Synthesis of novel dendritic 2,2′-bipyridine ligands and their application to lewis acid-catalyzed diels-alder and three-component condensation reactions

Article abstract of DOI:10.1021/jo070767a

(Chemical Equation Presented) A series of dendritic ligands with a 2,2′-bipyridine core was synthesized through the coupling of 4,4′-dihydroxy-2,2′-bipyridine with poly(aryl ether) dendrons. The corresponding dendritic Cu(OTf)₂ catalysts were used for Diels-Alder and three-component condensation reactions. The dendritic Cu(OTf) ₂-catalyzed Diels-Alder reaction proceeded smoothly, and these dendritic catalysts could be recycled without deactivation by reprecipitation. Three-component condensation reactions such as Mannich-type reactions also proceeded not only in dichloromethane but also in water. Furthermore, a positive dendritic effect on chemical yields was observed in both Diels-Alder reactions and aqueous-media three-component condensation reactions.

Full text of DOI:10.1021/jo070767a

Synthesis of Novel Dendritic 2,2′-Bipyridine Ligands and Their
Application to Lewis Acid-Catalyzed Diels-Alder and
Three-Component Condensation Reactions
Takahito Muraki, Ken-ichi Fujita,* and Masato Kujime
National Institute of AdVanced Industrial Science and Technology (AIST), AIST Tsukuba Central 5, 1-1-1,
Higashi, Tsukuba, Ibaraki 305-8565, Japan
k.fujita@aist.go.jp
ReceiVed April 12, 2007
A series of dendritic ligands with a 2,2′-bipyridine core was synthesized through the coupling of 4,4′-
dihydroxy-2,2′-bipyridine with poly(aryl ether) dendrons. The corresponding dendritic Cu(OTf)2 catalysts
were used for Diels-Alder and three-component condensation reactions. The dendritic Cu(OTf)2-catalyzed
Diels-Alder reaction proceeded smoothly, and these dendritic catalysts could be recycled without
deactivation by reprecipitation. Three-component condensation reactions such as Mannich-type reactions
also proceeded not only in dichloromethane but also in water. Furthermore, a positive dendritic effect on
chemical yields was observed in both Diels-Alder reactions and aqueous-media three-component
condensation reactions.
Introduction
powerful tools for synthetic organic chemistry. Furthermore,
the development of a recyclable catalyst has also been required.
One of the most certain methods of recycling a catalytic metal
is its immobilization on a macromolecular ligand.6 Polystyrene
beads containing Lewis acids are one of the better known
modalities for recyclable Lewis acid catalysts.7 However, they
are heterogeneous catalysts, and their reactivity is relatively low.
Therefore, novel synthetic macromolecular ligands, which have
good solubility for various solvents as well as recyclability, are
required.
In contemporary organic synthesis, Lewis acid-catalyzed
carbon-carbon bond formation reactions are among the most
important; therefore, more powerful and versatile Lewis acid
catalysts have been required.1 During the past two decades,
metal trifluoromethanesulfonates (triflates) such as Sc(OTf)3,2
Yb(OTf)3,3 and Cu(OTf)2,4 have been developed, because they
have excellent Lewis acidity. Furthermore, they also work well
even in aqueous media.5 From the perspectives of green
chemistry, these water-compatible Lewis acid catalysts are very
(5) (a) Li, C.-J.; Chen, L. Chem. Soc. ReV. 2006, 35, 68-82. (b) Li,
C.-J. Chem. ReV. 2005, 105, 3095-3165. (c) Manabe, K.; Kobayashi, S.
Chem. Eur. J. 2002, 8, 4095-4101. (d) Kobayashi, S.; Manabe, K. Acc.
Chem. Res. 2002, 35, 209-217.
* Author to whom correspondence should be addressed. Fax:
81-29-861-4577.
(1) Yamamoto, H., Ed. Lewis Acids in Organic Synthesis; Wiley-VCH:
Weinheim, 2000.
(2) Kobayashi, S.; Hachiya, I.; Araki, M.; Ishitani, H. Tetrahedrom Lett.
1993, 34, 3755-3758.
(6) Buchmeiser, M. R., Ed. Polymeric Materials in Organic Synthesis
and Catalysis; Wiley-VCH: Weinheim, 2003.
(3) Kobayashi, S.; Hachiya, I.; Takahori, T.; Araki, M.; Ishitani, H.
Tetrahedrom Lett. 1992, 33, 6815-6818.
(7) (a) Fujita, K.; Muraki, T.; Hashimoto, S.; Oishi, A.; Taguchi, Y. Bull.
Chem. Soc. Jpn. 2004, 77, 2097-2098. (b) Altava, B.; Burguete, M. I.;
Escuder, B.; Luis, S. V.; Salvador, R. V. J. Org. Chem. 1997, 62, 3126-
3134.
(4) (a) Evans, D. A.; Miller, S. J.; Lectka, T. J. Am. Chem. Soc. 1993,
115, 6460-6461. (b) Hertweck, C. J. Prakt. Chem. 2000, 342, 316-321.
10.1021/jo070767a CCC: $37.00 © 2007 American Chemical Society
Published on Web 09/21/2007
J. Org. Chem. 2007, 72, 7863-7870
7863

Muraki et al.
Dendrimers are fascinating macromolecules due to their
unique physical and chemical properties, which are caused by
their well-defined hyperbranched framework.8 Metallodendri-
mers, with catalytic sites immobilized within the dendrimer, are
particularly useful as synthetic catalysts not only because of
their unique reactivity and fair solubility but also because of
their recyclability by reprecipitation or by nanofiltration based
on their nano-order size.9 Presently, metallodendrimers with a
catalytic site at the core have received considerable attention
because of their unique selectivity and reactivity caused by
specific reaction fields constructed by the dendron.10 Some of
them show a positive dendritic effect on regio- or stereoselec-
tivity, meaning that their selectivity is enhanced by an increase
in the generation number of the dendrimer.11,12 However,
examples of a positive dendritic effect on reactivity are at present
limited.13,14
We report herein the simple synthesis of novel dendritic 2,2′-
bipyridine core ligands, the preparation of the corresponding
dendritic copper(II) triflate (Cu(OTf)2) and their application to
the Diels-Alder reaction and three-component condensation
reactions such as the Mannich-type reaction, and the positive
dendritic effect on chemical yields. A dendritic effect was
observed in aqueous-media reactions, especially in the case of
three-component condensation reactions. We found that in both
reactions these dendritic Cu(OTf)2 catalysts were recyclable by
reprecipitation.
