Direct allylation of α-aryl alcohols with allyltrimethylsilane catalyzed by heterogeneous tin ion-exchanged montmorillonite

Article abstract of DOI:10.1016/j.tetlet.2010.04.074

The direct allylation of α-aryl alcohols with allyltrimethylsilane efficiently proceeded in the presence of tin ion-exchanged montmorillonite under mild conditions according to the proper addition order of reactants and a catalyst.

Full text of DOI:10.1016/j.tetlet.2010.04.074

Tetrahedron Letters 51 (2010) 3300–3303
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Direct allylation of
a-aryl alcohols with allyltrimethylsilane catalyzed
by heterogeneous tin ion-exchanged montmorillonite
*
Jiacheng Wang, Yoichi Masui, Makoto Onaka
Department of Chemistry, Graduate School of Arts and Science, The University of Tokyo, Meguro, Tokyo 153-8902, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
The direct allylation of a-aryl alcohols with allyltrimethylsilane efficiently proceeded in the presence of
tin ion-exchanged montmorillonite under mild conditions according to the proper addition order of reac-
tants and a catalyst.
Received 18 March 2010
Revised 12 April 2010
Accepted 19 April 2010
Available online 22 April 2010
Ó 2010 Elsevier Ltd. All rights reserved.
During the past two decades, Lewis acid-catalyzed allylation of
carbonyl compounds and acetals with allyllic silanes, known as the
Sakurai–Hosomi reaction, has been extensively investigated.1
However, the direct allylation of alcohols as electrophiles received
less attention owing to the poor leaving ability of hydroxyl groups
in the alcohols. The nucleophilic allylation of alcohols is normally
performed through a prior transformation of a hydroxyl group to
a good leaving one such as halides, carbonates, phosphonates,
and sulfonates.
lonite (Sn-Mont) was considered the most acidic for the silylation
of alcohols.12 It has also been found that Sn-Mont was an effective
acid catalyst for the cyanosilylation of carbonyl compounds,15,17,20
the reduction of carbonyl compounds with hydrosilanes,16 the Mi-
chael reaction,13 and the one-pot Strecker synthesis of
a-amino
nitriles.21
In continuation of our work to the development of new syn-
thetic organic transformations, we herein report that water-resis-
tant and environmentally benign Sn-Mont, easily prepared from
natural Na-Mont,17 acted as a highly efficient heterogeneous cata-
There have been several precedents on the direct allylation of a-
aryl alcohols with allyltrimethylsilane, in which only homogeneous
catalysts were employed such as MX3 (M = Bi, In; X = Cl, Br),2–5
B(C6F5)3,6 bis(fluorosulfuryl)imide,7 phosphomolybdic acid,8 ZrCl4,9
and BF3.10 Most of these protocols often require long reaction times,
harsh reaction conditions, and tedious workup procedures. More-
over, Lewis acid catalysts such as metal halides are prone to hydro-
lysis and consequent deactivation, and hence are not reusable.
Thus, to overcome these disadvantages, the development of heter-
ogeneous catalyst systems is in high demand. Proton ion-ex-
changed montmorillonite (H-Mont) as a heterogeneous catalyst
lyst for the allylation of a wide range of a-aryl alcohols with allyl-
trimethylsilane in very short reaction times (0.1–0.5 h) at rt.
Moreover, it was surprisingly found that the addition order of
the reactants and Sn-Mont was definitely critical to the yields
and selectivity. The catalyst was easily recovered by filtration
and reused without an appreciable loss of activity and selectivity.
