Effect of phosphine and phosphine sulfide ligands on the cobalt-catalyzed reductive coupling of 2-iodobutane with n-butyl acrylate

Article abstract of DOI:10.1016/j.poly.2013.06.013

Tris[2-(diphenylphosphino)ethyl]phosphine disulfide (pp3S2), in which two terminal phosphino groups are selectively sulfidated, was prepared by utilizing the selective sulfidation reaction of [PdI(pp₃)]I. Co(II) complexes with bidentate, tridentate and tetradentate phosphines and pp3S2 were prepared from anhydrous CoI₂. X-ray crystal analyses revealed that the reaction of CoI₂ with 1 equivalent of 1,2-bis(diphen-ylphosphino) ethane (p₂) gave rise to partial oxidation of p₂ to give the dicationic octahedral [Co(p₂O₂)₂(CH ₃CN)₂]²⁺ (p₂O₂ = p ₂ dioxide) and dianionic p₂O-bridged tetrahedral dinuclear [CoI₃(p₂-O)CoI₃]² (p₂O = p₂ monooxide) complexes, while the reaction with 2 equivalents of p₂ gave the square-pyramidal [CoI(p₂)₂] ⁺ complex. The catalytic activity for the Co-catalyzed coupling reaction of 2-iodobutane with n-butyl acrylate was compared using multidentate phosphines and phosphine sulfides as ligands, and the efficiency of the phosphine sulfides was shown. The tendency for multidentate phosphine to deactivate the Co-catalysis can substantiate an oxidative addition driven mechanism in which the multidentate ligand should interfere with the formation of the alkyl halide Co(III) adduct and subsequent coordination of an alkene.

Full text of DOI:10.1016/j.poly.2013.06.013

Polyhedron 62 (2013) 37–41
Contents lists available at SciVerse ScienceDirect
Polyhedron
journal homepage: www.elsevier.com/locate/poly
Effect of phosphine and phosphine sulfide ligands on the
cobalt-catalyzed reductive coupling of 2-iodobutane with n-butyl
acrylate
a
b
Sen-ichi Aizawa a, , Koichi Fukumoto , Tatsuya Kawamoto
⇑
a Graduate School of Science and Engineering, University of Toyama, 3190 Gofuku, Toyama 930-8555, Japan
b Department of Chemistry, Faculty of Science, Kanagawa University, 2946 Tsuchiya, Hiratsuka, Kanagawa 259-1293, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Tris[2-(diphenylphosphino)ethyl]phosphine disulfide (pp3S2), in which two terminal phosphino groups
are selectively sulfidated, was prepared by utilizing the selective sulfidation reaction of [PdI(pp3)]I. Co(II)
complexes with bidentate, tridentate and tetradentate phosphines and pp3S2 were prepared from anhy-
drous CoI2. X-ray crystal analyses revealed that the reaction of CoI2 with 1 equivalent of 1,2-bis(diphen-
ylphosphino)ethane (p2) gave rise to partial oxidation of p2 to give the dicationic octahedral
[Co(p2O2)2(CH3CN)2]2+ (p2O2 = p2 dioxide) and dianionic p2O-bridged tetrahedral dinuclear [CoI3(p2-
O)CoI3]2ꢀ (p2O = p2 monooxide) complexes, while the reaction with 2 equivalents of p2 gave the
square-pyramidal [CoI(p2)2]+ complex. The catalytic activity for the Co-catalyzed coupling reaction of
2-iodobutane with n-butyl acrylate was compared using multidentate phosphines and phosphine sul-
fides as ligands, and the efficiency of the phosphine sulfides was shown. The tendency for multidentate
phosphine to deactivate the Co-catalysis can substantiate an oxidative addition driven mechanism in
which the multidentate ligand should interfere with the formation of the alkyl halide Co(III) adduct
and subsequent coordination of an alkene.
Received 22 April 2013
Accepted 3 June 2013
Available online 20 June 2013
Keywords:
Cobalt-catalyzed reductive coupling
Activated alkene
Oxidative addition driven mechanism
Phosphine sulfide ligand
Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction
Co(I) with an alkyl halide. Although PPh3 or bidentate phosphines
were used for the bound ligand in that work, the steric and elec-
Carbon–carbon bond formation reactions by transition metal
catalysts are now quite versatile and indispensable in organic syn-
theses. In particular, direct C–H bond transformation is a highly
attractive strategy because of the commonness of C–H bonds [1].