FIGURE 1. Structual formulas of Gn[R].
Results and Discussion
periphery groups R was used as a dendron (Figure 1).15 This is
because Gn[R] has a perfect, defect-free dendritic structure. We
noticed that bidentate ligands, which coordinate with nitrogen,
are the core units of the dendritic ligand. This is because they
can coordinate the various transition metals and ligate them more
strongly than monodentate ligands. We chose 4,4′-dihydroxy-
2,2′-bipyridine as the core unit of the dendritic ligands.16 The
2,2′-bipyridine unit has fair coordination ability in relation to
various transition metals, and the introduction of dendrons to
the 2,2′-bipyridine unit is easy because it has a phenolic
hydroxyl group. We expected that defect-free rigid dendrons
and strong bidentate coordination would enable us to effectively
construct the desired reaction field around the catalytic metal.
Dendritic 2,2′-bipyridine ligands 3(Gn[R]) were prepared
from 4,4′-dihydroxy-2,2′-bipyridine 1 and the corresponding
poly(aryl ether) dendritic benzyl bromide 2(Gn[R]) in the
presence of potassium carbonate and a catalytic amount of 18-
crown-6 in tetrahydrofuran. Chemical yields of 3(Gn[R]) are
shown in Table 1.14a
Preparation of Dendritic Ligands Having Bipyridine Core
3(Gn[R]). In the chemistry of dendritic catalysts, one of the
most attractive goals is the construction of a specific reaction
field for organic synthesis. To construct it effectively, Gn[R]
having a poly(aryl ether) skeleton and one of a variety of
(8) (a) Fre´chet, J. M. J.; Tomalia, D. A., Eds. Dendrimers and Other
Dendritic Polymers; John Wiley & Sons: New York, 2001. (b) Newkome,
G. R.; Moorfield, C. N.; Vo¨gtle, F. Dendrimers and Dendrons; Wiley-
VCH: Weinheim, 2001.
(9) (a) Gade, L. H., Ed. Dendrimer Catalyst; Springer-Verlag: Berlin,
Heidelberg, 2006. (b) Me´ry, D.; Astruc, D. Coord. Chem. ReV. 2006, 250,
1965-1979. (c) Mu¨ller, C.; Nijkamp, M. G.; Vogt, D. Eur. J. Inorg. Chem.
2005, 4011-4021. (d) Astruc, D.; Heuze´, K.; Gatard, S.; Me´ry, D.; Nlate,
S.; Plault, L. AdV. Synth. Catal. 2005, 347, 329-338. (e) Reek, J. N. H.;
de Groot, D.; Oosterom, G. E.; Kamer, P. C. J.; van Leeuwen, P. W. N. M.
C. R. Chem. 2003, 1061-1077. (f) Dijkstra, H. P.; van Klink, G. P. M.;
van Koten, G. Acc. Chem. Res. 2002, 35, 798-810.
(10) Twyman, L. J.; King, A. S. H.; Martin, I. M. Chem. Soc. ReV. 2002,
31, 69-82.
(11) Review of the dendritic effect: (a) Helms, B.; Fre´chet, J. M. J. AdV.
Synth. Catal. 2006, 348, 1125-1148. (b) Chow, H.-F.; Leung, C.-F.: Wang,
G.-X, Yang, Y.-Y. C. R. Chem. 2003, 6, 735-745.
(12) Examples of dendritic effect on selectivity: (a) Benito, J. M.; de
Jesu´s, E.; de la Mata, F. J.; Flores, J. C.; Go´mez, R. Chem. Commun. 2005,
5217-5219. (b) Fujita, K.; Muraki, T.; Sakurai, T.; Oishi, A.; Taguchi, Y.
Chem. Lett. 2005, 34, 1180-1181. (c) Muraki, T.; Fujita, K.; Oishi, A.;
Taguchi, Y. Polym. J. 2005, 37, 847-853. (d) Ribourdouille, Y.; Engel,
G. D.; Richard-Plouet, M.; Gade, L. H. Chem. Commun. 2003, 1228-1229.
(e) Mizugaki, T.; Murata, M.; Ooe, M.; Ebitani, K.; Kaneda, K. Chem.
Commun. 2002, 52-53. (f) Bhyrappa, P.; Young, J. K.; Moore, J. S.;
Suslick, K. S. J. Am. Chem. Soc. 1996, 118, 5708-5711.
All of the dendritic ligands 3(Gn[R]), except for 3(G1[Me])
(Table 1, entry 1), were obtained in fair yields. The reason for
the poor yield of entry 1 was probably due to the nucleophilic
attack of the nitrogen of bipyridine on 1, owing to the small
size of the G1[Me] dendron.
(15) (a) Tomoyose, Y.; Jiang, D.-L.; Jin, R.-H.; Aida, T.; Yamashita,
T.; Horie, K.; Yashima, E.; Okamoto, Y. Macromolecules 1996, 29, 5236-
5238. (b) Hawker, C. J.; Fre´chet, J. M. J. J. Am. Chem. Soc. 1990, 112,
7638-7647. (c) Brewis, M.; Helliwell, M.; Mckeown, N. B. Tetrahedron
2003, 59, 3863-3872.
(16) Previously reported 2,2′-bipyridine core dendrimers: (a) Hong, Y.-
R.; Gorman, C. B. J. Org. Chem. 2003, 68, 9019-9025. (b) Roy, R.; Kim,
J. M. Tetrahedron 2003, 59, 3881-3893. (c) Tyson, D. S.; Luman, C. R.;
Castellano, F. N. Inorg. Chem. 2002, 41, 3578-3586. (d) Ghaddar, T. H.;
Wishart, J. F.; Kirby, J. P.; Whitesell, J. K.; Fox, M. A. J. Am. Chem. Soc.
2001, 123, 12832-12836.
(13) Examples of dendritic effect on reactivity: (a) Helms, B.; Liang,
C. O.; Hawker, C. J.; Fre´chet, J. M. J. Macromolecules 2005, 38, 5411-
5415. (b) Fujiwara, T.; Obora, Y.; Tokunaga, M.; Sato, H.; Tsuji, Y. Chem.
