Trimming nanostructured walls while fluorinating their surfaces: A route to making and widening pores of nanoporous materials and efficient catalysts

Article abstract of DOI:10.1021/cm101041a

We report on the synthesis and characterization of highly ordered mesoporous fluorosilicas (OMFs) and nanoporous and corrugated fluorosilica nanospheres which contain 2.4-7.0 wt % F, corresponding to a loading of 1.3-3.7 mmol/g. Synthesis of these materia

Full text of DOI:10.1021/cm101041a

4950 Chem. Mater. 2010, 22, 4950–4963
DOI:10.1021/cm101041a
Trimming Nanostructured Walls While Fluorinating their Surfaces: A Route to
Making and Widening Pores of Nanoporous Materials and Efficient Catalysts
Cole T. Duncan,§ Ankush V. Biradar,†,‡ Sylvie Rangan,^ Richard E. Mishler II,†,‡ and
Tewodros Asefa*,†,‡
†Department of Chemistry and Chemical Biology, Rutgers, The State University of New Jersey, 610 Taylor Road,
Piscataway, New Jersey 08854, ‡Department of Chemical Engineering and Biochemical Engineering, Rutgers,
The State University of New Jersey, 98 Brett Road, Piscataway, New Jersey 08854, §Department of Chemistry,
Syracuse University, 111 College Place, Syracuse, New York 13244, and ^Department of Physics and Astronomy,
Rutgers, The State University of New Jersey, 136 Frelinghuysen Road, Piscataway, New Jersey 08854
Received April 14, 2010. Revised Manuscript Received June 27, 2010
We report on the synthesis and characterization of highly ordered mesoporous fluorosilicas (OMFs) and
nanoporous and corrugated fluorosilica nanospheres which contain 2.4-7.0 wt % F, corresponding to a
loading of 1.3-3.7 mmol/g. Synthesis of these materials is carried out from parent mesoporous silicas
(MCM-41 and SBA-15) and silica nanospheres under ambient conditions using dilute nonaqueous
solutions of triethyloxonium tetrafluoroborate (Et3OBF4). As evidenced by nitrogen physisorption
measurements, small angle powder X-ray diffraction (XRD), and transmission electron microscopy
(TEM), fluorination of the mesostructures is accomplished with only minor alteration of the materials’
overall order. Detailed compositional analyses before and after fluorination are carried out with the aid of
FTIR spectroscopy and X-ray photoelectron spectroscopy (XPS) as well as elemental analysis proving the
existence of silicon oxyfluoride species. FTIR studies show the appearance of new absorptions (730-750
cm-1) upon fluorination corresponding to [F6-nSi(OH)n]2-, O4/2SiF species or Si-O-Si stretches altered
by the presence of nearby Si-F species as well as general broadening of the Si-O-Si asymmetric stretching
region (1000-1200 cm-1). XPS analyses exhibit two distinct F species observed at 690.5 and 687.6 eV
corresponding to silicon oxyfluoride (SiOF) and silicon fluoride (SiF), respectively. Surface silanol content
is evaluated by 29Si HP-MAS NMR spectroscopy, which shows significant reductions in (Q2þQ3)/Q4 ratios
from 1.26, 0.61, and 1.07 to 0.22, 0.38, and 0.61 for SBA-15, trimethylsilyl-capped SBA-15, and MCM-41
mesoporous silicas, respectively, because of the formation of (SiO3)SiF or pentacoordinated (SiO4)SiF
species. The as-described synthetic procedure is also applied to sulfonic acid (-SO3H)-functionalized
SBA-15. Enhancement of the acid-catalyzed ring opening of styrene oxide by aniline to produce the
corresponding β-aminoalcohol by the fluorinated materials is presented. Ring opening of styrene oxide by
aniline at room temperature after 4 h using -SO3H functionalized SBA-15 result in 77% conversion,
whereas the as-synthesized OMF and OMF-SO3H materials exhibit 79% and 87% conversions, respectively.
Fluorination of silica nanospheres with Et3OBF4 under similar conditions also produces etched nanoporous
fluorosilica nanospheres with different degrees of exfoliation depending on the fluorination temperature.
1. Introduction
properties for various applications.1,2 Many of these materials
contain procedural elements which include etching steps
which partially dissolve silica from the material to increase
surface area or enhance adsorptive capabilities while other
methodologies aim to totally remove the silica to form novel
metal oxides with unique structural properties such as in the
formation hollow metal oxide nanospheres,3 nanowires or
nanotubes,4 or mesoporous materials5 where the silica was
used as a scaffold. Solution phase silica etching has been
widely reported to be accomplished by the use of either a
concentrated strong base (i.e., KOH or NaOH) or by hydro-
fluoric acid (HF), which must be handled with extreme care
due to inherent risks they impose during experimentation.
These harsh etching techniques form soluble H2SiF6 or
M2SiO3 (M = Na, K) salts, which are removed in the
In recent years, there have been many significant develop-
ments toward the fabrication of silica-based nanoscale mate-
rials such as functionalized mesoporous silicates (FMS), func-
tionalized SiO2 nanospheres, and a plethora of core-shell
nanomaterials with a wide range of physical and chemical
*Corresponding author. Tel: (732) 445-2970. Fax: (732) 445-5312. E-mail:
tasefa@rci.rutgers.edu.
(1) Asefa, T.; Anan, A.; Duncan, C. T.; Xie, Y. In Non-Magnetic Bi-
Metallic and Metal Oxide Nanomaterials for Life Sciences; Kumar, C.
S. S. R., Ed.; Wiley-VCH: Weinheim, Germany, 2009; Vol. 3, Chapter 9.
(2) (a) Yang, Q.; Liu, J.; Zhang, L.; Li, C. J. Mater. Chem. 2009, 19, 1945–
1955. (b) Huang, S.; Fan, Y.; Cheng, Z.; Kong, D.; Yang, P.; Quan, Z.;
Zhang, C.; Lin, J. J. Phys. Chem. C 2009, 113, 1775–1784. (c) Gao, Q.;
Xu, Y.; Wu, D.; Li, X. J. Phys. Chem. C 2009, 113, 12753–12758.
(d) Vallet-Regı, M.; Balas, F.; Arcos, D. Angew. Chem., Int. Ed. 2007,
€
46, 7548–7558. (e) Hoffman, F.; Cornelius, M.; Morell, J.; Froba, M.
Angew. Chem., Int. Ed. 2006, 45, 3216–3251. (f) De Palma, R.; Peeters,
S.; Van Bael, M. J.; Van den Rul, H.; Bonroy, K.; Laureyn, W.;
Mullens, J.; Borghs, G.; Maes, G. Chem. Mater. 2007, 19, 1821–1831.
(3) Pan, J. H.; Zhang, X.; Du, A. J.; Sun, D. D.; Leckie, J. O. J. Am.
Chem. Soc. 2008, 130, 11256–11257.
r
2010 American Chemical Society
pubs.acs.org/cm
Published on Web 08/05/2010

