An amphiphilic pillar[5]arene as efficient and substrate-selective phase-transfer catalyst

Article abstract of DOI:10.1021/ol4016546

An amphiphilic macrocyclic compound consisting of 10 tetra-alkyl phosphonium bromide groups and a pillar[5]arene core was prepared. This compound was soluble in both aqueous and organic media and acted as a highly efficient and substrate-selective phase-transfer catalyst. In particular, oxidation of the linear alkene1-hexene to 1-pentanal by KMnO₄ was >99%, whereas that of the branched alkene 4-methyl-1-hexene was only 31%, under ideal conditions.

Full text of DOI:10.1021/ol4016546

ORGANIC
LETTERS
2013
Vol. 15, No. 14
3742–3745
An Amphiphilic Pillar[5]arene as Efficient
and Substrate-Selective Phase-Transfer
Catalyst
Tomoki Ogoshi,* Naosuke Ueshima, and Tada-aki Yamagishi
Graduate School of Natural Science and Technology, Kanazawa University,
Kakuma-machi, Kanazawa, 920-1192, Japan
ogoshi@se.kanazawa-u.ac.jp
Received June 12, 2013
ABSTRACT
An amphiphilic macrocyclic compound consisting of 10 tetra-alkyl phosphonium bromide groups and a pillar[5]arene core was prepared. This
compound was soluble in both aqueous and organic media and acted as a highly efficient and substrate-selective phase-transfer catalyst. In
particular, oxidation of the linear alkene1-hexene to 1-pentanal by KMnO4 was >99%, whereas that of the branched alkene 4-methyl-1-hexene was
only 31%, under ideal conditions.
Phase transfer catalysts (PTCs) embody the ideals of
organic synthesis by providing a simple, cheap, and general
protocol for various reactions.1ꢀ4 PTCs allow reactions
to proceed through active transport of polar reagents
from one phase to another. Amphiphilic small organic
molecules such as quaternary ammonium compounds,
crown ethers, phosphonium compounds, and pyridinium
compounds are generally used as PTCs. Design of new
smart PTCs with additional advantages and functions is
a new research target.4 In this letter, we report a new PTC
containing a macrocyclic structure (pillar[5]arene) in the
core. Pillar[5]arenes5ꢀ9 are new macrocyclic hosts that were
first reported by our group5 and are good hosts for linear
guests in organic7 and aqueous media.8 Pillar[5]arene has
(5) Ogoshi, T.; Kanai, S.; Fujinami, S.; Yamagishi, T.; Nakamoto, Y.
J. Am. Chem. Soc. 2008, 130, 5022–5023.
(1) (a) Phase-Transfer Catalysis: New Chemistry, Catalysts, and
Applications; American Chemical Society: Washington, DC, 1985. (b)
Starks, C. M.; Liotta, C. L.; Halpern, M. Phase-Transfer Catalysis: Funda-
mentals, Applications and Industrial Perspectives; Chapman and Hall: New
York, 1994. (c) Phase-Transfer Catalysis: Mechanisms and Syntheses;
Halpern, M. E., Ed.; American Chemical Society: Washington, DC, 1997.
(d) Makosza, M.; Fedorynski, M. In Interfacial Catalysis; Volkov, A. G.,
Ed.; Marcel Dekker: New York, 2003; pp 150ꢀ201. (e) Chen, L. L.; Yan, Z.
Org. Lett. 2012, 14, 784–787. (f) Denmark, S. E.; Weintraub, R. C.;
Gould, N. D. J. Am. Chem. Soc. 2012, 134, 13415–13429.
(6) Reviews: (a) Ogoshi, T. J. Incl. Phenom. Macrocycl. Chem. 2012,
72, 247–262. (b) Ogoshi, T.; Yamagishi, T. Bull. Chem. Soc. Jpn. 2013,
86, 312–332. (c) Ogoshi, T.; Yamagishi, T. Eur. J. Org. Chem. 2013,
2961–2975. (d) Cragg, P. J.; Sharma, K. Chem. Soc. Rev. 2012, 41, 597–
607. (e) Xue, M.; Yang, Y.; Chi, X.; Zhang, Z.; Huang, F. Acc. Chem.
Res. 2012, 45, 1294–1308.
(7) (a) Ogoshi, T.; Demachi, K.; Kitajima, K.; Yamagishi, T. Chem.