Sn-Mont was prepared through the ion-exchange of Na-Mont
with an aqueous SnCl4 solution.17 Elemental analysis of Sn-Mont
by inductively coupled plasma (ICP) showed that the tin molar
content of the as-prepared Sn-Mont was 1.9 lmol/mg. The
was found to be efficient for the direct allylation of
a
-aryl alcohols
nitrogen sorption data showed that the specific surface area greatly
with allyltrimethylsilane, but the long reaction times and high reac-
increased from 12 m2 gꢀ1 of pristine Na-Mont to 280 m2 gꢀ1 of
tion temperatures of 60–100 °C were normally required.11
Sn-Mont with
a
porous structure. Transmission electron
We have been focusing on various metal ion-exchanged mont-
morillonites as the solid alternatives to homogeneous Brønsted
acid catalysts.12–17 Montmorillonite, a naturally occurring clay, is
composed of stacked, negatively-charged two-dimensional alumi-
nosilicate layers that hold exchangeable cationic species, mostly
sodium ions in the interlamellar space.18 Once multivalent metal
cations or protons are substituted for sodium ions in the montmo-
rillonite, the clay becomes acidic and has been utilized in various
acid-catalyzed organic reactions.12,19 Among the various metal
ion-exchanged montmorillonites, tin ion-exchanged montmoril-
microscopy (TEM) indicated that Sn-Mont was a nanocomposite
of nanoparticles of SnO2 surrounded by montmorillonite silicate
layers. FT-IR spectrum of Sn-Mont after the pyridine adsorption
proved that Sn-Mont had Brønsted acidity.21
Because it has been already shown that the addition order of
reactants had a great influence on the cyanation of a-aryl alcohols
with cyanotrimethylsilane in the presence of solid Brønsted acid
catalysts,22 the addition order was initially examined in the reac-
tion of benzhydrol (1a) with allyltrimethylsilane (ATMS) in CH2Cl2
in the presence of Sn-Mont (20 mg, 3.8 mol %), as shown in
Figure 1.24
When Method A in which ATMS was added to a suspended
mixture of 1a and Sn-Mont was adopted at rt, the desired 2a was
* Corresponding author. Tel.: +81 3 5454 6595; fax: +81 3 5454 6998.
E-mail address: conaka@mail.ecc.u-tokyo.ac.jp (M. Onaka).
0040-4039/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2010.04.074

J. Wang et al. / Tetrahedron Letters 51 (2010) 3300–3303
3301
hydrol (1a) and its derivatives 1b–d which have electron-with-
drawing or electron-donating substituents on the aromatic rings
were effectively allylated at rt (entries 1–4). No product was pro-
duced without the catalyst (entry 5).
Ph
Ph
OTMS
OH
Ph
Sn-Mont
+
+
+
TMS
(2 equiv.)
CH2Cl2, rt
O
Ph
Ph
Ph
Ph
Ph
Ph
Ph
1a
2a
3a
4a
(a trace amount)
80
100
80
60
40
20
0
A secondary alcohol, 1-phenylethanol (1e), yielded the corre-
sponding alkene 2e in moderate yield (61%), which might be due
to the formation of polystyrene as a side product (Table 1, entry
7).10 The direct allylation of 1e with ATMS failed by the use of
bis(fluorosulfuryl)imide,7 phosphomolybdic acid,8 and ZrCl4.9 BF3
gave a poor yield (22%) of the desired products with a large amount
of side products.10 The introduction of an electron-donating group
at the para-position of 1e resulted in higher yields: 75% for 1f in
entry 8 and 98% for 1g in entry 9. Similarly, 1-(40-Me or MeO-
substituted phenyl)propanols (1h–i) afforded a middle yield of
81% (entry 10) and a high yield of 94% (entry 11), respectively.
An acid-sensitive substrate (1j) was also applicable for the present
heterogeneous catalytic system and the desired allylated product
(2j) was gained in 73% yield (entry 12). Other 1-arylethanol deriv-
atives (1k–m) reacted with ATMS to produce the allylated prod-
ucts in middle to good yields (entries 13–15). Unfortunately,
primary alcohols such as benzyl alcohol and a simple aliphatic
alcohol like decan-3-ol were not suitable for the Sn-Mont-cata-
lyzed reactions and no desired products were obtained. Reactions
of allylic alcohols 1n and 1o also gave good yields (entries 16–
17). To our delight, two sterically hindered alcohols 1p and 1q
were also proper substrates for giving the corresponding allylated
products in excellent yields at rt (entries 18–19).