Commun. 2005, 4526-4528. (c) Dahan, A.; Portnoy, M. Org. Lett. 2003,
5, 1197-1200. (d) Fan, Q.-H.; Chen, Y.-M.; Chen, X.-M.; Jiang, D.-Z.;
Xi, F.; Chan, A. S. C. Chem. Commun. 2000, 789-790.
(14) Our recent reports on the dendritic effect on reactivity: (a) Fujita,
K.; Muraki, T.; Hattori, H.; Sakakura, T. Tetrahedron Lett. 2006, 47, 4831-
4843. (b) Muraki, T.; Fujita, K.; Terakado, D. Synlett 2006, 2646-2648.
7864 J. Org. Chem., Vol. 72, No. 21, 2007

Dendritic 2,2′-Bipyridine Ligands
TABLE 1. Preparation of 3(Gn[R])
TABLE 2. Diels-Alder Reaction with 3(G2[Me])-Cu(OTf)2 as the
Catalyst
entry
R
Gn
yield (%)
1
2
3
4
5
6
7
Me
Me
Me
Bn
Bn
Bn
G1
G2
G3
G1
G2
G3
G1
8
86
86
83
80
85
91
TEG
In the case of the preparation of 3(G1[Me]), it was found
that
1 was coupled with 3,5-dimethoxybenzyl alcohol
(G1[Me]-OH) 4 under Mitsunobu conditions17 to provide
3(G1[Me]) in a 50% yield (Scheme 1).
SCHEME 1. Preparation of 3(G1[Me])
Diels-Alder Reaction with Dendritic Cu(OTf)2 Catalyst.
In this study, Cu(OTf)2 was selected as the Lewis acid catalyst,
because this compound has a fair Lewis acidity not only in
organic solvents but also in aqueous media,4b and furthermore,
Cu(OTf)2 has better azaphilicity18 than other Lewis acids such
as aluminum chloride and scandium triflate. We examined the
utility of the dendritic 2,2′-bipyridine-Cu(OTf)2 core catalyst
by performing Lewis acid-catalyzed Diels-Alder reactions
(Table 2).19
First, by employing 10 mol % of the 3(G2[Me])-Cu(OTf)2
catalyst, which was prepared in situ in dichloromethane, Diels-
Alder reactions of cyclopentadiene with various dienophiles
proceeded at room temperature to afford the corresponding
adducts 5 in excellent yields (Table 2, entries 1-3). In these
reactions, the endo/exo ratios were almost similar to those
previously reported.2,20
1
a Determined by H NMR spectra.
On the other hand, when a dichloromethane solution of cyclo-
pentadiene and various dienophiles at room temperature was
stirred in the presence of 10 mol % Cu(OTf)2, which was not
immobilized on the dendritic ligand, the corresponding Diels-
Alder adduct was not obtained. Cu(OTf)2 promoted the cationic
polymerization of cyclopentadiene. Furthermore, even by em-
ploying the modified Cu(OTf)2 catalyst, which was coordinated
by nondendritic 4,4′-dimethoxy-2,2′-bipyridine () 3(G0[Me])),
the corresponding Diels-Alder adduct was obtained in only a
1% yield. Thus, the immobilization of Cu(OTf)2 on the
dendrimer is effective for the modulation of its Lewis acidity.
Encouraged by these results, we applied the dendritic
3(G2[Me])-Cu(OTf)2 catalyst to other dienes such as cyclo-
hexadiene and 2,3-dimethyl-1,3-butadiene; in these cases, only
the corresponding Diels-Alder reaction proceeded smoothly
(Table 2 entries 4-7).
(17) (a) Freeman, A. W.; Chrisstoffels, L. A. J.; Fre´chet, J. M. J. J. Org.
Chem. 2000, 65, 7612-7617. (b) Forier, B.; Dahaen, W. Tetrahedron 1999,
55, 9829-9846.
Recently, it has been discovered that the reactivities of
dendritic catalysts depend on the generation number of the
dendrimers; this unique phenomenon has been called a “dendritic
effect”.11 First, we performed a Diels-Alder reaction of
cyclopentadiene and 3-crotonyl-2-oxazolidinone, which had the
same substrates as entry 1 in Table 2, with 10 mol % of various-
generation dendritic catalysts 3(Gn[R])-Cu(OTf)2 to confirm the
dendritic effect on chemical yields (Table 3).
The second generation 3(G2[Me])-Cu(OTf)2 afforded a better
yield of the Diels-Alder adduct 5a than the first generation
3(G1[Me])-Cu(OTf)2 (Table 3, entries 1 and 2); however, the
yield of 5a in the case of 3(G3[Me])-Cu(OTf)2 was lower than
(18) Johnson, J. S.; Evans, D. A. Acc. Chem. Res. 2000, 33, 325-335.
(19) Previously reported Cu(OTf)2-catalyzed Diels-Alder reaction: (a)
Sagasser, I.; Helmchen, G. Tetrahedron Lett. 1998, 39, 261-264. (b) Brunel,
J. M.; Campo, B. D.; Buono, G. Tetrahedron Lett. 1998, 39, 9663-9666.
(c) Carbone, P.; Desimoni, G.; Faita, G.; Filippone, S.; Righetti, P.
Tetrahedron 1998, 54, 6099-6110.
(20) (a) Evans, D. A.; Barnes, D. M.; Johnson, J. S.; Lectka, T.; von
Matt, P.; Miller, S. J.; Murry, J. A.; Norcross, R. D.; Shaughnessy, E. A.;
Campos, K. R. J. Am. Chem. Soc. 1999, 121, 7582-7594. (b) Kobayashi,
S.; Ishitani, H.; Hachiya, I.; Araki, M. Tetrahedron 1994, 50, 11623-11636.
(c) Kobayashi, S.; Hachiya, I.; Araki, M.; Ishitani, H. Tetrahedron Lett.
1993, 34, 3755-3758. (d) Narasaka, K.; Iwasawa, N.; Inoue, M.; Yamada,
T.; Nakashima, M.; Sugimori, J. J. Am. Chem. Soc. 1989, 111, 5340-5345.
(e) Allen, C. F. H.; Bell, A. Org. Synth. 1942, 22, 37-38. (f) Diels, O.;
Alder, K.; Stein, G. Ber. Deutsch. Chem. Ges. 1929, 62, 2337-2372.
J. Org. Chem, Vol. 72, No. 21, 2007 7865

Muraki et al.