Article
Chem. Mater., Vol. 22, No. 17, 2010 4951
washing procedure leaving the remaining etched silica. As will
be discussed, there are also many examples where fluorine
containing species, such as ionic liquids or gaseous mixtures,
have been utilized as etching and fluorinating agents. Such
methodologies produce a substantial amount of fluoride ions
which, when in the presence of a silica species, lead to the
formation of insoluble fluorosilicate or organofluorosilicate
species. The fabrication of such materials has been studied
extensively as useful components of low dielectric (k) micro-
electronic devices.6 Fluorinated silicas have also found utility
in amide dehydration reactions7 and alkylation reactions.8
The synthesis of fluoro- and organofluoro-silicate species
has been accomplished using a variety of different methodo-
logies. For example, the fabrication of thin films via plasma-
enhanced chemical vapor deposition (PECVD), using gas-
eous mixtures consisting of fluorine containing species such
as SiF4, C4F8, or SF6, oxygen (O2), inert gases such as Ar or
He, and a silica source [i.e., Si(100) substrates, tetraethy-
lorthosilicate (TEOS) or trimethylsilane (TMS)] has shown
to result in potential fluorosilicate species useful for low k
devices.9 Fabrication using PECVD requires careful manip-
ulation of reagent ratios by varying gas flow rate during
deposition producing thin films with reported F content
ranging from 2-25 at %. Direct fluorination of mesoporous
silica gel using F2 gas has also been proven to produce F
containing materials with 4-13 wt % F (2.1-6.8 mmol/g)
without the formation of unwanted SiF4 or SiF62- species.
However, the latter synthetic method involving F2 gas signi-
ficantly reduces the surface area (14-46%) of the parent
material and is extremely dangerous, necessitating highly
specialized equipment that may inhibit broad application.10
Other procedures utilize aqueous or nonaqueous fluor-
ide containing salt solutions to etch and/or fluorinate
silica. For instance, dilute NH4F in aqueous media has
been shown to promote the formation of Si-F species via
nucleophilic substitution of surface silanols.11 This pro-
cedure, however, generates species with relatively low
fluorine contents of 0.7-1.6 wt % (0.4-0.8 mmol/g)
whereas higher fluorine content is due to the formation
1
of (NH4)2SiFx(OH)6-x salts as evidenced by H high-
spinning frequency MAS NMR spectroscopy.12 Tetra-
fluoroborate (BF4-), hexafluorophosphate (PF6-), and
other similar ionic liquids (ILs) have also been used to
produce F- ions in situ in a slow and controlled fashion.
For example, the dissociation of tetrafluoroborate (eq 1)
has a significantly low dissociation constant (K);13 how-
ever, because of the high affinity that silicon has for
fluoride ion, fluoride ion is rapidly quenched resulting
in an equilibrium shift to the right.
BF4 ðaqÞTBF3ðaqÞ þ F- ðaqÞ K ¼ 1 ꢀ 10- 39 ð1Þ
-
Using this motif, 1-hexyl-3-methyl-imidazolium, 1-butyl-
-
4-methyl-pyridinium, and tetrabutylammonium BF4
/
-
PF6 compounds were used in the etching and growth
of InP nanocrystals by microwave induction.14 Similar
fluorine-containing ILs have also been shown to be a
suitable alternative to HF for electrochemically and
photoelectrochemically induced etching of n-type and
p-type silicon wafers.
(4) (a) Arnold, D. C.; O’callaghan, J. M.; Sexton, A.; Tobin, J. M.;
Amenitsch, H.; Holmes, J. D.; Morris, M. A. Appl. Surf. Sci. 2009, 255,
9333–9342. (b) Megouda, N.; Hadjersi, T.; Piret, G.; Boukherroub, R.;
Elkechai, O. Appl. Surf. Sci. 2009, 255, 6210–6216. (c) Douani, R.;
Hadjersi, T.; Boukherroub, R.; Adour, L.; Manseri, A. Appl. Surf. Sci.
2008, 254, 7219–7222. (d) Asefa, T.; Shi, Y.-L. J. Mater. Chem. 2008, 18,
5604–5614. (e) Ge, J.; Zhang, Q.; Zhang, T.; Yin, Y. Angew. Chem., Int.
Ed. 2008, 47, 8924–8928. (f) Mulvihill, M. J.; Rupert, B. L.; He, R.;
Hochbaum, A.; Arnold, A.; Yang, P. J. Am. Chem. Soc. 2005, 127,
16040–16041. (f) Whitsitt, E. A.; Barron, A. R. Nano Lett. 2003, 3, 775–
778.
On the other hand, although there are many examples of
fluorinated nanoscale low k materials, the use of fluorinated
and/or functionalized fluorinated silicates in catalysis is still
relatively uncharted. One property of interest relating to the
fluorination of silicates or aluminosilicates is the enhance-
ment of surface acid catalyzed reactions. Recently, the dea-
lumination of the zeolite ferrierite using ammonium hexa-
fluorosilicate (NH4)2SiF6 yielded a fluorinated zeolite that
exhibited selectivity in the skeletal isomerization of n-butene
to isobutene.15 Ammonia temperature programmed desorp-
tion (TPD) and pyridine chemisorption studies showed the
surface acidity was reduced because of the fluorination of
mainly the stronger acid sites. This, in turn, led to a lower
surface density of acid sites, which lowered the possibility
of competing active sites and reduced the local reagent
concentration, thus inhibiting the formation of unwanted
byproducts. Although the fluorinated materials exhibited a
(5) Yan, J.; Wang, A.; Kim, D.-P. J. Phys. Chem. B 2006, 110, 5429–
5433.
(6) (a) Cardinaud, C.; Peignon, M.-C.; Tessier, P.-Y. Appl. Surf. Sci.
2000, 164, 72–83. (b) Mocella, M. T. J. Fluorine Chem. 2003, 122,
87. (c) Urbansky, E. T. Chem. Rev. 2002, 102, 2837–2854.
(d) Katsuo, T.; Satohito, Y.; Kimio, T. Bull. Jpn. Pet. Inst. 1970,
12, 136–141.
(7) Jacques, R.; Reppelin, M.; Seigneurin, L. Fluorinated silica cata-
lyst and preparation of aromatic/aliphatic nitriles in the presence
thereof. U.S. Patent 4 822 903, May 27, 2006.
(8) (a) Mantilla-Ramırez, A.; Ferrat-Torres, G.; Domınguez, J. M.;
Aldana-Rivero, C.; Bernal, M. Appl. Catal., A 1996, 143, 203–214.
(b) Han, S.; Shihabi, D. S.; Chang, C. D. J. Catal. 2000, 196, 375–
~
378. (c) Canizares, P.; Carrero, A. Appl. Catal., A 2003, 248, 227–
237.
(9) (a) Pankov, V.; Alonso, J. C.; Ortiz, A. J. Appl. Phys. 1999, 86, 275–
280. (b) Wang, P.-F.; Ding, S.-J.; Zhang, W.; Zhang, J.-Y.; Wang,
J.-T.; Wei, W.-L. Chin. Phys. Lett. 2000, 17, 912–914. (c) Kim, S. P.;
Choi, S. K. Thin Solid Films 2000, 379, 259–264. (d) Wang, P.-F.;
Ding, S.-J; Zhang, D. W.; Wang, J.-T.; Lee, W. W. Thin Solid Films
2001, 385, 115–119. (e) Park, Y.-B.; Rhee, S.-W. J. Non-Cryst.
Solids 2004, 343, 33–38. (f) Cheng, Y. L.; Wang, Y. L.; Juang, Y.;
O’Neill, M. L.; Lukas, A. S.; Karwacki, E. J.; McGuian, S. A.;
Tang, A.; Wu, C. L. J. Phys. Chem. Solids 2008, 69, 518–522.
(g) Pereira, J.; Pichon, E.; Dussart, R.; Cardinaud, C.; Duluard,
C. Y.; Oubensaid, E. H.; Lefaucheux, P.; Boufnichel, M.; Ranson,
P. Appl. Phys. Lett. 2009, 94, 071501–071503.
(11) (a) Jungblut, H.; Lewerenz, H. J. Appl. Surf. Sci. 2000, 168, 194–
197. (b) Loius, B.; Kiwi-Minsker, L. Microporous Mesoporous
Mater. 2004, 74, 1–10. (c) Hartmeyer, G.; Marichal, C.; Lebeau,
B.; Caullet, P.; Hernandez, J. J. Phys. Chem. C 2007, 111, 6634–
6644. (d) Barabash, R. M.; Zaitsev, V. N.; Kovalchuk, T. V.; Sfihi,
H.; Fraissard, J. J. Phys. Chem. A 2003, 107, 4497–4505.
(12) Hartmeyer, G.; Marichal, C.; Lebeau, B.; Rigolet, S.; Caullet, P.;
Hernandez, J. J. Phys. Chem. C 2007, 111, 9066–9071.
(13) Stacey, M.; Tatlow, J. C.; Sharpe, A. G. In Advances in Fluorine
Chemistry; Academic Press: New York, 1960; Vol. 1, p 72.
ꢀ
(10) (a) Lataste, E.; Legein, C.; Body, M.; Buzare, J.-Y.; Tressaud, A.;
Demourgues, A. J. Phys. Chem. C 2009, 113, 18652–18660.
(b) Lataste, E.; Demourgues, A.; Leclerc, H.; Goupil, J.-M.;
ꢁ
Durand, E.; Labrugere, C.; Benalla, H.; Tressaud, A. J. Phys.
(14) Lovingood, D. D.; Strouse, G. F. Nano Lett. 2008, 8, 3394–3397.
~
(15) Canizares, P.; Carrero, A. Appl. Catal., A 2003, 248, 227–248.
Chem. C 2008, 112, 10943–10951.

