Microporous and Mesoporous Materials 129 (2010) 189–199

Contents lists available at ScienceDirect

Microporous and Mesoporous Materials

journal homepage: www.elsevier .com/locate /micromeso

Fluorocarbon and hydrocarbon functional group incorporation into nanoporoussilica employing fluorinated and hydrocarbon surfactants as templates

Gifty Osei-Prempeh a, Hans-Joachim Lehmler b, Anne-Frances Miller c, Barbara L. Knutson a,Stephen E. Rankin a,*

a Department of Chemical and Materials Engineering, University of Kentucky, Lexington, KY 40506-0046, USAb Department of Occupational and Environmental Health, University of Iowa, College of Public Health, Iowa City, IA 52242-5000, USAc Department of Chemistry, University of Kentucky, Lexington, KY 40506-0055, USA

a r t i c l e i n f o a b s t r a c t

Article history:Received 3 February 2009Received in revised form 10 August 2009Accepted 21 September 2009Available online 24 September 2009

Keywords:Functionalized nanoporous silicaFluorocarbon surfactantsDecyltriethoxysilaneHeptadecafluoro-1,1,2,2-tetrahydro-decyltriethoxysilanen-DecylPerfluoro-decyl

1387-1811/$ - see front matter � 2009 Elsevier Inc. Adoi:10.1016/j.micromeso.2009.09.015

* Corresponding author. Tel.: +1 859 257 9799; faxE-mail address: srankin@engr.uky.edu (S.E. Rankin

Ordered mesoporous hydrocarbon functionalized (n-decyl) silica samples are synthesized by the ‘one-pot’ (direct) synthesis method using two cationic fluorinated surfactants, C6F13C2H2NC5H5Cl (HFOPC)and C8F17C2H2NC5H5Cl (HFDePC), and a typical hydrocarbon surfactant, C16H33N(CH3)3Br (CTAB), as tem-plates. The properties of the materials are compared to those of silica samples functionalized with a fluo-rocarbon functional group, heptadecafluoro-1,1,2,2-tetrahydro-decyl, whose fluorocarbon separatingability was reported earlier [33]. The pore characteristics, organic loading, and wetting properties ofthe resulting materials are determined for the following combinations of surfactant/functional groups:hydrocarbon/hydrocarbon, hydrocarbon/fluorocarbon, fluorocarbon/hydrocarbon and fluorocarbon/fluo-rocarbon. Synthesis using the longer chain fluoro-surfactant (HFDePC) template results in the highestincorporation of both n-decyl and fluorocarbon functional groups, with a corresponding loss of long-range pore order in the fluorinated material. Materials synthesized using the HFOPC template have verylow levels of functional group incorporation compared to the HFDePC-templated materials. CTAB-tem-plated materials display greater long-range pore order than the fluorocarbon templated materials. Theincorporation of the fluorocarbon functional precursor is more effective (on a % yield basis) than thehydrocarbon functional precursor for silica materials synthesized using CTAB. Similarly, the use of fluo-rinated surfactant templates enhances fluorocarbon incorporation relative to hydrocarbon incorporationin the mesoporous material. Solubility of the precursors (n-decyltriethoxysilane and heptadecafluoro-1,1,2,2-tetrahydro-decyltriethoxysilane) in the synthesis medium and favorable aggregation behaviorof the alkane/fluoroalkane functional groups with the surfactant micelles improves the incorporationof the functional group.

� 2009 Elsevier Inc. All rights reserved.

1. Introduction

Synthesis of organically functionalized nanoporous silica is anactive research area because of the broad range of potential appli-cations of these high surface area materials in sensing, catalysisand separation. Surfactant templating of sol–gel derived silica re-sults in large surface areas, large pore volumes and narrow poresize distributions. Also, the pore size and structure can be tailoredfor specific applications by using surfactants with different hydro-phobic tails and varying the synthesis conditions. Since the discov-ery of the M41S family of nanostructured materials [1], variousorganic functional groups have been incorporated into mesoporoussilica by means of post-synthesis grafting or direct co-condensa-tion of organofunctional silanes and silica precursors [2].

ll rights reserved.

: +1 859 323 1929.).

Synthesis of functionalized mesoporous silica by co-condensa-tion results in high degrees of functional group incorporation [3].Lim and Stein [3] and Kruk et al. [4] investigated vinyl incorpora-tion by co-condensation in the presence of cetyltrimethylammo-nium bromide (CTAB) as a pore template. An increase in vinylcontent decreases the pore size (moving from mesoporous tomicroporous materials), pore volume and order of the materials.This is consistent with our observations of reduced long-range or-der, pore size, surface area and pore volume with increasing vinylcontent when partially fluorinated surfactants were used as tem-plates to synthesize vinyl functionalized nanoporous silica [5]. Pri-mary amines and secondary amines have also been incorporatedinto mesoporous silica by employing cationic surfactants (e.g.CTAB) [6], anionic surfactant (sodium dodecylsulfate, SDS) [7],nonionic surfactants [8] and block co-polymers [9] as templates.In addition, other functional groups, such as thiol groups (mainlymercaptopropyl functionalization) [10–14] and, more recently,

190 G. Osei-Prempeh et al. / Microporous and Mesoporous Materials 129 (2010) 189–199

organic bridging groups [15,16] have been incorporated intomesoporous silica through co-condensation of their correspondingsilica precursors with tetraalkoxysilanes. Bi-functionalized nano-porous silica materials (e.g., ureidopropyl with 3-[2-(2-aminoeth-ylamino)ethylamino]-propyl [17], sulfonic acid-amino groups[18] and carboxyl-amino groups [19] functionalized silica) havealso been prepared by co-condensation synthesis. Post-synthesisattachment to reactive organic sites in these organic/inorganiccomposites has been used to further functionalize mesoporous sil-ica, for example, with biomolecules [20–22].

Numerous investigations have examined hydrocarbon-basedtemplates [2,23–25] and, more recently, fluorocarbon surfactanttemplates [26–31] for the synthesis of mesoporous silica. However,organic functionalization has focused primarily on the incorpora-tion of hydrocarbon, not fluorocarbon, functional groups. Thereare limited examples of fluorocarbon incorporation into mesopor-ous silica [32–37]. Fluorocarbon functionalization of silica resultsin a surface with increased hydrophobicity and lipophobicity[36]. The applications of fluorinated surfaces, which are ‘‘solventresponsive” include their use in chromatographic purification andseparation of fluorinated compounds [33,38–40], in fluorous bi-phasic catalysis [40,41] and as hydrophobic surface coatings[36,42]. Low-surface tension mobile phases such as fluorocarbonsolvents and supercritical carbon dioxide effectively wet the sur-faces of fluorocarbon-functionalized silica, allowing for their effec-tive use for separation processes. Fluorocarbon-incorporatedporous silica can also be used as a low-k dielectric for semiconduc-tor devices [35].

