Role of base in palladium-catalyzed arylation of carbanions

Article abstract of DOI:10.1016/j.jorganchem.2003.12.039

The arylation reaction of carbanions, derived from certain sulfones, cyanoacetic ester and malononitrile, with aryl bromides (using the catalytic system of Pd₂dba₃/3L, L=PPh₃, P^tBu₃) as well as the reaction of the carbanions with one equivalent of 4-CF₃C₆ H₄ Pd(PPh₃)₂Br has been studied. These reactions proceed smoothly provided that the base stronger than the initial carbanion is present in the reaction mixture. In the absence of the above type of base the reactions do not proceed at all. Taking that into account we have proposed a novel mechanism of palladium-catalyzed arylation of CH-acids. The main feature of this mechanism is the accelaration of the reductive elimination due to the deprotonation of the intermediate ArPdL₂CHXY. The correlation between the carbanion reactivity and the pK_a values for related CH-acids as well as the ligand effect are discussed in the framework of the proposed mechanism.

Full text of DOI:10.1016/j.jorganchem.2003.12.039

Journal of Organometallic Chemistry 689 (2004) 1085–1090
www.elsevier.com/locate/jorganchem
Role of base in palladium-catalyzed arylation of carbanions
*
Anton V. Mitin, Alexander N. Kashin , Irina P. Beletskaya
Department of Chemistry, Moscow State University, Leninskie Gory 1/3, Moscow 119992, Russia
Received 1 October 2003; accepted 2 December 2003
Abstract
The arylation reaction of carbanions, derived from certain sulfones, cyanoacetic ester and malononitrile, with aryl bromides
(using the catalytic system of Pd2dba3/3L, L ¼ PPh3, PtBu3) as well as the reaction of the carbanions with one equivalent of 4-
CF3C6H4 Pd(PPh3)2Br has been studied. These reactions proceed smoothly provided that the base stronger than the initial carb-
anion is present in the reaction mixture. In the absence of the above type of base the reactions do not proceed at all. Taking that into
account we have proposed a novel mechanism of palladium-catalyzed arylation of CH-acids. The main feature of this mechanism is
the accelaration of the reductive elimination due to the deprotonation of the intermediate ArPdL2CHXY. The correlation between
the carbanion reactivity and the pKa values for related CH-acids as well as the ligand effect are discussed in the framework of the
proposed mechanism.
Ó 2004 Elsevier B.V. All rights reserved.
Keywords: Palladium; Catalysis; Arylation; Sulfone; Phosphine ligand; Proton transfer
1. Introduction
novel mechanism of the palladium-catalyzed arylation
of CH-acids.
The palladium-catalyzed arylation of CH-acids by
aryl halides is a direct method of the introduction of a
aryl group in the a-position of nitriles [1], ketones [2],
aldehydes [3], amides [4], esters [5], nitroalkanes [5b,6],
sulfones [7]. In most cases, the palladium-catalyzed
arylation proceeds effectively only in the presence of
electron-rich bulky phosphine ligands, which are sup-
posed to reflect their influence on the reductive elimi-
nation stage [2c,8]. Meanwhile, some relatively strong
CH-acids – functionalized nitriles [1a–1d] and sulfones
[7] in particular – can be arylated just using triphenyl-
phosphine as a ligand.
Studying the influence of the base upon the reactivity
of sulfones in the arylation and using PPh3 as a ligand,
we unexpectantly found, that the reaction does not
proceed without certain excess of the base. Thus, the
role of the base in these reactions is more complicated
than just a carbanion generation. In this article, an at-
tempt has been made to explain this unusual phenom-
enon and, judging by the data obtained, to propose a
2. Results and discussion
It is known, that the arylation of CH-acids by aryl
halides is generally carried out using at least two
equivalents of the base. One equivalent of the base is
consumed in the deprotonation of the reaction product,
which is a stronger CH-acid than the starting com-
pound. It turned out, however, that if two equivalents
(relative to the aryl halide) of the preformed anion of the
starting CH-acid are used instead of two equivalents of
the base, the palladium-catalyzed arylation of sulfones
by aryl halides (4-bromobenzotrifluoride and 1-bro-
monaphtalene) using the PPh3 ligand does not proceed
at all. Meanwhile, in the presence of one equivalent of
the CH-acid, two equivalents of the base, such as NaH,
KOtBu or NaHMDS, and PPh3 almost complete con-
version of the aryl bromide is achieved and the arylation
product is formed (Scheme 1, Table 1).