4952 Chem. Mater., Vol. 22, No. 17, 2010
Duncan et al.
lower % conversion, they reduced the formation of bypro-
ducts such as propane and n-butane. The oligomerization of
propene was also studied using an (NH4)2SiF6 trþeated zeolite
HZSM-5.16 The formation of unbranched C20 hydrocar-
bons was enhanced upon fluorination compared to that of
the nonfluorinated zeolite. Additionally, the formation of
unwanted branched oligomers was reduced. The fluorinated
materials were shown to have isolated acidic sites and
lowered surface acidity than that of the parent material,
and result in limited oligomer branching. Sulphated TiO2-
SiO2 mixed oxide species that could be used as a Brønsted
acid catalyst for cumene cracking and esterification reac-
tions showed increased reactivity upon fluorination.17
Fluorination was accomplished using a solution of NH4F
in dilute sulfuric acid resulting in materials with 0.5-2 wt %
F (0.3-1.1 mmol/g) and enhanced catalytic activity because
that of the corresponding nonfluorinated materials. β-Ami-
no alcohols have shown promise as biological/medicinal
agents or chiral auxiliaries,18 insecticides,19 and metal che-
lates in catalysis.20 To date, several methodologies have
been shown to catalytically form β-amino alcohols from
epoxides and aromatic/aliphatic amines. Metal halides,21
triflates,22 nitrates23 and perchlorates,24 sulfamic acid,25
hexafluoro-2-propanol (HFIP),26 1,4-diazabicyclic[2,2,2]-
octane (DABCO),27 ionic liquids,28 and even water29 have
all been shown capable of catalyzing the ring opening
reaction of epoxides by amines. Heterogeneous catalysts,
such as aluminosilicates,30 functionalized and non-
functionalized mesoporous silicas,31 polyoxometalates,32
silica and titanosilicate nanoparticles,33 amberlyst-15,34
(19) (a) Ferrarini, S. R.; Duarte, M. O.; da Rosa, R. G.; Rolim, V.;
ꢀ
of higher Brønsted acidity and favorable inductive interac-
Eifler-Lima, Gilsane von Poser, V. L.; Sarda Ribeiro, V. L. Vet.
2-
Parasitol. 2008, 157, 149–153. (b) Graebin, C. S.; Madeira, M;
Yokoyama-Yasunaka, J. K. U.; Miguel, D. C.; Uliana, S. R. B.;
tions with nearby SO4
.
Herein, we report on the synthesis of mesoporous SBA-
15 and MCM-41 type highly ordered mesoporous fluor-
osilicas (OMFs) and nanoporous and corrugated fluorosi-
lica nanospheres (CFNSs) using triethyloxonium tetrafluo-
roborate (Et3OBF4), a commonly used alkylating agent in
organic syntheses. Using facile synthetic protocols, OMFs
and CFNSs were prepared without the use of HF, toxic gas
mixtures, or specialized equipment necessary for electro-
chemical, PECVD or microwave induced procedures.
Although many of the other procedures mentioned use
either silica wafers or as-synthesized thin layers, this pro-
cedure is capable of producing OMF’s on a large scale. The
composition of synthesized OMF species were examined
via FTIR spectroscopy, X-ray photoelectron spectroscopy
(XPS), and elemental and thermogravimetric (TGA) ana-
lyses and found to contain 2.4-7.0 wt % F corresponding
to a loading of 1.3-3.7 mmol/g. The morphology of the
materials was investigated using nitrogen physisorption mea-
surements, small angle powder X-ray diffraction (XRD),
transmission electron microscopy (TEM), and solid state
29Si HP-MAS and CP-MAS NMR spectroscopy. This
synthetic method enables widening and restructuring of
the pore diameters of mesoporous silica without compro-
mising its ordered structure while at the same time per-
mitting fluorination of the material’s surface. Fluorinated
materials exhibit larger pore volumes and diameters with
only minor resultant pore constriction. To further study
the effects of our presented fluorination methodology, we
assessed the catalytic activity of OMF and sulfonic-acid-
functionalized mesoporous fluorosilica materials (OMF-
SO3H) in the ring opening of styrene oxide by aniline form-
ing the corresponding β-amino alcohol and compared it to
ꢀ
Benitez, D.; Cerecetto, H.; Mercedes Gonzalez, M.; Gomes da
Rosa, R.; Eifler-Lima, V. L. Eur. J. Med. Chem. 2010, 45, 1524–
1528.
(20) (a) Yan, W.; Mao, B.; Zhu, S.; Jiang, X.; Liu, Z.; Wang., R. Eur. J.
Org. Chem. 2009, 3790–3794. (b) Zhu, S.; Yan, W.; Mao, B.; Jiang, X.;
Wang, R. J. Org. Chem. 2009, 74, 6980–6985. (c) Mahadik, G. S.;
Knott, S. A.; Szczepura, L. F.; Peters, S. J.; Standard, J. M.;
Hitchcock, S. R. J. Org. Chem. 2009, 74, 8164–8173. (d) Dahlenburg,
L.; Treffert, H.; Farr, C.; Heinemann, F. W.; Zahl, A. Eur. J. Inorg.
Chem. 2007, 1738–1751. (e) Li, Y.; He, B.; Qin, B.; Feng, X.; Zhang, G.
J. Org. Chem. 2004, 69, 7910–7913. (f) Rasmussen, T.; Norrby, P.
J. Am. Chem. Soc. 2003, 125, 5130–5138.
(21) (a) Mojtahedi, M. M.; Abaee, M. S.; Hamidi, V. Catal. Commun.
2007, 8, 1671–1674. (b) Chakraborti, A. K.; Rudrawar, S.; Kondaskar,
A. Eur. J. Org. Chem. 2004, 3597–3600. (c) Sundararajan, G.;
Vijayakrishna, K.; Varghese, B. Tetrahedron Lett. 2004, 45, 8253–
ꢀ
8256. (d) Pachon, L. D.; Gamez, P.; van Brussel, J. J. M.; Reedijk, J.
Tetrahedron Lett. 2003, 44, 6025–6027. (e) Chakraborti, A. K.;
Kondaskar, A. Tetrahedron Lett. 2003, 44, 8315–8319.
(22) (a) Procopio, A.; Gaspari, M.; Nardi, M.; Oliverio, M.; Rosati, O.
Tet. Lett. 2008, 49, 2289–2293. (b) Placzek, A. T.; Donelson, J. L.;
Trivedi, R. Tetrahedron Lett. 2005, 46, 9029–9034. (c) Gibbs, R. A.;
De, S. K. Chem. Lett. 2004, 33, 304–305.
(23) Bhanushali, M. J.; Nandurkar, N. S.; Bhor, M. D.; Bhanage, B. M.
Tetrahedron Lett. 2008, 49, 3672–3676.
(24) Pujala, S. B.; Chakraborti., A. K. J. Org. Chem. 2007, 72, 3713–
3722.
(25) Kamal, A.; Rajendra Prasad, B.; Malla Reddy, A.; Naseer, M.;
Khan, A. Catal. Commun. 2007, 8, 1876–1880.
(26) Das, U.; Crousse, B.; Kesavan, V.; Bonnet-Delpon, D.; Begue, J. P.
J. Org. Chem. 2000, 65, 6749–6751.
(27) Wu, J.; Xia, H. G. Green Chem. 2005, 7, 708–710.
(28) Yadav, J. S.; Reddy, B. V. S.; Basak, A. K.; Narsaiah, A. V. Tet.
Lett. 2003, 44, 1047–1050.
(29) (a) Wang, Z.; Cui, Y. T.; Xu, Z. B.; Qu., J. J. Org. Chem. 2008, 73,
2270–2274. (b) Bonollo, S.; Fringuelli, F.; Pizzo, F.; Vaccaro, L.
Green Chem. 2006, 8, 960–964. (c) Azizi, N.; Saidi, M. R. Org. Lett.
2005, 7(17), 3649–3651.
(30) (a) Kureshy, R. I.; Agrawal, S.; Kumar, M.; Khan, N. H.; Abdi,
S. H. R.; Bajaj, H. C. Catal. Lett. 2010, 134, 318–323. (b) Chakravarti,
R.; Oveisi, H.; Kalita, P.; Pal, R. R.; Halligudi, S. B.; Kantam,
M. L.; Vinu, A. Microporous Mesoporous Mater. 2009, 123, 338–
344. (c) Robinson, M. W. C.; Timms, D. A.; Williams, S. M.;
Graham, A. E. Tetrahedron Lett. 2007, 48, 6249–6251. (d) Kureshy,
R. I.; Singh, S.; Khan, N. H.; Abdi, S. H. R.; Suresh, E.; Jasra, R. V.
J. Mol. Catal., A 2007, 264, 162–169. (e) Onaka, M.; Kawai, M.;
Izumi, Y. Chem. Lett. 1985, 779–782.
(31) (a) Bordoloi, A.; Hwang, Y. K.; Hwang, J. S.; Halligudi, S. B.
Catal. Commun. 2009, 10, 1398–1403. (b) Heravi, M. M.; Baghernejad,
B.; Oskooie, H. A. Catal. Lett. 2009, 130, 547–550. (c) Saikia, L.;
Satyarthi, J. K.; Srinivas, D.; Ratnasamy, P. J. Catal. 2007, 252,
148–160.
(32) Danafar, H.; Yadollahi, B. Catal. Commun. 2009, 10, 842–847.
(33) (a) Sreedhar, B.; Radhika, P.; Neelima, B.; Hebalkar, N. J. Mol.
Cat. A: Chem. 2007, 272, 159–163. (b) Satyarthi, J. K.; Saikia, L.;
Srinivas, D.; Ratnasamy, P. Appl. Catal., A 2007, 330, 145–151.
(34) Vijender, M.; Kishore, P.; Narender, P.; Satyanarayana, B. J. Mol.
Catal., A 2007, 266, 290–293.
(16) (a) Han, S.; Shihabi, D. S.; Chang., C. D. J. Catal. 2000, 196, 375–
378.
(17) Samantaray, S. K.; Parida, K. M. Appl. Catal., A 2001, 211, 175–
187.
ꢀ
ꢀ
ꢀ
(18) (a) Cordova, I.; Leon, L. G.; Leon, F.; San Andres, L.; Luis, J. G.;
ꢀ
Padron, J. M. Eur. J. Med. Chem. 2006, 41, 1327–1332. (b) Clarkson,
C.; Musonda, C. M.; Chibale, K.; Campbell, W. E.; Smith, P.
Bioorg. Med. Chem. 2003, 11, 4417–4422. (c) Danieli, E.; Trabocchi,
A.; Menchi, G.; Guarna, A. Eur. J. Org. Chem. 2005, 4372–4381.
(d) Ager, D. J.; Prakash, I.; Schaad, D. R. Chem. Rev. 1996, 96, 835–
875.

Article
Chem. Mater., Vol. 22, No. 17, 2010 4953
and sulfated zirconia35 have also been used as catalysts
for the epoxide ring-opening reaction. Many of these
catalysts, however, suffer from extended reaction times,
elevated temperatures, large amounts of catalyst, or
assistance of a solvent to perform efficiently. Herein, we
also report on the use of OMF’s as solid acid catalysts for
the formation of β-amino alcohols. The fluorination
methodology outlined in this paper was applied to the
fluorination SBA-15 and sulfonic acid functionalized
SBA-15 catalysts. The outlined catalytic procedure is
solvent-free and does not require elevated temperatures,
which is necessary to prevent the formation of unwanted
byproducts and to avoid the loss or degradation of
organic materials functionalized on the surface.9d This
fluorination and etching method has also been proven to
be versatile as it also allows etching of nanosized silica
spheres of ∼250-300 nm producing corrugated nanopor-
ous fluorosilica nanospheres (CFSNs). This new method
may serve as a facile route for tuning structures and
fluorinating surfaces of a wide range of nanostructured
and nanoporous metal oxides as well as provide a method
for improved preparation of functionalized mesoporous
silicates or similar nanoscale materials and nanocatalysts.
decanted. The precipitate was washed three times with a solution
of deionized (DI) H2O (20 mL) and ethanol (20 mL) under
sonication. The precipitate was dispersed in a 1:1 water/ethanol
solution (40 mL overall) until further use. The resulting silica
nanospheres (NS) were found by TEM to be 250 ( 10 nm in
diameter.
Etching and Fluorination of Mesoporous Materials. Typically,
mesoporous silica (0.2 g) was dispersed in CH2Cl2 (100 mL). 1.0
M Et3OBF4 (1.5-6 mL) was then added and the solution was
allowed to stir for 19 h at room temperature or under reflux
(50 ꢀC). Upon completion, the mesoporous silica was recovered
via vacuum filtration and washed with CH2Cl2 (200 mL) and
EtOH (200 mL). The material was dried overnight at 80 ꢀC and
stored in a desiccator for future analysis. This experiment was
performed on MCM-41, SBA-15 and CSBA-15. Furthermore,
different concentrations: low (1.5 ꢀ 10-2 M), medium (2.9 ꢀ
10-2 M), and high (5.7 ꢀ 10-2 M) concentrations of Et3OBF4
and different etching temperatures of room temperature (RT)
and 50 ꢀC were used in each case. The resulting etched samples
were given designations based on the etching condition em-
ployed (etching temperature and concentration). For instance,
SBA-15 samples after etching at RT in low (1.5 ꢀ 10-2 M),
medium (2.9 ꢀ 10-2 M), and high (5.7 ꢀ 10-2 M) concentrations
of Et3OBF4 were designated as SBA-L-RT, SBA-M-RT, and
SBA-H-RT, respectively, where L, M, and H stand for low,
medium, and high, respectively. Similarly, SBA-15 samples after
etching at 50 ꢀC in low, medium, and high concentrations of
Et3OBF4 were designated as SBA-L-50, SBA-M-50, and SBA-
H-50, respectively. Similar sample designation was also applied
for the MCM-41, CSBA-15, and silica nanospheres.
2. Experimental Section
Materials and Reagents. All reactions were performed under
standard atmospheric conditions. Ethanol (EtOH, 95%) was
obtained from Pharmco-AAPER. Aniline (g99.5%), dichloro-
methane (CH2Cl2, g99.5%) hexamethyldisilazane (g99%),
styrene oxide (97%), tetraethyl orthosilicate (98%) (TEOS),
toluene (g99.5%), and triethyloxonium tetrafluoroborate
(Et3OBF4, 1.0 M in CH2Cl2) were purchased from Sigma-
Aldrich. 3-Mercaptopropyltrimethoxysilane (MPTS) was pur-
chased from Gelest. Hydrochloric acid (12.1 N), sulphuric acid
(36 N), and H2O2 (30 wt %) were obtained from Fisher
Scientific. All the chemicals were used as received.
Synthesis of Solid Acid Catalyst. Synthesis of sulfonic-acid-
functionalized SBA-15 was carried out in a manner similar to
previously reported procedures.39 Briefly, as-synthesized SBA-
15 was calcined at 550 ꢀC for 5 h. SBA-15 (1.6 g) was dispersed in
dry toluene. MPTS (2.2 mL, 11.8 mmol) was then added to the
solution and stirred at 80 ꢀC for 5 h. The MPTS-modified SBA-
15 was then separated via vacuum filtration and washed co-
piously with EtOH (95%) and dried overnight at 80 ꢀC. Oxida-
tion of surface mercaptopropyl groups was performed by
dissolving the material (0.6 g) in H2O2 (30 wt %, 20 mL) and
stirring at RT for 24 h. The material was then separated via
vacuum filtration and washed copiously with DI H2O. The
material was then resuspended in 60 mL of 1 M H2SO4 for 2 h
to ensure complete oxidation. The formed sulfonic acid cata-
lyst (SBA-15-SO3H) was then separated via vacuum filtration
and washed with 2 L DI H2O followed by 500 mL EtOH (95%)
and dried.
Synthesis of SBA-15 and MCM-41 Mesoporous Silica Mate-
rials. Synthesis of mesoporous SBA-15 and surfactant extraction
via solvent washing was performed as reported in previously
published literature.36 Synthesis of mesoporous MCM-41 and
surfactant extraction via calcination at 550 ꢀC for 5 h was
performed as reported previously.37
Synthesis of Organic Capped SBA-15 (CSBA-15). A solution
of extracted SBA-15 (1.0 g) in toluene (200 mL) was prepared.
Excess hexamethyldisilazane (HMDS, 5 mL) was added to the
solution and allowed to stir at room temperature for 24 h as
performed previously.38 The solution was filtered and the solid
was washed with ethanol and then dried. The resulting tri-
methylsilyl-capped sample was named as CSBA-15.
Fluorination of the as-synthesized solid acid catalyst was
performed as above. SBA-15-SO3H (0.200 g) was dispersed in
CH2Cl2 (100 mL) via sonication. A 1.0 M Et3OBF4 solution
(6 mL) was added via syringe and the reaction was stirred
overnight at RT. The fluorinated solid acid catalyst (OMF-
SO3H) was separated via vacuum filtration and washed with
CH2Cl2 (500 mL) and dried for 3 h at 80 ꢀC and stored in a
desiccator until further use.
Synthesis of 250 nm Silica Nanospheres (NS). A solution of
ethanol (100 mL) and water (3.6 g) was stirred vigorously for
2 min. TEOS (2.8 g) was then added into it under moderate
stirring. After 3 h of stirring, a milky solution was formed. The
solution was centrifuged and the supernatant was carefully
Catalytic Experiments. Aniline (450 μL, 5 mmol) and styrene
oxide (570 μL, 5 mmol) were combined in a 5 mL round-bottom
(35) Reddy, B. M.; Patil, M. K.; Reddy, B. T.; Park, S. E. Catal.
Commun. 2008, 9, 950–954.
(36) Zhao, D. Y.; Huo, Q. S.; Feng, J. L.; Chmelka, B. F.; Stucky, G. D.
J. Am. Chem. Soc. 1998, 120, 6024–6036.
(37) Kresge, C. T.; Leonowicz, M. E.; Roth, W. J.; Vartuli, J. C.; Beck,
J. S. Nature 1992, 359, 710–712.
(38) (a) McKittrick, M. W.; Jones, C. W. Chem. Mater. 2005, 17, 4758–
4761. (b) Xie, Y.; Quinlivan, S.; Asefa, T. J. Phys. Chem. C. 2008,
112, 9996–10003.
(39) (a) Dhainaut, J.; Dacquin, J.; Lee, A. F.; Wilson, K. Green Chem.
2010, 12, 296–303. (b) Kureshy, R. I.; Ahmad, I.; Pathak, K.; Khan,
N. H.; Abdi, S. H. R.; Jasra, R. V. Catal. Commun. 2009, 10, 572–
575. (c) Reddy, S. S.; Raju, B. D.; Kumar, S.; V.; Padmasri, A. H.;
Narayanan, S. K.; Rao, S. R. Catal. Commun. 2007, 8, 261–266.
(d) Saikia, L.; Satyarthi, J. K.; Srinivas, D.; Ratnasamy, P. J. Catal.
2007, 252, 148–160.

