filtering or centrifugation.

IntroductionThere is considerable current interest in the use of

micelles, microemulsions, surfactant liquid crystals, andother microheterogeneous liquids as media for organicsynthesis.Suchsystemsmaygive largerateenhancementsas compared to reactions in conventional media, mainlydue to high local concentration of the reactants in theinterfacial zone where the reaction occurs.1-6 The differenttypes of self-organized surfactant systems are particularlyuseful as reaction media when one of the reactants is alipophilic organic compound, soluble in hydrophobic mediabut insoluble in water and other polar solvents, and theother reactant is a polar compound, such as an inorganicsalt, that is insoluble in most organic solvents. Thesemicroheterogeneous, single-phase media all have bothpolar and nonpolar domains in which reactants of differentsolubility characteristics can dissolve. The high reactivityobtained is mainly due to the large interface between thedomains. The incompatible reactants will meet at theinterface, and it has been demonstrated that there is agood correlation between the total interfacial area of amicroemulsion and the rate of a substitution reaction.7Use of microheterogeneous media can be seen as analternative to phase transfer catalysis, that is, use of a

two-phase system with added phase transfer agent. Thetwo approaches can also be combined, in which case highreactivity may be obtained.8

All reactions performed in organized surfactant systemssuffer from one major drawback, however, the sometimescomplicated procedure to separate the surfactant fromthe product. The presence of the surfactant makes theworkup tedious because both extractions and chromato-graphic separations become complicated when a surfaceactive compound is present in high concentration. To acertain extent, the same problem exists for reactionsperformed using phase transfer catalysis. Removal of thephase transfer agent is sometimes a nontrivial issue. Theaim of the present work is to explore the possibility ofperforming an organic reaction in a surfactant-freenanostructured medium. The medium is mesoporous silicamade from a block copolymer self-assembly. The solid,mesoporous material can be said to be the replica of thesurfactant liquid crystal, and the dimensions are roughlythe same. A bimolecular substitution reaction, involvingone polar and one nonpolar reactant, is performed in themesoporous material, and the rate obtained is comparedto that obtained in the liquid crystalline phase used astemplate.

Experimental Section

ReagentsandAnalysisTechniques.The triblock copolymerPluronic 105, which is poly(ethylene oxide)-block-poly(propyleneoxide)-block-poly(ethylene oxide) with the composition (EO)37-(PO)58(EO)37, was obtained from BASF. 4-tert-Butylbenzyl bro-mide (4-TBBB, purity 97%), tetraethyl orthosilicate (TEOS),CDCl3, and HCl were supplied by Aldrich. Potassium iodide (KI)was purchased from Merck. n-Butanol was purchased from Fluka.All reagents were used as received.

Transmission electron microscopy (TEM) was performed on aJEOL 1200 EX II at 120 kV microscope, and specimens wereprepared by crushing of the material, dispersing the powder in

* Corresponding author. Tel.: +46 317722969. Fax: +4631160062. E-mail: kh@chem.chalmers.se.

(1) Starks, C. M.; Liotta, C. L.; Halpern, M. Phase TransferCatalysis: Fundamentals, Applications and Industrial Perspectives;Chapman & Hall: New York, 1994.

(2) Hager, M.; Currie, F.; Holmberg, K. Organic Reactions inMicroemulsions. In Colloid Chemistry II; Antonietti, M., Ed.; Topics inCurrent Chemistry 227; Springer-Verlag: Heidelberg, 2003; p 53.

(3) Palmqvist, A. E. C. Curr. Opin. Colloid Interface Sci. 2003, 8,145.

(4) Menger, F. M.; Erlington, A. R. J. Am. Chem. Soc. 1991, 113,9621.

(5) Schomacker, R. Mikroemulsionen als Medium fur chemischeReaktionen. Nachr. Chem., Tech. Lab. 1992, 40, 1344.

(6) Sjoblom, J.; Lindberg, R.; Friberg, S. E. Adv. Colloid InterfaceSci. 1996, 95, 125.

(7) Bode, G.; Lade, M.; Schomacker, R. Chem. Eng. Technol. 2000,23, 405. (8) Hager, M.; Holmberg, K. Chem.sEur. J. 2004, 10, 5460.

3782 Langmuir 2005, 21, 3782-3785

10.1021/la0472504 CCC: $30.25 © 2005 American Chemical SocietyPublished on Web 04/02/2005

ethanol, and placing a drop of the dispersion onto a Holey carbongrid followed by drying at room temperature.

Small-angle X-ray scattering (SAXS) measurements wereperformed using a Kratky compact small angle system (HECUSMBraun, Graz, Austria) instrument equipped with a position-sensitive detector containing 1024 channels of 55 µm width. Amonochromator with a nickel filter was used to select the Cu KRradiation (λ ) 1.542 Å) provided by the generator at 50 kV and40 mA. The sample to detector distance was 282 mm. The sampleswere prepared in a paste holder with thin mica windows. Thecamera volume and the sample were held under vacuum tominimize the scatter from the air.