This base effect phenomenon is observed not only
with sulfones, but also with other CH-acids, such as
cyanoacetic ester and malononitrile (Table 1, entries 5,
* Corresponding author. Tel.: +7-095-9393618; fax: +7-0959-393618.
E-mail address: ankashin@org.chem.msu.ru (A.N. Kashin).
0022-328X/$ - see front matter Ó 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2003.12.039

1086
A.V. Mitin et al. / Journal of Organometallic Chemistry 689 (2004) 1085–1090
difference being, that in the absence of NaH the starting
Pd2dba3, PPh3
NaH, dme
X
SO2Ph
X
SO2Ph
+
ArBr
complex slowly decomposed, but did not give the ary-
lation product. This product was formed only when
NaH was added to the reaction mixture (Scheme 3).
Thus, the effect of the base in the catalytic and
stoichiometric reactions turns out to be the same,
namely – its presence affects the arylation reaction of the
carbanion, provided that the base is stronger than the
initial carbanion [CHXY]ꢀ. This fact could be explained
by assuming that the rate limiting stage of the arylation
reaction is the reductive elimination, which is hampered
in the neutral complex ArPd(PPh3)2CHXY formed by
weak sulfonic CH-acids. We suppose that, in the pres-
ence of a strong base, deprotonation takes place, leading
to the formation of anionic palladium complex
[ArPd(PPh3)2CXY]ꢀ which undergoes the reductive
elimination. The Scheme 4 presents the catalytic cycle
included the proton transfer stage (c).
According to the proposed arylation mechanism,
however, it is not possible to explain why bromide in the
ArPdL2Br complex is not exchanged in the stoichiom-
etric reaction (Schemes 2 and 3). It is assumed that in
case of stable carbanions used in this study, the step of
the halogen substitution in the palladium complex is
reversible (Scheme 4, Stage b) and the equilibrium is
shifted towards the carbanion, and only the subsequent
reductive elimination shifts this equilibrium. The similar
equlibria, in which the order of ligand affinity to tran-
sition metal is opposite to the order of ligand pKa values
are known [9].
Ar
X=CO2Et, SO2Ph
Scheme 1.
6), that is a double quantity of the base is required to
carry out the reaction with these compounds.
Why does the arylation of CH-acid RH not proceed
with their preformed salts? There could be two possible
solutions of this problem: either the palladium complex
ArPdL2Br formed in the oxidation addition does not
react with the salt of the CH-acid under these condi-
tions, or the product of the ligand substitution – Ar-
PdL2R – is resistant to the reductive elimination.
In order to clarify this problem the stoichiometric
reaction of the palladium complex 4-CF3C6H4
Pd(PPh3)2Br (prepared beforehand) and the salt
[PhSO2CHCO2Et]ꢀ Naþ in dioxane at 70 °C has been
studied. Surprisingly, no ligand exchange has occurred.
The 1H NMR and 19F NMR displayed no visible
changes, only the signals of starting materials being
observed. However, upon the addition of the base (NaH
or NaHMDS) to the reaction mixture, the reaction did
take place, giving the salt of the arylated CH-acid. When
NaH was used as a base, large quantities of PhCF3,
which is the product of the palladium complex reduc-
tion, were also formed (Scheme 2).