4954 Chem. Mater., Vol. 22, No. 17, 2010
Duncan et al.
flask. The catalyst (13 mg) was then added to the reaction
mixture and dispersed via sonication. The reaction was vigor-
ously stirred under ambient conditions. The reaction yields and
compositions were monitored by gas chromatography (GC) and
gas chromatography-mass spectrometry (GC-MS). The pro-
duct was purified using column chromatography and analyzed
via 1H NMR. For 2-phenylamino-2-phenyl ethanol (Regioi-
Scheme 1. Possible Mechanisms of Mesoporous Fluorosilicate
Formation: (A) Fluorination of Surface Silanols via Nucleophilic
Substitution and/or (B) Fluorination of the Fully Condensed Silica
within the Mesoporous Framework
1
somer B): H NMR (CDCl3, 300 MHz): δ 3.72 (1H, dd), 3.92
(1H, dd), 4.52 (1H, dd), 6.65 (2H, d), 6.79 (1H, t), 7.20 (2H, t),
and 7.30-7.44 (5H, m). 13C-NMR (CDCl3, 300 MHz): δ 60.4,
67.7, 114.4, 118.4, 127.3, 128.0, 129.2, 129.7, 140.7, 141.1, 142.5,
147.9, 149.7. GC-MS (m/e): 214.1, 213.0, and 182.0.
Instrumentation. Solid-state 29Si CP-MAS and HP-MAS
€
NMR spectra were obtained using a 300 MHz Bruker Avance
NMR spectrometer. For 29Si CP-MAS NMR experiments,
7.0 kHz spin rate, 10 s recycle delay, 10 ms contact time, π/2
1
pulse width of 5.6 μs, and 256 - 1024 scans using TPPM H
decoupling were employed. 29Si HP-MAS experiments were
done with a 7.0 kHz spin rate, 100 s recycle delay, π/6 pulse
width of 1.9 μs, and 700-4000 scans using high-power contin-
uous-wave 1H decoupling. Solution state 1H and 13C NMR
€
spectra were obtained using a Bruker DPX-300 NMR. FTIR
spectra were obtained using a Thermo-Nicolet IR200 spectro-
meter with samples loaded onto a KBr disk (25 ꢀ 4 mm).
Thermogravimetric analysis (TGA) or decomposition profiles
were acquired for mesoporous materials with a Q500 series
thermogravimetric analyzer (TA Instruments). The TGA data
were collected under nitrogen atmosphere (60 cm3/min) in the
temperature range of 25-800 ꢀC at a rate of 10 ꢀC/min. Gas
adsorption isotherms were obtained using a Micromeritics
Tristar 3000 surface area and porosity analyzer. Samples were
degassed at 160 ꢀC for 12-16 h and run at 77 K under nitrogen
adsorption/desorption. The specific surface area (SBET) was
calculated according to the Brunauer-Emmett-Teller (BET)
method from the adsorption branch of the isotherm. Pore size
distributions were obtained using the Barrett-Joyner-Halenda
(BJH) method from the adsorption branch of the N2 sorption
isotherm. Powder X-ray diffraction (XRD) profiles were ob-
tained using a Scintag powder diffractometer. TEM images were
taken with a FEI Tecnai T-12 S/TEM instrument. X-ray photo-
electron spectroscopy (XPS) was performed using a Surface
Science Instrument SSX-100. The reaction mixtures of catalytic
studies were analyzed by GC (Agilent equipped with HP 6850,
FID detector and HP-1, 30 m long ꢀ 0.25 mm ID column). The
products were confirmed by GC-MS (HP-5971 equipped with a
HP-5 MS 50 m long ꢀ 0.200 mm ꢀ 0.33 μm capillary column).
Elemental analyses of all the materials were carried out by QTI-
Intertek (Whitehouse, NJ).
NMR spectroscopy, TGA, TEM, powder XRD, XPS
spectroscopy, and elemental analyses.
The fluorination process is believed to take place
according to two different possible mechanisms shown in
Scheme 1. These mechanisms are supported by detailed
characterization methods such as solid-state 29Si NMR,
elemental analysis, FTIR and XPS which will be discussed
in the proceeding sections. As shown in mechanism A of
Scheme 1, the reaction is believed to be carried by the
following steps: (1) formation of oxonium ion via intro-
duction of Etþ, (2) nucleophilic attack by fluoride ion to
silica surface, and (3) displacement of ethanol or, in the
case of CSBA-15, ethoxytrimethylsilane. The coordination
of F- and dissociation of EtOR (R = H or SiMe3) most
likely takes place through an associative mechanism with
the formation of a pentavalent silicon species, which is the
commonly accepted mechanism for substitution for sili-
con. The second possible route (mechanism B) shows a
similar mechanism; however, in this case, formation of the
oxonium ion occurs within the fully condensed silica
framework. Uponnucleophilicattackof F-, a correspond-
ing ethoxysilane moiety is formed on the surface which
may further react, as in mechanism A, to form another
fluorosilica species and one equivalent of Et2O. Because of
the fact that fluorination occurs in a nonaqueous solvent,
3. Results and Discussion
The preparation of highly ordered OMFs was accom-
plished from SBA-15 and MCM-41 using Et3OBF4, a
common alkylating agent in many organic syntheses. To
study the effect of concentration and temperature, prepara-
tion was carried out in dilute Et3OBF4 solutions of 1.5, 2.9,
and 5.7 ꢀ 10-2 M (in CH2Cl2), respectively, at room
temperature (RT) and reflux (50 ꢀC) under atmospheric
conditions. These concentrations of Et3OBF4 were labeled
as low (L), medium (M), and high (H) concentrations,
respectively. The materials before and after fluorination
with Et3OBF4 were characterized by nitrogen physisorp-
tion measurements, FT-IR spectroscopy, solid-state 29Si

