Palladium-catalyzed allylation of malonic acid derivatives in heterogeneous systems containing ionic liquids

Article abstract of DOI:10.1016/j.mencom.2013.12.007

Palladium-catalyzed potassium carbonate-assisted reaction between diethyl malonate and allyl acetate in the presence of 1,3-dialkyl-imidazolium ionic liquids (ILs) as solvents or phase transfer catalysts affords major monoallylation product, whereas in the presence of 1,2,3-trialkylimidazolium or quaternary ammonium/phosphonium ILs diallylation product is preferably formed. These procedures are extended to some other CH-acids and allylic acetates.

Full text of DOI:10.1016/j.mencom.2013.12.007

Mendeleev
Communications
Mendeleev Commun., 2014, 24, 23–25
Palladium-catalyzed allylation of malonic acid derivatives
in heterogeneous systems containing ionic liquids
Andrei A. Vasil’ev and Sergei G. Zlotin*
N. D. Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, 119991 Moscow,
Russian Federation. Fax: +7 499 135 5328; e-mail: zlotin@ioc.ac.ru
DOI: 10.1016/j.mencom.2013.12.007
Palladium-catalyzed potassium carbonate-assisted reaction between diethyl malonate and allyl acetate in the presence of 1,3-dialkyl-
imidazolium ionic liquids (ILs) as solvents or phase transfer catalysts affords major monoallylation product, whereas in the presence
of 1,2,3-trialkylimidazolium or quaternary ammonium/phosphonium ILs diallylation product is preferably formed. These procedures
are extended to some other CH-acids and allylic acetates.
Ionic liquids (ILs), the low-melting (<100°C) salts comprising
organic cations, have been recognized in the last decade as
perspective green media for chemical reactions due to their
recoverability, negligible volatility and special solvation effects.1
Certain usually ineffective in conventional organic solvents reac-
tions have been successfully accomplished in IL media.2,3 Among
variety of reactions performed in ILs, transition metal-catalyzed
processes attract considerable attention. In distinction to the
Suzuki–Miyaura, Heck and Sonogashira C–C cross-coupling,4
the Tsuji–Trost reaction5 (TTR, a palladium-catalyzed reaction
between nucleophiles and allylic alcohol derivatives), is far less
explored in IL media. In cases of 1-butyl-3-methylimidazolium
tetrafluoroborates or hexafluorophosphates ([bmim][BF4],
[bmim][PF6]), the palladium source and phosphine ligand were
preliminarily heated in these ILs to supposedly furnish N-hetero-
cyclic carbene (NHC) which acted as the additional ligand,6 the
further cross-coupling having been performed at room tempera-
ture. The stoichiometric amounts of 1,3-disubstituted imidazolium
derivatives were used to purposefully prepare Pd–NHC com-
plexes, followed by coupling in a usual manner.7 Microwave
irradiation in a water–1-ethyl-3-methylimidazolium tetrafluoro-
borate ([emim][BF4]) mixture on using water-soluble trisodium
salt of tris(3-sulfophenyl)phosphine as the ligand effected the
complete coupling within several minutes for a large scope of
reactants.8 These data limited mostly to coupling between highly
active 1,3-diphenylprop-2-enyl or cinnamyl acetates and typical
CH acids do not clarify which type of ligands, NHC or phosphine,
can be preferable in the TTR.
N
N
N
N
N
N
Me
Bu
Me
Bu
Me
Hex
BF4
TfO
Tf2N
[bmim][BF4]
[bmim][OTf]
[bmim][NTf2]
IL1
IL2
IL3
N
N
N
N
N
N
Me
Hex
Me
Et
BF4
Me
Bu
BF4
PF3(C2F5)3
Me
Me
[hmim][PF3(C2F5)3]
[memim][BF4]
[mbmim][BF4]
IL4
IL5
IL6
BF4
TfO
Tf2N
PF3(C2F5)3
N
N
N
N
Me Bu
Me Bu
Me Bu
Me Bu
IL7
IL8
IL9
IL10
Hex3PC14H29 PF3(C2F5)3
IL11
out on using potassium carbonate. We anticipated that similarly to
other types of reactions ILs would promote better deprotonation of
the CH acids and stabilization of their carbanions1,4 and influence
the Pd-mediated catalytic cycle.