The hydrocarbon–hydrocarbon interaction of traditional surfac-tant templates and organic functional groups, as well as the hydro-phobic or hydrophilic nature of the functional group, determinethe orientation of the functional group within the pores during di-rect synthesis of organic functionalized mesoporous silica [43–45].The interaction of organic functional groups with the surfactantcan be compared to alcohol interaction with cationic surfactants.For example, the role of alcohol in the formation of CnTAB (alkyltri-methylammonium bromide) aggregates changes from co-solventto co-aggregate to co-surfactant with increasing alcohol chainlength [46]. Matching the chain length and chemical nature of ahydrophobic group and the surfactant hydrophobic tail promotesthe incorporation of the hydrophobic group in surfactant micelles,improving solubilization and stability of the aggregate [47].

Post-synthesis attachment has been the primary method toprepare fluorocarbon-functionalized silica [48]. When extendingdirect synthesis approaches to fluorocarbon-functionalized mate-rial, the nature of the surfactant/organic precursor interactions(e.g., hydrocarbon surfactant template/fluorinated silica precursor)is expected to influence the self assembly process. Mixtures ofhydrocarbon and fluorocarbon surfactants may form uniformlymixed aggregates or segregated hydrocarbon rich and fluorocarbonrich micelles or mixed micelles with de-mixed hydrocarbon andfluorocarbon regions [49,50]. Aggregation behavior of hydrocarbonand fluorocarbon surfactant mixtures is a function of the surfactantchain lengths, hydrophobic chain structure (e.g. aromaticity andbranching), concentration, and temperature [51,52].

The role of perfluoroalkyl chain length in surfactant assemblyand its effect on the silica mesostructure was investigated in thesynthesis of fluorinated organosilicate films by solvent evaporationusing CTAC as template [35]. The fluorocarbon-functionalized silicafilms display both hexagonal and cubic mesostructure for shortchain functional group (Rfh2-) and only hexagonal structure forthe long chain functional groups (Rf6h2- and Rf8h2-). An increasein the hexagonal cell parameter and pore wall thickness withincreasing perfluoroalkyl chain length is attributed to the exten-sion of fluorocarbon groups in the pore channel [35]. Porcherie etal. [37] have also examined templating with hydrocarbon and fluo-

rocarbon surfactants in the synthesis of fluorocarbon-functional-ized silica, thus providing the first investigation of the effect offluorocarbon surfactant/fluorocarbon functional group combina-tion on silica materials properties. The use of cationic fluorinatedsurfactant template resulted in disordered porous silica with lowfluorocarbon incorporation for a long chain fluorinated functionalgroup (Rf6h2-), even at a high TEOS to functional precursor ratioof 19:1. Improved fluorocarbon incorporation and material orderwere observed when a short chain fluorocarbon (CF3(CH2)2- orRfh2-) functionalized silica was synthesized, suggesting that incor-poration of the long chain functional group disrupts the self assem-bly of the cationic fluorinated surfactant micelles. In a study of thepreparation of fluoroalkane-functional materials for fluorophilicseparations, we found contrary evidence that pore order could beretained even with a large amount of fluorocarbon-based silane[33]. However, the degree of incorporation of the functional groupand the effects of the nature of the functional group (hydrocarbonvs. fluorocarbon) required more systematic investigation to under-stand how they influence materials physical properties.

This work examines the effect of combinations of hydrocarbonand fluorocarbon templates and functionalized silica precursorson the material textural properties, pore structure, long-range or-der and functional group loading of functionalized silica formedby direct synthesis. The incorporation of fluorocarbon functionalgroup (heptadecafluoro-1,1,2,2-tetrahydro-decyl; perfluoro-decyl)and its hydrocarbon analogue (n-decyl) is investigated using CTAB(cetyltrimethylammonium bromide), HFOPC (tridecafluoro-1,1,2,2-tetrahydro-octylpyridinium chloride) and HFDePC (hepta-decafluoro-1,1,2,2-tetrahydro-decylpyridinium chloride) as tem-plates. We demonstrate the incorporation of a longer chainfluorinated group with a much higher ratio of the functionalizedprecursor to tetraethoxysilane than has been examined previously[37]. The influence of hydrocarbon/fluorocarbon surfactant/tem-plate interactions on functional group incorporation, mesostruc-ture and physical properties of the silica is examined.

2. Materials and methods

Tetraethoxysilane (TEOS; Fig. 1) with a purity of 99% was ob-tained from Fluka Chemika. n-Decyltriethoxysilane (H10TES, 95%purity; Fig. 1) and heptadecafluoro-1,1,2,2-tetrahydro-decyltrieth-oxysilane (Rf8h2TES, 95% purity; Fig. 1) were purchased fromGelest Inc. CTAB (99% purity; Fig. 2) was obtained from Sigma.The fluorinated surfactants (Fig. 2), C6F13C2H4NC5H5Cl andC8F17C2H4NC5H5Cl, labeled HFOPC and HFDePC, respectively, weresynthesized as previously described [5].

2.1. Synthesis of mesoporous silica

Functionalized mesoporous silica was synthesized by utilizing a4:1 mole ratio of TEOS to Rf8h2TES or H10TES. Particles were pre-pared from aqueous solutions for CTAB and HFOPC-templatedmaterials. HFDePC-templated materials were prepared fromhomogeneous water/ethanol solution, where the addition of etha-nol was required to obtain ordered structures. Non-functionalizedmaterials were synthesized as previously described [5].

Materials synthesis began by adding the surfactant to de-ion-ized ultra-filtered water (DIUFW) and ethanol (if used), and stirringthe mixture for 5 min. Concentrated ammonia (catalyst, 28–30%solution from Malinckrodt) was then added with continuous stir-ring of the mixture for an additional 10 min, after which TEOS ora mixture of TEOS and Rf8h2TES (or H10TES), in the desired propor-tions, was slowly added. The mixture was aged with stirring atroom temperature for 24 h. The molar ratios of the reactants usedin the syntheses were 186 DIUFW:0.184 CTAB (or 0.082

N+

CF2

F2CCF2

F2CCF2

CF3

Cl- N

+

CF2

F2CCF2

F2CCF2

F2CCF2

CF3

Cl-

A B

C

N+

C3 CH3

C3CH3

Br-H

H

Fig. 2. Structures of surfactants: (A) tridecafluoro-1,1,2,2-tetrahydro-octylpyridinium chloride (HFOPC), (B) heptadecafluoro-1,1,2,2-tetrahydro-decylpyridinium chloride(HFDePC) and (C) cetyltrimethylammonium bromide (CTAB).