TABLE 3. Diels-Alder Reaction of Cyclopentadiene with
3(Gn[R])-Cu(OTf)2 as the Catalyst
TABLE 4. Diels-Alder Reaction of Cyclohexadiene with
3(Gn[R])-Cu(OTf)2 as the Catalyst
entry
R
Gn
yield (%)
entry
R
Gn
solvent
yield (%)
1
2
3
4
5
6
Me
Me
Me
Bn
Bn
Bn
G1
G2
G3
G1
G2
G3
5
56
80
77
87
94
1
2
3
4
5
6
7
Me
Me
Me
Me
Bn
Bn
Bn
G1
G2
G3
G3
G1
G2
G2
CH2Cl2
CH2Cl2
CH2Cl2
PhCH3
CH2Cl2
CH2Cl2
PhCH3
20
96
58
85
45
57
93
On the other hand, the negative dendritic effect on reaction
rate in the Diels-Alder reaction caused by employing the
dendritic bis(oxazoline)-Cu(OTf)2 catalyst has been reported by
Chow’s group.21 They reported that the reason of their negative
dendritic effect was due to the steric hindrance of its bulky
dendritic skeleton. We assume that our profound dendritic effect
is probably derived from the increase of the Lewis acidity due
to the distorted bipyridine skeleton of the 3(Gn[R])-Cu(OTf)2
complex by the steric repulsion of the dendritic wedge,22 thus
affording better chemical yield while increasing the generation
number of the dendrimer. Exactly, by a comparison of the
chemical yields in the same generation, 3(Gn[Bn]), which
possesses bulkier groups at its periphery, affords better chemical
yields than 3(Gn[Me]) in all generations.
that in the case of 3(G2[Me])-Cu(OTf)2 (Table 3, entries 2 and
3). In entry 3, the polymerization product of cyclopentadiene
was also obtained. The reason for the low yield in entry 3 was
probably that 3(G3[Me])-Cu(OTf)2 catalyzed not only the
Diels-Alder reaction but also the cationic polymerization of
cyclopentadiene. To retard the cationic polymerization of
cyclopentadiene, the solvent was changed from dichloromethane
to toluene. As a result, the Diels-Alder adduct 5a was obtained
in an 85% yield without the polymerization of cyclopentadiene
(Table 3, entry 4). In the case of having a peripheral benzyl
substituent, the first generation 3(G1[Bn])-Cu(OTf)2 gave 5a
in a 45% yield exclusively, without the polymerization of
cyclopentadiene. On the other hand, the second generation
3(G2[Bn])-Cu(OTf)2 catalyzed not only the Diels-Alder reac-
tion but also the cationic polymerization of cyclopentadiene.
Then, by changing the solvent from dichloromethane to toluene,
again using 3(G2[Bn])-Cu(OTf)2, the chemical yield of 5a was
significantly enhanced, in a way similar to 3(G3[Me])-Cu(OTf)2
(Table 3, entries 6 and 7). These results show that a reactive
diene such as cyclopentadiene is easily polymerized by a higher
generation 3(Gn[R])-Cu(OTf)2 catalyst. Therefore, in Table 3,
we could not confirm the dendritic effect on the chemical yields
of the Diels-Alder adduct 5a.
Since dendrimers have a polymer-like character, the repre-
cipitation of dendrimers is often used to recover them.23 In other
groups’ reports, several dendritic catalysts were recovered by
reprecipitation and subsequently reused.13c,24 In this study, we
examined the recycling of the dendritic catalyst in Diels-Alder
TABLE 5. Catalyst Recycling in Diels-Alder Reactions
Generally, cyclohexadiene is less reactive than cyclopenta-
diene for cationic polymerization. By employing cyclohexadiene
and 3-crotonyl-2-oxazolidinone with 3(G2[Me])-Cu(OTf)2 as
a catalyst, even the corresponding Diels-Alder reaction did not
proceed. We next performed the Diels-Alder reaction of
cyclohexadiene and 3-acryloyl-2-oxazolidinone, which was more
reactive than crotonyl derivatives, with 10 mol % of the various-
generation dendritic catalysts 3(Gn[R])-Cu(OTf)2. All reactions
were carried out in dichloromethane for 17 h at room temper-
ature (Table 4).
yield (%)
Gn
-
firsta
second
third
fourth
G2b
G3c
88
94
98
94
99
99
95
99
88
99
As a result, only the corresponding Diels-Alder reaction
proceeded, and the cationic polymerization did not proceed.
Furthermore, the chemical yield of the Diels-Alder adduct 5d
was enhanced by increasing the generation number of the
dendritic catalysts 3(Gn[R])-Cu(OTf)2 in the case of both
3(Gn[Me]) and 3(Gn[Bn]).14a This relationship between the
generation number of the dendritic catalyst and the chemical
yield is one of the positive dendritic effects.11,13
a Number of recycle. b Carried out in CH2Cl2. c Carried out in toluene.
(21) (a) Chow, H.-F.; Mak, C. C. J. Org. Chem. 1997, 62, 5116-5127.
(b) Mak, C. C.; Chow, H.-F. Macromolecules 1997, 30, 1228-1230.
(22) It has been reported that Lewis acid complexes having a distorted
skeleton afforded better Lewis acidity: (a) Brunner, H.; Bluchel, C.; Doyle,
M. P. J. Organomet. Chem. 1997, 541, 89-95. (b) Koide, Y.; Bott, S. G.;
Barron, A. R. Organometallics 1996, 15, 5514-5518.
7866 J. Org. Chem., Vol. 72, No. 21, 2007

Dendritic 2,2′-Bipyridine Ligands
TABLE 6. Three-Component Condensation Reaction in CH2Cl2
entry
Gn
R1
nucleophile
X
product
yield (%)
1
2
3
4
5
6
7
8
9
10
G1
G2
G3
G2
G2
G2
G2
G2
G2
G3
Ph
Ph
Ph
CH2dC(OTMS)Ph 6
6
6
6
CMe2dC(OTMS)OMe 7
7
P(OEt)3 8
8
8
8
CH2COPh
CH2COPh
CH2COPh
CH2COPh
CMe2CO2Me
CMe2CO2Me
PO(OEt)2
PO(OEt)2
PO(OEt)2
PO(OEt)2
9a
9a
9a
9b
10a
10b
11a
11b
11c
11c
31
47
49
38
34
36
63
53
72
73
PhCH2CH2
Ph
PhCH2CH2
Ph
PhCH2
PhCH2CH2
PhCH2CH2
reactions of cyclopentadiene and acryloyl derivatives with 10
mol % of the 3(Gn[Me])-Cu(OTf)2 catalyst. After the reaction
had proceeded for 30 min, the reaction mixture was poured
into hexane and was stirred to precipitate the dendritic
catalyst 3(Gn[Me])-Cu(OTf)2, which was collected by filtration
and was subsequently reused. Both the second- and the third-
generation dendritic Cu(OTf)2 catalysts could be used five
times without deactivation (Table 5). However the precipitated
dendritic catalyst included a small amount of the Diels-Alder
adduct 5c, and yields were slightly lower when virgin
catalysts, which had not been recycled, were used (G2: 88%,
G3: 94%).
