Diphenylphosphino functionalization of mesoporous silica using tripodal linker units

Article abstract of DOI:10.1016/j.jorganchem.2010.12.030

The immobilization of diphenyl phosphine onto ordered mesoporous silicas using a tripodal linker unit possessing one bromopropyl group and three anchoring silicon atoms was investigated. Solid-state ³¹P, ²⁹Si, and ¹³C cros

Full text of DOI:10.1016/j.jorganchem.2010.12.030

Journal of Organometallic Chemistry 696 (2011) 1565e1569
Contents lists available at ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
Diphenylphosphino functionalization of mesoporous silica using tripodal
linker units
,
b
b
a
Takayuki Miyaji a, Shun-ya Onozawa b , Norihisa Fukaya , Masae Ueda , Yukio Takagi ,
*
Toshiyasu Sakakura b, Hiroyuki Yasuda b
,
*
a N.E. CHEMCAT Corporation, 678 Ipponmatsu, Numazu, Shizuoka 410-0314, Japan
b National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba Central 5, 1-1-1 Higashi, Tsukuba, Ibaraki 305-8565, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
The immobilization of diphenyl phosphine onto ordered mesoporous silicas using a tripodal linker unit
possessing one bromopropyl group and three anchoring silicon atoms was investigated. Solid-state 31P,
29Si, and 13C cross-polarization/magic angle spinning (CP/MAS) NMR spectroscopic studies as well as
isothermal nitrogen adsorption/desorption measurements revealed that grafting the tripodal linker unit
and a subsequent reaction with potassium diphenylphosphide (the “bottom-up” method) successfully
realized diphenylphosphino functionalization of silica while maintaining the mesoporous structure. In
contrast, directly grafting tripodal diphenylphosphino ligands pre-synthesized from the tripodal linker
unit onto silica (the “top-down” method) was unsuccessful.
Received 28 October 2010
Received in revised form
17 December 2010
Accepted 24 December 2010
Keywords:
Immobilized catalysts
Mesoporous silica
Tripodal linker unit
Aryl phosphine
Ó 2011 Elsevier B.V. All rights reserved.
CP/MAS NMR spectroscopy
1. Introduction
have been reported [9e13]. Therefore, it is interesting to investigate
the immobilization of aryl phosphines onto silica via the tripodal
We have recently developed a tripodal linker unit possessing
three leaving allylsilyl or isopropoxysilyl groups (1a and 1b in Fig.1)
[1]. This linker unit can be tightly bound to a silica surface via three
independent SieOeSi bonds. In fact, organo-modified silica with
a grafted tripodal linker unit exhibits a higher hydrothermal
stability against leaching of organic moieties than silica modified
with conventional trialkoxysilanes. Additionally, the tripodal linker
unit contains a bromopropyl moiety to which various functional
groups can be attached. Therefore, diverse organic functional
molecules, including auxiliary ligands for metal complexes, can be
firmly immobilized onto a silica surface employing the linker unit.
Immobilization of molecular catalysts like transition metal
complexes onto inorganic porous supports such as ordered meso-
porous silicas has been actively researched because immobilized
catalysts are advantageous in terms of separating and recycling the
catalyst [2e8]. Aryl phosphines are the most common ligands in
transition metal complex catalyzed reactions, and numerous
examples of immobilized metal aryl phosphine complex catalysts
linker unit. There are two preparative approaches to functionalize
silica by diarylphosphino groups via the linker: the “bottom-up”
(route 1 in Fig. 1) and “top-down” (route 2) methods [14]. In the
former, the tripodal linker unit is grafted onto silica and the bromo
group is subsequently converted into diaryl phosphine. In the latter,
a phosphorous compound containing a diarylphosphinopropyl
moiety and three allylsilyl or isopropoxysilyl groups (tripodal dia-
rylphosphino ligand) is pre-synthesized from the tripodal linker
unit, and then the compound is directly grafted onto silica. Herein
we examine these two methods to determine an effective strategy
to tightly bind aryl phosphine ligands to silica using the tripodal
linker unit.