The coupling 1 + 2a in the essence of 1,3-dialkylimidazolium
salts IL1–4 (see Scheme 1, conditions i)† never brought about
full diallylation (Table 1, entries 1–4) and usually monoallyl
derivative 3a was obtained. In cases of quaternary ammonium
IL8–10 and phosphonium IL11 salts diallylated compound 4a
was the sole product (entries 5–8). Apparently, imidazolium salts
IL1–4 on the action of the base form the corresponding NHCs
Note that our recent studies on TTR under phase transfer
conditions9 allowed us to replace traditional expensive and/or
flammable BSA–AcOH and NaH by cheap and convenient in
handling potassium carbonate and phosphate. However, the appli-
cation of such a procedure was limited by a substrate type. Taking
these results into account, we pioneered in performing the TTR
in supercritical carbon dioxide.10
The purpose of the present study was to thoroughly examine
effect of the IL nature on the outcome of the TTR. Ionic liquids
IL1–11 differing in both cation and anion structure were tested,
with their subdivision into 1,3-dialkylimidazolium IL1–4,
1,2,3-trialkylimidazolium IL5,6, pyrrolidinium IL7–10 and tetra-
alkylphosphonium IL11 chemotypes. Reactions were performed
in IL media as well as in ordinary solvents with the use of ILs
as phase transfer catalysts (PTC). The model coupling between
diethyl malonate 1 and allyl acetate 2a leading generally to
mono- (3a) and diallylation (4a) products (Scheme 1) was carried
OAc
2a
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
i or ii
CO2Et
1
3a
4a
Scheme 1 Reagents and conditions: i, AllOAc (2.4 equiv.), Pd(dba)2 (2 mol%)
,
PPh3 (8 mol%), K2CO3 (2.5 equiv.), IL, 20°C, 18–24 h; ii, AllOAc (2.4 equiv.),
Pd(dba)2 (2 mol%), PR3 (8 mol%), K2CO3 (2.5 equiv.), IL (~13 mol%),
CH2Cl2, 20°C, 18–24 h.
© 2014 Mendeleev Communications. All rights reserved.
– 23 –

Mendeleev Commun., 2014, 24, 23–25
Table 1 Coupling between diethyl malonate 1 and allyl acetate 2a in IL
Table 2 Coupling between diethyl malonate 1 and allyl acetate 2a in CH2Cl2
media.a
under phase transfer conditions.a
Product composition (GC) (%)
Product composition (GC), %
Entry
Ionic liquid
Entry
PTC
Ligand
3a
4a
unreacted 1
3a
4a
unreacted 1
1
2
IL1
PPh3
PPh3
PBu3
PPh3
PBu3
PPh3
PBu3
PPh3
PBu3
PCy3
PPh3
PBu3
PPh3
PBu3
PPh3
PPh3
PPh3
PPh3
PBu3
PPh3
PPh3
—
87
47
2
13
0
—
53
98
11
52
25
85
0
1
2
3
4
5
6
7
8
IL1
IL2
IL3
IL4
IL8
IL9
IL10
IL11
49
85
69
78
0
51
0
0
15
0
IL2
IL2
IL3
IL3
IL4
IL4
IL5
IL5
IL5
IL6
IL6
IL7
IL7
IL8
IL9
3
0
31
4
89
48
75
15
0
0
22
0
5
0
100
100
100
100
0
6
0
0
0
7
0
0
0
8
100
100
0
0
0
9
0
0
a Conditions:AllOAc (2.4 equiv.), Pd(dba)2 (2 mol%), PPh3 (8 mol%), K2CO3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
4
96
0
(2.5 equiv.), IL (2 g mmol–1 ), 20°C, 18–24 h.
0
100
100
100
100
99
98
100
100
100
100
25
0
0
0
which would coordinate on palladium thus giving less active
catalytic system. Nature of the IL anion was in whole less crucial.