Reactions with another CH-acid – C6F5CH2SO2CF3
– followed a similar pattern (DME, 80 °C), the only
Table 1
Palladium-catalyzed arylation of CH-acids with aryl bromidesa
Entry
CH-acid
Conditions
Stoichiometric amount of the base (NaH)
Base (NaH) is present in excess
Time (h)
Conversion of ArBr, %
(Yield, %)
Time (h)
Conversion of
ArBr, % (Yield, %)
1
2
3
4
5
6
7
8
PhSO2CH2CO2Et
PhSO2CH2CO2Et
(PhSO2)2CH2
1-bromonaphthalene,
dioxane, L ¼ PPh3
4-CF3C6H4Br,
12
8
5 (0)
12
4
82 (61)b
100 (77)
100 (86)
87 (81)
2
13
0
dioxane, L ¼ DPPF
4-CF3C6H4Br,
12
12
12
12
4
4
DME, L ¼ PPh3
4-CF3C6H4Br,
CF3SO2CH2C6F5
NCCH2COOEt
NCCH2CN
12
15
6
DME, L ¼ PPh3
4-CF3C6H4Br,
0
86 (72)
DME, L ¼ PPh3
4-CF3C6H4Br,
0
100 (84)
100 (70)
100 (56)
DME, L ¼ PPh3
4-CF3C6H4Br,
PhSO2CH2CO2Et
PhSO2CH2CO2Et
0
4
c
DME, L ¼ PPh3
4-CF3C6H4Br,
4
0
4
d
DME, L ¼ PPh3
a Heating at 70 °C in dioxane or DME, aryl bromide – 0.5 mmol, Pd2dba3 – 2 mol%, L/Pd ¼ 3:1, solvent – 4 ml; in the presence of the base excess:
NaH – 1.25 mmol, CH-acid – 0.65 mmol; without of the base excess: NaH – 1.0 mmol, CH-acid – 1.25 mmol, see Section 3 for details.
b Naphthalene forms in yield 20–24% as a consequence of side reaction – the reduction of ArBr.
c KOtBu used as a base.
d NaHMDS used as a base.

A.V. Mitin et al. / Journal of Organometallic Chemistry 689 (2004) 1085–1090
1087
CO2Et
Na+
CF3
CF3
Pd(PPh3)2Br+
No reaction
°
dioxane,70 C
SO2Ph
CO2Et
Na+ + PhCF3
SO2Ph
CO2Et
NaH
+
CF3
+
+
Pd(PPh3)2Br
Pd(PPh3)2Br
Na
°
dioxane,70 C
SO2Ph
46%yield
33% yield
CO2Et
Na+
SO2Ph
65% yield
CO2Et
SO2Ph
NaHMDS
Na+
CF3
CF3
°
dioxane,70 C
Scheme 2.
C6F5
No reaction
C6F5
+
CF3
Pd(PPh3)2Br Na+
°
dme, 80 C
SO2CF3
NaH
C6F5
+
Na+
+
Pd(PPh3)2Br Na
CF3
CF3
°
dme,
80 C
SO2CF3
SO2CF3
48%yield
Scheme 3.
ArBr
In the framework of the proposed mechanism the
unusual dependence the reactivity of sulfones upon their
pKa could be also clearly explained. As it was shown
earlier, the arylation reaction proceeds only with sulf-
ones, which are relatively strong CH-acids (pKa < 13)
and correspondingly produce poorly nucleophilic car-
banions. Meanwhile, weak CH-acids, such as
CH3SO2Ph, PhCH2SO2Ph and C6F5CH2SO2Ph do not
enter the reaction. The lack of reaction is not related
with the difficulty of deprotonation of these compounds,
because their sodium salts do not react either [7]. We
suppose that this is caused by the difficulty of deproto-
nation of the palladium intermediate, because the latter
is even a weaker CH-acid, than the corresponding initial
sulfone. Deprotonation of the palladium intermediate
can be effective only in the case of sulfones, which are
strong CH-acids.
(a)
ArPdL2Br
Pd(0)L2
[ArCXY]
[CHXY]
Br
(b)
(d)
[ArPdL2CXY]
ArPdL2CHXY
(c)
base
Scheme 4.