In order to confirm that the observed catalysis is truly caused by
the solid Sn-Mont rather than the Sn species leaching into the solu-
tion, the following experiments were performed: The allylation of
1a with ATMS was carried out at rt as shown in entry 1 of Table 1.
After the complete consumption of 1a, Sn-Mont was removed from
the reaction mixture by filtration. Another substrate 1d and ATMS
were then added to the filtrate at rt. Even in 1 h, no 2d was
produced, indicating that the solution phase did not include any
catalytically active species. Moreover, Sn-Mont is superior to
homogeneous catalysts in the way that Sn-Mont can be easily sep-
arated from the reaction mixture and be reused without an appre-
ciable loss of activity and selectivity (Table 1, entry 6).
(A)
(B)
2a
3a
2a
60
40
20
0
0.0
0.5
1.0
1.5
2.0
0.0 0.1 0.2 0.3
Time (h)
Time (h)
Figure 1. The effects of the addition orders of reactants and a catalyst on the yields
of the allylated product 2a and ether 3a for the reaction of benzhydrol 1a and
allyltrimethylsilane catalyzed by Sn-Mont at rt. Reaction conditions: Sn-Mont
(20 mg, 3.8 mol %), 1a (1 mmol), allyltrimethylsilane (2 mmol), CH2Cl2 (2 mL), rt.
(A) Method A; (B) Method B.
obtained in only 16% yield, together with a large amount of 3a
(57%) in 0.1 h, as indicated in Figure 1A. By-product 3a gradually
changed into 2a, and a moderate yield (75%) of 2a was gained with
remaining 3a (24%) when the reaction time was extended to 2 h. It
is evident that the formation of side product 3a reduced the yield
of 2a.
In order to suppress the intermolecular condensation of 1a to 3a
by the acid catalysis of Sn-Mont, another addition procedure called
Method B was employed, in which a CH2Cl2 solution of 1a was
slowly added to a suspended mixture of Sn-Mont and ATMS. Fol-
lowing this method, to our surprise, the reaction became very fast
and completed in only 0.1 h at rt as shown in Figure 1B. The yield
of 2a was dramatically improved to 98% without the formation of
3a. In both Methods A and B, only trace amounts of 4a were found
based on 1H NMR and GC.
Sn-Mont is a nanoporous material with strong Brønsted acid-
ity.21 Thus, the most probable reaction pathway in the allylation
of alcohols with allyltrimethylsilane should involve a carbocation
intermediate dehydrated by the Brønsted acidity of Sn-Mont, fol-
lowed by the direct attack with allyltrimethylsilane to form the
allylated products.11 In fact, during the allylation of 1a with ATMS,
the ether (3a) and a trace amount of diphenylmethyl trimethylsilyl
ether (4a) were observed, together with the allylated product 2a.
Thus, 3a and 4a are probable intermediates which further react
with ATMS to afford 2e.
It is noteworthy that the turn-over frequency (TOF) of the reac-
tion of 1a with ATMS catalyzed by Sn-Mont is as high as 258 hꢀ1
;
this value is far greater than the preceding ones for the same reac-
tion using various homogeneous catalysts such as BiCl3 36.8 hꢀ1 2
,
Brønsted acid, bis (fluorosulfuryl)imide HN(SO2F)2 115 hꢀ1 7
, InCl3
6.7 hꢀ1,4 InBr3 34 hꢀ1,5 and ZrCl4 196 hꢀ1.9 The activity of Sn-Mont
for the allylation of 1a is also far superior to that of the commer-
cially available Amberlyst-15. And the reaction rate
(490 mmol hꢀ1
g
) based on the weight of Sn-Mont was far higher
ꢀ1
cat
Sn-Mont
than those (Method A, 1.5 mmol hꢀ1
g
;
Method B,
ꢀ1
1a
3a
4a
3a
Yield: 71%
ð1Þ
ð2Þ
ð3Þ
cat
CH2Cl2, rt, 1 h
34.5 mmol hꢀ1
g
) for Amberlyst-15. Moreover in our case the
Sn-Mont catalyst was easily separable from the reaction mixture
by a simple filtration. Although H-Mont was reported to be effi-
25
ꢀ1
cat
Sn-Mont
AT MS
(4 equiv.)