Then, we studied reactions of diethyl malonate 1 with allyl
acetate 2a in dichloromethane using ILs as PTCs since they
contained bulky lipophilic cations (Scheme 1, conditions ii,
Table 2).‡ On using 1,3-dialkylimidazolium salts IL1–4 and
PPh3 as the ligand (entries 1,2,4,6), monoallylated product 3a
at incomplete conversion of the starting compound 1 was mostly
formed, analogously to the phenomenon observed when such
ILs were used as reaction media (see Table 1, entries 1–4). When
PBu3 was used instead of PPh3 the yield was significantly dropped
(Table 2, entries 3,5,7); most probably, this stronger ligand in
combination with the corresponding NHCs afforded the Pd com-
plexes less suitable for TTR. On moving to 1,2,3-trisubstituted
imidazolium salts IL5,6 (entries 8,9,11,12), which cannot form
carbenes under the action of bases, the yield of diallylated pro-
duct 4a was 100% regardless PPh3 or PBu3 having been applied.
However, in the case of tricyclohexylphosphine (entry 10), the
yield of monoallylation product 3a was negligible, apparently,
due to bulkiness of this electron-rich phosphine which completely
covers the palladium atom thus retarding formation of p-allylic
complexes. The close results were obtained with pyrrolidinium
IL7–10 (entries 13–17), phosphonium IL11 (entries 18,19) and
tetrabutylammonium (entry 20) salts. Control experiment without
0
0
0
0
1
0
2
0
IL10
IL11
IL11
Bu4NBr9(b),10(c)
—
0
0
0
0
0
0
0
0
75
3
0
IL1
IL5
97
98
—
2
0
a Conditions:AllOAc (2.4 equiv.), Pd(dba)2 (2 mol%), PR3 (~8 mol%), K2CO3
(2.5 equiv.), IL (~13 mol%), CH2Cl2 (3 ml mmol–1), 20°C, 18–24 h.
PTC (entry 21) showed that the reaction proceeded at incomplete
conversion into 4a, the total conversion having been deeper than
that in the cases of IL1–4 (entries 1–7) thus confirming the
inhibition effect of 1,3-dialkylimidazolium salts. Experiments
performed in the absence of phosphine (entries 22,23) gave
almost nothing of the products and were accompanied by rapid
precipitation of Pd black from Pd(dba)2. Although this was not
unexpected, we anticipated somehow that IL1-derived NHC
(entry 22) should have stabilized palladium in the solution.
The PTC conditions were found applicable to some other
CH-acids and allylic acetates (Scheme 2, Table 3). Experiments
with but-3-en-2-yl acetate 2b were carried out with equimolar
ratio of reactants (entries 1–4), otherwise easy formation of
diallylation products, namely diethyl di(but-2-en-1-yl)malonate
and diethyl (but-2-en-1-yl)(but-1-en-3-yl)malonate,7(b) would
†
Reaction between diethyl malonate 1 and allyl acetate 2a in ILs.
A Schlenk tube was charged with diethyl malonate (40 mg, 0.25 mmol) and
corresponding IL (0.5 g), and the mixture was three times deaerated by
evacuation and filling with argon. PPh3 (5.5 mg, 20 mmol) and Pd(dba)2
(3 mg, 5 mmol) were added, and the mixture was stirred at ambient tem-
perature for ca. 30 min which resulted in colour change due to the forma-
tion of palladium phosphine complex. Allyl acetate (0.075 ml, ~0.6 mmol)
was added, followed in 10 min by freshly powdered K2CO3 (104 mg,
0.75 mmol). Each opening of the Schenk tube was followed by evacuation
and filling with argon. The mixture was stirred at ambient temperature
for 18–24 h. A probe was triturated with hexane, and the hexane extract
was analyzed by GC, comparing peak retention times with those of the
authentic samples.
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
1
R1
+
CO2Et
R1 R2
3'b,c
R2
OAc
R1 R2
2a–c
3a–c
‡
Reactions of CH-acids and allylic acetates under PTC conditions.