2. Results and discussion
2.1. Diphenylphosphino functionalization via route 1 (the “bottom-
up” method)
Grafting of 3-bromopropyltris[3-(allyldimethylsilyl)propyl]silane
(1a) onto an ordered mesoporous silica TMPS-4 (denoted as SIO-1) in
heptane at reflux afforded bromopropyl-functionalized silica 2
(hereafter, 2 for SIO-1 is denoted as 2-1) [15]. Likewise, grafting of 1a
onto another ordered mesoporous silica TMPS-7 (denoted as SIO-2)
* Corresponding authors. Tel.: þ81 29 861 9399; fax: þ81 29 861 4580.
E-mail addresses: s-onozawa@aist.go.jp (S.-y. Onozawa), h.yasuda@aist.go.jp
(H. Yasuda).
0022-328X/$ e see front matter Ó 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2010.12.030

1566
T. Miyaji et al. / Journal of Organometallic Chemistry 696 (2011) 1565e1569
Fig. 1. Synthetic routes of diphenylphosphino-functionalized mesoporous silica using tripodal linker units.
Table 1
(corresponding to 0.013 mmol/g), indicating that the Br group was
almost completely converted into PPh2 groups. Fig. 2 shows the 31P,
29Si, and 13C CP/MAS spectra of 4(2)-1 obtained under the condition
in entry 4 of Table 1. In the 31P CP/MAS spectrum, a major peak
Summary of diphenylphosphino functionalization of mesoporous silica via route 1.
Entry Modified Silica Conditions
Loading
SBET
Vp
DBJH
silica
(mmol/g)a (m2/g)b (cm3/g)c (nm)d
KPPh2
Temp.
assigned to PPh2 species was observed at
a minor peak due to the oxidized PPh2 species was also detected at
40 ppm. In addition to the silica peaks in the 29Si CP/MAS spec-
trum, two signals corresponding to one tetraalkyl-coordinated
d
ꢀ15 ppm, although
1
2
3
4
5
6
2-1
SIO-1
e
e
0.48
813
570
675
734
876
778
0.97
0.58
0.77
0.85
1.68
1.56
3.1
2.7
2.8
2.8
5.8
5.8
4(2)-1
4(2)-1
4(2)-1
2-2
SIO-1 7 equiv. Reflux 0.32
SIO-1 7 equiv. RT 0.34
SIO-1 5 equiv. 323 K 0.35
SIO-2 0.39
SIO-2 5 equiv. 323 K 0.31
d
e
e
silicon (
d
2 ppm) and three trialkylmonooxygen-coordinated silicon
4(2)-2
atoms (
d
14 ppm) [16] in 4(2) were observed, suggesting that the
a
b
c
Determined by elemental analysis of carbon.
BET specific surface area.
Pore volume determined at a relative pressure (P/P0) of 0.99.
Average pore diameter calculated by the BJH method from the adsorption
tripodal anchoring structure of 2 was maintained even after
phosphine was attached. The 13C CP/MAS spectrum exhibited
several peaks, which corresponded to the carbon skeleton of 4(2).
The peaks assigned to the carbons of the benzene ring in the PPh2
d
branch.
group appeared near
into a PPh2 group caused the peak attributed to the carbon at the
-position to shift from 30 ppm (2-1) to approximately 20 ppm
d 130 ppm. The transformation of the Br group
gave 2 (hereafter, 2 for SIO-2 is denoted as 2-2). Similar to 2-1 [1], the
29Si and 13C CP/MAS spectra of 2-2 confirmed successful grafting via
three covalent SieOeSi bonds (see Supplementary data, Figs. S1 and
S2). The amounts of organic moieties in 2-1 (prepared in this study)
and 2-2, which were determined by elemental analysis of carbon,
were 0.48 and 0.39mmol/g(0.61 and 0.48 mmol/g-SiO2), respectively
(entries 1 and 5 of Table 1).