Article
Chem. Mater., Vol. 22, No. 17, 2010 4955
Table 1. Structural Data of Mesoporous Materials before and after
Fluorination
average pore
diameter
(A)c
BET BJH average
surface pore volume
unit
cell (A)b
samplea
SBA-15 (A)
SBA-M-RT (B)
SBA-H-RT (C)
SBA-L-50 (D)
SBA-M-50 (E)
SBA-H-50 (F)
MCM-41 (A)
MCM-L-RT (B)
MCM-M-RT (C)
MCM-H-RT (D)
MCM-M-50 (E)
MCM-H-50 (F)
CSBA-15 (A)
area (m2/g)
(cm3/g)c
104
119
111
113
116
106
42
43
42
42
43
74
72
73
85
73
81
23
21
22
21
21
24
72
78
85
78
581
527
0.44
0.47
0.51
0.55
0.50
0.69
0.55
0.51
0.52
0.46
0.64
0.70
0.68
0.81
0.86
0.91
584
388
578
497
991
1137
1147
1053
1051
1092
579
43
119
119
121
119
CSBA-15-M-RT (B)
CSBA-15-H-RT (C)
CSBA-15-M-50 (D)
641
675
788
a Sample nomenclature (i.e., SBA-15 (A)) to give reference to relating
figures. b ao = 2d100/31/2 (A) for 2D hexagonally ordered materials.
ꢀ
c Obtained from adsorption branch of the nitrogen adsorption iso-
therms.
The XRD patterns show that all the materials maintained
their ordered mesoporous structures indicating that the
fluorination did not result in collapse of the channel walls
or the formation of macropores. In fact, as can be seen in
Figure 1, fluorination actually resulted in increase in the
Bragg reflection peaks of the samples. This may be a result
of either pore channel reconstruction into more ordered
structures or increase in electron contrast between the
channel pores and channel walls upon etching. The latter
was a resultof the mild etching of the walls of the materials,
broadening of the pores (see below). The interplanar
spacing of the mesoporous structures did not exhibit
significant changes except subtle increases to higher lattice
spacing or lattice expansion upon fluorination. The TEM
images of the samples also showed that the mesoporous
structures of the organic capped and uncapped SBA-15
samples remained intact after fluorination and etching (see
the Supporting Information, Figure S1). In the case of the
uncapped SBA-15 many of the pore walls appear to be
somewhat corrugated, with regions containing irregular
bulges, which may be consistent with a broadened pore size
distribution and/or increased pore interconnectivity. As will
be discussed later, corrugation is also observed to clearly
take place on the surface of 250 nm silica nanospheres.
Nitrogen sorption studies were performed to study how
the fluorination process altered the surface area, pore
diameter and volume, and sorption isotherm (Figure 2
Figure 1. XRD patterns of (I) SBA-15 (A) before and (B-G) after
fluorination under varying conditions: (B) SBA-L-RT, (C) SBA-M-RT,
(D) SBA-H-RT, (E) SBA-L-50, (F) SBA-M-50, and (G) SBA-H-50;
(II) MCM-41 (A) before and (B-F) after fluorination: (B) MCM-L-
RT, (C) MCM-M-RT, (D) MCM-H-RT, (E) MCM-M-50, and
(F) MCM-H-50; and (III) CSBA-15 (A) before and (B-E) after treat-
ment: (B) CSBA-L-RT, (C) CSBA-M-RT, (D) CSBA-H-RT, and
(E) CSBA-M-50.
ꢀ
and Table 1). Parent SBA-15 material exhibited ∼47 A
pore diameter with a surface area of 581 m2/g and a pore
volume of 0.44 cm3/g. The isotherm shows sharp capillary
condensation step with good hysteresis indicating uni-
form mesopore size. Upon etching and fluorination, pore
volume increases of 7-16% and 14-57% resulted for RT
and 50 ꢀC treatments, respectively. BJH pore size dis-
tributions generally show a smaller pore size distribution
(PSD) for samples fluorinated at RT while a broadening
occurs for those fluorinated at 50 ꢀC (see the Supporting
however, formation of new surface silanols in this mechan-
ism is not as likely. The detailed characterization results to
support these mechanisms are discussed below.
3.1. Structural Analyses. Figure 1 displays the XRD
patterns for (I) extracted SBA-15, (II) calcined MCM-41,
(III) CSBA-15, and their fluorinated products. The struc-
tural data about the materials are also compiled in Table 1.

4956 Chem. Mater., Vol. 22, No. 17, 2010
Duncan et al.
diameters (Figure 2-II and the Supporting Information,
Figure S2-II). Fluorination, in almost all cases, resulted in
lowered average diameters, whereas the surface areas and
pore volumes were increased by 6-16% and 6-27%,
respectively (see the Supporting Information, Figure S2-
II). This could be indicative of increased order of the
mesopores by possible restructuring of the silicate frame-
work. The only exception is treatment under the highest
[BF4-] at elevated temperature where the PSD became
larger and broadened in a fashion similar to that of SBA-
15 series samples.
The HMDS-capped (or CSBA-15) sample exhibits
some interesting characteristics after treatment with
-
BF4 (Figures 2-III and the Supporting Information,
Figure S2-III). The surface area and pore volume in-
creases gradually with increasing [BF4-] as well as with
increasing temperature. As seen in Table 1, the surface
area of the parent CSBA-15 (579 m2/g) increases to 641,
675, and 788 m2/g for samples etched with [BF4-] con-
centration of 2.9 ꢀ 10-2 M at RT; 5.7 ꢀ 10-2 M at RT;
and 2.9 ꢀ 10-2 M at 50 ꢀC, respectively. The samples with
doubled concentration of [BF4-] and elevated tempera-
tures showed pore volume increases of 19 - 34%. BJH
pore size distributions (see Figure S2-III in the Support-
ing Information) also showed that as concentration and
temperature increased, shifts toward larger diameters and
broadened PSDs resulted for the fluorinated samples
compared to that of the parent CSBA-15.
The majority of etched materials exhibited larger sur-
face areas than their parent materials. Pore diameter
increase in the fluorinated SBA-15 materials may be due
to the large density of surface silanols and microprosity of
solvent-extracted SBA-15 material and their removal
upon fluorination. In the case of the CSBA-15 materials
this is due to, as will be shown, desilylation of the TMS
capping agents as well as the removal of remaining
uncapped silanols upon fluorinaton. A certain extent of
pore reconstruction is observed but there is relatively little
loss of order or integrity (i.e., pore channel collapse).
3.2. Compositional Analyses. The weight content of
trimethylsilyl and silanol species of parent and fluorinated
CSBA-15, extracted SBA-15, and calcined MCM-41 mate-
rials were obtained from thermogravimetric traces (see the
Supporting Information, Table S1 and Figure S3). As can
be seen with the CSBA-15 and extracted SBA-15 samples,
the overall weight percent loss before and after fluorination
is not drastically changed. The TGA profiles of the extracted
SBA-15 and its fluorinated samples do show subtle differ-
ences, however (see the Supporting Information, Figure S3-
I). As can be seen, the extracted materials shows three main
regions of weight loss: (1) 25-100 ꢀC, where physisorbed
H2O was removed, (2) 100-400 ꢀC, corresponding to
condensation of geminal or closely adjacent silanols parti-
cipating in H-bonding, and (3) >400 ꢀC where isolated
silanols are removed.8b Upon fluorination at room tem-
perature or low [BF4-] concentration at elevated tempera-
Figure 2. Nitrogen gas adsorption isotherms of mesoporous silica (I)
SBA-15 (A) before and (B-F) after its treatment under varying condi-
tions: (B) SBA-M-RT, (C) SBA-H-RT, (D) SBA-L-50, (E) SBA-M-50,
and (F) SBA-H-50; (II) MCM-41 (A) before and (B-F) after fluorina-
tion: (B) MCM-L-RT, (C) MCM-M-RT, (D) MCM-H-RT, (E) MCM-
M-50, and (F) MCM-H-50; and (III) CSBA-15 (A) before and (B-D) after
fluorination: (B) CSBA-M-RT, (C) CSBA-H-RT, and (D) CSBA-M-50.
Materials, Figure S2-I). The corresponding isotherms of
all fluorinated SBA-15 materials (Figure 2-I) exhibit a
shift toward larger partial pressures (P/Po), showing an
enlargement of the primary mesopores. As shown in
Figure 2-I, the formation of a tail during the desorption
branch, indicative of partial blocking or clogging of the
mesopores, also results, proving pore reconstruction in
extracted SBA-15 samples leading toward the formation
of some restricted mesopores.16 As shown by PSDs of the
materials, RT treatment makes a more highly ordered
structure, whereas elevated temperature etches the mate-
rials more significantly.
ture (50 ꢀC), the second region extends to 450 - 550 ꢀC with
-
The calcined MCM-41 samples, meanwhile, did not
undergo significant changes of isotherm or average pore
steep loss at the end of this region. When higher [BF4
concentrations are used at elevated temperature, the second
]

Article
Chem. Mater., Vol. 22, No. 17, 2010 4957
region extends and experiences a smoother, less steep,
weight percent loss. One possible reason for this occurrence
may be due to the presence of more unreacted silanols
trapped within the mesostructure which would not be
reactive during the fluorination process at room tempera-
ture or low [BF4-] at elevated temperature (50 ꢀC). So, these
samples still have higher weight loss in the 450-550 ꢀC
region compared to the latter samples whose Si-OH groups
were replaced by Si-F. It is seen that some of the fluori-
nated CSBA-15 samples lose a lower wt% in the organic
region than the parent sample (see the Supporting Informa-
tion, Table S1 and Figure S3-II). This may be attributed to
the loss of capping agent during BF4- treatment in addition
to the reduced number of Si-OH groups due to the capping
and subsequent fluorination steps. Nevertheless, the weight
percent does not change drastically, because some of the
new weight loss may also be due to the presence of residual
unreacted silanols trapped within the mesostructure or the
loss of certain fluorosilicate species that may have formed
(i.e., (SiO)1SiF3), which condense with nearby silanols.9d The
loss of these types of species have been shown to occur
between 250-350 ꢀC and have similar masses as the capping
agent. A slightly steeper weight percent loss over this
temperature range for fluorinated CSBA-15 than that of
the parent CSBA-15 material is observed. The calcined
MCM-41 samples, however, show a marked decrease in
silanol (and organic, if any residual surfactant exists) con-
tent. During calcination, surface silanols condense to form a
more rigid and interconnected Si-O-Si network, which
decreases the content of (SiO)2-Si(OH)2 and (SiO)3-Si-
OH species, leading to reduced weight loss from 100 to
600 ꢀC where these silanols condense, expelling H2O. The
weight loss in this region for calcined MCM-41 corresponds
to only ∼1 wt % (see the Supporting Information, Table S1
and Figure S2-III). Upon fluorination, this value increases
to ∼4-8 wt %. This weight loss is due to loss of new
silanols/ethoxysilanes, HF, and fluorosilicate species that
are formed by the fluorination steps. The lowered surface
silanol density of calcined materials limits available fluori-
nation sites, inhibiting the mechanism proposed in
Scheme 1A, and consequently promotes the second possi-
blepathway(Scheme1B) wherecleavageofsiloxanebridges
occurs. From the TGA data shown, a general broadening
and elongation to higher temperatures of the 100-400 ꢀC
implies a lower surface concentration of silanols. Addition-
ally, the higher temperatures necessary for removal of such
species proves that the fluorinated materials exhibit en-
hanced stability and resistance to thermal destruction.
Using solid-state 29Si{1H} cross-polarization magic angle
spinning (CP-MAS, Figure 3) and high power magic angle
spinning (HP-MAS) NMR spectroscopy (Figure 4), the
degree of condensation within the silica framework was
determined for samples before and after fluorination. The
CP-MAS spectra, when compared to that of the parent
materials and their fluorinated products, show many differ-
ent properties. The spectrum for the parent SBA-15 material
(Figure 3A) shows well-defined (SiO)2Si(OH)2 (Q2), (SiO)3-
SiOH (Q3), and (SiO)4Si (Q4) silicate species, appearing at
-94, -103, and -112 ppm, respectively. Upon fluorination
Figure 3. 29Si CP-MAS NMR spectra of (A) SBA-15, (B) SBA-M-RT,
(C) SBA-M-50, (D) CSBA-15, (E) CSBA-M-RT, (F) CSBA-M-50,
(G) calcined MCM-41, (H) MCM-M-RT, and (I) MCM-M-50.
Figure 4. 29Si HP-MAS NMR spectra of (A) SBA-15, (B) SBA-M-RT,
(C) SBA-M-50, (D) CSBA-15, (E) CSBA-M-RT, (F) CSBA-M-50,
(G) calcined MCM-41, (H) MCM-M-RT, and (I) MCM-M-50.
(Figure 3B,C), the Q2 and Q3 sites are almost completely
removed, and a dominant Q4 peak remains. The Q4 peak
exhibit slightly upfield shifts from -112 to -114 ppm. The
spectrum for CSBA-15 parent material (Figure 3D) shows
D1 species at 11 ppm corresponding to the trimethylsilyl
(TMS) species (formed from capping with HMDS), Q3
(-103 ppm), and Q4 (-112 ppm). Upon fluorination at
RT (Figure 3E), D1 sites are reduced substantially and are
almost completely removed at 50 ꢀC (Figure 4F). The Q3 and
Q4 peaks shift upfield to -105 and -114, respectively. The
spectra for MCM-41 (Figures 4G-4I) samples exhibit a trend
similar to that of the SBA-15 samples. Well-defined Q2 (-93
ppm), Q3 (-103 ppm), and Q4 (-112 ppm) silicate species
can be seen in the parent MCM-41 material. Upon fluorina-
tion, the Q2 peak disappears. The Q3 and Q4 species shift
upfield to -105 and -113 ppm, respectively, for samples
treated BF4- at RT and 50 ꢀC. Previously performed studies
have reported (SiO)2Si(OH)F and (SiO)2SiF2 to correspond
to visible signals at -95.5 to -97 ppm and (SiO)3SiF to be
present at -106.5 ppm as well as the formation of penta-
coordinated species (SiO)4SiF and (SiO)3SiF2 to occur from
-118 to -125 ppm.10a,11c When these values are compared to
the spectra for the fluorinated materials in this study, it is
noticed that all spectra exhibit some small intensity contribu-
tion from -95 to -100 ppm indicating the possibility of
(SiO)2Si(OH)F and (SiO)2SiF2 species. Also, in all spectra,
the Q4 peaks baseline at approximately -120 ppm which