Although some of the second row transition metal catalysts, such
as Rh, Pd and Ru complexes, have proven to be highly effective
in the direct coupling processes [2], the development of new
cost-effective catalysts is significant for industrial use. In line with
the above challenge, several Co-catalyzed reactions using a Grig-
nard reagent have been reported, where a radical generation mech-
anism was proposed [3]. However, these catalytic reactions cannot
be used for the coupling of an alkyl halide with an electron-
withdrawing alkene because Grignard reagents are quite reactive
to the activated alkene. Recently, Cheng et al. reported efficient
Co-catalyzed Csp3 -Csp3 bond forming reactions of primary, second-
ary and tertiary alkyl halides with electron-withdrawing alkenes
under mild conditions without Grignard reagents [4], and they pro-
posed an oxidative addition driven mechanism in which an alkene
is bound to a Co(III) adduct formed by the oxidative addition of
tronic properties of the ligand on the Co center should be quite sig-
nificant factors affecting the catalytic activity in the oxidative
addition driven mechanism. Thus, in the present report, the steric
and electronic effects of ligands are investigated by using some ste-
rically different multidentate phosphines and more
p-accepting
and polarizable phosphine sulfides, which are electronically more
efficient for the oxidative addition–reductive elimination mecha-
nism in the Pd(0) catalytic cycle [5]. The observed differences in
the catalytic activity provide further insight into the reaction
mechanism of the catalytic cycle from the view point of coordina-
tion chemistry.
2. Experimental
2.1. Reagents
Tris[2-(diphenylphosphino)ethyl]phosphine (pp3, Aldrich),
bis[2-(diphenylphosphino)ethyl]phenylphosphine (p3, Aldrich),
1,2-bis(diphenylphosphino)ethane (p2, Wako), anhydrous cobal-
t(II) iodide (Kishida Chemical), tetrakis(acetonitrile)palladium(II)
tetrafluoroborate ([Pd(CH3CN)4](BF4)2, Aldrich), tetra(n-butyl)
ammonium iodide (Wako), sulfur (Wako), zinc powder (Wako),
⇑
Corresponding author. Tel./fax: +81 76 445 6980.
E-mail address: saizawa@eng.u-toyama.ac.jp (S.-i. Aizawa).
0277-5387/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.poly.2013.06.013

38
S.-i. Aizawa et al. / Polyhedron 62 (2013) 37–41
2-iodobutane (Tokyo Chemical), n-butyl acrylate (Kishida Chemi-
cal) and dibenzyl ether (Kishida Chemical) were used for the prep-
aration and catalytic reactions without further purification.
2.2.5. [CoI2(L)] (L = pp3 (4), p3 (5), and pp3S2 (6))
These phosphine and phosphine disulfide complexes were ob-
tained by a procedure similar to that described for 1, using 1 equiv-
alent of pp3 for 4, p3 for 5 and pp3S2 for 6 instead of 2 equivalents of
p2. Yields: 0.14 g (74%) for 4, 0.10 g (63%) for 5 and 0.16 g (72%) for
6. Anal. Calc. for C42H42P4CoI2 (4)ꢁ1.5CHCl3: C, 44.94; H, 3.77; N,
0.00; S, 0.00. Found: C, 44.82; H, 3.88; N, 0.04; S, 0.00%. Anal. Calc.
for C34H33P3CoI2 (5)ꢁCHCl3: C, 43.49; H, 3.55; N, 0.00; S, 0.00.
Found: C, 44.19; H, 3.60; N, 0.00; S, 0.00%. Anal. Calc. for C42H42P4-
S2CoI2 (6)ꢁ2C4H4Oꢁ1.5CHCl3: C, 45.39; H, 3.81; N, 0.00; S, 4.71.
Found: C, 45.16; H, 3.88; N, 0.08; S, 4.59%.