Commun. 2011, 47, 10290–10292. (b) Zhang, Z.; Luo, Y.; Chen, J.;
Dong, S.; Yu, Y.; Ma, Z.; Huang, F. Angew. Chem., Int. Ed. 2011, 50,
1397–1401. (c) Li, C.; Xu, Q.; Li, J.; Yao, F.; Jia, X. Org. Biomol. Chem.
2010, 8, 1568–1576. (d) Shu, X.; Fan, J.; Li, J.; Wang, X.; Chen, W.; Jia,
X.; Li, C. Org. Biomol. Chem. 2012, 10, 3393–3397. (e) Strutt, N. L.;
Forgan, R. S.; Spruell, J. M.; Botros, Y. Y.; Stoddart, J. F. J. Am. Chem.
Soc. 2011, 133, 5668–5671.
(8) (a) Ogoshi, T.; Hashizume, M.; Yamagishi, T.; Nakamoto, Y.
Chem. Commun. 2010, 3708–3710. (b) Hu, X. B.; Chen, L.; Si, W.; Yu,
Y.; Hou, J. L. Chem. Commun. 2011, 47, 4694–4696. (c) Ma, Y.; Ji, X.;
Xiang, F.; Chi, X.; Han, C.; He, J.; Abliz, Z.; Chen, W.; Huang, F. Chem.
Commun. 2011, 47, 12340–12342. (d) Yao, Y.; Xue, M.; Chi, X.; Ma, Y.;
He, J.; Abliz, Z.; Huang, F. Chem. Commun. 2012, 48, 6505–6507.
ꢀ
(2) (a) Eller, L. R.; Stepien, M.; Fowler, C. J.; Lee, J. T.; Sessler, J. L.;
)
Moyer, B. A. J. Am. Chem. Soc. 2007, 129, 11020–11021. (b) Shin, J. A.;
Lim, Y. G.; Lee, K. H. J. Org. Chem. 2012, 77, 4117–4122. (c) Sala,
G. D.; Nardone, B.; Riccardis, F. D.; Izzo, I. Org. Biomol. Chem. 2013,
11, 726–731. (d) Hooley, R. J.; Biros, S. M.; Rebek, J. Angew. Chem., Int.
Ed. 2006, 45, 3517–3519.
(3) (a) Lee, D. G.; Chang, V. S. J. Org. Chem. 1978, 43, 1532–1536.
(b) Starks, C. M. J. Am. Chem. Soc. 1971, 93, 195–199.
(4) (a) Hashimoto, T.; Maruoka, K. Chem. Rev. 2007, 107, 5656–
5682. (b) Maruoka, K.; Ooi, T. Chem. Rev. 2003, 103, 3013–3028.
r
10.1021/ol4016546
Published on Web 07/01/2013
2013 American Chemical Society

10 reactive sites at its rims, and introduction of functional
groups on these rims affects the physical properties of the
pillar[5]arene because these groups cover the core.9 In this
study, we prepared a pillar[5]arene derivative containing 10
phosphonium cations 1 (Figure 1). Thanks to the phos-
phonium cations, 1was soluble in a wide varietyof solvents.
Here, we report the very efficient PTC properties and
substrate selectivity of 1 based on the hostꢀguest chemistry
of pillar[5]arene.
The new amphiphilic pillar[5]arene 1 was obtained as
a white solid in 95% yield by reacting 210 with excess tri-
butylphosphine. 1 was soluble in a wide variety of solvents
including water, alcohols, acetone, tetrahydrofuran,
chloroform, acetonitrile, DMF, and DMSO and insoluble
in hexane and diethyl ether. Based on this amphiphilicity,
we investigated the oxidation of alkenes by KMnO4 in a
two-phase chloroformꢀwater solvent system using 1 as a
PTC. First, we investigated hostꢀguest complexation of 1
with alkene substrates such as the linear alkene 1-hexene
and branched alkene 4-methyl-1-hexene, in CDCl3 by
1H NMR. As the concentration of 1 increased, upfield
Figure 2. 1H NMR titrations of (a) 1-hexene (3 mM) and (b)
4-methyl-1-hexene (3 mM) on addition of 1 (0ꢀ5 equiv) in
CDCl3 at 25 °C. Resonances are labeled in Figure 1. A peak
with an asterisk derives from host 1.