The specific surface areas were determined by the BET method(after Brunauer, Emmett, and Teller) using an ASAP 2010instrument (accelarated surface area porosimetry system). Thepore size distributions were calculated from the N2 desorptionusing the BJH (after Barrett-Joyner-Halenda) method.9

Preparation of the Hexagonal Mesoporous Material.Pluronic 105 (4.0 g) was dissolved in water (30 g) and a 2 M HClsolution (120 g) under stirring at 35 °C. Tetraethyl orthosilicate(8.50 g) was added, and stirring was continued at 35 °C for 20h. The mixture was aged at 80 °C overnight without stirring.The solid product was recovered (1.8 g), washed with water, andair-dried at room temperature. Calcination was carried out byslowly increasing the temperature from room temperature to500 °C during 2 h followed by heating at 500 °C for 17 h.

Substitution Reaction in the Hexagonal Liquid Crys-talline Phase. For a Ratio 1:1 between KI and 4-TBBB: 4-tert-Butylbenzyl bromide (0.275 g, 1.2 mmol, 1 equiv) was added toPluronic 105 (0.632 g) and placed in a closed bottle, which washeated to above the melting point (about 40 °C). A solution ofpotassium iodide (0.195 g, 1.2 mmol, 1 equiv) in water (0.535 g)was added, and the mixture was vigorously stirred for 1 min.

For a Ratio 10:1 between KI and 4-TBBB: 4-tert-Butylbenzylbromide (0.0275 g, 1.2 mmol, 1 equiv) was added to Pluronic 105(0.632 g) and placed in a closed bottle, which was heated to abovethe melting point (about 40 °C). A solution of potassium iodide(0.195 g, 1.2 mmol, 1 equiv) in water (0.535 g) was added, andthe mixture was vigorously stirred for 1 min.

Samples of the reaction mixture were taken at different timesand added to a 10-fold volume excess of CDCl3. The aqueousphase was removed, and the organic phase was dried with MgSO4.After filtration, the CDCl3 solution was analyzed by NMR.

Substitution Reaction in the Hexagonal MesoporousMaterial. For a 1:1 Ratio between KI and 4-TBBB: Potassiumiodide (0.12 g of a 50% aqueous solution, 1 equiv) was addeddropwise to the mesoporous material (0.1 g), and the slurry wasleft for 24 h. 4-tert-Butylbenzyl bromide (0.082 g, 1 equiv) wasadded, and the suspension was continuously shaken at roomtemperature. Samples taken at various time intervals were addedto a 10-fold volume excess of CDCl3. The organic phase wasseparated and dried with MgSO4. After filtration, the CDCl3solution was analyzed by NMR. To assess the mass balance ofthe reaction, one experiment was carried out for 80 h withoutcollection of samples. The CDCl3 solution obtained after removingthe mesoporous material by filtration was evaporated, the residuewas weighed, and the ratio of starting material (4-tert-butyl-benzylbromide) and product (4-tert-butylbenzyliodide) was de-termined by NMR; see below.

For a 10:1 Ratio between KI and 4-TBBB: Potassium iodide(0.12 g of a 50% aqueous solution, 10 equiv) was added dropwiseto the mesoporous material (0.1 g), and the slurry was left for24 h. 4-tert-Butylbenzyl bromide was dissolved in hexane, anamount corresponding to 0.0082 g, 1 equiv of 4-tert-butylbenzylbromide in 0.03 g of hexane was added, and the suspension wascontinuously shaken at room temperature. Samples taken atvarious intervals were added to a 10-fold volume excess of CDCl3.The organic phase was separated and dried with MgSO4. Afterfiltration, the CDCl3 solution was analyzed by NMR. To assessthe mass balance of the reaction, one experiment was carried outfor80hwithout collectionof samples.TheCDCl3 solutionobtainedafter removing the mesoporous material by filtration wasevaporated, the residue was weighed, and the ratio of starting

material (4-tert-butylbenzylbromide) and product (4-tert-butyl-benzyliodide) was determined by NMR; see below.

The 4-tert-butylbenzyl bromide to 4-tert-butylbenzyl iodideratio was analyzed by means of 1H NMR using a 400 MHz Varianspectrometer. The chemical shifts for -CH2I and -CH2Br were4.46 and 4.50 ppm, respectively. All measurements wereperformed at room temperature.