2.2. Preparation
2.2.1. Tris[2-(diphenylphosphino)ethyl]phosphine disulfide (pp3S2)
The disulfide complex [PdI2(pp3S2)] was prepared by the reac-
tion of [PdI(pp3)]I [6] (0.362 g, 0.351 mmol) with sulfur (0.0225 g,
0.702 mmol) in degassed chloroform (ca. 10 cm3) under N2 at
40 °C for 3 days. The reaction solution was concentrated to a small
volume, and to this was added diethyl ether. The resultant yellow
solid was filtered and air-dried. Yield: 0.28 g (70%). Anal. Calc. for
2.2.6. [CoI(pp3)] (7)
This Co(I) complex was obtained by a procedure similar to that
C
42H42I2P4S2Pdꢁ0.5CHCl3: C, 44.20; H, 3.71; N, 0.00; S, 5.55. Found:
described in the literature [8]. 31P{1H} NMR (CHCl3) d: 59.5 (d, 3P,
C, 44.48; H, 3.75; N, 0.00; S, 5.57%. 31P{1H} NMR (CHCl3) d: 44.4 (d,
3
terminal), 157.9 (q, 1P, center); JP(center)-P(terminal) = 37.2 Hz.
3
2P, sulfide), 70.0 (d, 1P, terminal), 78.1 (m, 1P, center); JP(sulfide)-
P(center) = 55.0 Hz, JP(terminal)-P(center) = 17.8 Hz.
3
2.3. Crystal structure determination
The disulfide Pd(II) complex prepared above, [PdI2(pp3S2)]-
ꢁ0.5CHCl3 (0.190 g, 0.165 mmol), and pp3 (0.122 g, 0.182 mmol)
were dissolved in degassed chloroform (ca. 5 cm3) and reacted un-
der N2 at 50 °C for 1 day. The formed red pp3 complex [PdI(pp3)]I
[6] was removed by silica gel column chromatography with chloro-
form as the eluent. The 31P NMR spectrum of the concentrated col-
orless eluate indicated that the eluate contained unreacted pp3 and
dissociated free pp3S2. The concentrated chloroform eluate was
mixed with an aqueous solution of K2[PdCl4] to form red
[PdCl(pp3)]Cl [6,7] in the chloroform layer. Aqueous K2[PdCl4]
was added until the 31P NMR signals of pp3 completely disap-
peared, showing only the signals of pp3S2 and [PdCl(pp3)]Cl [6,7].
Then, the red Pd(II) complex was removed by silica gel column
chromatography, eluting with chloroform. The colorless eluate
was concentrated and to this was added diethyl ether to obtain
pure pp3S2. Yield: 0.049 g (40%). Anal. Calc. for C42H42P4S2ꢁ0.5H2O:
C, 67.82; H, 5.82; N, 0.00; S, 8.62. Found: C, 67.81; H, 5.67; N, 0.00;
S, 8.36%. 31P{1H} NMR (CHCl3) d: ꢀ19.7 (q, 1P, center), ꢀ13.1 (d, 1P,
The measurements for 1ꢁ3CHCl3 and 2 were made on a Rigaku
Mercury CCD X-ray diffractometer using graphite-monochromated
Mo Ka (k = 0.71075 Å) radiation at 200 K. Cell parameters were re-
fined using the program CRYSTALCLEAR (Rigaku and MSC, version 1.3,
2001) and the collected data were reduced using the program CRYS-
TALSTRUCTURE (Rigaku and MSC, version 3.8, 2006). Empirical absorp-
tion corrections were applied. The structures were solved by direct
methods using SIR 92 [9] and refined by full-matrix least-squares
techniques using SHELXL-97 [10]. All non-hydrogen atoms for
1ꢁ3CHCl3 and 2 were refined anisotropically and hydrogen atoms
were included in calculated positions. The structure of 2 contains
CoI3 moieties disordered over two different orientations for I1
and three different orientations for I2, I3, I4, I5, and I61
.
2.4. General procedure for the cobalt-catalyzed coupling reaction
3
3
Reactions of n-butyl acrylate (0.128 g, 1.00 mmol) with 2-iodo-
butane (0.184 g, 1.00 mmol) were carried out in a sealed tube con-
taining degassed acetonitrile (2 cm3), the cobalt catalyst
(0.040 mmol), zinc powder (0.164 g, 2.50 mmol), water (0.018 g,
1.0 mmol) and dibenzyl ether (0.015 g, 0.075 mmol) under N2 at
80 °C. The yields were calculated by the 1H NMR intensity of the
methylene protons of the formed n-butyl-4-methylhexanoate on
the basis of the intensity of the methylene protons of dibenzyl
ether contained as an internal reference and followed as a function
of time.
terminal), 44.4 (d, 2P, sulfide); JP(center)-P(terminal) = 27.5 Hz, JP(cen-
ter)-P(sulfide) = 30.7 Hz.