shiftsofthe peaks for the methylene for 1-hexene (Figure 1,
proton Ha) were observed (Figure 2a),11 which suggests
complexation occurs between 1 and 1-hexene. The driving
force of the complexation is the CHꢀπ interaction, which
is also observed in the hostꢀguest complexes between
hydrocarbons and pillar[5]arenes.7 The stoichiometry of
the complex determined from a Job plot (Figure S3) was
1:1, and the association constant K for the binding of
1-hexene into the cavity of 1 was calculated as 65 ( 16 Mꢀ1
using nonlinear curve fitting analysis (Figure S4). By
contrast, no peak shifts of 4-methyl-1-hexene were ob-
served even with excess 1 (Figure 2b), which indicates that
the branched alkene 4-methyl-1-hexane did not interact
with 1. The compact linear shape of 1-hexene is favorable
for complexation, but the branched shape of 4-methyl-1-
hexene is not. This is because the cavity of pillar[5]arene
˚
is very narrow (∼5 A) and has a well-defined cylindrical
architecture.5 The same size-selective hostꢀguest system
has been reported in hydrocarbon recognition using
pillar[5]arenes.7
We investigated the oxidation of alkenes in a two-phase
D2OꢀCDCl3 solvent system using PTCs and the oxidant
KMnO4. In the presence of 1, the chloroform phase (lower
phase) quickly became colored (Figure 3a-i), and took on
the original deep purple color of the aqueous solution
(upper phase). However, phase transfer did not take place
in the absence of 1 (Figure 3a-ii). These observations
ꢀ
suggest phase transfer of MnO4 from the aqueous to
chloroform phase. The oxidation reaction of the guest
1-hexene by KMnO4 in the two-phase D2OꢀCDCl3 sol-
vent system in the presence of 1 (Figure 3b) was monitored
Figure 1. Phase-transfer catalysts: pillar[5]arene modified with
10 tetra-alkylphosphonium bromide groups (1), unit model (3),
dicyclohexano-18-crown-6 (4), and tetrabutyl phosphonium
bromide (5). Substrates: 1-hexene and 4-methyl-1-hexene.
1
by H NMR. We found the new proton peak from the
aldehyde moiety of 1-pentanal along with the proton peaks
from remaining substance 1-hexene in the CDCl3 phase,
but not in the D2O phase (Figure S5), which indicates that
the oxidation reaction proceeded in the CDCl3 phase, but
not in the D2O phase. The conversions from 1-hexene to
1-pentanal were calculated using the integration ratio of
(9) (a) Ogoshi, T.; Shiga, R.; Yamagishi, T. J. Am. Chem. Soc. 2012,
134, 4577–4580. (b) Ogoshi, T.; Kida, K.; Yamagishi, T. J. Am. Chem.
Soc. 2012, 134, 20146–20150. (c) Ogoshi, T.; Aoki, T.; Shiga, R.; Iizuka,
R.; Ueda, S.; Demachi, K.; Yamafuji, D.; Kayama, H.; Yamagishi, T.
J. Am. Chem. Soc. 2012, 134, 20322–20325. (d) Nierengarten, I.; Guerra,
S.; Holler, M.; Nierengarten, J. F.; Deschenaux, R. Chem. Commun.
2012, 48, 8072–8074.
(11) The peak shifts of the vinyl proton peaks with 1 were not
observed. The other methylene peaks were overlapped with 1; thus the
methylene peaks (Figure 1, Ha) were used for investigation of the
complexation.
(10) Ogoshi, T.; Ueshima, N.; Yamagishi, T.; Toyota, Y.; Matsumi,
N. Chem. Commun. 2012, 48, 3536–3538.
Org. Lett., Vol. 15, No. 14, 2013
3743

To determine why 1 has excellent PTC properties, we
examined the oxidation of 1-hexene using 1 in the presence
of the competitive guest (Table 1).
Table 1. Oxidation Reaction of Alkenes Using PTCsa
alkene
conversion
(%)b
run
PTC
(0.13 mmol)
1
ꢀ
1-hexene
0
2
1 (1.2 μmol)
1 (1.2 μmol)
1 (1.2 μmol)
1 (1.2 μmol)
1 (1.2 μmol)
1-hexene
>99
32
3c
4d
5
1-hexene
1-hexene
7
4-methyl-1-hexene
1-hexene þ
4-methyl-1-hexene
1-heptene
1-octene
31
85e þ 7f
6
7
1 (1.2 μmol)
1 (1.2 μmol)
1 (1.2 μmol)
3 (6.0 μmol)g
3 (6.0 μmol)g
3 (6.0 μmol)g
3 (6.0 μmol)g
3 (6.0 μmol)g
4 (1.2 μmol)
5 (12 μmol)h
>99
14
9
8
9
1-decene
10
11d
12
13
14
15
16
1-hexene
35
18
40
38
30
0
1-hexene
1-heptene
1-octene
1-decene
1-hexene
1-hexene
20
a The reactions were carried out with KMnO4 (0.40 mmol) and PTC
in a mixture of D2O (4 mL) and CDCl3 (4 mL) under vigorous stirring at
25 °C for 12 h. b Conversion to alkanal. c 1.4-Dicyanobutane (0.13 mmol)
was added as a competitive guest. d A mixture of D2O (4 mL) and CD2Cl2
(4 mL) was used for the reaction. e Conversion from 1-hexene to 1-penta-
nal. f Conversion from 4-methyl-1-hexene to 3-methyl-1-pentanal.