Following the proposed concept, it could be expected
that increasing the acidity of sulfones, with the PhSO2
group being changed to the more electron-withdrawing
CF3SO2 group, would accelerate the arylation and allow
to expand the range of substrates entering this reaction.
And indeed, benzyltrifluoromethylsulfone and its de-
rivatives (Scheme 5, Table 2, entries 1–5), unlike the
corresponding benzylphenylsulfones [7], turned out to
be active under the same conditions. The yields of
monoarylated products are presented in Table 2.
The suggestion that the ArPd(PPh3)2CHXY complex
in the case of the studied CH-acids has indeed to be
deprotonated prior to the reductive elimination, is sup-
ported by the fact, that the reaction does not proceed
with the CH-acids having tertiary carbon RSO2-
C(R0)HX, when the palladium complex cannot be de-
protonated [7]. The fact, that the arylation reaction of
sulfones proceeds exclusively as monoarylation, is due,
in our opinion, to the same reason.
The role of the base in the reaction is apparently
connected with the activation of the reductive elimina-
tion stage via the deprotonation of the ArPd(PPh3)2
CHXY complex, but it is not connected with the cis–
trans isomerization of the complex, because the reaction
of PhSO2CH2CO2Et and 4-CF3C6H4Br using DPPF as
a bidentate ligand also requires a certain excess of the
base to proceed (Table 1, entry 2).
Pd2dba3, PPh3
R
SO2CF3
R
SO2CF3
+
ArBr
NaH, dme
Ar
Scheme 5.

1088
A.V. Mitin et al. / Journal of Organometallic Chemistry 689 (2004) 1085–1090
Table 2
Palladium-catalyzed arylation of trifluoromethyl sulfones by aryl bromides
Entry
1
Sulfone
Aryl bromide
Time (h)
20
Yield (%)
67
SO2CF3
F3C
Br
-ꢁ-
2
3
40
14
36
30
SO2CF3
Br
Br
-ꢁ-
-ꢁ-
SO2CF3
4
5
12
81
F
F
F
F
SO2CF3
F
-ꢁ-
Cl
Br
24
24
48
0
F3C
Br
6
7
SO2CF3
-ꢁ-
24
0
SO2CF3
EtO2C
The reaction was carried out in DME at 70 °C; aryl bromide – 0.5 mmol, sulfone – 0.5 mmol, NaH – 1.5 mmol, Pd2dba3 – 2 mol%, PPh3/Pd ¼ 3:1,
DME – 4 ml, see Section 3 for details.
Table 3
Ligand effect in palladium-catalyzed arylation of pentafluorobenzyl
triflone with 4-bromobenzotrifluoridea
However, the acidity of PhCH2CH2SO2CF3 is too
low to make the reaction proceed (Table 2, entry 6). The
reaction also does not proceed with a very strong CH-
Ligand L
Conversion of ArBr (%)
Yield (%)
acid – CF3SO2CH2CO2Et – which gives the weakest
nucleophile after deprotonation (Table 2, entry 7).
In the light of the mechanism proposed, it seems
important to study the influence of the ligands on the
reactions. Using the palladium-catalyzed reaction of
pentafluorobenzyl trifluoromethyl sulfone with 4-bro-
mobenzotrifluoride as a model, it has been found, that
even a significant change in the donating properties of
the ligand only slightly influences the conversion of
aryl halide. The only exception is the case of the bid-
entate ligand DPPBz when the reaction proceeds very
slowly, this apparently being connected with a small
value of the bite angle of this ligand (Table 3).
In general, both the yield of the arylated product, and
the conversion of aryl bromide are only slightly influ-
enced by the ligand nature, but the yield is somewhat
lower when an electron-rich ligand PtBu3 is used. We
can point out, that such effect of an electron-rich ligand
is markedly different from that observed in the arylation
of many other CH-acids. A large number of CH-acidic
compounds, ketones, nitriles, esters, etc., were success-
fully arylated with high product yields using PtBu3 as a
ligand [5a,5d,5e].