2.2.2. [CoI(p2)2]I (1)
Anhydrous CoI2 (0.0510 g, 0.163 mmol) and p2 (0.130 g,
0.326 mmol) were dissolved in degassed THF and reacted under
N2 at room temperature for 2 h. The reaction solution was concen-
trated under N2, and to this was added degassed diethyl ether to
obtain the resultant brown solid by filtration. The solid was recrys-
tallized from chloroform, and single crystals suitable for an X-ray
analysis were obtained. Yield: 0.13 g (55%). Anal. Calc. for C52H48P4-
CoI2 (1)ꢁCHCl3: C, 50.82; H, 3.94; N, 0.00. Found: C, 50.58; H, 3.87;
N, 0.00%.
2.5. Measurements
31P and 1H NMR spectra for solutions were recorded on a JEOL
JNM-A400 FT-NMR spectrometer operating at 160.70 and
399.65 MHz, respectively. In order to determine the chemical shifts
of the 31P NMR signals, a 3-mm-o.d. NMR tube containing the sam-
ple solution was coaxially mounted in a 5-mm-o.d. NMR tube con-
taining deuterated water as a lock solvent and phosphoric acid as a
reference.
2.2.3. [Co(p2O2)2(CH3CN)2][CoI3(p2O)CoI3] (2)
This complex salt was obtained by a procedure similar to that
described for 1, using 1 equivalent of p2 instead of 2 equivalents.
Single crystals suitable for an X-ray analysis were obtained by
recrystallization from acetonitrile. Yield: 0.050 g (40%). Anal. Calc.
for C82H78N2O5P6Co3I6 (2): C, 42.91; H, 3.42; N, 1.22. Found: C,
42.67; H, 3.45; N, 1.09%.
1
ꢀ
Crystal data for 1ꢁ3CHCl3: C55H51Cl9CoI2P4, Mr = 1467.62, triclinic, space group P1,
a = 14.4437(15) Å, b = 20.182(2) Å, c = 20.612(2) Å,
a = 93.404(3)° b = 91.004(3)°
c
= 93.107(3)° V = 5987.8(11) Å3, Z = 4, qcalcd = 1.628 Mg mꢀ3
,
crystal size
0.26 ꢂ 0.23 ꢂ 0.12 mm3, F(000) = 2908,
l(Mo Ka , GOF = 1.080,
) = 1.862 mmꢀ1
2.2.4. [CoI(p2O2)2]I (3)
26866 independent reflections (Rint = 0.0588). The final R indicates were R1 = 0.0635
(I > 2
r(I)) and wR2 = 0.1641 (all data). Crystal data for 2: C82H78Co3I6N2O5P6,
This bis(P2O2) complex was obtained by a procedure similar to
that described for 1, using P2O2 prepared by the oxidation of p2
with 10% aqueous H2O2 in chloroform instead of P2. Yield: 0.16 g
(65%). Anal. Calc. for C64H60O4P4CoI2 (3)ꢁ3C4H4O: C, 55.79; H,
4.39; N, 0.00; S, 0.00. Found: C, 55.90; H, 4.37; N, 0.07; S, 0.00%.
Mr = 2295.47, monoclinic, space group P21, a = 11.583(3) Å, b = 20.038(5) Å,
c = 20.039(5) Å, b = 105.090(8)° V = 4491(2) Å3, Z = 2,
q
calcd = 1.698 Mg mꢀ3, crystal
size 0.25 ꢂ 0.10 ꢂ 0.05 mm3, F(000) = 2226,
l(Mo Ka
) = 2.764 mmꢀ1, GOF = 1.117,
20075 independent reflections (Rint = 0.0577). The final R indicates were R1 = 0.0819
(I > 2r(I)) and wR2 = 0.1411 (all data).