Results and Discussion

Mesoporous silica with hexagonal geometry was syn-thesized from a micellar solution of the block copolymer(EO)37(PO)58(EO)37, where EO and PO denote oxyethyleneand oxypropylene, respectively. This amphiphilic polymeris known to form a hexagonal liquid crystalline phasebetween 48 and 66 wt % polymer in water.10 Inspectionin a polarized microscope showed that the existence regionof this liquid crystalline phase did not change much neitherwhen pure water was replaced by an aqueous solution ofthe silica precursor (for preparation of the mesoporoussolid) nor when the reactants were added to the water-amphiphilic polymer system (for reaction in the liquidcrystalline phase).

The silica precursor, tetraethyl orthosilicate, underwentspontaneous polymerization in the surfactant solution,forming a silica-block copolymer composite material. Thepolymer was removed by washing. After calcination, anentirely organics-free porous silica was obtained. Thematerial obtained was characterized by powder X-raydiffraction, transmission electron microscopy (TEM), anddetermination of specific surface area and pore size. Thepowder diffraction measurement gave the diffractionpattern with peaks at 2Θ: 1.12, 1.94, and 2.22. Figure 1shows that both the liquid crystalline phase and themesoporous material exhibit hexagonal geometry.

The BET surface area was found to be 809 m2/g. Thepore volume was 0.93 cm3/g, and the average pore diameterwas 46 Å. Figure 2 shows that the pore size distributionis narrow, indicating proper control of the synthesis.Because the pores correspond to the hydrophobic domainsof the liquid crystalline phase, the pore diameter shouldbe approximately equal to the hydrophobic segment ofthe block copolymer. The polyoxypropylene block of thepolymer contains an average of 58 oxypropylene units.This means less than 1 Å per -O-CH(CH3)-CH2- unit.This is of course much less than the theoretical value andindicates that the polyoxypropylene chains are far fromlinear. Similar dimensions have been reported before forrelated systems.11,12

(9) Barrett, E. P.; Joyner, L. G.; Halenda, P. P. J. Am. Chem. Soc.1951, 73, 373.

(10) Alexandridis, P.; Zhou, D.; Khan, A. Langmuir 1996, 12, 2690.(11) Flodstrom, K.; Alfredsson, V. Microporous Mesoporous Mater.

2003, 59, 167.(12) Flodstrom, K.; Alfredsson, V.; Kallrot, N. J. Am. Chem. Soc.

2003, 125, 4402.

Figure 1. Micrographs of the liquid crystalline phase obtainedby a microscope equipped with a polarization filter (left), andthe mesoporous material obtained by TEM (right). Both imagesare typical of a hexagonal geometry.

Use of a Mesoporous Material for Organic Synthesis Langmuir, Vol. 21, No. 9, 2005 3783

The mesoporous material, as well as the correspondinghexagonal liquid crystalline phase, was used as media foranucleophilic substitutionreaction.Tworeferencesystemswereused: amicroemulsionbasedonanonionic surfactantand a surfactant-free oil-water two-phase system. Thereaction profiles are shown in Figure 3. The curves for themicroemulsion and the two-phase reactions are taken fromprevious work in our laboratory.13

The reaction studied is a typical nucleophilic substitu-tion involving a lipophilic organic compound and aninorganic salt: reaction between 4-tert-butylbenzyl bro-

mide (4-TBBB) and potassium iodide (KI). We havepreviously investigated this reaction in some detail usingdifferent types of surfactant self-assembly systems asreaction media, and we have shown that the reactionproceeds well in such microheterogeneous systems.13,14

We have also performed the reaction in a range of organicsolvents with different polarity (and without surfactant),and by comparing the reaction rates obtained in thedifferent solvents we could establish that the mechanismis that of a second-order nucleophilic substitution reac-tion.13

The reaction was monitored by 1H NMR, following therise of the -CH2I signal and the decay of the -CH2Brsignal, as illustrated in Figure 4. There was good cor-respondence between the increase of the -CH2I signaland the decrease of the -CH2Br signal, which indicatesthat side reactions are not important. The most likelyside reaction is probably hydrolysis of 4-TBBA into thecorresponding benzyl alcohol. If this reaction would occurin parallel to the formation of the benzyl iodide, themonitoring of the latter reaction would be distortedbecause there would then not be full equivalence in termsofNMRsignal intensitybetweendisappearanceof-CH2Brand appearance of -CH2I. The peak of -CH2OH meth-ylene protons appears somewhat more downfield than thatof -CH2Br, at 4.58 ppm. No peak appeared at thatfrequency in either the system based on mesoporousmaterial or the liquid crystal-based system.