Si

O

O

O

H3CCH3

H3C

CH3Si

O

OO

O

CH3

CH3H3C

CH3

SiO

O

O

H3C

CF2

CH3

H3C

CF2

CF2CF2

CF2CF2

CF2CF3

A B

C

Fig. 1. Structures of precursors: (A) tetraethoxysilane (TEOS), (B) n-decyl triethoxysilane (H10TES) and (C) heptadecafluoro-1,1,2,2-tetrahydro-decyltriethoxysilane (F8H2TES).

G. Osei-Prempeh et al. / Microporous and Mesoporous Materials 129 (2010) 189–199 191

HFOPC):3.31 NH3:1 TEOS:0.25 H10TES (or Rf8h2TES) and 136DIUFW:64 ethanol: 0.197 HFDePC:10.6 NH3:1 TEOS:0.25 H10TES(or Rf8h2TES). The synthesis procedure is based on the room-tem-perature synthesis of unmodified silica reported by Kumar et al.[53]. The mixture was vacuum-filtered after the aging period andleft to air-dry for 24 h. An ethanol/HCl solution of 150 ml ethanoland 5 g concentrated aqueous HCl was used for the extraction ofthe surfactant from all samples. Extraction was accomplished bystirring the dried sample in acidic ethanol for 24 h. This extractionprocess was repeated once.

Synthesized materials were labeled H10-CTAB and Rf8h2-CTAB;H10-HFOPC and Rf8h2-HFOPC; and H10-HFDePC and Rf8h2-HFDePC.The naming convention for the functionalized silica is the func-tional group incorporated followed by the template used. In tables

and figures, non-functionalized materials were labeled simplyaccording to their surfactant templates (i.e., CTAB, HFOPC andHFDePC).

2.2. Materials characterization

Nitrogen adsorption measurements were performed using aMicromeritics Tristar 3000 automated gas adsorption instrument.The materials were purged at 150 �C using flowing nitrogen gasfor 4 h before performing the adsorption analysis. Data from nitro-gen sorption were analyzed to obtain material total surface area(St), pore volume (vp) and pore size distribution (dp) using as-plotsand the modified BJH method proposed by Kruk, Jaroneic andSayari (KJS) [54–56].

192 G. Osei-Prempeh et al. / Microporous and Mesoporous Materials 129 (2010) 189–199

Fourier transform infrared spectroscopy (FTIR) analysis wasperformed to verify surfactant removal and functional group incor-poration. A small amount (approximately 1 wt.%) of the silica sam-ple was pressed into a pellet with a KBr matrix and analyzed usinga Thermo Nicolet Nexus 470 FT-IR analyzer.

Thermogravimetric analysis (TGA) was carried out using a Net-zsch STA 449 C instrument to quantify organic group incorpora-tion. The materials were analyzed under air purge with a startingtemperature of 22 �C and heated at 10 �C/min up to 800 �C.

Powder X-ray diffraction (XRD) patterns were obtained using aSiemens 5000 diffractometer operating with Cu Ka radiation ofwavelength 0.154098 nm and a graphite monochromator.

Transmission electron micrograph (TEM) images of the mate-rials, after surfactant extraction, were recorded with a JEOL2000FX transmission electron microscope. The TEM samples wereprepared by grinding the sample and mixing with a smallamount of acetone. The solution was allowed to settle and laceycarbon TEM grids (Ted Pella Inc.) were used to scoop some of theclear liquid and then allowed to dry. This was done so that thesmallest particles could be mounted on the grid for better TEManalysis.

Solid-state nuclear magnetic resonance (NMR) experimentswere performed on a Varian Inova NMR spectrometer at the fre-quencies of 79.5 and 400.3 MHz for 29Si and 1H nuclei, respectively.Direct polarization (DP) or cross polarization (CP) from the nearby1H nuclei were used to observe the 29Si nuclei. The samples werepacked in 5-mm zirconia rotors and spun at 4 kHz in a doublytuned chemagnetics probe. Saturation recovery experiments werecarried out for H10- and Rf8h2- samples which produced 29Si longi-tudinal relaxation times in the range of 30–65 s. Quantitative spec-tra were obtained for the samples from 29Si MAS measurements.Excitation with a single 45� pulse of 2.6 ls followed by data acqui-sition under continuous wave (CW) 1H decoupling at 45 kHz wasused. Typically 100–600 scans were completed using a pulse delayof 180 s. To selectively observe 29Si nuclei near 1H nuclei, magneti-zation was transferred from 1H to 29Si by cross polarization at50 kHz. Different rates of 29Si resonance build-up most likely re-flect different distances to nearby 1H nuclei. The maximum1H–29Si polarization transfer was achieved using a contact timeof approximately 10 ms. Usually, 64 scans for 1H–29Si CPMAS weregathered. All chemical shifts were referenced to TSP (trimethylsi-lylpropionic acid) which was set to a chemical shift of 0 ppm.

3. Results and discussion

The incorporation of fluorocarbon and hydrocarbon functional-ity in mesoporous silica by direct synthesis is demonstrated forCTAB, HFOPC, and HFDePC-templated materials. FTIR providesqualitative evidence for organic functionalization, while incorpora-tion is quantified using NMR and TGA, which is interpreted as afunction of the fluorocarbon/hydrocarbon template and precursorcombinations. The effects of fluorocarbon incorporation on poresize, pore order, and surface area are determined using XRD andN2 adsorption analysis. Changes in the surface hydrophobicityare observed with hydrocarbon and fluorocarbon functionalization.

3.1. Functional group incorporation

Fourier transform infrared spectroscopy (FTIR) analysis of thematerials demonstrates successful removal of surfactant templatesand also provides evidence of functional group incorporation.Peaks at 2856 cm�1 and 2925 cm�1 in the as-synthesized (unex-tracted, denoted with the suffix A) CTAB-templated materials aswell as the H10-functionalized materials (H10-CTAB, H10-HFOPC,H10-HFDePC) correspond to the characteristic stretching of CH2

groups in the hydrocarbon H10-functional group and CTAB, thehydrocarbon surfactant template (Fig. 3). The magnitude of thepeak is reduced in H10-CTAB, for example, after surfactant extrac-tion. A significant reduction of the peak at 1473 cm�1 (C–H bend-ing) and disappearance of the 1493 cm�1 peak (N–C stretching;attributed to the presence of CTAB surfactant) is also observed fol-lowing extraction for H10-CTAB. The as-synthesized H10-function-alized materials templated with the fluoro-surfactants (H10-HFOPC-A (Fig. 3B) and H10-HFDePC-A (Fig. 3C)) display peaks cor-responding to the n-decyl (H10) functional group (1473, 2856 and2925 cm�1) as well as additional peaks which disappear after sur-factant extraction (1502 cm�1, due to the pyridinium head group;and 1211 and 1145 cm�1, due to –CF2– stretching of surfactant[57]). The low peak intensities at 2856 and 2925 cm�1 followingsurfactant extraction suggest low n-decyl incorporation in H10-HFOPC relative to H10-HFDePC and H10-CTAB.