Next we examined the transformation of the Br group in 2 into
a PPh2 group. Reactions of 2-1 with varying amounts of KPPh2
(5ꢀ7 equiv.) were carried out at several temperatures. Entries 2e4
of Table 1 list the amounts of organic groups loaded in the resulting
PPh2-functionalized SIO-1 (4(2)-1). Treatment of 2-1 with KPPh2
reduced the loading amounts, presumably because the treatment
partially eliminated the grafted tripodal linker unit. The bromine
contents of all the 4(2)-1 samples were less than 0.1 wt%
b
d
d
(4(2)-1). The CP/MAS spectra of 4(2)-1 (entries 2 and 3 of Table 1) as
well as those of PPh2-functionalized SIO-2 (4(2)-2) (entry 6) were
essentially the same as Fig. 2 (see Supplementary data, Figs. S3eS5).
Hence, the treatment of 2 with KPPh2 efficiently introduced a PPh2
group into 2 while maintaining the tripodal anchoring structure.
Fig. 3 displays the N2 adsorption/desorption isotherms of 2 and
4(2). Table 1 includes the textural properties. All the samples
provided type IV isotherms, which are typical for mesoporous
structures. However, 4(2)-1, which was prepared while refluxing
(entry 2 of Table 1), showed a smaller step in the relative pressure
(P/P0) range of 0.3e0.5, suggesting the ordered mesopores were
partially destroyed. This is also supported by the significant
decrease in the surface area and the pore volume. On the other
hand, the isotherms of 4(2)-1 (entries 3 and 4 of Table 1) as well as
Fig. 2. (A) 31P, (B) 29Si, and (C) 13C CP/MAS spectra of PPh2-functionalized SIO-1 (4(2)-1) (entry 4 of Table 1). (*): spinning side band.

T. Miyaji et al. / Journal of Organometallic Chemistry 696 (2011) 1565e1569
1567
Fig. 3. Nitrogen adsorption/desorption isotherms of functionalized SIO-1 (left panel) and SIO-2 (right panel). (a) Bromopropyl-functionalized SIO-1 (2-1) (entry 1 of Table 1); (b)
PPh2-functionalized SIO-1 (4(2)-1) (entry 2); (c) PPh2-functionalized SIO-1 (4(2)-1) (entry 3); (d) PPh2-functionalized SIO-1 (4(2)-1) (entry 4); (e) bromopropyl-functionalized SIO-2
(2-2) (entry 5); (f) PPh2-functionalized SIO-2 (4(2)-2) (entry 6). Open and closed symbols correspond to the adsorption and desorption branches, respectively. (a), (b), (c), and (e) are
offset by 600, 400, 200, and 200 cm3/g (STP), respectively.
4(2)-2 (entry 6) demonstrate that the PPh2 group was introduced
while preserving the mesoporous structure. Thus, the diphenyl-
phosphino functionalization via the tripodal linker unit was
accomplished for both SIO-1 and SIO-2 using route 1 (the “bottom-
up” method).
13C CP/MAS spectrum. These observations suggest that the allyl
groups in 3a did not fully react with the surface silanol group. This
is probably because the interaction of the phosphorous group in 3a
with the silica surface was preferential over the reaction between
the allylsilyl groups and the silanol groups.
Next, we grafted the tripodal diphenylphosphino ligand con-
taining three isopropoxysilyl moieties 3b onto SIO-1 and SIO-2 in
heptane. Entries 8 and 9 of Table 2 show the results of the obtained
PPh2-functionalized silica 4(3b) (hereafter, 4(3b) for SIO-1 and SIO-
2 are denoted as 4(3b)-1 and 4(3b)-2, respectively). Several times
the amount was loaded with 3b than with 3a. Fig. 4(b) shows the
CP/MAS spectra of 4(3b)-2. Besides the major peak from PPh2
2.2. Diphenylphosphino functionalization via route 2 (the “top-
down” method)
We initially grafted the tripodal diphenylphosphino ligand
containing three allylsilyl moieties 3a onto SIO-1 in heptane while
refluxing. Entry 7 of Table 2 shows the loading amount and textural
property of the resulting PPh2-functionalized silica (4(3a)-1).