S.-i. Aizawa et al. / Polyhedron 62 (2013) 37–41
39
3. Results and discussion
3.1. Synthesis
The selective sulfidation of two terminal phosphino groups of
the tetradentate tris[2-(diphenylphosphino)ethyl]phosphine (pp3)
was reported previously for the five-coordinated trigonal-bipyra-
midal Pd(II) complex, [PdCl(pp3)]Cl [11]. The disulfide complex ex-
ists in equilibrium between the five-coordinate pseudo-trigonal-
bipyramidal complex and the four-coordinate square-planar one
with two pendant phosphine sulfide groups in chloroform [11a].
In this work, such reactivity of the pp3 complex was applied to
the preparation of pp3S2 (pp3 disulfide) (Scheme 1). The selective
sulfidation of two terminal phosphino groups of pp3 was carried
out with [PdI(pp3)]I instead of the corresponding chloro complex,
[PdCl(pp3)]Cl. The obtained disulfide complex exhibited 31P NMR
behavior similar to that of the chloro complex, with broadening
of the signals, suggesting an equilibrium between the five and four
coordinate structures as observed for the chloro complex [11a].
The pp3S2 ligand was successfully dissociated from the Pd(II) com-
plex by competitive coordination of pp3 to form the more stable
five coordinate [PdI(pp3)]I [6]. The reason for use of the iodo com-
plex instead of the chloro one is that the electron-accepting ability
of the iodo ligand is greater than that of chloro one, which may
Fig. 1. ORTEP diagram of 1ꢁ3CHCl3.
3.2. Crystal structure
The perspective views of 1ꢁ3CHCl3 and 2 are shown in Figs. 1
and 2, respectively. The bis(p2) complex 1, prepared with a 1:2
CoI2/p2 ratio, takes a five-coordinate square-pyramidal geometry
with one iodide group in the apical position. The other iodide ion
is surrounded by three chloroform molecules. The chelate bite an-
gles of the p2 bidentate ligands are considerably smaller than 90°
(P1-Co1-P2 = 82.09(5)° and P3-Co1-P4 = 81.57(5)°). This structure
indicates that the five-membered chelate of p2 is too small to form
a regular square plane for the Co(II) centre. It is obvious that forma-
tion of a tetrahedral geometry, expected for the mono(p2) complex
[CoI2(p2)], would give rise to further distortion because the chelate
bite angle should be 109° for a regular tetrahedral geometry.
Complex 2, which was obtained from a 1:1 CoI2/p2 ratio, con-
sists of the dicationic octahedral complex [Co(p2O2)2(CH3CN)2]2+
and the dianionic p2O-bridged dinuclear complex with two tetra-
hedral Co(II)I3 moieties linked by phosphino and phosphine oxide
groups. The complex cation takes a nearly regular octahedral
geometry in which the chelate bite angles of the p2O2 ligands are
almost 90° (89.6–90.7°). This is attributed to the flexibility and ste-
ric adjustability of the seven-membered p2O2 chelate. Though large
deviations of the bond distances and angles were observed in the
two distorted tetrahedral Co(II) moieties, the averaged I–Co–I bond
angles for the phosphine complex moiety (114.1°) and the phos-
phine oxide complex moiety (113.8°) are obviously larger than
promote coordination of phosphine
r donors to form the five coor-
dinate pp3 complex [6], and accordingly the pp3S2 ligand may dis-
sociate more easily from the pp3S2 complex. Furthermore, taking
advantage of the stability of the five-coordinate Pd(II) complex
with the tetradentate pp3 ligand compared to the pp3S2 complex
equilibrated between the four and five-coordinate structures,
unreacted pp3 was successfully removed from the mixture of free
pp3 and pp3S2 by complexation with K2[PdCl4], followed by
separation with a silica gel column (see Section 2).
For a comparison of the catalytic activity of the Co(II) complexes
with different bound ligands, the preparation of the 1:1 complexes
with a phosphine or phosphine chalcogenide was attempted.
However, the reaction of anhydrous CoI2 with 1 equiv of 1,2-
bis(diphenylphosphino)ethane (p2) gave rise to partial oxidation
of p2, probably during recrystallization, to give 2 which consists
of the dicationic octahedral bis(p2O2) complex, [Co(p2O2)2(CH3-
CN)2]2+ (p2O2 = p2 dioxide), and the dianionic p2O-bridged dinucle-
ar tetrahedral complex, [CoI3(p2O)CoI3]2ꢀ (p2O = p2 monooxide), in
which p2O does not act as a bidentate chelate (see below). This fact
indicates that the 1:1 complex, [CoI2(p2)], is not stable because the
p2 chelate bite distance is too small to give the tetrahedral Co(II)
complex (see below). Consequently, the 1:2 complexes were pre-
pared using 2 equivalents of p2 or p2O2 to give 1 and 3.