As can be seen from Figure 3, top, where equimolarconcentrations of the reactants have been used, the ratesare in the order liquid crystal . mesoporous material )microemulsion. Using a 10:1 ratio of KI to 4-TBBB (Figure3, bottom), the order is liquid crystal . mesoporousmaterial > microemulsion > two-phase system. The curvefor reaction in the slurry of mesoporous materials at a10:1 molar ratio of reactants deviates considerably froman ideal curve of loss of starting material for an SN2reaction. We believe that the deviation is due to difficultiesin collecting representative samples from the reactionmixture, which is in the form of a concentrated suspension.Even if this curve cannot be used to determine the rateconstant in a quantitative way, the set of curves shownin Figure 3 allows one to make the qualitative statementthat the liquid crystalline phase is an extremely efficient

(13) Hager, M.; Olsson, U.; Holmberg, K. Langmuir 2004, 20, 6107.(14) Hager, M.; Currie, F.; Holmberg, K. Colloids Surf., A 2004, 250,

163.

Figure 2. Pore size distribution of the mesoporous material.

Figure 3. Reaction profiles for reaction between 4-tert-butylbenzyl bromide and potassium iodide using a 1:1 (top) ora 1:10 (bottom) molar ratio of the reactants. The reactions wereperformed in a hexagonal liquid crystalline phase (LC), amicroemulsion (microemulsion), a mesoporous material withhexagonal symmetry (mesoporous), and, for the reaction wherea 1:10 molar ratio of reactants is used, a two-phase system.

Figure 4. The reaction between 4-tert-butylbenzyl bromideand potassium iodide was studied by 1H NMR, monitoring thedecrease of the -CH2Br signal and the increase of the -CH2Isignal.

3784 Langmuir, Vol. 21, No. 9, 2005 Witula and Holmberg

reaction medium and that the slurry of mesoporousmaterial and the microemulsion gives approximately thesame reaction rate. To provide a quantitative assessmentof the extent of the reaction in the slurry of mesoporousmaterial, the absolute amounts of the starting material,TBBB, and of the product, 4-tert-butylbenzyl iodide (4-TBBI), were determined. The total recoveries of substrate+ product for the reactions with 1:1 and 10:1 molar ratiosof reactants were 95% and 92%, respectively.

It is interesting that the reaction runs so well in theslurry of mesoporous material. The surfactant-basedreaction media, and in particular microemulsions, aredynamic systems, and their usefulness as media fororganic reactions may be due both to the large oil-waterinterface and to the dynamics of the interface, withsurfactants going in and out and surfactant monolayersdisintegrating and reforming. No such dynamics arepresent in the medium based on the solid silica material.The porous material is filled with the aqueous solution ofKI, and the material is mechanically dispersed in anonpolar medium, which is either the lipophilic reactantas such or a solution of the lipophilic reactant inhydrocarbon. Potassium iodide has negligible solubilityin the nonpolar medium and 4-tert-butylbenzyl bromideis virtually insoluble in water,13 which means that thereactants must meet and react at the pore openings. Theproduct obtained, 4-TBBI, is lipophilic and will partitioninto the nonpolar medium.

An attractive feature of the reaction in the slurry ofmesoporous material is the ease of the workup procedure.After completed reaction, the solid particles are filteredoff (and washed with hydrocarbon to remove adsorbedproduct and/or starting material). The residue remainingafter evaporation of the solvent is the reaction mixturewith no contamination of auxiliary substances, such assurfactants or phase transfer agents. As will be demon-strated in a separate paper, the mesoporous solid can bereused several times without much loss of efficiency. Afacile workup procedure may seem trivial but should not

be ignored. Surfactant self-assembly systems, such asmicroemulsions and liquid crystalline phases, contain highconcentrations of amphiphiles that may give rise tofoaming or persistent emulsions during the workup.Likewise, phase transfer agents are sometimes difficultto quantitatively remove from the product. If separationof surfactants, quaternary ammonium compounds, crownethers, or other reaction aids involves time-consumingoperations, the organic chemist will appreciate thisalternativeapproachtoovercomereactant incompatibility.Further work to elucidate in more detail the scope andlimitations of slurries of mesoporous materials as mediafor reactions between incompatible reactants is underwayin our laboratory.

ConclusionsThe reaction between 4-TBBB and KI, which is sluggish

in a simple two-phase system, is fast in the system basedon a slurry of a mesoporous material with hexagonalgeometry. It was found to be fast also in a microemulsionand fast when the hexagonal liquid crystalline phase thatcorresponded to the mesoporous material was used asreaction medium. Even if the slurry of mesoporousmaterial did not give a higher reaction rate than thesystems based on surfactant self-assembly, the formerreaction medium is more attractive from a practical pointof view. The extremely facile workup after completedreaction, filtering followed by evaporation of the solvent,makes it a useful tool in preparative organic synthesis.

Mesoporous materials are today being explored for avariety of applications, out of which catalysis is the mostprominent. We here demonstrate that such materials alsohave a potential for use as medium for reactions betweenincompatible reactants.

Acknowledgment. We thank the Swedish Foundationfor Strategic Research through its Colloid and InterfaceTechnology Program for financial support for T.W.

LA0472504

Use of a Mesoporous Material for Organic Synthesis Langmuir, Vol. 21, No. 9, 2005 3785