AsPh3
PFu3
PPh3
PtBu3
DBPBPb
DPPBzc
66
65
63
75
51
13
55
54
51
37
40
8
a The reaction was carried out at 75 °C for 5 h in DME; 4-bro-
mobenzotrifluoride – 0.125 mmol, C6F5CH2SO2CF3 – 0.16 mmol,
NaH – 0.375, Pd2dba3 – 2 mol%, L/Pd ¼ 3:1, DME – 1 ml.
b DBPBP – 2-(di- t-butylphosphino)biphenyl.
c DPPBz – 1,2-bis(diphenylphosphino)benzene.
In order to determine whether the features observed
in the arylation reactions in the presence of PPh3 also
apply to the reactions using electron-rich bulky PtBu3
ligand, we studied the arylation reaction of several CH-
acids using this ligand and varying the amount of the
base – NaH (Table 4).
As it is shown above, in the presence of PPh3 the
studied CH-acids do not enter the arylation reaction
without a certain excess of the base. When electron-
rich bulky PtBu3 is used, some of the CH-acids behave
in a similar way, that is, they are not arylated without

A.V. Mitin et al. / Journal of Organometallic Chemistry 689 (2004) 1085–1090
1089
Table 4
Palladium-catalyzed arylation of CH-acids with aryl bromidesa
Entry CH-acid
Conditions
Stoichiometric amount of the base (NaH) Base (NaH) is present in excess
Time (h)
Conversion of ArBr,
% (Yield, %)
Time (h)
Conversion of
ArBr, % (Yield, %)
1
2
3
4
5
PhSO2CH2CO2Et
1-bromonaphthalene,
L ¼ PtBu3
17
15
12
12
–
55 (0)
40 (0)
16 (15)
7
12
12
12
12
10
100 (41)
PhSO2CH2CO2Et
NCCH2COOEt
NCCH2CN
1-bromonaphthalene,
b
100 (49)
100 (80)
100 (82)
22 (0)
L ¼ DBPBP
4-CF3C6H4Br,
L ¼ PtBu3
4-CF3C6H4Br,
L ¼ PtBu3
PhSO2CH(CH3)CO2Et
4-CF3C6H4Br, L ¼ PtBu3
–
a Heating at 70 °C in dioxane, aryl bromide – 0.5 mmol, Pd2dba3 – 2 mol%, L/Pd ¼ 3:1, dioxane – 4 ml; in the presence of the base excess: NaH –
1.25 mmol, CH-acid – 0.65 mmol; without of the base excess: NaH – 1.0 mmol, CH-acid – 1.25 mmol, see Section 3 for details.
b DBPBP – 2-(di-t-butylphosphino)biphenyl.
a certain excess of the base. The situation is somewhat
different, however, in the case of cyanoacetic ester: a
slow arylation reaction does take place without a cer-
tain excess of the base (Table 4, entry 3). The number
of CH-acids was arylated with no excess of the base
when the electron-rich ligands were used [1e,2b,5e]. The
application of electron-rich bulky PtBu3 also enables
one to carry out the arylation of some CH-acids having
tertiary carbon, such as the derivatives of esters [5e],
cyanoacetic ester, nitriles and fluoromalonic ester [10],
though the corresponding sulfones do not enter this
reaction (Table 4, entry 5). This data shows, that in
principle, the reductive elimination from the Ar-
PdL2CHXY complex without a deprotonation step is
possible for some CH-acids, provided that the electron-
rich bulky phosphine ligands are used; however, when
triphenylphosphine is used as a ligand, the deproto-
nation step is essential.
(3 ꢃ 10 ml). The extract was dried with Na2SO4 and was
filtered. The solvent was removed under reduced pres-
sure and the residue was chromatographed on silica gel
40–100 lm, eluting with light petroleum ether:ethyl ac-
etate mixture (v/v 4:1).
All products were purified with chromatography on
silica gel and were characterized by elemental analysis,
1
and H, 19F NMR and IR spectroscopy.
Acknowledgements
Financial support by the Russian Foundation for
Basic Research (Project Nos. 00-03-32766 and AO115)
is gratefully acknowledged.