The FTIR of the perfluoro-decyl (Rf8h2)-functionalized silicamaterials display peaks which suggest incorporation of thefluoro-functional group and complete removal of surfactants asdiscussed in Osei-Prempeh et al. [33]. The prominent peaks ob-served in all the materials at 1100 cm�1 and the shoulder at1200 cm�1 are typical of silica. The broad band at wavenumbers< 3200 cm�1 is due to adsorbed water and silanols [58].

TGA was used to quantify the extent of organic incorporation inthe synthesized materials. TGA analyses of all materials show aninitial weight loss at temperatures less than 200 �C (Fig. 4) due tothe loss of water and solvent from the materials and possible con-densation of available silanols [59]. A steep transition in weight isobserved between 200 and 800 �C for all samples. The weight lossin the region between 200 and 800 �C indicates the decompositionof the organic functional group [60], giving the percentage of or-ganic functional group in the silica material (Table 1). The theoret-ical organic content, assuming complete hydrolysis and siloxanebond formation during synthesis (i.e., 100% yield), is also presentedin Table 1. The organic content measured by TGA is less than thetheoretical values, which suggests incomplete hydrolysis of theprecursors or the formation of unattached, soluble silsesquioxaneswhich are removed during surfactant extraction.

The incorporation of the fluoro-functional group is greatest inRf8h2-HFDePC relative to Rf8h2-HFOPC but comparable to that ofRf8h2-CTAB, as measured by TGA analysis (Table 1). This higher de-gree of fluorocarbon incorporation in Rf8h2-HFDePC may be due tofavorable interaction between functional group and the surfactanttail, which have the same molecular structure (–C2H2C8F17).Surprisingly, the greatest extent of hydrocarbon functionalizationis also achieved using this fluoro-surfactant template (H10-HFDePC),although the extent of incorporation (60% on a basis of the theoret-ical yield) is significantly less than that of the fluorocarbon func-tional group (Rf8h2-HFDePC; 95%). This high degree of n-decylincorporation may be due to matching the number of carbonatoms in the precursor and template for H10-HFDePC. While hydro-carbons are more soluble in hydrocarbons than fluorocarbons,matching fluorinated chain length and hydrocarbon chain lengthhas been demonstrated to significantly increase the miscibilityrange [61]. Alternatively, solubilization of both the fluorocarbonand hydrocarbon precursors may be enhanced in the micelleaggregates of the ethanol/water media used to template theHFDePC materials, leading to a high degree of functional groupincorporation.

In the absence of added ethanol, the amount of fluorocarbonfunctionalization is higher (1.23 mmol/g) in the CTAB-templatedmaterial (Rf8h2-CTAB) than the HFOPC-templated material(Rf8h2-HFOPC: 0.78 mmol/g). The higher fluorocarbon incorpora-tion in Rf8h2-CTAB may be due to the solubilization of a shorterchain functional group in the longer chain surfactant micelles ofCTAB. The incorporation of the hydrocarbon precursor is minimal

0.0

0.4

0.8

1.2

1.6

2.0

2.4

2.8

CTAB

H10

-CTAB-A

H10

-CTAB

F8H

2-CTAB-A

F8H

2-CTAB

Wavenumber, cm-1

HFOPC

H10

-HFOPC-A

H10

-HFOPC

F8H

2-HFOPC-A

F8H

2-HFOPC

HFDePC

H10

-HFDePC-A

H10

-HFDePC

F8H

2-HFDePC-A

F8H

2-HFDePC

4000 3200 2400 1600 80040.0

0.4

0.8

1.2

1.6

2.0

2.4

2.8A

CTAB

H10

-CTAB-A

H10

-CTAB

F8H

2-CTAB-A

F8H

2-CTAB

Abs

orba

nce

0.0

0.4

0.8

1.2

1.6

2.0

2.4

2.8

0.0

0.4

0.8

1.2

1.6

2.0

2.4

2.8

Abs

orba

nce

0.0

0.4

0.8

1.2

1.6

2.0

2.4

2.8

0.0

0.4

0.8

1.2

1.6

2.0

2.4

2.8

Abs

orba

nce

Wavenumber, cm-1 Wavenumber, cm-14000 3200 2400 1600 8004

Wavenumber, cm-1

Wavenumber, cm-14000 3200 2400 1600 8004

Wavenumber, cm-1

HFOPC

H10

-HFOPC-A

H10

-HFOPC

F8H

2-HFOPC-A

F8H

2-HFOPC

B

HFDePC

H10

-HFDePC-A

H10

-HFDePC

F8H

2-HFDePC-A

F8H

2-HFDePC

C

Rf8h2-CTAB

Rf8h2-CTAB-A

Rf8h2-HFOPC

Rf8h2-HFOPC-A

Rf8h2-HFDePC

Rf8h2-HFDePC-A

Fig. 3. FTIR plots of as-synthesized and extracted (A) CTAB-templated, (B) HFOPC-templated and (C) HFDePC-templated silica. The thin boxes outline the surfactant andorganic bands discussed in the text.

Fig. 4. Sample TGA plots of HFDePC-templated silica.

G. Osei-Prempeh et al. / Microporous and Mesoporous Materials 129 (2010) 189–199 193

in HFOPC (H10-HFOPC, 33% of theoretical maximum). In addition,efficiency of hydrocarbon incorporation in the hydrocarbon-tem-

plated material (H10-CTAB) is reduced relative to fluorocarbonincorporation (47% and 91% theoretical yields, respectively) at aTEOS to functional precursor ratio of 4:1. Fluorocarbon groupsare more hydrophobic than hydrocarbons, which may promotethe incorporation of the fluorocarbon precursor within the micellesof the hydrocarbon surfactant template in spite of its dissimilarchemical nature.

The degree of silica condensation and the extent of functionalgroup incorporation for the materials were confirmed by solidstate 29Si and 29Si 1H-CP NMR of the non-functionalized, H10 andRf8h2 functionalized silica (Fig. 5). The resonance peaks observedat chemical shifts of d = �102 ppm and �111 ppm are attributedto Q4 (siloxane, Si(OSi„)4) and Q3 (one silanol; (OH)–Si(OSi@)3)Si environments, respectively [6,62]. These peaks are seen in bothfunctionalized and non-functionalized silica. In addition both thefluorocarbon and hydrocarbon functionalized silica display tworesonance peaks at d = �58 and �67 ppm, which are attributed toT3 (R–Si(OSi„)3) and T2 ((R)2Si(OSi„)2) Si environments respec-tively [6,62] (where R is either H10 or Rf8h2). The presence of T2

and T3 peaks confirm the covalent attachment of the functionalgroup to Si on the silica surface. The amounts of functional groupincorporation are calculated based on the ratio (T3 + T2)/(Q4 + Q3 + T3 + T2) [6] and are given in Table 1. Similar to the organ-ic content values from TGA, all the NMR values obtained are lowerthan the theoretical values. However, the trend of higher extent offluorocarbon group incorporation compared to hydrocarbon groupseen in the data from TGA analysis is also observed in the NMR

Fig. 5. NMR plots of HFDePC-templated silica (A) observations of cross polarization of 29Si from 1H and (B) direct observations of 29Si.