A Schlenk tube was charged with the corresponding CH-acid (0.5 mmol),
IL (20 mg), and dichloromethane (1.5 ml). The mixture was three times
deaerated by careful evacuation (300 Torr, dichloromethane started boiling)
and filling with argon. Phosphine (0.04 mmol) and Pd(dba)2 (6 mg,
0.01 mmol) were added, and the mixture was stirred until the colour grew
pale. Allylic acetate (0.5–1.5 mmol) was added, and the mixture was
stirred for 10 min, followed by freshly powdered K2CO3 (138–207 mg,
1–1.5 mmol). Each opening of the Schenk tube was followed by evacua-
tion and filling with argon. The mixture was stirred at ambient temperature
for 18–24 h. A probe was treated with water and extracted with hexane
and analyzed by GC, comparing peak retention times with those of the
authentic samples obtained and characterized previously.9(a),(b),10(c)
CO2Et
CO2Et
Me
CO2Et
CO2Et
CO2Et
CO2Et
Me
Me
R1
+
5
R2
Me
a R1 = R2 = H
b R1 = Me, R2 = H
c R1 = R2 = Me
6a,b
6'b
Scheme 2 Reagents and conditions: Pd(dba)2, phosphine, K2CO3, CH2Cl2,
IL (cat.), 20°C.
– 24 –

Mendeleev Commun., 2014, 24, 23–25
Table 3 Coupling between CH acids and allylic acetates in CH2Cl2 using
ILs as PTCs.a
2 (a) N. I. Simirskaya, N. V. Ignat’ev, M. Schulte and S. G. Zlotin,
Mendeleev Commun., 2011, 21, 94; (b) N. I. Simirskaya, N. V. Ignat’ev,
M. Schulte and S. G. Zlotin, Mendeleev Commun., 2012, 22, 317.
3 (a) M. A. Epishina, I. V. Ovchinnikov, A. S. Kulikov, N. N. Makhova
and V. A. Tartakovsky, Mendeleev Commun., 2011, 21, 21; (b) M. A.
Epishina, A. S. Kulikov, N. V. Ignat’ev, M. Schulte and N. N. Makhova,
Mendeleev Commun., 2011, 21, 334; (c) M. A. Epishina, A. S. Kulikov,
M. I. Struchkova, N.V. Ignat’ev, M. Schulte and N. N. Makhova, Mendeleev
Commun., 2012, 22, 267.
4 (a) H. Zhao and S. V. Malhotra, Aldrichim. Acta, 2002, 35, 75;
(b) P. Wasserscheid and P. Schulz, in Ionic Liquids in Synthesis, 2nd edn.,
eds. P. Wasserscheid and T. Welton, Wiley-VCH Verlag, Weinheim, 2008.
5 (a) B. M. Trost, Acc. Chem. Res., 1996, 29, 355; (b) B. M. Trost and
D. L. Van Vranken, Chem. Rev., 1996, 96, 395; (c) Z. Lu and S. Ma,
Angew. Chem. Int. Ed., 2008, 47, 258; (d) G. Helmchen, U. Kazmaier
and S. Förster, in Catalytic Asymmetric Synthesis, 3rd edn., ed. I. Ojima,
Wiley, Hoboken, New Jersey, 2010, pp. 497–641; (e) L. Milhau and
P. J. Guiry, Top. Organomet. Chem., 2012, 38, 95.
6 (a) W. Chen, L. Xu, C. Chatterton and J. Xiao, Chem. Commun., 1999,
1247; (b) Š. Toma, B. Gotov, I. Kmentová and E. Solcˇániová, Green
Chem., 2000, 2, 149; (c) I. Kmentová, B. Gotov, E. Solcˇániová and
Š. Toma, Green Chem., 2002, 4, 103; (d) J. Ross, W. Chen, L. Xu and
J. Xiao, Organometallics, 2001, 20, 138; (e) C. de Bellefon, E. Pollet
and P. Grenouillet, J. Mol. Catal. A: Chem., 1999, 145, 121.
7 (a) Y. Sato, T. Yoshino and M. Mori, Org. Lett., 2003, 5, 31; (b) Y. Sato,
T.Yoshino and M. Mori, J. Organomet. Chem., 2005, 690, 5753; (c) L. G.
Bonnet and R. E. Douthwaite, Organometallics, 2003, 22, 4187.