References
3. Experimental
[1] (a) M. Uno, K. Seto, S. Takahashi, J. Chem. Soc. Chem.
Commun. (1984) 932;
(b) M. Uno, K. Seto, W. Ueda, M. Masuda, S. Takahashi,
Synthesis (1985) 506;
3.1. Typical procedure for the palladium-catalyzed reac-
tion of CH-acids with aryl bromides
(c) T. Sakamoto, E. Katoh, Y. Kondo, H. Yamanaka, Chem.
Pharm. Bull. 36 (1988) 1664;
The argon filled reactor equipped with condenser
cooler and stirring bar was charged with 0.65 mmol of
the CH-acid (Tables 1 and 2), 0.5 mmol of the aryl
bromide, 1.25 mmol of NaH, Pd2dba3 ꢂ CHCl3 – 2 mol%
and 0.06 mmol of the ligand (ligand/Pd ¼ 3:1). The
reaction mixture diluted with the degassed solvent (di-
oxane or DME, 4 ml) and degassed by three freeze-
pump-thaw cycles. The reactor was refilled with argon
and the reaction was carried out with stirring at 70 °C
under positive argon pressure. In a few hours later GLC
showed the absence of the starting aryl bromide. The
reaction mixture was cooled to room temperature,
mixed with brine, extracted with ether three times
(d) T. Sakamoto, E. Katoh, Y. Kondo, H. Yamanaka, Chem.
Pharm. Bull. 38 (1990) 1513;
(e) J. You, J.G. Verkade, Angew. Chem. Int. Ed. 42 (2003) 5051.
[2] (a) T. Satoh, Y. Kawamura, M. Miura, M. Nomura, Angew.
Chem. Int. Ed. Engl. 36 (1997) 1740;
(b) Y. Terao, Y. Kametani, H. Wakui, T. Satoh, M. Miura, M.
Nomura, Tetrahedron 57 (2001) 5967;
(c) B.C. Hamann, J.F. Hartwig, J. Am. Chem. Soc. 119 (1997)
12382;
(d) M. Palucki, S.L. Buchwald, J. Am. Chem. Soc. 119 (1997)
11108.
[3] Y. Tekao, Y. Fukuoka, T. Satoh, M. Miura, M. Nomura,
Tetrahedron Lett. 43 (2002) 101.
[4] K.H. Shaughnessy, B.C. Hamann, J.F. Hartwig, J. Org. Chem. 63
(1998) 6546.

1090
A.V. Mitin et al. / Journal of Organometallic Chemistry 689 (2004) 1085–1090
[5] (a) M. Kawatsura, J.F. Hartwig, J. Am. Chem. Soc. 121 (1999)
1473;
(f) D.A. Culkin, J.F. Hartwig, J. Am. Chem. Soc. 124 (2002)
9330.
(b) J.M. Fox, X. Huang, A. Chieffi, S.L. Buchwald, J. Am. Chem.
Soc. 122 (2000) 1360;
[6] E.M. Vog, S.L. Buchwald, J. Org. Chem. 67 (2002) 106.
[7] A.N. Kashin, A.V. Mitin, I.P. Beletskaya, R. Wife, Tetrahedron
Lett. 43 (2002) 2539.
(c) W.A. Moradi, S.L. Buchwald, J. Am. Chem. Soc. 123 (2001)
7996;
[8] J.P. Wolkowski, J.F. Hartwig, Angew. Chem. Int. Ed. 41 (2002)
4289.
(d) S. Lee, N.A. Beare, J.F. Hartwig, J. Am. Chem. Soc. 123 (2001)
8410;
[9] J.F. Hartwig, Angew. Chem. Int. Ed. 37 (1998) 2046.
[10] N.A. Beare, J.F. Hartwig, J. Org. Chem. 67 (2002) 541.
(e) M. Jørgensen, S. Lee, X. Liu, J.P. Wolkowski, J.F. Hartwig, J.
Am. Chem. Soc. 124 (2002) 12557;