rather low yields (Table 6, entries 1-6). In entries 1-6, the
corresponding aldol product was not obtained. The low yields
of 9 and 10 were due to the decomposition of silyl enolates 6
and 7 to afford the corresponding carbonyl compounds catalyzed
by 3(Gn[Me])-Cu(OTf)2. Contrary to our expectations, no
remarkable dendritic effect on chemical yields was observed
(Table 6, entries 1-3).
Next, by employing triethyl phosphite 8 as a nucleophile,
the corresponding R-aminophosphonate 11 was obtained in fair
yields (Table 6, entries 7-10).27 Also in the case of 8, no
dendritic effect on chemical yields was observed (Table 6 entries
9 and 10).
Three-Component Condensation Reaction with Dendritic
Cu(OTf)2 Catalyst. Lewis acid-catalyzed condensation reac-
tions, such as aldol and Mannich-type reactions, have been very
important transformations for organic synthesis. Today, aqueous-
media condensation reactions have attracted a great deal of
attention not only because water is an ideal solvent from the
perspective of safety, but also because unique reactivities are
often observed in aqueous media.25
In the various Lewis acid-catalyzed condensation reactions,
three-component condensation reactions such as Mannich-type
reactions are basic and useful transformations for the synthesis
of â-carbonyl amines and R-aminophosphonates, which are
important synthetic intermediates for various pharmaceuticals
and natural products.26 Thus, we employed dendritic Cu(OTf)2
catalysts to perform these three-component condensation reac-
tions not only in dichloromethane but also in water.
In this transformation, we tried the catalyst recycling experi-
ment. 3-Phenylpropanal, o-anisidine, and triethyl phosphite,
which were the same substrates as entries 9 and 10 in Table 6,
were stirred for 2 h in the presence of 3(G2[Me])-Cu(OTf)2 or
3(G3[Me])-Cu(OTf)2 as a catalyst in dichloromethane (Table
7). After the reaction was completed, the reaction mixture was
poured into hexane. The dendritic Cu(OTf)2 catalyst was
recovered as a tar-like oil, and subsequently reused. Both the
second- and third-generation dendritic catalysts were used five
times without deactivation.
TABLE 7. Catalyst Recycling in Three-Component Condensation
Reactions
According to Table 6, three-component condensation reac-
tions in dichloromethane were carried out for 2 h by use of
aldehyde, o-anisidine, and nucleophiles 6-8, with 10 mol %
of various generations of the dendritic catalyst 3(Gn[Me])-
Cu(OTf)2.
yield (%)
First, by employing silyl enolates 6 and 7 as nucleophiles,
the corresponding Mannich-type products were obtained in
Gn
-
firsta
second
third
fourth
G2
G3
0
63
63
71
68
66
66
70
68
67
(23) (a) Yamazaki, N.; Washio, I.; Shibasaki, Y.; Ueda, M. Org. Lett.
2006, 8, 2321-2324. (b) Catalano, V. J.; Parodi, N. Inorg. Chem. 1997,
36, 537-541.
a Number of recycle.
(24) (a) Ji, B.; Yuan, Y.; Ding, K.; Meng, J. Chem. Eur. J. 2003, 9,
5989-5996. (b) Yang, B.-Y.; Chen, X.-M.; Deng, G.-J.; Zhang, Y.-L.; Fan,
Q.-H. Tetrahedron Lett. 2003, 44, 3535-3538.
Recently, novel Lewis acid catalysts such as surfactant-
supported scandium triflate and polymer-micelle incarcerated
(25) (a) Azizi, N.; Torkiyan, L.; Saidi, M. R. Org. Lett. 2006, 8, 2079-
2082. (b) Loncaric, C.; Manabe, K.; Kobayashi, S. AdV. Synth. Catal. 2003,
345, 1187-1189. (c) Iimura, S.; Nobutou, D.; Manabe, K.; Kobayashi, S.
Chem. Commun. 2003, 1644-1645.
(26) (a) Ollevier, T.; Nadeau, E. Synlett 2006, 219-222. (b) Co´rdova,
A. Acc. Chem. Res. 2004, 37, 102-112. (c) Arend, M.; Westermann, B.;
Risch, N. Angew. Chem., Int. Ed. 1998, 37, 1044-1070.
(27) (a) Lee, S.-g.; Park, J. H.; Kang, J. Lee, J. K. Chem. Commun. 2001,
1698-1699. (b) Manabe, K.; Kobayashi, S. Chem. Commun. 2000, 669-
670. (c) Ranu, B. C.; Hajra, A.; Jana, U. Org. Lett. 1999, 1, 1141-1143.
(d) Qian, C.; Huang, T. J. Org. Chem. 1998, 63, 4125-4128. (e) Kukhar,
V. P.; Solodenko, V. A. Russ. Chem. ReV. 1987, 56, 859-874.
J. Org. Chem, Vol. 72, No. 21, 2007 7867

Muraki et al.
TABLE 9. Three-Component Condensation Reactions in Water
scandium triflate, which promoted aqueous-phase three-
component condensation reactions such as a Mannich-type
reaction were reported.28 Then, we applied the dendritic catalysts
3(Gn[R])-Cu(OTf)2 to an aqueous-media Mannich-type reaction.