g 1 contains 10 tetrabutylphosponium groups, while unit model 3 contains
only 2 of them. Therefore, in order to have an identical concentration
of phosponium groups, a 5 times molar amount of 3 over 1 was used.
h 1 contains 10 tetrabutylphosponium groups, while 5 contains only 1 of
them. Therefore, in order to have an identical concentration of phospo-
nium groups, a 10 times molar amount of 5 over 1 was used.
Figure 3. (a) Photographs of KMnO4 in a two-phase D2O (upper
phase)/CDCl3 (bottom phase) solvent system (i) with and (ii)
without 1. (b) Percentage conversion to 1-pentanal versus molar
amounts of pillar[5]arene 1 (red circles) and unit model 3 (blue
triangles); the reaction time is 12 h. (c) Kinetic studies: percen-
tage conversion to 1-pentanal versus reaction time with pillar-
[5]arene 1 (1.2 μmol, red circles) and unit model 3 (6 μmol, blue
triangles).
1,4-Dicyanobutane forms very highly stable 1:1 hostꢀ
guest complexes with pillar[5]arenes. The association con-
stants K of the complexes between 1,4-dicyanobutane
and pillar[5]arenes are generally greater than 104 Mꢀ1 13
,
the proton peaks between the aldehyde moiety of 1-pentanal
(Figure 3, Hb0) and the vinyl moiety of 1-hexene (Figure 3,
Hb) in CDCl3. In the absence of PTC, the reaction did not
proceed. The percentage converted by the oxidation sharply
increased as the amount of 1 increased, and the conversion
to 1-pentanal was >99% with 1.2 μmol of 1 (Figure 3b, red
circles, turnover number (TON) = 108.3). By contrast, the
conversions were low with a similar amount of the unit
model 3 and only reached moderate yields (42%, TON =
3.0) even with a large excess of 3 (Figure 3b, blue triangles).12
Kinetic studies for the PTC system were also carried out
(Figure 3c). Complete oxidation of 1-hexene with 1 required
∼7 h (t1/2 = 4.51 ( 0.12 h). The turnover frequency (TOF)
was calculated to be 15.5 hꢀ1. By contrast, the oxidation with
3 required ∼12 h to reach equilibrium (t1/2 = 5.84 ( 0.12 h).
The TOF was calculated to be 0.25 hꢀ1. These data indicate
that 1 has better PTC properties compared to unit model 3.
while the K of the hostꢀguest complexes between linear
hydrocarbons and pillar[5]arenes are very weak (K =
10ꢀ100 Mꢀ1).7 The conversion dramatically decreased
with the competitive guest 1,4-dicyanobutane (run 3,
1 equiv of 1,4-dicyanobutane to 1-hexene), while the
reaction went to completion without 1,4-dicyanobutane
(run 2). The association constant K of the 1:1 1 1,4-
dicyanobutane complex was higher (K = (1.62 ( 0.026)
3
ꢁ 103 Mꢀ1, Job plot (Figure S6) and determination of
1
K from H NMR) than that of the 1 1-hexene complex
3
(K = 65 ( 16 Mꢀ1). Therefore, 1,4-dicyanobutane acted
as a competitive guest and induced dissociation of the
1 1-hexene complex. The conversion was also extremely
3
low (7%) in a two-phase D2O/CD2Cl2 solvent system
using 1 (run 4), which was lower than that in a two-phase
D2O/CDCl3 solvent systemusing 1 (>99%, run 2) and in a
two-phase D2O/CD2Cl2 solvent system using unit model 3
(12) 1 contains 10 tetrabutylphosponium groups, while unit model 3
contains only 2 of them. Therefore, in order to have an identical
concentration of phosponium groups, a 5 times molar amount of 3 over
1 was used.
(13) Shu, X.; Chen, S.; Li, J.; Chen, Z.; Weng, L.; Jia, X.; Li, C. Chem.