4958 Chem. Mater., Vol. 22, No. 17, 2010
Duncan et al.
Figure 5. TEM images ofsilica nanospheres before and after fluorination. Images show(A) silica nanospheres (200 and 500 nm scales), (B) NS-M-RT, and
(C) NS-M-50.
implicates the presence of (SiO)3SiF but not (SiO)2SiF2. The
overall upfield shift of the remaining Q3 peaks and conver-
gence of signal intensity toward the Q4 range is evidence of
the formation of (SiO)3SiF species. Additionally, the overall
upfield shift of the Q4 peaks would also provide evidence for
the formation of the pentacoordinated species.
The 29Si HP-MAS solid-state NMR data presented
would be consistent with either increased silica condensa-
tion, which is not directly indicated by the small angle
XRD or nitrogen gas sorption data, or decrease in surface
silanol content due to nucleophilic substitution by fluor-
ide ion. In all cases, the amount of silanol sites is reduced,
possibly to a similar degree for the parent materials SBA-
15 and CSBA-15. In the CSBA-15 samples desilylation
takes place without the formation of new Q2 and Q3
silicon sites. The calcined MCM-41 samples most likely
did form some amount of new silanols or ethoxysilyl
groups from the cleavage/reconstruction of Si-O-Si
bonds due to the lack of silanol sites for F- ions to react
with promoting the reaction, with the Q4 sites being
substituted by some fluorinated sites (as shown by TGA).
Even though the materials exhibit different TGA profiles
and the parent materials have distinctive solid state NMR
spectra, there is no immediately discernible differences
between the spectra of the final fluorinated OMF analo-
gues whether their parent materials are extracted, capped,
or calcined. This is an indication that fluorination of both
surface silanols and capping species occurs resulting in
products with similar composition when analyzed by
TGA and solid state NMR techniques. Therefore, fluor-
ination mainly occurs on the surface of the materials
without extensive etching deep within the pore walls
unless the surface silanol content is insufficient.
Elemental analyses performed before and after fluor-
ination showed the presence of a substantial concentra-
tion of fluorinated species (Table 2). After fluorination,
the extracted SBA-15 samples showed the highest F
content ranging from 5.0-7.0 wt %, whereas CSBA-15
samples showed a slightly lower loading of 4.0-4.7 wt %,
and calcined MCM-41 samples exhibited only 2.4-3.6
wt % F. These results are actually quite high considering
the fact that the fluorination process predominantly
occurs on the surface of the material by substituting
surface silanols (and surface organic groups in case
of CSBA-15), indicated by solid state NMR and struc-
tural analyses, leading to a material with a very high
surface density of silicon oxyfluoride species. These
values indicate that these materials contain 1.3-1.9,
2.1-2.5, and 2.6-3.7 mmol/g F for fluorinated MCM-
41, CSBA-15, and SBA-15 samples, respectively. Even
though MCM-41 has a significantly higher surface area,
and is therefore expected to have a larger surface reacti-
vity than the SBA-15 and CSBA-15 materials, the lower
values of mmol/g F in the former are not surprising due to
As seen in the HP-MAS solid-state NMR spectrum of
extracted SBA-15 (Figure 4A), (SiO)2Si(OH)2 (Q2),
(SiO)3SiOH (Q3), and (SiO)4Si (Q4) silicate species are
present, appearing at -93, -103, and -112 ppm, respec-
tively, with a (Q2þQ3)/Q4 ratio of 1.26, indicative of
incomplete hydrolysis and condensation of the silanol
species during synthesis. Upon fluorination, however, this
ratio substantially drops to 0.22 and 0.24 for samples
fluorinated at room temperature (Figure 4B) and 50 ꢀC
(Figure 4C), respectively, using a 2.9 ꢀ 10-2 M Et3OBF4
solution. Q2 and Q3 species present in the parent material
are almost completely removed after fluorination at RT and
50 ꢀC. The fluorinated products exhibit a dominant Q4 peak
that shifts upfield to -114 and -115 ppm for the samples
treated at RT and 50 ꢀC, respectively. CSBA-15 samples
(Figures 4 D-F) not only showed reduction in the percentage
of Q2 and Q3 silicates upon fluorination, but also concomi-
tant reduction of D1 trimethylsilane (TMS) capping species.
The 29Si HP-MAS NMR spectrum of CSBA-15 (Figure 4D)
exhibited a (Q2þQ3)/Q4 ratio of 0.61 and 12% TMS silicate
species. Upon fluorination, the (Q2þQ3)/Q4 ratio decreases
as before, to 0.39 and 0.38 at RT (Figure 4E) and 50 ꢀC
(Figure 4F), respectively. A reduction of capped species is
also observed in these samples leaving only 2.2 and 0.8%
TMS silica sites upon fluorination at RT and 50 ꢀC,
respectively, with subsequent decrease in Q2 and Q3 indicat-
ing fluorination without extensive etching, but rather a more
rigid mesostructure, is evidenced. The Q4 species exhibit the
same upfield shift as seen with the extracted SBA-15 samples.
Additionally, as shown by BET (see Figure 2-III and the
Supporting Information, Figure S2-III), pore reconstruction
is shown to occur leading to a more rigid material with a
broader PSD than its parent material. 29Si HP-MAS NMR
spectra also showed the formation of Q4 species with upfield
shifts from -111 ppm in calcined MCM-41 sample before
fluorination to -113 and -114 ppm for its fluorinated
samples at RT and 50 ꢀC, respectively. Additionally, the
Q3 region is reduced and slightly shifted upfield from -103
to -106 ppm. The calcined material exhibited a (Q2þQ3)/Q4
ratio of 1.07 (Figure 4G), which was reduced to 0.71 and 0.61
after fluorination at RT (Figure 5H) and 50 ꢀC (Figure 4I),
respectively.