Compared to 2-1, less 3a was loaded onto SIO-1. Fig. 4(a) shows the
31P, 29Si, and 13C CP/MAS spectra of 4(3a)-1. In addition to the PPh2
(
d
ꢀ15 ppm) and the minor peak from OPPh2
around 25 ppm (possibly due to PeOeSi) was distinctly observed
in the 31P CP/MAS spectrum. Moreover, a new silicon peak arising
from 3b emerged near
20 ppm in the 29Si CP/MAS spectrum,
(d 40 ppm), a peak
d
d
(
d
ꢀ15 ppm) and OPPh2
appeared near
25 ppm in the 31P CP/MAS spectrum. The peak near
25 ppm might be due to the PeOeSi species formed by the
(d 40 ppm) peaks, a shoulder peak
suggesting the existence of silicon species that did not participate
in binding to silica. In the 13C CP/MAS spectrum, the peaks assigned
to an isopropoxy group were detected. It is assumed that the iso-
propoxy peaks resulted from the unreacted isopropoxy group in 3b
rather than from the isopropoxide species (iPrOeSi) formed by the
reaction of the eliminated isopropyl alcohol and the surface silanol
group because such iPrOeSi species were not observed when the
tripodal linker unit containing leaving isopropoxy groups 1b was
grafted onto SIO-1 [1]. 4(3b)-1 gave the similar 31P, 29Si, and 13C
CP/MAS spectra as Fig. 4(b). These CP/MAS NMR results demon-
strate that 3b was not successfully grafted onto the silica surface via
three covalent SieOeSi bonds.
d
d
interaction between the phosphorous group and the surface silanol
group, similar to the findings of Blümel et al. [17,18]. In the 29Si
CP/MAS spectrum, the peak assigned to the tetraalkyl-coordinated
silicon (d 2 ppm) was detected, whereas the peak due to three
trialkylmonooxygen-coordinated silicon atoms was negligible. The
peaks assigned to the allyl group in 3a were clearly observed in the
Table 2
Summary of diphenylphosphino functionalization of mesoporous silica via route 2.
Furthermore, we attempted to graft 3b onto SIO-2 in toluene
and isopropyl alcohol (entries 10 and 11 of Table 2). Compared to
grafting in heptane, less grafting occurred in toluene. In addition,
there was little difference in the CP/MAS spectra between both
solvents. The 13C CP/MAS spectrum revealed that the surface of SIO-
2 was isopropoxylated when isopropyl alcohol was used for the
grafting (spectrum (c) in Fig. 4(C)). Thus, the loading amount could
not be determined by carbon analysis, but a smaller loading
amount than toluene was inferred from the fact that the signals
Entry Modified Silica Solvent
Loading
SBET
Vp
DBJH
silica
(mmol/g)a (m2/g)a (cm3/g)a (nm)a
7
8
9
10
11
4(3a)-1
4(3b)-1
4(3b)-2
4(3b)-2
4(3b)-2
SIO-1 Heptane
SIO-1 Heptane
SIO-2 Heptane
SIO-2 Toluene
SIO-2 Isopropyl
alcohol
0.07
0.23
0.30
0.21
e
1002
879
802
933
945
1.33
1.14
1.79
1.99
2.01
3.7
3.4
5.8
6.2
6.2
a
See Table 1.

1568
T. Miyaji et al. / Journal of Organometallic Chemistry 696 (2011) 1565e1569
Fig. 4. (A) 31P, (B) 29Si, and (C) 13C CP/MAS spectra of PPh2-functionalized silica. (a) PPh2-functionalized SIO-1 (4(3a)-1) (entry 7 of Table 2); (b) PPh2-functionalized SIO-2 (4(3b)-2)
(entry 9); (c) PPh2-functionalized SIO-2 (4(3b)-2) (entry 11). (B): allyl group; (C): isopropoxy group; (*): spinning side band.
derived from 3b were weaker in the 29Si CP/MAS spectrum (spec-
trum (c) in Fig. 4(B)).