2a Yield: 20%
2a Yield: 84%
cient for the same reaction, but heating at 60 or 100 °C was re-
CH2Cl2, rt, 6 h
quired,11 and the reaction rate of 48 mmol hꢀ1
g
based on the
ꢀ1
cat
weight of H-Mont was also far lower than that of 490 mmol hꢀ1
g
Sn-Mont
ꢀ1
cat
AT MS
(2 equiv.)
of Sn-Mont which promoted the reaction even at ambient
temperature.
CH2Cl2, rt, 1 h
Secondly, to investigate the scope and limitations of the present
catalytic protocol with Sn-Mont following the optimal addition or-
der (Method B), various alcohols were examined and the results
are summarized in Table 1.26 Most reactions gave the desired prod-
ucts with good to excellent yields in the presence of Sn-Mont
(20 mg; 3.8 mol %) in very short reaction times (0.1–0.5 h). Benz-
Here we independently examined whether or not these two
side products (3a and 4a) would be further transformed into 2a.
Firstly, 3a was in advance prepared in a 71% isolated yield by sim-
ply mixing Sn-Mont and 1a in CH2Cl2 as shown in Eq. 1. Secondly,
the reaction of 3a with ATMS afforded 2a, but in a poor yield of 20%
even after 6 h (Eq. 2). Moreover, another side product 4a, prepared

3302
J. Wang et al. / Tetrahedron Letters 51 (2010) 3300–3303
Table 1
Allylation of various alcohols with allyltrimethylsilane using Sn-Monta,b
R1
R2
OH
Sn-Mont
+
Me3Si
R2
R1
CH2Cl2, rt, 0.1 - 0.5 h
1
2
Entry
Alcohol 1
OH
R
Product 2
Yieldc (%)
1
2
H
Cl
1a
1b
2a
2b
95
92
3
4
5d
6e
Me
OMe
1c
1d
2c
2d
96
98
R
R
1a
1a
2a
2a
0
96
7
8
9
H
Me
OMe
1e
1f
1g
2e
2f
2g
61
75
98
HO
R
R
R
R
10
11
Me
OMe
1h
1i
2h
2i
81
94
HO
OH
O
O
O
O
12
1j
2j
73
71
OH
13f
1k
2k
14f
15f
H
OMe
1l
1m
2l
2m
66
85
OH
R
R
OH
16
17
1n
1o
1p
1q
2n
2o
2p
2q
93
72
89
91
OH
OH
18f
OH
Ph
Ph
19f,g
a
b
c
d
e
f
Alcohol (1 mmol), allyltrimethylsilane (2 mmol), Sn-Mont (20 mg, 3.8 mol %), dichloromethane (2 mL), rt.
Method B was employed.
Isolated yields.
The reaction was performed without catalyst.
With the recovered catalyst.
Alcohol (0.5 mmol).
g
1.5 h.
according to the literature,23 efficiently reacted with ATMS in a
good yield of 84% (Eq. 3). Thus, it was concluded that in the allyla-
tion of 1a, 3a and 4a were both intermediates which further re-
acted with ATMS to give 2a. Since the reaction of 3a with ATMS
is very slow, the production of 2a should be avoided by optimizing
the addition order of the reactants and the catalyst.