Table 1Organic content of functionalized materials.

Material Organic content by TGA (mmol/g)a Organic content by NMR (mmol/g) Theoretical organic content (mmol/g)b

Rf8h2-CTAB 1.23 (91%) 1.23 (91%)8<:

100%Rf8h2-HFOPC 0.78 (58%) 0.72 (53%) 1.35 100%Rf8h2-HFDePC 1.28 (95%) 1.29 (96%) 100%H10-CTAB 1.09 (47%) 0.65 (28%)

8<:

100%H10-HFOPC 0.76 (33%) 0.41 (18%) 2.31 100%H10-HFDePC 1.37 (60%) 1.73 (75%) 100%

a Percentage of organic content relative to theoretical yield is in parenthesis.b Theoretical organic content is based on complete hydrolysis and siloxane bond formation ((SiO2)4–Si(R)O3/2, where R is the functional group) (i.e., 100% yield).

194 G. Osei-Prempeh et al. / Microporous and Mesoporous Materials 129 (2010) 189–199

data. Additionally, NMR analysis confirms earlier observation ofgreater functional group incorporation in HFDePC template mate-rials. The NMR-derived yields are 96% (Rf8h2) and 75% (H10) infunctionalized silica templated with HFDePC, whereas CTAB-tem-plated materials show incorporation of 91% (Rf8h2) and 28% (H10)of theoretical yield and 53% (Rf8h2) and 18% (H10) for HFOPC tem-plate materials. The latter sample (H10-HFOPC) gave a very broadpeak that was difficult to detect and integrate, which is consistentwith the low level of organic functionalization reported. The de-gree of incorporation in the fluorocarbon-functionalized materialsobtained from the NMR analysis is comparable to the values ob-tained from TGA. However, for the hydrocarbon functionalizedmaterials the NMR values obtained are either lower (aqueousbased materials; H10-CTAB and H10-HFOPC) or higher (homoge-neous ethanol/water solution based material; H10-HFDePC) thanvalues obtained from TGA (Table 1). The former results (higherapparent organic content via TGA) can be explained by unextractedorganic material (silane oligomers or surfactant) which is not con-nected to the silica framework. This would be consistent with thelow reactivity expected for a long-chain alkyltriethoxysilanemonomer (which may have hydrolyzed independently of TEOS).The latter (less apparent organic content via TGA) is more unusualbut would be consistent with collapse or sintering of the pores dur-ing heating (before TGA, these samples have the narrowest pores –

see below) which causes some organics to be retained during ther-mal treatment.

3.2. Hydrophobicity of functionalized silica materials

The functionalization of silica, which is hydrophilic due to thepresence of surface hydroxyl groups, can result in dramatically dif-ferent surface properties. The hydrophobicity of the functionalizedmaterials due to the incorporation of hydrocarbon and fluorocar-bon groups is confirmed by observing wetting properties of thefunctionalized and non-functionalized silica powder. Representa-tive images of a water droplet on the surface of powder silica(HFDePC, H10-HFDePC and Rf8h2-HFDePC) spread on glass are pre-sented in Fig. 6. The images show the droplets on the functional-ized silica possess very large contact angles (h > 90�), whichsuggest the materials are highly hydrophobic. Hydrocarbon func-tionalized silica particles are able to disperse on the outer surfaceof the water droplet (Fig. 6B), but in the presence of the morehydrophobic fluorocarbon-functionalized silica the water dropletsitting on the silica surface remains clear (Fig. 6C). In contrast,water droplets are immediately absorbed into the non-functional-ized silica material (Fig 6A). The dramatic change in surface prop-erties of the silica powders not only indicates functionalization, butalso suggests the potential application of fluoro-functionalized

Fig. 6. Images of water droplet on (A) non-functionalized HFDePC silica, (B) H10-HFDePC and (C) Rf8h2-HFDePC.

G. Osei-Prempeh et al. / Microporous and Mesoporous Materials 129 (2010) 189–199 195

materials when high lipophobicity and hydrophobicity arerequired.

3.3. Structural properties of functionalized silica

The pore structure and order of the silica material after surfac-tant extraction was examined by powder X-ray diffraction (Fig. 7).For the decyl functionalized silica, 2-D hexagonal pore structure isobserved for the CTAB-templated material (H10-CTAB), as inter-preted from the presence of (1 0 0), (1 1 0), (2 0 0) and (2 1 0)reflections in the XRD profile. Silica templated with the fluorinatedsurfactants (H10-HFOPC and H10-HFDePC) is characterized by onlythe first three reflections, which also correspond to a 2-D hexago-nal pore structure. These reflections are of lower intensity than thehydrocarbon-templated material, suggesting better long-range or-der in the CTAB-templated material. TEM images (Fig. 8) obtainedfor these materials confirm the 2-D hexagonal pore structure in-ferred from the XRD patterns. However, the image of H10-HFDePC(Fig. 8E) shows particles of spherical morphology with radial pores,which is characteristic of materials synthesized in homogeneouswater/ethanol solution [27].

Similar to the trend observed in hydrocarbon functionalized sil-ica, hydrocarbon surfactant templating results in fluoro-function-alized materials with greater order than fluoro-surfactanttemplated materials. While both Rf8h2-CTAB and Rf8h2-HFOPC dis-

1 2 3 4 5 6 7 82-

Rel

ativ

e In

tens

ity

F8H2-CTAB

CTAB

(100)

(110)(200)

(110)(200)

(210)

H10-CTAB

A

2 3 4 5

Rel

ativ

e In

tens

ity

F8H

HHFOPC

(100)

(110

(11

Rf8h2-CTAB

Rf8h

θ 2

Fig. 7. Powder X-ray diffraction plots of (A) CTAB templated, (B) H

play XRD peaks characteristic of hexagonal pore structure, theintensities of the reflections are much higher in the CTAB-tem-plated material. The single broad peak in the XRD pattern forRf8h2-HFDePC suggests materials of low order. The peak widthcould also be due to low electron density contrast because of thehigh incorporation of the fluorocarbon functional group in thismaterial (Table 1). Increasing the incorporation of functionalgroups in mesoporous silica results in decreased pore order[3,4,63]. Spherical particles are also observed for the fluoro-func-tionalized silica synthesized in a homogenous water/ethanol solu-tion (Rf8h2-HFDePC; Fig. 8F). Pores are not observable in Rf8h2-HFDePC at a higher magnification, possibly due the low contrastin the presence of fluorocarbons in the pores.