S
P
+
I
+
P
Pd
I
P
Pd
I
P
S
P
P
S
S
P
2 S
P
I
P
P
P
P
I
Pd
I
[PdI(pp3S2)]I
[PdI(pp3)]I
[PdI2(pp3S2)]
+
+
I
P
P
Pd
I
K2[PdCl4]
pp3
P
P
P
P
pp3
pp3S2
+
pp3S2
P
Pd
Cl
+
Cl
+
P
(isolated)
[PdCl(pp3)]Cl
[PdI(pp3)]I
removed by column
chromatography
removed by column
chromatography
Scheme 1. Isolation of pp3S2.

40
S.-i. Aizawa et al. / Polyhedron 62 (2013) 37–41
Fig. 3. Change in the yield of butyl-4-methylhexanoate with time in the cobalt-
catalyzed coupling reaction using 1 (s), 3 (d), 4 (4), 5 (N), 6 (h) and a mixture of
CoI2 and pp3S4 (j).
3.3. Cobalt-catalyzed coupling reaction
The catalytic activities of the p2, pp3 and bis[2-(diphenylphos-
phino)ethyl]phenylphosphine (p3,) phosphine complexes, 1,
4
and 5, and the pp3S2 phosphine sulfide complex 6 for the cobalt-
catalyzed coupling reaction of an alkyl halide with an alkene were
compared using 2-iodobutane and n-butyl acrylate as the sub-
strates. The changes in the yield of the formed n-butyl-4-methyl-
hexanoate are plotted in Fig. 3. If an excess of n-butyl acrylate
(more than 2 equivalents) and more Co catalyst (0.1 equivalents)
were used, as described in the literature [4], a much higher yield
was obtained for each ligand. However, in the present experiments,
equimolar amounts of substrates and 0.04 equivalents of the Co
catalysts were used in order to make the difference in the activity
clearer. The catalytic activity of the phosphine sulfide complex 6 is
definitely higher than those of the phosphine complexes 1, 4 and 5.
The efficiency of phosphine sulfides was also shown by the reac-
tion with a mixture of pp3S4 (pp3 tetrasulfide) [5a] and CoI2
(Fig. 3). Among the phosphine complexes, complex 4, with the tet-
radentate phosphine ligand pp3, showed the lowest activity, and
the bidentate phosphine, p2, does not deactivate the catalysis even
in the case of the bis(p2) complex 1. On the other hand, the phos-
phine oxide group gives rise to deactivation, as observed for the
reaction using the bis(p2O2) complex 3 as an analogue of 1.
Phosphines have been reported to be quite versatile ligands so
far, especially for Pd catalyzed coupling reactions [12]. However,
phosphines are usually susceptible to oxidation to give inactive
Pd black and addition of excess phosphines inhibits the catalytic
reactions by blocking the substrate coordination. More recently,
it has been reported that phosphine sulfides are more effective li-
gands for some Pd catalyzed C–C coupling reactions because phos-
phine sulfides can stabilize the active Pd(0) species, even in the air,
and do not interfere with the oxidative addition of Pd(0) and sub-
sequent coordination of substrates on Pd(II) [5]. The nature of P@X
(X = O, S, Se) bonds has been extensively discussed and reviewed
Fig. 2. ORTEP diagram of
2
showing [Co(p2O2)2(CH3CN)2]2+ (a) and [CoI3(p2-
O)CoI3]2ꢀ (b), individually.
the averaged I–Co–P (103.5°) and I–Co–O (104.7°) bond angles, due
to the repulsion between the large iodide anions. It should be
noted that the averaged Co–I bond distance for the phosphine com-
plex moiety (2.52 Å) is considerably shorter than that for the phos-
phine oxide complex moiety (2.61 Å), in spite of the larger steric
bulkiness of the diphenylphosphino group in the coordination
sphere compared to the phosphine oxide group. Such a difference
in bond distances may suggest that the electron donation of the
soft diphenylphosphino group to the hard tetrahedral Co(II) ion
is relatively weaker than that of the hard phosphine oxide group,
resulting in smaller electronic repulsion on the Co(II) center with
a phosphino group rather than that with a phosphine oxide group.