First, in order to examine the catalytic activity in water and
the solubility of the dendritic Cu(OTf)2 catalysts, we employed
various peripheral group-containing dendritic catalysts 3(Gn[R])-
Cu(OTf)2 (R ) Me, Bn, TEG) in aqueous-media Mannich-type
reactions (Table 8).
entry
R1
nucleophile
X
product yield (%)
1
2
3
4
5
6
7
Ph
CH2dC(OTMS)Ph 6
6
CMe2dC(OTMS)OMe 7 CMe2CO2Me
7
P(OEt)3 8
8
8
CH2COPh
CH2COPh
9a
9b
53
51
78
97
61
83
99
PhCH2CH2
Ph
PhCH2CH2
Ph
PhCH2
PhCH2CH2
10a
10b
11a
11b
11c
CMe2CO2Me
PO(OEt)2
PO(OEt)2
PO(OEt)2
TABLE 8. Reactivity of Various Cu(OTf)2 Catalysts in Water
The high yields obtained in water are probably due to the
hydrophobic effect owing to the cohesion of organic substrates
in water.30 However, the yields of â-amino ketones 9 from silyl
enolate 6 were not so high (Table 9, 53% for entry 1 and 51%
for entry 2).
We subsequently performed these Mannich-type reactions
with a silyl enolate 6 using 10 mol % of various generations of
the dendritic catalyst 3(Gn[Me])-Cu(OTf)2; these reactions were
carried out in water for 17 h (Table 10). For both benzaldehyde
and 3-phenylpropanal, the chemical yields of the Mannich
products 9 were enhanced by increasing the generation numbers
of the dendritic Cu(OTf)2 catalysts. This relationship between
the generation number of the dendritic catalyst and the chemical
yield shows a positive dendritic effect.14b
entry
3(Gn[R])
yield (%)
1
2
3
4
5
nonea
23
18
47
48
32
G0[Me]
G1[Me]
G1[Bn]
G1[TEG]
a Only Cu(OTf)2 was used.
TABLE 10. Dendritic Effect on Chemical Yields of 9
Benzaldehyde, o-anisidine, and silyl enolate 6 were treated
with 10 mol % of 3(G1[R])-Cu(OTf)2 catalysts in water, and
the reaction mixture was stirred at room temperature for 17 h.
3(G1[Me])-Cu(OTf)2 and 3(G1[Bn])-Cu(OTf)2, which were
hardly dissolved in water to afford a dispersion mixture, gave
better chemical yields of 9a than Cu(OTf)2 alone, which was
easily soluble in water (Table 8, entries 1, 3, and 4). In contrast,
the nondendritic catalyst 3(G0[Me])-Cu(OTf)2 was not effective
for this Mannich-type reaction in water (Table 8, entry 2).
Meanwhile, although 3(G1[TEG])-Cu(OTf)2 showed good
solubility in water, the yield was rather low, contrary to our
expectations (Table 8, entry 5). This low yield may be caused
by reduction in the Lewis acidity due to coordination of the
ethylene glycol unit to a copper atom. The results in Table 8
suggest that dendritic Cu(OTf)2 catalysts containing hydrophobic
dendrons with peripheral alkyl groups are effective for carrying
out Mannich-type reactions in water.29
Then, the second-generation dendritic catalyst 3(G2[Me])-
Cu(OTf)2, which had peripheral methyl substituents, was used
in various aqueous-media three-component condensation reac-
tions with several aldehydes and nucleophiles 6-8 (Table 9).
Interestingly, compared with the chemical yields in Table 6,
which were carried out in dichloromethane, almost all substrates
in Table 9 gave better yields. When using ketene silyl acetal 7
or triethyl phosphite 8 as the nucleophile, the corresponding
three-component condensation reaction proceeded smoothly in
water in fair yields (Table 9, entries 3-7). In particular,
3-phenylpropanal gave the corresponding products 10b, 11c in
97% and 99% yields, respectively (Table 9, entries 4 and 7).
entry
R1
Gn
9
yield (%)
1
2
3
4
5
6
7
8
Ph
Ph
Ph
Ph
G0
G1
G2
G3
G0
G1
G2
G3
9a
9a
9a
9a
9b
9b
9b
9b
18
47
53
61
37
47
51
61
PhCH2CH2
PhCH2CH2
PhCH2CH2
PhCH2CH2
On the basis of these results, we have hypothesized that the
hydrophobic reaction field constructed by the dendritic Cu(OTf)2
catalyst is essential to increasing the chemical yield of organic
synthesis in water.
We finally confirmed the effects of the hydrophobic dendron
in the dendritic Cu(OTf)2 catalysts. According to Scheme 2, a
Mannich-type reaction was carried out in water by employing
10 mol % of the nondendritic catalyst 3(G0[Me])-Cu(OTf)2 in
the presence of 20 mol % G2[Me]-OPh 12, which possessed a
hydrophobic dendritic group. The yield of the Mannich product
9a was only 16%;31 thus, the addition of 12 was not effective
in increasing the chemical yield of 9a. It can be concluded that
(28) (a) Manabe, K.; Mori, Y.; Wakabayashi, T.; Nagayama, S.;
Kobayashi, S. J. Am. Chem. Soc. 2000, 122, 7202-7207. (b) Takeuchi,
M.; Akiyama, R.; Kobayashi, S. J. Am. Chem. Soc. 2005, 127, 13096-
13097.
(29) The dendritic ligand 3G1[Bn] did not catalyze an aqueous-media
Mannich-type reaction.
(30) (a) Blokzijl, W.; Engberts, J. B. F. N. Angew. Chem., Int. Ed. Engl.
1993, 32, 1545-1579. (b) Breslow, R. Acc. Chem. Res. 1991, 24, 164-
170.
7868 J. Org. Chem., Vol. 72, No. 21, 2007

Dendritic 2,2′-Bipyridine Ligands
SCHEME 2. Mannich-Type Reaction in Water by
Employing Nondendritic Cu(OTf)2 Catalyst in the Presence
of 12
Cl2) 3001, 2937, 2838, 1597, 1458, 1374, 1296, 1230, 1204, 1155,
1
1055, 833 cm-1; H NMR (CDCl3) δ (ppm) 8.47 (d, J ) 5.6 Hz,
2H), 8.06 (d, J ) 2.7 Hz, 2H), 6.88 (dd, J ) 5.6, 2.7 Hz, 2H), 6.69
(d, J ) 2.2 Hz, 4H), 6.59-6.57 (m, 10H), 6.41 (t, J ) 2.2 Hz,
4H), 5.15 (s, 4H), 4.99 (s, 8H), 3.79 (s, 24H); 13C NMR (CDCl3)
δ (ppm) 165.7, 161.0, 160.1, 157.8, 150.2, 139.0, 138.2, 111.4,
107.1, 106.4, 105.2, 101.9, 100.0, 70.1, 69.7, 55.3; FAB MS for
C60H61N2O14 m/z: Calcd: 1033.4 [(M + H)+]; Found: 1033.5;
Anal. Calcd for C60H60N2O14: C, 69.75; H, 5.85; N, 2.71%.