The physical properties (pore size, dp; total surface area, St; andpore volume, vp) of the silica materials were analyzed by means ofnitrogen adsorption. All materials display type IV adsorptionisotherms (Fig. 9), characteristic of mesoporous materials [64].The H10-functionalized materials possess higher sorption capacity(Fig. 9A–C) than the corresponding fluoro-functionalized materialsowing to the more ordered pore structure of the H10 materials. Forboth the hydrocarbon (H10) and fluorocarbon (Rf8h2)-functional-ized materials, the HFOPC-templated silica materials display sig-nificant hysteresis in the adsorption isotherm and largeadsorption values at relative pressures close to one. This significanthysteresis is due to non-uniformity in the pores and the presenceof a large number of macropores.

6 7 8

2-HFOPC

10-HFOPC

)(200)

(200)0)

B

2 3 4 5 6 7

Rel

ativ

e In

tens

ity

HFDePC

F8H2-HFDePC

H10-HFDePC

(100)

(110) (200)

C

Rf8h2-HFDePC 2-HFOPC

-θ 2-θ

FOPC templated and (C) HFDePC-templated silica materials.

Fig. 8. Representative TEM images of (A) H10-CTAB, (B) Rf8h2-CTAB, (C) H10-HFOPC, (D) Rf8h2-HFOPC, (E) H10-HFDePC and (F) Rf8h2-HFDePC.

196 G. Osei-Prempeh et al. / Microporous and Mesoporous Materials 129 (2010) 189–199

Pore size distributions (Fig. 9a–c) in the synthesized silica weredetermined from the adsorption isotherms using the KJS method[54–56]. A unimodal pore size distribution was observed in therange of 1.5–10 nm for mixed surfactant/functional group systems(hydrocarbon/fluorocarbon and fluorocarbon/hydrocarbon surfac-tants and precursors). A single pore size is consistent with node-mixing into hydrocarbon rich and fluorocarbon rich micellesduring the synthesis of the materials. The similar pore sizes(Table 2) for H10-HFOPC and Rf8h2-HFOPC (2.44 nm) suggest that thefunctional group chain length, rather than the type of functionalgroup, controls the micelle size captured in the synthesis process.Functionalization results in a slight reduction in pore size as com-pared to non-functionalized HFOPC (2.60 nm). In contrast to thisslight reduction of pore size with alkyl or perfluoroalkyl function-alization, we previously observed that the direct synthesis of vinylfunctionalized silica using HFOPC (4:1, TEOS:vinyl triethoxysilane

synthesis conditions) resulted in marked reduction of pore sizefrom 2.60 nm (HFOPC) to 1.66 nm (v4-HFOPC) [5]. The vinyl chainmay act as a co-aggregate during micellar formation due to itsmuch shorter chain length relative to the surfactant templates,whereas the longer Rf8h2-tail will act as a co-surfactant.

The 10-carbon chain hydrocarbon and fluorocarbon functionalgroups (H10 and Rf8h2, respectively) are not expected to fully ex-tend into the hydrophobic core of the CTAB (16-carbon hydropho-bic tail) aggregate. However, the Rf8h2-functional groups will beable to replace some of the CTAB molecules during aggregation.This effect may be accentuated by the larger volume occupied bythe bulky fluorocarbon Rf8h2-group compared with the H10 group.This is consistent with the larger pore size of H10-CTAB (2.76 nm)relative to Rf8h2-CTAB (2.58 nm). On the other hand, functionaliza-tion of HFDePC-templated silica results in a hydrocarbon function-alized material (H10-HFDePC) with a smaller pore size (2.12 nm)

0

100

200

300

400

500

600

700

0.0 0.2 0.4 0.6 0.8 1.0

Vol

ume

Ads

orbe

d, c

m3 /

g ST

P

Relative Pressure, P/Po

0.0 0.2 0.4 0.6 0.8 1.0

0.0 0.2 0.4 0.6 0.8 1.0

Relative Pressure, P/Po

H10-CTAB

CTAB

A

Rf8h2-CTAB

0

100

200

300

400

500

600

700

800

900

Vol

ume

Ads

orbe

d, c

m3 /

g ST

P

Rf8h2-HFOPC

H10-HFOPC

HFOPC

B

0

50

100

150

200

250

300

350

400

Vol

ume

Ads

orbe

d, c

m3 /

g ST

P

Relative Pressure, P/Po

H10-HFDePC

HFDePC

Rf8h2-HFDePC

C

0.0

0.2

0.4

0.6

0.8

1.0

2.0 3.0 4.0 5.0 6.0

Por

e Si

ze D

istr

ibut

ion,

dV

/dD

, cm

3 /g-

nm

Pore Size, nm

CTAB

H10

-CT

AB

Rf8h2-CTAB

a

0.0

0.1

0.2

0.3

0.4

1.5 2.5 3.5 4.5

Por

e Si

ze D

istr

ibut

ion,

dV

/dD

, cm

3 /g-

nm

Pore Size, nm

1.5 2.5 3.5 4.5Pore Size, nm

HFOPC

H10

-HF

OPC

Rf8h2-HFOPC

b

0.0

0.1

0.2

0.3

0.4

Por

e Si

ze D

istr

ibut

ion,

dV

/dD

, cm

3 /g-

nm

HFDePC

Rf 8

h 2-H

FDeP

C

H10-HFDePC

c

Fig. 9. Nitrogen adsorption isotherm and pore size distribution of (A) CTAB templated, (B) HFOPC templated and (C) HFDePC-templated silica materials.

Table 2Structural properties of fluorocarbon and hydrocarbon functionalized materials.

Material d100

(nm)St

(m2/g)Sexternal

a

(m2/g)vp

(cm3/g)dp

(nm)ao

a

(nm)tpw

a

(nm)

CTAB 4.01 995 106 0.76 3.67 4.63 0.96HFOPC 2.87 811 323 0.28 2.60 3.31 0.71HFDePC 3.29 739 41 0.40 2.77 3.80 1.03Rf8h2-CTAB 4.5 819 45.7 0.34 2.58 5.20 2.62Rf8h2-HFOPC 3.07 444 192 0.10 2.44 3.54 1.10Rf8h2-HFDePC 3.87 370 9.48 0.12 2.32 4.47 2.15H10-CTAB 4.14 947 98.6 0.45 2.76 4.78 2.02H10-HFOPC 2.88 618 234 0.13 2.44 3.33 0.89H10-HFDePC 3.23 899 24.4 0.32 2.16 3.73 1.57

a Sexternal is the external specific surface area, ao is the hexagonal cell parameter(2d100/31/2) and tpw is the pore wall thickness (ao � dp).