Article
Chem. Mater., Vol. 22, No. 17, 2010 4959
Table 2. Elemental Analysis Results of the Different Mesoporous Samples
Studied before and after Fluorination
band at 930-940 cm-1 (Si-OH stretching), and a mod-
erate to weak band at 790-800 cm-1 (Si-O-Si bending).
Upon fluorination, the formation of SiFx (x = 1-4) is
expected to yield absorptions from 830 to 1010 cm-1 and
450 cm-1 corresponding to stretching, bending, and
rocking vibrational modes.2a-c,e Because of the broad
nature of the Si-O-Si stretching mode, however, eluci-
dation of many of these absorptions are clearly difficult
due to overlapping. A general broadening of this region is
noticed, however, in many of the spectra. Absorptions at
1630 and 1700 cm-1 are observed because of the δOH
deformations of physisorbed water. The formation of
new ethoxyxilane species on the surface is evidenced by
new absorptions at approximately 1370 and 1420 cm-1
corresponding to CH3 δ symmetric, CH3 δ asymmetric,
and CH2 δ asymmetric stretching modes as well as the
formation of new absorptions at 2980 cm-1 correspond-
ing to sp3-hydridized C-H stretching modes.2f The for-
mation of a new peak in the 730-750 cm-1 region, which
was previously assigned to [F6-nSi(OH)n]2- species,10
O4/2SiF species, or Si-O-Si stretches altered or polarized
by the presence of nearby Si-F species,14 is also shown to
appear in many of the fluorinated SBA-15 samples. As
shown by the obtained FTIR spectra, upon fluorination,
every fluorinated product exhibits a new band in this
region that was not present in the parent materials,
CSBA-15, SBA-15, and MCM-41, indicative of new
fluorosilicate species within the mesoporous framework.
To further probe the surface properties of these materials,
we utilized X-ray photoelectron spectroscopy (XPS) and
obtained XPS spectra of SBA-15, CSBA-15, and MCM-41,
as well as representative fluorinated species (see the Support-
ing Information, Figure S5). From the spectra, the composi-
tion and degree of fluorination and/or fluorosilicate species
formed upon fluorination can be determined by deconvolu-
tion of the F1s spectra using a Gaussian distribution func-
tion. Upon fluorination, all materials exhibit Si 2p (105-
1.07 eV), O 1s (533.3-536.5 eV) and F 1s (683-693 eV)
signals consistent with previously reported literature.7,15
Further careful analysis reveals that two distinct F species
are observed at 690.5 and 687.6 eV corresponding to silicon
oxyfluoride [(SiO)3-xSiFx] and silicon fluoride (i.e., SiF4)
species, respectively, where the oxyfluoride species is domi-
nant (60-80%). In the case of the CSBA-15 materials (see
the Supporting Information, Figure S5D-F), the removal
of D1 species is completed with the same efficiency regardless
of reaction temperature. The parent CSBA-15 material is
shown to contain 11.2 at % C, which is reduced to 4.2-4.3 at
% C upon fluorination in agreement with observations
made from the results obtained via solid-state 29Si HP-
MAS NMR, nitrogen gas sorption studies (BET and BJH
pore size distributions), and elemental analysis.
sample
wt % Fa
wt % Cb
wt % Hb
SBA-15 Samples
0.1
SBA-15
SBA-L-50
SBA-M-RT
SBA-M-50
SBA-H-RT
SBA-H-50
5.98
7.02
5.00
5.23
5.44
MCM-41 Samples
0.1
MCM-41
MCM-L-RT
MCM-M-50
MCM-M-RT
MCM-H-50
MCM-H-RT
2.56
2.41
2.79
3.16
3.61
CSBA-15 Samples
0.1
4.08
4.15
4.68
CSBA-15
9.25
3.29
3.55
2.78
2.08
1.02
1.35
0.93
CSBA-M-RT
CSBA-M-50
CSBA-H-RT
a Determined via ion chromatography. b Determined via combustion
using a Perkin-Elmer 2400 Elemental Analyzer.
its reduced surface silanol concentration and lower con-
tent of reactive Q2 ((SiO)2Si(OH)2) or Q3 ((SiO)3SiOH)
species resulted because of its calcination. The F loading
of the SBA-15 materials compares well with the largest
possible amounts of surface organic groups that can be
placed on the surface of mesoporous materials, which is
often around 4 mmol/g.40 Typical surface organic density
values that can be obtained after postgrafting mesoporous
materials with organosilanes is actually ∼2 mmol/g,40 which
is well-below the surface fluorine density in our case. Desi-
lylation of TMS capping groups of CSBA-15 upon fluorina-
tion clearly evidenced by the marked drop in wt % C and wt
% H from 9.25 and 2.08 wt %, respectively, to 2.8-3.6 wt %
C and 0.9-1.4 wt % H. The loss of capping agent, however,
does not result in production of Q2 and Q3 sites, as shown by
the 29Si HP-MAS NMR spectra indicating that the fluorina-
tion reaction takes place predominantly on the amount of
reactive surface sites available. As shown in Table 2, the wt%
F of the OMFs is not directly related to changes in neither
reaction temperature nor the [BF4-] because multiple reac-
tion variations yield similar results with poor correlation.
However, the three types of materials do exhibit similar
surface fluorination and fluorinated species. These findings,
as well, are indications that the concentration of reactive
surface sites is the dominating factor.
To investigate the composition of the materials after
fluorination and etching, FTIR spectra of the samples
were also obtained (see the Supporting Information,
Figure S4). As reported extensively, the mesoporous silica
materials exhibit a strong, broad absorptions in the
3200-3700 cm-1 (Si-OH stretching) and 1000-1200
cm-1 (Si-O-Si asymmetric stretching) regions, a weak
The etching/fluorination process was also found to form
corrugated, or etched, fluorosilica nanospheres (FSNs)
upon using silica nanospheres (NS) of 250 ( 10 nm in
diameter. These microspheres were characterized before
and after treatment with Et3OBF4 (Figure 5) by TEM. The
TEM images of the silica nanospheres after treatment with
a 2.9 ꢀ 10-2 M Et3OBF4 at RT (Figure 5B) and 50 ꢀC
(40) (a) Sharma, K. K.; Asefa, T. Langmuir 2008, 24, 14306–14320.
(b) Sharma, K. K.; Anan, A.; Buckley, R. P.; Ouellette, W.; Asefa,
T. J. Am. Chem. Soc. 2008, 130, 218–228. (c) Sharma, K. K.; Asefa,
T. Angew. Chem., Int. Ed. 2007, 46, 2879–2882.

4960 Chem. Mater., Vol. 22, No. 17, 2010
Duncan et al.
Figure 7. Nitrogen gas adsorption isothermsof(A) SBA-15, (B) SBA-15-
SH, (C) SBA-15-SO3H, and (D) OMF-SO3H.
Figure 6. XRD patterns of (A) SBA-15, (B) OMF, (C) SBA-15-SH,
(D) SBA-15-SO3H, and (E) OMF-SO3H.
Table 4. Chemical Compositions of Solid Acid Catalysts
C (wt %)a
H (wt %)a
S (wt %)b
F (wt %)b
Table 3. Structural Data of Mesoporous Epoxidation Catalysts
sample
average
pore
BJH average
BET surface pore volume
SBA-15-SH
SBA-15-SO3H
OMF-SO3H
3.15
2.31
2.54
1.45
1.48
0.80
0.68
0.29
0.33
unit
sample
cell (A)a diameter (A)b area (m2/g)
(cm3/g) b
2.90
SBA-15 (A)
OMF (B)
SBA-15-SH (C)
SBA-15-SO3H (D)
OMF-SO3H (E)
108
104
106
108
113
73
860
0.84
a Determined via combustion using a Perkin-Elmer 2400 elemental
analyzer. b Determined via ion chromatography.
70
73
73
689
702
663
0.72
0.75
0.79
materials, the calcined SBA-15 remains highly ordered
after MPTS grafting, oxidation with H2O2/H2SO4, and
subsequent fluorination. All samples exhibit a clearly
defined (100) planes as well as higher order (110) and
(200) planes. In fact, the OMF-SO3H catalyst shows more
clearly defined (110) and (200) planes than those of the
calcined SBA-15 and -SO3H-functionalized parent ma-
terials, indicating that the fluorination process leads to
pore-reconstruction, yielding a more highly ordered
structure in agreement with the other fluorinated materi-
als presented in this paper.
a ao = 2d100/31/2 (A) for 2D hexagonally ordered materials. b Ob-
tained from adsorption branch of the N2(g) sorption isotherms.
ꢀ
(Figure 5C) indicated some etching of the surface of the
nanoparticles, which is dependent on the etching conditions.
As can be seen, no corrugation is shown to result when
treatment is performed at RT and the average diameter
shows only a slight decrease to 244 ( 13 nm. When the
nanospheres are treated at elevated temperature, however,
etching of the silica surface is clearly evident and the
diameters of these nanospheres are decreased by approxi-
mately 13% to 218 ( 8 nm. The latter materials have also
exhibited increased wt% F up to ∼2 wt %, proving applica-
tion of the [BF4-]-based fluorination/etching to nanostruc-
tured materials other than mesoporous silicates.
3.3. Catalytic Studies. To test if the fluorination pro-
cess enhanced the catalytic activity of the Brønsted acid
catalyzed ring opening reaction of styrene oxide by ani-
line forming the corresponding β-amino alcohol, a cal-
cined SBA-15 (weakly acidic) and a sulfonic acid func-
tionalized SBA-15 were prepared. Both materials were
subsequently fluorinated at RT using dilute Et3OBF4
solution (5.7 ꢀ 10-2 M in CH2Cl2) forming the corre-
sponding fluorinated solid acids, OMF and OMF-SO3H,
respectively. The materials were characterized by small-
angle XRD, nitrogen gas adsorption, TGA, FTIR, XPS,
and elemental analyses.
Nitrogen gas adsorption studies were performed to
study how the fluorination process altered the surface
area, pore diameter and volume, and sorption isotherm of
the fluorinated mesoporous catalysts (Figure 7 and Ta-
ble 3 and the Supporting Information, Figure S6). The
parent calcined SBA-15 material used for making the
catalysts has a BET surface area of 860 m2/g and an
ꢀ
average pore diameter of 73 A. The adsorption isotherm
shows a sharp capillary condensation step and a very
narrow PSD. The SBA-15-SO3H catalyst shows a slightly
lower surface area of 702 m2/g. Upon fluorination in
5.7 ꢀ 10-2 M EtO3BF4 at RT, the surface area decreases
further to 663 m2/g, whereas the average pore diameter is
slightly broadened but relatively unaltered. As shown in
Figure S6 in the Supporting Information, the PSD broad-
ens slightly toward larger diameters. Unlike the extracted
SBA-15 samples, where tail formation in the nitrogen
adsorption isotherms due to partial pore clogging or
bottlenecking is common, fluorination does not lower
the integrity of the sulfonic-acid-functionalized calcined
SBA-15. The OMF-SO3H catalyst exhibits an excellent
type IV isotherm, void of any substantial pore constric-
tions. Combined with the small angle XRD data, it is
Figure 6 displays the XRD patterns for the -SO3H-
functionalized SBA-15, OMF, and OMF-SO3H materials
and their parent samples (calcined and MPTS functiona-
lized SBA-15) used for the synthesis of β-amino alcohols.
The structural data is shown in Table 3. As seen with the
previously discussed fluorinated SBA-15 and MCM-41