Found: C, 69.74; H, 6.03; N, 2.60%.
4,4′-Bis[3,5-bis[3,5-bis(3,5-dimethoxybenzyloxy)benzyloxy]-
benzyloxy]-2,2′-bipyridine (3(G3[Me])): white powder; mp 149.7-
150.7 °C; IR (KBr) 3001, 2937, 1595, 1459, 1374, 1246, 830 cm-1
;
1H NMR (CDCl3) δ (ppm) 8.45 (d, J ) 5.8 Hz, 2H), 8.07 (d, J )
2.4 Hz, 2H), 6.87 (dd, J ) 5.8, 2.4 Hz, 2H), 6.68-6.67 (m, 12H),
6.57-6.55 (m, 22H), 6.40 (t, J ) 2.4 Hz, 4H), 5.13 (s, 4H), 4.97
(s, 8H), 4.96 (s, 16H), 3.77 (s, 48H); 13C NMR (CDCl3) δ (ppm)
165.7, 161.0, 160.1, 157.8, 150.3, 139.1, 138.2, 111.5, 107.1,
106.44, 106.38, 105.2, 101.9, 101.6, 100.0, 70.1, 70.0, 69.7, 55.3;
Anal. Calcd for C124H124N2O30: C, 70.17; H, 5.89; N, 1.32%.
Found: C, 70.07; H, 5.80; N, 1.30%.
4,4′-Bis(3,5-dibenzyloxybenzyloxy)-2,2′-bipyridine (3(G1[Bn])):
white powder; mp 166.5-167.5 °C; IR (CH2Cl2) 3000, 1585, 1451,
1376, 1295, 1159, 1027 cm-1; 1H NMR (CDCl3) δ (ppm) 8.47 (d,
J ) 5.8 Hz, 2H), 8.06 (d, J ) 2.5 Hz, 2H), 7.56-7.30 (m, 20H),
6.88 (dd, J ) 5.8, 2.5 Hz, 2H), 6.70 (d, J ) 2.4 Hz, 4H), 6.59 (t,
J ) 2.2 Hz, 2H), 5.15 (s, 4H), 5.04 (s, 8H); 13C NMR (CDCl3) δ
(ppm) 165.7, 160.2, 157.9, 150.3, 138.2, 136.7, 128.6, 128.0, 127.5,
111.5, 107.1, 106.4, 101.9, 70.2, 69.7; FAB MS for C52H44N2O6
m/z: Calcd: 792.9 M+; Found: 793; Anal. Calcd for C52H44N2O6:
C, 78.77; H, 5.59; N, 3.53%. Found: C, 78.75; H, 5.53; N, 3.44%.
the hydrophobic dendron surrounding Cu(OTf)2 is essential to
organic syntheses in water as the hydrophobic reaction fields.
Conclusion
We synthesized novel bipyridine-core dendritic ligands for
the preparation of dendritic Cu(OTf)2 complexes to be used as
Lewis acid catalysts. In Diels-Alder reactions, they showed a
positive dendritic effect on the chemical yields of adducts.
Furthermore, these dendritic Cu(OTf)2 catalysts were easily
recovered by reprecipitation and could be reused without
deactivation. By immobilization of Lewis acids on dendrimers
at the core position, unique reactivities such as the positive
dendritic effect caused by dendritic reaction fields can be
expected.
Furthermore, in Mannich-type reactions, a positive dendritic
effect on chemical yields was observed in water, because the
dendron of the dendritic Cu(OTf)2 catalyst acted as an effective
hydrophobic reaction field in water. It was found that the
immobilization of Lewis acids on dendrimers at the core position
is suitable for the design of water-compatible Lewis acid
catalysts.
4,4′-Bis[3,5-bis(3,5-dibenzyloxybenzyloxy)benzyloxy]-2,2′-bi-
pyridine (3(G2[Bn])): white powder; mp 149.0-150.0 °C; IR
1
(KBr) 3031, 2867, 1595, 1448, 1375, 1297, 1158, 827 cm-1; H
NMR (CDCl3) δ (ppm) 8.45 (d, J ) 5.6 Hz, 2H), 8.07 (d, J ) 2.7
Hz, 2H), 7.41-7.27 (m, 40H), 6.87 (dd, J ) 5.6, 2.5 Hz, 2H),
6.68 (bs, 12H), 6.57-6.55 (m, 6H), 5.13 (s, 4H), 5.02 (s, 16H),
4.98 (s, 8H); 13C NMR (CDCl3) δ (ppm) 165.7, 160.2, 160.1, 157.8,
150.3, 139.1, 138.2, 136.8, 129.0, 128.6, 128.0, 127.6, 111.5, 107.1,
106.5, 106.4, 101.9, 101.6, 70.1, 70.0, 69.7; FAB MS for
C
108H92N2O14 m/z: Calcd: 1640.7 M+; Found: 1641; Anal. Calcd
for C108H92N2O14: C, 79.00; H, 5.65; N, 1.71%. Found: C, 78.96;
H, 5.49; N, 1.61%.
4,4′-Bis[3,5-bis[3,5-bis(3,5-dibenzyloxybenzyloxy)benzyloxy]-
benzyloxy]-2,2′ -bipyridine (3(G3[Bn])): colorless glass; IR (KBr)
1
3031, 2870, 1595, 1448, 1375, 1295, 1160, 1051, 829 cm-1; H
NMR (CDCl3) δ (ppm) 8.42 (d, J ) 5.7 Hz, 2H), 8.05 (d, J ) 2.5
Hz, 2H), 7.39-7.26 (m, 80H), 6.82 (dd, J ) 5.7, 2.5 Hz, 2H),
6.67-6.65 (m, 28H), 6.57 (t, J ) 2.1 Hz, 2H) 6.54 (t, J ) 2.1 Hz,
8H), 6.53 (t, J ) 2.1 Hz, 4H), 5.07 (s, 4H), 4.98 (s, 32H), 4.95 (s,
8H), 4.93 (s, 16H); 13C NMR (CDCl3) δ (ppm) 165.7, 160.15,
160.07, 157.8, 150.3, 139.2, 139.1, 138.3, 136.8, 128.6, 128.0,
127.5, 111.5, 107.1, 106.4, 101.6, 70.1, 70.0, 69.7; MALDI-TOF
MS for C220H189N2O30 m/z: Calcd: 3338.3 (M + H)+; Found:
3338.4; Anal. Calcd for C220H188N2O30: C, 79.12; H, 5.67; N,
0.84%. Found: C, 79.13; H, 5.48; N, 0.78%.