G. Osei-Prempeh et al. / Microporous and Mesoporous Materials 129 (2010) 189–199 197

than the corresponding fluoro-functionalized material (Rf8h2-HFDePC, 2.32 nm). The functional groups in the more orderedH10-HFDePC might have caused a decrease in the effective poresize compared to the poorly ordered Rf8h2-HFDePC. The smallerpore sizes obtained for the HFDePC-templated materials than forthe HFOPC materials may be due to the combined effect of highfunctional group incorporation in the HFDePC-templated materialsand synthesis in homogeneous water/ethanol medium [65,66].Addition of ethanol tends to decrease micellar size due to reduc-tion in both the hydrophobic driving force for aggregation andhead group repulsion [46].

Pore wall thickness is observed to increase after functionalgroup incorporation (Table 2). The increase in pore wall thicknessis due to the covalent attachment of the functional groups to the

198 G. Osei-Prempeh et al. / Microporous and Mesoporous Materials 129 (2010) 189–199

silica walls [35]. Because of the relative size of the functionalgroups and the level of functional group incorporation, the fluoro-carbon-functionalized silica materials have thicker pore walls thanthe hydrocarbon functionalized materials. Comparing pore wallthickness across surfactants templates, CTAB (hydrocarbon) tem-plated materials possess the largest increase in pore wall thicknessrelative to non-functionalized silica, while the increase in pore wallthickness is less in HFOPC-templated silica possibly due to lowfunctional group incorporation. The incorporation of the functionalgroups in the silica walls generally decreases the total surface areaand pore volume (Table 2). Across all surfactants, the largest de-crease is observed for the fluorocarbon-functionalized silica mate-rials which are less ordered than the hydrocarbon functionalizedsilica.

4. Conclusions

Cationic fluorocarbon surfactants with two chain lengths (tri-decafluoro-1,1,2,2-tetrahydro-octylpyridinium chloride (HFOPC)and heptadecafluoro-1,1,2,2-tetrahydro-decylpyridinium chloride(HFDePC)) and a hydrocarbon surfactant (CTAB) are demonstratedas templates for the synthesis of 2-D hexagonal or disordered porestructured, fluorocarbon (perfluoro-decyl) functionalized meso-porous silica by direct (‘one-pot’) synthesis. Corresponding hydro-carbon functional group (n-decyl) incorporation is alsoinvestigated. For all surfactant/n-decyl precursor combinations,n-decyl functionalized silica with 2-D hexagonal pore structure isobtained. Synthesis in homogeneous water/ethanol medium leadsto high degree of functional group incorporation, and results in dis-ordered pore structure for the fluorocarbon-functionalizedHFDePC-templated silica. A higher yield of fluorocarbon functionalgroup incorporation than the hydrocarbon group was obtained forall surfactants. The dramatic increase in surface hydrophobicity ofthe silica after functional group incorporation, particularly for thefluorocarbon-functionalized silica, is consistent with the analysisby FTIR, NMR and TGA that demonstrated functional group incor-poration. Based on the unimodal pore size distributions achievedfor combinations of fluorinated and hydrocarbon precursors andtemplates, there is no evidence of segregation of hydrocarbon-richand fluorocarbon-rich aggregates during synthesis.

The length of the functional group relative to the surfactant tailand the type of incorporated functional group affect pore size.Structural order, pore size and surface area generally decrease withfunctionalization. The increase in pore wall thickness for function-alized materials, greater for fluorocarbon-functionalized materialsrelative to hydrocarbon functionalized materials, is consistent withthe attachment of the functional groups to the silica pore walls.The degree of fluorocarbon incorporation and the silica surfaceareas are considerably higher than reported previously for fluoro-functionalized silica formed by direct synthesis [37].

Acknowledgments

This material is based on work sponsored by the National Sci-ence Foundation under Grant Nos. DMR-0210517 and CBET-0348234 and the US Department of Agriculture, Grant No. 68-3A75-7-608. We thank Dr. Sandhya M. Vyas of the University ofIowa, IA, for the synthesis and characterization of the fluorinatedpyridinium surfactants.

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) 10834–10843.

[2] A.P. Wight, M.E. Davis, Chem. Rev. 102 (2002) 3589–3614.[3] M.H. Lim, A. Stein, Chem. Mater. 11 (1999) 3285–3295.

[4] M. Kruk, T. Asefa, M. Jaroniec, G.A. Ozin, J. Am. Chem. Soc. 124 (2002) 6383–6392.

[5] G. Osei-Prempeh, J.H. Lehmler, B.L. Knutson, S.E. Rankin, Micropor. Mesopor.Mater. 85 (2005) 16–24.

[6] T. Yokoi, H. Yoshitake, T. Tatsumi, J. Mater. Chem. 14 (2004) 951–957.[7] T. Yokoi, H. Yoshitake, T. Tatsumi, Chem. Mater. 15 (2003) 4536–4538.[8] M.A. Markowitz, J. Klaehn, R.A. Hendel, S.B. Qadriq, S.L. Golledge, D.G. Castner,

B.P. Gaber, J. Phys. Chem. B 104 (2000) 10820–10826.[9] R.K. Zeidan, S.-J. Hwang, M.E. Davis, Angew. Chem. Int. Ed. 45 (2006) 6332–

6335.[10] Y. Guari, C. Thieuleux, A. Mehdi, C. Reye, R.J.P. Corriu, S. Gomez-Gallardo, K.

Philippot, B. Chaudret, Chem. Mater. 15 (2003) 2017–2024.[11] V. Ganesan, A. Walcarius, Langmuir 20 (2004) 3632–3640.[12] K. Kosuge, T. Murakami, N. Kikukawa, M. Takemori, Chem. Mater. 15 (2003)

3184–3189.[13] A. Walcarius, C. Delacote, Chem. Mater. 15 (2003) 4181–4192.[14] A. Walcarius, M. Etienne, B. Lebeau, Chem. Mater. 15 (2003) 2161–2173.[15] K.J. Shea, D.A. Loy, Chem. Mater. 13 (2001) 3306–3319.[16] M.P. Kapoor, O. Yang, S. Inagaki, J. Am. Chem. Soc. 124 (2002) 15176–

15177.[17] S. Huh, H.-T. Chen, J.W. Wiench, M. Pruski, V.S.-Y. Lin, Angew. Chem. Int. Ed. 44

(2005) 1826–1830.[18] R.K. Zeidan, S.-J. Hwang, M.E. Davis, Angew. Chem. Int. Ed. 45 (2006) 6332–

6335.[19] L. Han, J. Ruan, Y. Li, O. Terasaki, S. Che, Chem. Mater. 19 (2007) 2860–2867.[20] C. Lei, Y. Shin, J. Liu, E.J. Ackerman, J. Am. Chem. Soc. 124 (2002) 11242–