Article
Chem. Mater., Vol. 22, No. 17, 2010 4961
Scheme 2. Epoxidation of Styrene Oxide with Aniline Forming β-Amino Alcohol Regioisomers, A and B
evident that fluorination using this methodology is accom-
plished predominantly on the surface of the silica while
maintaining the overall order and integrity of the catalyst.
The thermogravimetric analysis of the OMF-SO3H
catalyst (see Figure S7 in the Supporting Information)
showed significantly more weight loss in the 100 - 600 ꢀC
region compared to its nonfunctionalized counterpart
(OMF). However, when taking into consideration the
propylsulphonic acid groups in the former, the trend in
the weight loss due to fluorine species is quite similar. The
weight percent losses in the temperature range 100-
650 ꢀC for the SBA-15-SH and SBA-15-SO3H samples are
comparable. However, the OMF-SO3H sample showed
slightly higher weight loss in the same temperature range.
Elemental analyses (Table 4) were used to determine the
concentration of active sulfonic acid catalytic sites to be
0.09 and 0.10 mmol/g for the SBA-15-SO3H and OMF-
SO3H respectively. The F content of the OMF-SO3H was
found to be 1.5 mmol/g. Upon thiol oxidation there is a
significant loss of weight percent sulfur, possibly because
of cleavage during oxidation and subsequent washings.
Upon fluorination, however, sulfur and the carbon con-
tents are not further reduced, implying that this fluorina-
tion methodology does not cause cleavage of immobilized
functional groups as seen with surface silanols or capping
agents. The presence of fluorine in the OMF-SO3H sample
was further evidenced by FTIR spectroscopy (see the
Supporting Information, Figure S8). As can be seen, the
fluorinated solid acid catalyst exhibits a peak at 730-750
cm-1, consistent with the previously discussed fluorinated
silicas. Because of the low loading content of MPTS, the
characteristic -SH stretching band (2575 cm-1), is not seen
in the SBA-15-SH material. This signal is usually very
weak and barely visible even when the sample exhibits a
larger S wt%.41 XPS (see the Supporting Information,
Figure S9) shows that the S 2p of SBA-15 has a binding
energy of -163.7 eV. Upon oxidation (SBA-15-SO3H),
this values shifts to -168.5 eV in a manner consistent with
previously published data.42 The fluorinated product,
OMF-SO3H, has a S 2p binding energy which is shifted
slightly further to -168.0 eV, which proves that the
fluorination processdoesnotalter thesulfonicacidgroups.
The F 1s peak of OMF-SO3H has a binding energy of
-687.1 eV, which is consistent with that of the fluorinated
MCM-41 and SBA-15 samples reported in this work.
The effect of fluorination and etching on the catalytic
properties of the materials is investigated with a ring-opening
reaction of styrene oxide with aniline forming the corre-
sponding β-amino alcohol (Scheme 2). Control reactions
using SBA-15 and thiol (-SH)-modified SBA-15 resulted in
percent conversions of 40 and 46%, respectively, whereas
SBA-15-SO3H exhibited 77% conversion at RT (Table 5).
Upon fluorination, the resulting fluorinated material (OMF)
gave a 79% conversion, which was substantially better than
its nonfluorinated counterpart (SBA-15), indicating a higher
catalytic activity by OMF compared to its nonfluorinated
counterpart (SBA-15). On the other hand, OMF-SO3H
yielded an 87% conversion, a 10% increase from its parent
material (SBA-15-SO3H) (Table 5). As shown, the OMF and
OMF-SO3H catalysts performed competitively with other
heterogeneous catalysts while, in our case, using a lower
catalyst:reagent ratio of 13 mg:5 mmol (compared to as high
as 120 mg:1 mmol in other cases, Table 5) and/or a simplified
catalyst synthetic protocol compared to that of materials
requiring the incorporation of more expensive metal oxides,
metal-chelate complexes, or syntheses requiring metal
incorporation in the mesostructure (see Table 5). When
the reaction temperature is increased, all three catalysts
perform similarly in terms of percent conversion. How-
ever, as shown in Table 6, the regioselectivity of the SBA-
15-SO3H is reduced, whereas the fluorinated samples
suffer only minor, if any, reduction. Retention of regio-
chemistry at elevated temperatures may be due to the
removal of weakly acidic geminal silanols, leading to a
higher overall surface acidity. The fluorination process
may form site-isolated catalytic sites by increasing the
distance between reactive sites, inhibiting silanol-silanol
or silanol-SO3H hydrogen-bonding interactions and
reducing the chances of competition for available reagent.
Recyclability studies performed at RT show that reduc-
tions in catalytic efficiencies are observed for all three
catalysts (Table 7). Loss in catalytic activity may be due
to the loss of catalyst between trials, active site leaching,
or degradation of the active sites leading to a loss of
(41) Saikia, L.; Satyarthi, J. K.; Srinivas, D.; Ratnasamy, P. J. Catal.
2007, 252, 148–160.
(42) (a) Shen, J. G. C.; Herman, R. G.; Klier, K. J. Phys. Chem. B 2002,
106, 9976–9978. (b) Shylesh, S.; Sharma, S.; Mirajkar, S. P.; Singh,
A. P. J. Mol. Catal., A: Chem. 2004, 212, 219–228. (c) Shylesh, S.;
Samuel, P. P.; Srilakshmi, C.; Parischa, R.; Singh, A. P. J. Mol.
Catal., A: Chem. 2007, 274, 153–158.

4962 Chem. Mater., Vol. 22, No. 17, 2010
Table 5. Comparative Data of Catalytic Efficiency of As-Synthesized Solid Acids and Their Parent Materials
Duncan et al.
epoxide
conversion (%)a
selectivity
(A:B)
amine:epoxide
ratio (mmol)
catalyst
mass (mg)
catalyst
OMF-SO3H
OMF
-SO3H
-SH
SBA-15
mesoporous aluminosilicate
AlKIT-5
SBA-15/Co complex
Fe-MCM-41
H14[NaP5W29MoO110
time (h)
ref
87
79
77
46
10: 90
10:90
11:89
13: 87
11: 89
5: 95
NRc
4
4
4
4
4
5:5
5:5
5:5
5:5
5:5
1:1
1:1
100:100
1:1
5:5
10:10
2:2
2:2
13
13
13
13
13
120
50
200
NRd
10
50
100
60
this work
this work
this work
this work
this work
30c
30b
31a
31b
32b
33b, 42c
34
35
43a
43b
40
70b
87b
77
6
0.5
12
4
0.2
4
2
0.75
4
11:89
NRc
91b
92b
81
]
NRc
Ti-MCM-41
amberlyst-15
6: 94
NRc
1: 99
2: 98
NRc
92b
97b
95
SO42-/ZrO2
Fe-Zn double metal cyanide
phosphomolybdic acid-Al2O3
10:10
2:2
50
200
93b
0.5
a Determined via GC analysis. b Reported as yield (%). c Not reported. d Authors reported the use of 0.05 mmol of Fe-MCM-41 but did not report the
corresponding mass.
Table 6. Influence of Temperature on the Ring Opening of Styrene Oxide
with Aniline
ordered mesoporous fluorosilicas (OMFs) using dilute
concentrations of Et3OBF4, a common alkylating reagent,
temp
(ꢀC)
time
(h)
epoxide
conversion (%)
selectivity
(A:B)
under ambient conditions in a nonaqueous environment.
This procedure generally increases the surface area and pore
volume of the mesoporous materials. This fluorination and
etching method is shown to create a slightly broader pore size
distribution and, in some cases, partial pore clogging or
bottlenecking. In spite of minor structural changes, however,
the materials still retain a highly ordered mesoporous struc-
ture as shown by BET, XRD and TEM. It is quite common
for functionalized mesoporous silica (FMS) materials
formed by postgrafting and postchemical modification tech-
niques to suffer from reductions in surface area, pore dia-
meter and volume, and/or overall porosity. The appearance
silicon oxyfluoride (major) and silicon fluoride (minor) spe-
cies is evidenced by the formation of new peak at 730-
750 cm-1 in the FTIR spectra. The formation of these species
is further identified by characteristic Si-F binding energies in
the XPS spectra from 687 to 690 eV. 2.4-7.0 wt % F can be
attained with minor variation based upon [BF4-] while the
type of material used seemed to be a deciding factor. Calcined
MCM-41 samples, in spite of higher reactive surface areas,
were found to have the lowest wt % F, containing 2.4-3.6
wt %, followed by CSBA-15 samples (4.1-4.7 wt %), and
finally the SBA-15 samples, which contained 5.0-7.0 wt %
F. Because of the low atomic weight of fluorine and evidence
that fluorination predominantly takes place on the surface of
the silica, the density (wt %) fluorine obtained is actually very
high. Overall, samples containing 5-7 wt % are equivalent
to a silica containing 2.6-3.7 mmol/g F, which is mainly
present on the silica surface as opposed to doped within the
walls. This compares well with the largest possible amounts
of organic groups that can be placed on the surface of
mesoporous materials, which is often around 4 mmol/g,
whereas typical surface organic density values obtained after
postgrafting procedures result with organic functionality of
∼2 mmol/g,40 which is well-below the surface fluorine density
in our case. The as-described synthetic procedure was also
catalyst
OMF-SO3H
22
40
4
1
3
1
3
4
1
3
1
3
4
1
3
1
3
87
91
94
93
97
79
84
95
94
97
77
88
95
96
98
10:90
9:91
9:91
60
9:91
12:88
10:90
9:91
9:91
9:91
11:89
11:89
8:92
9:91
16:84
16:84
OMF
22
40
60
SBA-15-
SO3H
22
40
60
Table 7. Catalyst Recyclability Studies of Synthesized Solid Acids
catalyst
epoxide conversion (%)
selectivity (A:B)
OMF-SO3H
1st run
2nd run
3rd run
4th run
88
58
46
51
10: 90
8: 92
9: 91
9: 91
OMF
1st run
2nd run
3rd run
4th run
81
66
50
34
10: 90
8: 92
9:91
8: 92
-SO3H
1st run
2nd run
3rd run
4th run
65
52
41
37
9: 91
9: 91
10: 90
9: 91
surface acidity. In addition to be useful as a synthetic tool to
create nanomaterials with enlarged pores and fluorinated
surface, this method is clearly also proven to be useful to
make nanocatalysts with improved catalytic properties.
4. Conclusions
(43) (a) Saikia, L.; Satyarthi, J. K.; Gonnade, R.; Srinivas, D.;Ratnasamy,
P. Catal. Lett. 2008, 123, 24–31. (b) Kumar, S. H.; Leelavathi, P. J.
Mol. Catal., A: Chem. 2007, 266, 65–68.
We have developed a facile synthetic etching and
fluorination protocol toward the production of highly

Article
Chem. Mater., Vol. 22, No. 17, 2010 4963
applied to sulfonic acid (-SO3H) functionalized SBA-15.
Enhancement of the acid-catalyzed ring opening of styrene
oxide with aniline to produce the corresponding β-amino
alcohol was shown. Epoxidation at room temperature after
4 h using -SO3H-functionalized SBA-15 resulted in 77%
conversion, whereas the as-synthesized OMF and OMF-
SO3H materials exhibit 79 and 87% conversions, respec-
tively, because of their enhanced solid acid catalytic activity.
XPS and elemental analyses serve as proof that this nonaqu-
eous fluorination methodology does not alter the structure of
the -SO3H groups. BET and BJH studies and small-angle
XRD confirm that the fluorinated catalyst exhibits larger
pore diameter and volume than its precursor as well as
increased order. The reported synthetic approach is proven
to be versatile as this fluorination and etching method also
fluorinates and etches silica nanospheres (FNSs). These
findings may give new insights toward the large scale synthesis
of mesoporous fluoro- or organofluoro-silica nanomaterials
for industrial use in various applications such as silica-
based microelectronic devices, nanocatalysts, or materials
for chemical separations.
Acknowledgment. We gratefully acknowledge the financial
assistance by the U.S. National Science Foundation (NSF),
CAREER Grant CHE-0645348 and NSF DMR-0804846,
for this work. We thank Dr. Jonathan Shu at Cornell
University for his help with XPS measurements and Dave
Kiemle at SUNY-ESF for his assistance with solid-state
NMR measurements. We also thank Professor Robert A.
Bartynski in Physics Department at Rutgers University for
valuable discussion regarding XPS data and for offering us
an XPS instrument to use.
Supporting Information Available: Transmission electron
micrographs, thermogravimetric traces, XPS spectra, and FTIR
spectra. This material is available free of charge via the Internet
at http://pubs.acs.org.