Experimental Section
Preparation of Dendritic Ligands 3(Gn[R]) from 1 and
2(Gn[R]): Typical Procedure. A dry THF solution (15 mL) of
4,4′-dihydroxy-2,2′-bipyridine 1 (0.189 g, 1.00 mmol), 3,5-bis(3,5-
dimethoxybenzyloxy)benzyl bromide 2(G2[Me]) (1.057 g, 2.10
mmol), anhydrous potassium carbonate (0.375 g, 2.71 mmol), and
18-crown-6 (75.1 mg, 0.284 mmol) was refluxed for 7 h under an
argon atmosphere. The reaction mixture was filtered with Celite to
remove inorganic salts, and the filtrate was evaporated to dryness.
The residue was purified with silica gel column chromatography
(chloroform/methanol/triethylamine ) 95/4/1 as eluent) to obtain
3(G2[Me]) (0.890 g, 0.861 mmol) in an 86% yield.
4,4′-Bis[3,5-bis[2-[2-(2-methoxyethoxy)ethoxy]ethoxy]ben-
zyloxy]-2,2′-bipyridine (3(G1[TEG])): white powder; mp 44.0-
1
44.5 °C; IR (neat) 2876, 1585, 1458, 1295, 1109, 848 cm-1; H
NMR (CDCl3) δ (ppm) 8.48 (d, J ) 5.6 Hz, 2H), 8.06 (d, J ) 2.5
Hz, 2H), 6.89 (dd, J ) 5.6, 2.5 Hz, 2H), 6.61 (d, J ) 2.3 Hz, 4H),
6.47 (t, J ) 2.3 Hz, 4H), 5.14 (s, 4H), 4.12 (t, J ) 4.8 Hz, 8H),
3.85 (t, J ) 4.9 Hz, 8H), 3.75-3.73 (m, 8H), 3.70-3.65 (m, 16
H), 3.55 (dd, J ) 5.8, 3.7 Hz, 8H), 3.38 (s, 12H); 13C NMR (CDCl3)
δ (ppm) 165.7, 160.2, 150.2, 138.1, 111.5, 107.1, 106.1, 101.4,
72.0, 70.8, 70.7, 70.6, 69.8, 69.7, 67.6, 59.0; FAB MS for
C52H76N2O18 m/z: Calcd: 1016.5 M+; Found: 1017; Anal. Calcd
4,4′-Bis[3,5-bis(3,5-dimethoxybenzyloxy)benzyloxy]-2,2′-bipy-
ridine (3(G2[Me])): white powder; mp 151.5-152.5 °C; IR (CH2-
(31) Scheme 2 was carried out by use of 1.5 equiv of silyl enolate 6 as
compared with benzaldehyde according to the procedure, which was shown
in the Supporting Information. 6 was recovered in 59%, and acetophenone
was also obtained in 15%. These ratios are based on the amount of used
silyl enolate 6.
J. Org. Chem, Vol. 72, No. 21, 2007 7869

Muraki et al.
for C52H76N2O18: C, 61.40; H, 7.53; N, 2.75%. Found: C, 61.10;
H, 7.57; N, 2.63%.
2.2 Hz, 2H), 5.17 (s, 4H), 3.81 (s, 12H); 13C NMR (CDCl3) δ (ppm)
165.8, 161.1, 157.7, 150.2, 111.5, 107.2, 105.3, 69.8, 55.4; EI MS
for C28H28N2O6 m/z: Calcd: 488.1947 M+; Found: 488.1939; Anal.
Calcd for C28H28N2O6‚0.5H2O: C, 67.59; H, 5.87; N, 5.63%.
Found: C, 67.61; H, 5.63; N, 5.56%.
Preparation of Dendritic Ligand 3(G1[Me]) from 1 and 4.
The dry THF solution (3 mL) of diisopropyl azodicarboxylate (0.765
g, 3.78 mmol) was added to the THF solution (7 mL) of
4,4′-dihydroxy-2,2′-bipyridine 1 (0.283 g, 1.50 mmol), 3,5-
dimethoxybenzyl alcohol 4 (0.600 g, 3.57 mmol), and tri-
phenylphosphine (1.01 g, 3.85 mmol) at 0 °C under an argon
atmosphere. The resulting mixture was stirred for 24 h at room
temperature, and 5 mL of dichloromethane was added to dissolve
the newly formed triphenylphosphine oxide. After filtration of the
reaction mixture, the white powder that was collected was poured
into 100 mL of chloroform and was stirred for 6 h to dissolve
3(G1[Me]) completely. The mixture solution was filtered to remove
unreacted 1, and the filtrate was concentrated to obtain 3(G1[Me])
(0.369 g, 0.755 mmol) in a 50% yield.
Acknowledgment. This work was supported by Industrial
Technology Research Grant Program in ‘04 from New Energy
and Industrial Technology Development Organization (NEDO)
of Japan and by a Grant-in-Aid for Scientific Research from
Japan Society for the Promotion of Science.
Supporting Information Available: Procedures for Lewis acid-
catalyzed reactions and the preparation of 12, characterization data
for new compounds 5e, 5g, 9b, 10b, 11a, 11b, 11c, and 12, and
4,4′-Bis(3,5-dimethoxybenzyloxy)-2,2′-bipyridine (3(G1[Me])):
white powder; mp 197.0-198.0 °C; IR (KBr) 3016, 2942, 1586,
1
copies of H NMR and 13C NMR spectra for all new compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
1
1455, 1374, 1292, 1235, 1209, 890, 872 cm-1; H NMR (CDCl3)
δ (ppm) 8.49 (d, J ) 5.8 Hz, 2H), 8.08 (d, J ) 2.5 Hz, 2H), 6.91
(dd, J ) 5.8, 2.5 Hz, 2H), 6.61 (d, J ) 2.2 Hz, 4H), 6.44 (t, J )
JO070767A
7870 J. Org. Chem., Vol. 72, No. 21, 2007