11243.[21] Y. Wang, F. Caruso, Chem. Mater. 17 (2005) 953–961.[22] R. Ravindra, S. Zhao, H. Gies, R. Winter, J. Am. Chem. Soc. 126 (2004) 12224–

12225.[23] M.J. Kim, R. Ryoo, Chem. Mater. 11 (1999) 487–491.[24] K. Flodstrom, V. Alfredsson, N. Kallrot, J. Am. Chem. Soc. 125 (2003) 4402–

4403.[25] A. Berggren, A.E.C. Palmqvist, K. Holmberg, Soft Matter 1 (2005) 219–226.[26] B. Tan, A. Dozier, H.-J. Lehmler, B.L. Knutson, S.E. Rankin, Langmuir 20 (2004)

6981–6984.[27] B. Tan, H.-J. Lehmler, S.M. Vyas, B.L. Knutson, S.E. Rankin, Chem. Mater. 17

(2005) 916–925.[28] B. Tan, S.M. Vyas, H.-J. Lehmler, B.L. Knutson, S.E. Rankin, Adv. Funct. Mater. 17

(2007) 2500–2508.[29] J.L. Blin, M.J. Stebe, Micropor. Mesopor. Mater. 87 (2005) 67–76.[30] X. Meng, Y. Di, L. Zhao, D. Jiang, S. Li, F.-S. Xiao, Chem. Mater. 16 (2004) 5518–

5526.[31] C. Rodrıguez-Abreu, P.M. Botta, J. Rivas, K. Aramaki, M.A.L. Quintela, Colloids

Surf. A: Physicochem. Eng. Aspects 298 (2007) 284–286.[32] J.Y. Bae, J.-I. Jung, B.-S. Bae, J. Mater. Res. 19 (2004) 2503–2509.[33] G. Osei-Prempeh, B.L. Knutson, S.E. Rankin, J.H. Lehmler, Ind. Eng. Chem. Res.

47 (2008) 530–538.[34] B. Lebeau, C. Marichal, A. Mirjol, G.J. de A.A. Soler-Illia, R. Buestrich, M. Popall,

L. Mazerolles, C. Sanchez, New J. Chem. 27 (2003) 166–171.[35] J.-I. Jung, J.Y. Bae, B.-S. Bae, J. Mater. Chem. 14 (2004) 1988–1994.[36] M. Pagliaro, R. Ciriminna, J. Mater. Chem. 15 (2005) 4981–4991.[37] O. Porcherie, Y. Guari, C. Reye, New J. Chem. 29 (2005) 538–543.[38] T. Kamiusuki, T. Monde, N. Nakayama, K. Yano, T. Yoko, T. Konakahara, J.

Colloid Interface Sci. 220 (1999) 123–127.[39] S.M. Brittain, S.B. Ficarro, A. Brock, E.C. Peters, Nature Biotechnol. 23 (2005)

463–468.[40] J.A. Gladysz, D.P. Curran, I.T. Horvath, Handbook of Fluorous Chemistry, Wiley-

VCH, Germany, 2004.[41] D. Curran, Z. Lee, Green Chem. 3 (2001) G3–G7.[42] Z.-Z. Gu, H. Uetsuka, K. Takahashi, R. Nakajima, H. Onishi, A. Fujishima, O. Sato,

Angew. Chem. Int. Ed. 42 (8) (2003) 894–897.[43] C.-H. Tung, H.-F. Ji, J. Phys. Chem. 99 (1995) 8311–8316.[44] L. Mercier, T.J. Pinnavaia, Chem. Mater. 12 (2000) 188–196.[45] C.E. Fowler, S.L. Burkett, S. Mann, Chem. Commun. (1997) 1769–1770.[46] K.K. Karukstis, N.D. D’Angelo, C.T. Loftus, J. Phys. Chem. B 101 (1997) 1968–

1973.[47] A. Patist, V. Chhabra, R. Pagidipati, R. Shah, D.O. Shah, Langmuir 13 (1997)

432–434.[48] T. Kamiusuki, T. Monde, K. Yano, T. Yoko, T. Konakahara, Chromatographia 49

(11/12) (1999) 649–656.[49] M. Kadi, P. Hansson, M. Almgren, Langmuir 18 (2002) 9243–9249.[50] M. Almgren, V.M. Garamus, J. Phys. Chem. B 109 (2005) 11348–11353.[51] P. Lo Nostro, L. Scalise, P. Baglioni, J. Chem. Eng. Data 50 (2005) 1148–

1152.[52] T. Asakawa, H. Hisamatsu, S. Miyagishi, Langmuir 12 (1996) 1204–1207.[53] D. Kumar, K. Schumacher, C. Fresne von Hohenesche, M. Grün, K.K. Unger,

Colloids Surf. A 187–188 (2001) 109–116.[54] M. Kruk, M. Jaroneic, A. Sayari, Langmuir 13 (1997) 6267–6273.[55] M. Jaroneic, M. Kruk, J.P. Olivier, Langmuir 15 (1999) 5410–5413.[56] M. Kruk, M. Jaroneic, Chem. Mater. 13 (2001) 3169–3183.[57] D. Lin-Vien, N.B. Colthup, W.G. Fateley, J.G. Grasselli, The Handbook of Infrared

and Raman Characteristic Frequencies of Organic Molecules, Academic PressInc., San Diego, CA, 1991.

[58] S.M. Holmes, V.L. Zholobenko, A. Thursfield, R.J. Plaisted, C.S. Curdy, J. Dweyer,J. Chem. Soc. Faraday Trans. (1998) 2025–2032.

G. Osei-Prempeh et al. / Microporous and Mesoporous Materials 129 (2010) 189–199 199

[59] C.P. Jaroniec, R.K. Gilpin, M. Jaroniec, J. Phys. Chem. B 101 (1997) 6861–6866.[60] T. Hatakeyama, F.X. Quinn, Thermal Analysis Fundamentals and Applications

to Polymer Science, second ed., John Wiley and Sons, England, 1999.[61] T. Asakawa, H. Hisamatsu, S. Miyagishi, Langmuir 11 (1995) 478–482.[62] S. Huh, J.W. Wiench, J.-C. Yoo, M. Pruski, V.S.-Y. Lin, Chem. Mater. 15 (2003)

4247–4256.

[63] A. Sayari, S. Hamoudi, Chem. Mater. 13 (2001) 3151–3168.[64] S.J. Gregg, K.S.W. Sing, Adsorption, Surface Area and Porosity, Academic Press,

London, 1967.[65] B. Tan, S.E. Rankin, J. Phys. Chem. B 108 (2004) 20122–20129.[66] S. Liu, P. Cool, O. Collart, P. Van, E.F. der Voort, O.I. Vansant, G. Lebedev, M. Van

Tendeloo, J. Jiang, J. Phys. Chem. B 107 (2003) 10405–10411.