4.3. Diphenylphosphino functionalization via route 1 (the “bottom-
up” method)
To a suspension of silica (1.0 g) in dry heptane (25 cm3) was
added 1a (0.45e0.55 mmol), and the mixture was refluxed for 20 h.
The resulting solid was filtered, washed successively with THF,
ethyl acetate, and hexane (five times for each solvent (10 cm3)) to
remove unreacted 1a, and then dried at 353 K under a reduced
pressure for 3 h to give bromopropyl-functionalized silica (2).
KPPh2 (5ꢀ7 equiv.) in THF (0.5 M) was added, dropwise, to
a suspension of 2 (1.0 g) in dry THF (30 cm3), and the mixture was
stirred for 20 h. The resulting solid was filtered, washed succes-
sively with THF, methanol, and hexane (five times for each solvent
(10 cm3)) to eliminate unreacted KPPh2 as well as any by-products,
and then dried at 353 K under a reduced pressure for 3 h to give
PPh2-functionalized silica (4(2)).
3. Conclusions
The diphenylphosphino functionalization of ordered meso-
porous silicas via the tripodal linker unit could be realized by
grafting of 1a and a subsequent reaction with KPPh2 under the
appropriate conditions (7 equiv. at room temperature or 5 equiv. at
323 K) (the “bottom-up” method). On the other hand, directly
grafting the tripodal diphenylphosphino ligands (3a and 3b) (the
“top-down” method) suffered from PPh2 functionalization, prob-
ably because the interaction of the phosphorous groups with the
silica surface was preferential over the reaction between the
leaving groups and the silanol groups. Applications of the resulting
PPh2-functionalized mesoporous silica to transition metal complex
catalyzed reactions are currently underway.
4.4. Diphenylphosphino functionalization via route 2 (the “top-
down” method)
4. Experimental
To a suspension of silica (1.0 g) in dry solvent (25 cm3) was
added 3a or 3b (0.40e0.45 mmol), and the mixture was refluxed for
20 h. Dry heptane, toluene, and isopropyl alcohol were used as the
solvents. The resulting solid was filtered, washed successively with
THF, ethyl acetate, and hexane (five times for each solvent (10 cm3))
to remove unreacted 3a or 3b, and then dried at 353 K under
a reduced pressure for 3 h to give PPh2-functionalized silica (4(3a)
or 4(3b)).
4.1. General
All chemicals were reagent grade and used without further
purification. Two types of ordered mesoporous silicas [19] (Taiyo
Kagaku Co., Ltd., TMPS-4 and TMPS-7, denoted as SIO-1 and SIO-2,
respectively) were used as supports. The surface area, pore volume,
and average pore diameter were 1039 m2/g, 1.46 cm3/g, and 3.8 nm
for SIO-1 and 1050 m2/g, 2.39 cm3/g, and 6.4 nm for SIO-2,
respectively. SIO-1 and SIO-2 were dried at 353 K under a vacuum
for 3 h prior to use. All experiments were performed under inert gas
using standard Schlenk techniques.
4.5. Characterization
Solid-state 31P, 29Si, and 13C CP/MAS NMR spectra were recorded
on a Bruker AVANCE 400WB spectrometer equipped with a 4 mm
MAS probehead. Isothermal nitrogen adsorption/desorption
measurements were carried out at 77 K on a BEL Japan BELSORP
mini II after the sample was degassed at 353 K for 3 h. Elemental
compositions of carbon and bromine were determined using a CE
Instruments EA 1112 elemental analyzer and a DIONEX ICS-2000
ion chromatograph, respectively.