In summary, we have demonstrated the direct allylation of sec-
benzylic or sec-allylic alcohols with allyltrimethylsilane in the
presence of a catalytic amount of heterogeneous acid, Sn-Mont
by employing the proper addition order of the reactants and the
catalyst under mild conditions. Sn-Mont provided the high activity
for the allylation compared with those of the previously reported
homogeneous catalysts. In all cases, the desired allylated products
were obtained from a wide range of the alcohols in middle to
excellent yields. Furthermore, the catalyst could be reused without
any loss of activity and selectivity.
Acknowledgment
The present work was partially supported by a Grant-in-Aid for
Scientific Research from the Ministry of Education, Culture, Sports,
Science, and Technology.
References and notes
1. Sakurai, H. Pure Appl. Chem. 1982, 54, 1–22; Hosomi, A. Acc. Chem. Res. 1988, 21,
200–206; Marshall, J. A. Chem. Rev. 1996, 96, 31–47.
2. De, S. K.; Gibbs, R. A. Tetrahedron Lett. 2005, 46, 8345–8350.
3. Saito, T.; Nishimoto, Y.; Yasuda, M.; Baba, A. J. Org. Chem. 2006, 71, 8516–8522.

J. Wang et al. / Tetrahedron Letters 51 (2010) 3300–3303
3303
4. Yasuda, M.; Saito, T.; Ueba, M.; Baba, A. Angew. Chem., Int. Ed. 2004, 43, 1414–
1416.
24. Two different addition orders of the reactants and the catalyst were
investigated for the allylation of alcohols with allyltrimethylsilane as follows:
5. Kim, S. H.; Shin, C.; Pae, A. N.; Koh, H. Y.; Chang, M. H.; Chung, B. Y.; Cho, Y. S.
Synthesis-Stuttgart 2004, 1581–1584.
Method A: Into a flask were placed Sn-Mont (20 mg, 3.8 mol %), alcohol
(1 mmol), and CH2Cl2 (2 mL). The obtained solution was vigorously stirred and
then allyltrimethylsilane (2 mmol) was added. After stirred for a time interval
at rt, the solution was analyzed by GC.
6. Rubin, M.; Gevorgyan, V. Org. Lett. 2001, 3, 2705–2707.
7. Kaur, G.; Kaushik, M.; Trehan, S. Tetrahedron Lett. 1997, 38, 2521–2524.
8. Kadam, S. T.; Lee, H.; Kim, S. S. Appl. Organomet. Chem. 2009, 24, 67–70.
9. Sharma, G. V. M.; Reddy, K. L.; Lakshmi, P. S.; Ravi, R.; Kunwar, A. C. J. Org. Chem.
2006, 71, 3967–3969.
10. Cella, J. A. J. Org. Chem. 1982, 47, 2125–2130.
11. Motokura, K.; Nakagiri, N.; Mizugaki, T.; Ebitani, K.; Kaneda, K. J. Org. Chem.
2007, 72, 6006–6015.
12. Onaka, M.; Hosokawa, Y.; Higuchi, K.; Izumi, Y. Tetrahedron Lett. 1993, 34,
1171–1172.
13. Kawai, M.; Onaka, M.; Izumi, Y. Bull. Chem. Soc. Jpn. 1988, 61, 2157–2164.
14. Kawai, M.; Onaka, M.; Izumi, Y. Bull. Chem. Soc. Jpn. 1988, 61, 1237–1245;
Onaka, M.; Ohno, R.; Kawai, M.; Izumi, Y. Bull. Chem. Soc. Jpn. 1987, 60, 2689–
2691.
15. Higuchi, K.; Onaka, M.; Izumi, Y. Bull. Chem. Soc. Jpn. 1993, 66, 2016–2032.
16. Onaka, M.; Higuchi, K.; Nanami, H.; Izumi, Y. Bull. Chem. Soc. Jpn. 1993, 66,
2638–2645.
17. Wang, J. C.; Masui, Y.; Watanabe, K.; Onaka, M. Adv. Synth. Catal. 2009, 351,
553–557.
18. Pinnavaia, T. J. Science 1983, 220, 365–371; Laszlo, P. Science 1987, 235, 1473–
1477.