Commun. 2012, 48, 2967–2969.
3744
Org. Lett., Vol. 15, No. 14, 2013

The oxidations of the alkenes using 1 are highly substrate
selective. On the other hand, the conversions of the alkenes
using unit model 3 were 30ꢀ40% (runs 10, 12ꢀ14), in-
dicating that the substrate selectivity was observed in the
pillar[5]arene 1but not in the unit model 3. The difference in
substrate selectivity between pillar[5]arene 1 and unit model
3 also supports the proposed mechanism of the oxidation
reaction.
Crown ether and tetra-alkyl phosphonium bromide are
useful PTCs for the oxidation of alkenes. Thus, we com-
pared the conversions with these PTCs in the oxidation
reactions under the same conditions. The oxidation reac-
tion using dicyclohexano-18-crown-6 4 (Figure 1) did not
proceed (run 15). With a ten times molar amount of the
simple tetrabutyl phosphonium bromide 5 (Figure 1) the
conversion was only 20% (run 16), while the oxidation
reaction with 1 went to completion (run 2). Therefore,
compared to previously reported PTCs, 1 is an excellent
PTC for the oxidation of alkenes. Complexation of the
substrate with 1 and increasing the local concentration of
MnO4ꢀ around rims increased the efficiency of the oxida-
tion reaction.
In conclusion, the new PTC pillar[5]arene 1 was synthe-
sized and applied with high efficiency and substrate selec-
tivity. Oxidation reactions reached completion even with
very small amounts of PTC 1 (amount of substance
fraction 0.90%). This is a significant advantage compared
with typical PTCs, which generally require an amount of
substance fraction 5ꢀ100% for oxidation of alkenes.3
Oxidation of the linear alkenes 1-hexene and 1-heptene
completely proceeded, but that of the branched alkene
4-methyl-1-hexene and long chain alkenes did not. Sub-
strate selective PTC properties have not been observed in
other PTCs. The hostꢀguest properties of pillar[5]arene
contributed to the efficient and substrate-selective PTC
properties demonstrated in this study.
Figure 4. Proposed mechanism of the PTC oxidation reaction
based on hostꢀguest complexation.
(18%, run 11). CH2Cl2 was included in the cavity of
pillar[5]arenes, which was completely characterized by
single X-ray crystalline analysis.14 Thus, the hostꢀguest
complexation between1 and 1-hexene extremely weakened
in CD2Cl2 because CH2Cl2 acts as a competitive guest.
The percentage conversion in the oxidation reaction of
4-methyl-1-hexene (run 5) was 31%, which was lower than
that of 1-hexene (run 2). This is because branched
4-methyl-1-hexene cannot be included in the pillar[5]arene
cavity of 1, which is confirmed by 1H NMR measurements
(Figure 2b). We also carried out the oxidation of the
mixture of 1-hexene and 4-metyhl-1-hexene using 1 (run 6).
The conversion of 1-hexene to 1-pentanal was 85%, while
that of 4-methyl-1-hexene to 3-methyl-1-pentanal was 7%.
Thus, the substrate selectivity of 1 based on the conversion
was high (∼12:1). These results indicate that inclusion of the
substrate in the cavity of 1 is important for the oxidation of
alkenes.
Figure 4 shows the proposed mechanism of the oxidation
reaction using 1 as the PTC. Equilibrium counteranion
exchange from bromide to MnO4ꢀ occurs first. These ion
pairs are then transferred across the aqueousꢀorganic
ꢀ
interface. In the organic phase, the MnO4 anion reacts
Acknowledgment. This work was partly supported by
The Japan Securities Scholarship Foundation and Tech-
nology (MEXT), Japan and JSPS KAKENHI 23655210
(a Grant-in-Aid for Challenging Exploratory Research).
with 1-hexene. At this stage, 1-hexene within the cavity of 1
is fixed in a close position for efficient oxidation. When
1-heptene was used as a substrate, the oxidation reac-
tion proceeded completely (conversion >99%, run 7). By
contrast, the conversion of the oxidation of the long
chain alkenes were low (runs 8 (14%) and 9 (9%)). The
distance between the CdC bond and the alkylphospho-
nium bromide groups is very important for the oxidation.
Supporting Information Available. Experimental sec-
tion, characterization data, Job plots, and 1H NMR
titrations. This material is available free of charge via
the Internet at http://pubs.acs.org.
(14) Boinski, T.; Szumna, A. Tetrahedron 2012, 68, 9419–9422.
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 14, 2013
3745