8 M. Liao, X. Duan and Y. Liang, Tetrahedron Lett., 2005, 46, 3469.
9 (a) A. A. Vasil’ev, S. E. Lyubimov, E. P. Serebryakov, V. A. Davankov
and S. G. Zlotin, Mendeleev Commun., 2009, 19, 103; (b) A. A. Vasil’ev,
S. E. Lyubimov, E. P. Serebryakov, V. A. Davankov, M. I. Struchkova
and S. G. Zlotin, Russ. Chem. Bull., Int. Ed., 2010, 59, 605 (Izv. Akad.
Nauk, Ser. Khim., 2010, 592); (c) A. A. Vasil’ev, S. E. Lyubimov, E. P.
Serebryakov, V. A. Davankov and S. G. Zlotin, Mendeleev Commun.,
2012, 22, 39.
Product composition (GC) (%)
CH Allylic
acid acetate
Entry
PTC Ligand
products
unreacted 1, 5
1
2
3
4
5
6
7
8
1
1
1
1
1
1
5
5
2bb
2bb
2bb
2bb
2c
IL5 PPh3
IL7 PPh3
IL7 PBu3
IL11 PPh3
IL7 PPh3
IL7 PBu3
IL5 PPh3
IL5 PPh3
47 (3b) + 43 (3'b)
35 (3b) + 46 (3'b)
24 (3b) + 59 (3'b)
42 (3b) + 42 (3'b)
4 (3c) + 96 (3'c)
14 (3c) + 67 (3'c)
100 (6a)
16
19
17
10
—
19
—
41
2c
2a
2b
54 (6b) + 5 (6'b)
a Conditions: allylic acetate (2.4 equiv.), Pd(dba)2 (2 mol%), PR3 (~8 mol%),
K2CO3 (2.5 equiv.), IL (~13 mol%), CH2Cl2 (3 ml mmol–1), 20°C, 18–24 h.
b 1 equiv. 2b.
bring complications to analysis of the reaction mixture. In general,
reaction between 1 and 2b gave mixture of regioisomers 3b and
3'b in close amounts. In case of 2-methylbut-3-en-2-yl acetate
2c the ‘branched’ product 3'c was mostly formed (entries 5,6,
cf. ref. 6), the fraction of the ‘normal’ product 3c being somewhat
higher on application of tributylphosphine as the ligand (entry 6).
Disubstitution products never formed even with the use of excess
2c, moreover, the conversion of malonate 1 was often incomplete.
Less acidic diethyl methylmalonate 5 reacted smoothly with
allyl acetate 2a to give product 6a (Scheme 2, Table 3, entry 7).
In the case of but-3-en-2-yl acetate 2b the conversion of 5 was
incomplete (entry 8), while 2-methylbut-3-en-2-yl acetate 2c
gave none of the expected product.
In summary, based on extended number of both ILs and
substrates one may conclude that ILs can serve both as media
and PTC for TTR. The better reaction outcome is provided by
chemically inert ILs, while base-sensitive 1,3-dialkylimidazolium
ones form NHCs which reduce the catalyst activity. Most probably,
phosphine ligands should be preferable for the TTR, although
testing other 1,3-disubstituted imidazolium salts as additives,
especially with bulky substituents, and other reaction conditions
seems not undesirable (the NHC complexes were reported7 to
provide good yields under some other conditions and with other
substrates).
10 (a) S. E. Lyubimov, I. V. Kuchurov, A. A. Vasil’ev, A. A. Tyutyunov,
V. N. Kalinin, V. A. Davankov and S. G. Zlotin, J. Organomet. Chem.,
2009, 694, 3047; (b) S. E. Lyubimov, I. V. Kuchurov, A. A. Vasil’ev,
S. G. Zlotin and V. A. Davankov, Mendeleev Commun., 2010, 20, 143;
(c) A. A. Vasil’ev, I. V. Kuchurov and S. G. Zlotin, Mendeleev Commun.,
2013, 23, 84.
References
1 (a) T. Welton, Chem. Rev., 1999, 99, 2071; (b) S. Chowdhury, R. S. Mohan
and J. L. Scott, Tetrahedron, 2007, 63, 2363; (c) J. P. Hallett and T. Welton,
Chem. Rev., 2011, 111, 3508.
Received: 20th August 2013; Com. 13/4188
– 25 –