19. Kaneda, K. Synlett 2007, 999–1015; Kaneda, K.; Ebitani, K.; Mizugaki, T.; Mori,
K. Bull. Chem. Soc. Jpn. 2006, 79, 981–1016.
20. Izumi, Y.; Onaka, M. J. Mol. Catal. 1992, 74, 35–42.
21. Wang, J. C.; Masui, Y.; Onaka, M. Eur. J. Org. Chem. 2010, 1763–1771.
22. Wang, J. C.; Masui, Y.; Onaka, M. Submitted for publication.
23. Zareyee, D.; Karimi, B. Tetrahedron Lett. 2007, 48, 1277–1280; Azizi, N.; Yousefi,
R.; Saidi, M. R. J. Organomet. Chem. 2006, 691, 817–820.
Method B: Into
a flask were added Sn-Mont (20 mg, 3.8 mol %), CH2Cl2
(0.5 mL), and allyltrimethylsilane (2 mmol). Under stirring vigorously,
alcohol (1 mmol) dissolved in CH2Cl2 (1.5 mL) was then introduced to the
above reaction system dropwise (about 5 s per drop) using a syringe. After the
completion of the reaction checked by GC, the catalyst was removed by
filtration and the solvent was evaporated to give the crude product. Further
purification was carried out with silica chromatography.
25. We performed the direct allylation of 1a (1 mmol) with allyltrimethylsilane
(2 mmol) in CH2Cl2 (2 mL) catalyzed by Amberlyst-15 (20 mg) at rt. In 1 h, the
yields of the desired product are 3% for Method A and 69% for Method B,
respectively.
26. The spectral data of the selected alkenes: 4-(4-methylphenyl)-4-phenyl-1-
butene (2c), 1H NMR (500 MHz, CDCl3) d 7.42–7.10 (m, 9H), 5.79 (ddt, J = 6.7,
10.4, 17.1, 1H), 5.06 (ddd, J = 5.5, 9.5, 37.8, 2H), 4.04 (t, J = 7.9, 1H), 2.87 (dd,
J = 6.7, 7.9, 2H), 2.36 (s, 3H). 13C NMR (126 MHz, CDCl3) d 144.8, 141.5, 137.0,
135.7, 129.1, 128.4, 127.9, 127.8, 126.1, 116.2, 50.8, 39.9, 20.9. HRMS calcd for
C17H18 222.1409, found 222.1423. 4-(4-Methoxyphenyl)-1-pentene (2g), 1H
NMR (500 MHz, CDCl3) d 7.12 (d, J = 8.5, 2H), 6.84 (d, J = 8.5, 2H), 5.77–5.63 (m,
1H), 5.04–4.88 (m, 2H), 3.79 (s, 3H), 2.80–2.69 (m, 1H), 2.40–2.21 (m, 2H), 1.23
(d, J = 6.7, 3H). 13C NMR (126 MHz, CDCl3) d 157.8, 139.2, 137.4, 127.9, 115.8,
113.7, 55.2, 42.8, 38.9, 21.7. HRMS calcd for C12H16
O 176.1201, found
176.1203. 4-(4-Methoxyphenyl)-1-hexene (2i), 1H NMR (500 MHz, CDCl3) d
7.10–7.03 (m, 2H), 6.88–6.81 (m, 2H), 5.75–5.57 (m, 1H), 5.06–4.83 (m, 2H),
3.78 (s, 3H), 2.46 (ddd, J = 5.8, 8.5, 11.6, 1H), 2.40–2.24 (m, 2H), 1.77–1.65 (m,
1H), 1.51 (tdd, J = 6.1, 10.7, 14.6, 1H), 0.76 (t, J = 7.3, 3H). 13C NMR (126 MHz,
CDCl3) d 157.8, 137.4, 137.3, 128.6, 115.6, 113.6, 55.1, 46.7, 41.0, 28.9, 11.9.
HRMS calcd for C13H18O 190.1358, found 190.1364.