4.2. Preparation of tripodal linker units and tripodal
diphenylphosphino ligands
3-Bromopropyltris[3-(allyldimethylsilyl)propyl]silane (1a) and 3-
bromopropyltris[3-(dimethylisopropoxysilyl)propyl]silane (1b) were
prepared according to the previously described procedure [1].
3-Diphenylphosphinotris[3-(allyldimethylsilyl)propyl]silane (3a)
and 3-diphenylphosphinotris[3-(dimethylisopropoxysilyl)propyl]
silane (3b) were prepared by reacting 1a and 1b, respectively,
with potassium diphenylphosphide in tetrahydrofuran at room
temperature for 20 h.
Acknowledgment
This research was financially supported by the Project of
“Development of Microspace and Nanospace Reaction Environment

T. Miyaji et al. / Journal of Organometallic Chemistry 696 (2011) 1565e1569
1569
[8] X.S. Zhao, X.Y. Bao, W. Guo, F.Y. Lee, Mater. Today 9 (2006) 32e39.
[9] M. Cai, G. Zheng, L. Zha, J. Peng, Eur. J. Org. Chem. (2009) 1585e1591.
Technology for Functional Materials” of New Energy and Industrial
Technology Development Organization (NEDO), Japan.
[10] S.-G. Shyu, S.-W. Cheng, D.-L. Tzou, Chem. Commun. (1999) 2337e2338.
[11] W. Zhou, Y. Li, D. He, Appl. Catal. A 377 (2010) 114e120.
[12] C. Kang, J. Huang, W. He, F. Zhang, J. Organomet. Chem. 695 (2010) 120e127.
[13] K. Melis, D.D. Vos, P. Jacobs, F. Verpoort, J. Mol. Catal. A: Chem. 169 (2001)
47e56.
[14] L. Bemi, H.C. Clark, J.A. Davies, C.A. Fyfe, R.E. Wasylishen, J. Am. Chem. Soc. 104
(1982) 438e445.
[15] A method that modifies the silica surface utilizing deallylation of allylsilane
derivatives has been reported, see (a) T. Shimada, K. Aoki, Y. Shinoda,
T. Nakamura, N. Tokunaga, S. Inagaki, T. Hayashi, J. Am. Chem. Soc. 125 (2003)
4688e4689;
Appendix. Supplementary data
Supplementary data related to this article can be found online at
doi:10.1016/j.jorganchem.2010.12.030.
References
[1] N. Fukaya, S. Onozawa, M. Ueda, K. Saitou, Y. Takagi, T. Sakakura, H. Yasuda,
Chem. Lett. 39 (2010) 402e403.
(b) K. Aoki, T. Shimada, T. Hayashi, Tetrahedron: Asymmetry 15 (2004)
1771e1777.
[2] A.P. Wight, M.E. Davis, Chem. Rev. 102 (2002) 3589e3614.
[3] D.E. de Vos, M. Dams, B.F. Sels, P.A. Jacobs, Chem. Rev. 102 (2002) 3615e3640.
[4] J.Y. Ying, C.P. Mehnert, M.S. Wong, Angew. Chem. Int. Ed. 38 (1999) 56e77.
[5] A. Corma, H. Garcia, Adv. Synth. Catal. 348 (2006) 1391e1412.
[6] D. Brunel, A.C. Blanc, A. Galarneau, F. Fajula, Catal. Today 73 (2002) 139e152.
[7] J.H. Clark, D.J. Macquarrie, Chem. Commun. (1998) 853e860.
[16] G. Engelhardt, H. Jancke, E. Lippmaa, A. Samoson, J. Organomet. Chem. 210
(1981) 295e301.
[17] J. Sommer, Y. Yang, D. Rambow, J. Blümel, Inorg. Chem. 43 (2004) 7561e7563.
[18] J. Blümel, Coord. Chem. Rev. 252 (2008) 2410e2423.
[19] M.P. Kapoor, W. Fujii, M. Yanagi, Y. Kasama, T. Kimura, H. Nanbu, L.R. Juneja,
Microporous Mesoporous Mater. 116 (2008) 370e377.