This is consistent with the fact that the Co–P distance in the tetra-
hedral Co(II) moiety in 2 (2.463(5) Å) is significantly longer than
the averaged Co–P distance of the five-coordinate Co(II) complex
1 (2.283 Å), though the ionic radius of four-coordinate tetrahedral
Co(II) should be smaller than those for five- or six-coordinate
Co(II). On the other hand, the Co–O distance in the tetrahedral
Co(II) moiety in 2 (1.892(8) Å) is reasonably shorter than the aver-
aged Co–O distance in the six-coordinate octahedral Co(II) moiety
in 2 (2.036 Å).
over the years [13]. In line with these discussions, the
p accepting
ability of the P@S bond of phosphine sulfides can stabilize low-va-
lent metal centers to promote the generation of the catalytically
active species. Also the polarizability of the P@S bond can be effec-
tive for coordination to higher positively charged metal ions, so as
not to interfere with the oxidative addition of the active species.
Furthermore, because phosphine sulfides are weak
r donors, form-
ing relatively weak bonds, the coordination of another substrate
r
and subsequent migration on the oxidized metal center are not
blocked. The polarization of P@S bonds by coordination is observed
even in the solid state, that is the P@S distances for coordinated
p2S2 and pOp2S3 ligands (2.009–2.018 Å for [PdCl2(p2S2)] [14],

S.-i. Aizawa et al. / Polyhedron 62 (2013) 37–41
41
Zn(II)
X
the bis(p2) complex 1 may be attributed to blocking the formation
of the alkyl halide adduct and subsequent coordination of an al-
kene by the robust coordination structure with the tetradentate li-
gand. On the other hand, the tetradentate phosphine sulfides pp3S2
and pp3S4 do not interfere with the substrate coordination, proba-
Co(I) L
R1
X
Zn
Oxidative
addition
Reduction
of Co(III)
X Co(III) L
X
R1
bly due to partial dissociation of the weakly
r-donating phosphine
R1 Co(III) L
sulfide groups on Co(III). These results are consistent with an oxi-
dative addition driven mechanism with alkene coordination to the
alkyl halide adduct.
OH
Elimination with
protonation
Alkene
coordination
R2
Insersion
of alkene
H2O
R1
X
X
R1 Co(III) L
Acknowledgments
Co(III) L
R2
R2
R2
This research was supported by Grant-in-Aid for Scientific Re-
search (No. 23550068) from the Japan Society for the Promotion
of Science.
R
1 = Alkyl group
R2 = Electron-withdrawing group
Scheme 2. Oxidative addition driven mechanism.
Appendix A. Supplementary data
CCDC 923329 and 923330 contain the supplementary crystallo-
graphic data for 1 and 2, respectively. These data can be obtained
free of charge via http://www.ccdc.cam.ac.uk/conts/retriev-
ing.html, or from the Cambridge Crystallographic Data Centre, 12
Union Road, Cambridge CB2 1EZ, UK; fax: +44 1223 336 033; or
e-mail: deposit@ccdc.cam.ac.uk.
Zn(II)
X
R1
X
Co(I) L
Zn
Reduction
of Co(III)
Electron
transfer
X
R1
X
X Co(III) L
Co(II) L
R1
Elimination with
protonation
R1
Radical
addition
OH
References
R2
Co(III) L
R1
[1] (a) D. Alberico, M.E. Scott, M. Lautens, Chem. Rev. 107 (2007) 174;
(b) M. Miura, T. Satoh, Top. Organomet. Chem. 14 (2005) 55;
(c) L.-C. Campeau, K. Fagnou, Chem. Commun. (2006) 1253;
(d) I.Y. Seregin, V. Gevorgyan, Chem. Soc. Rev. 36 (2007) 1173;
(e) J. Ellman, Science 316 (2007) 1131.
[2] (a) S. Proch, R. Kempe, Angew. Chem., Int. Ed. 46 (2007) 3135;
(b) M. Lafrance, K. Fagnou, J. Am. Chem. Soc. 128 (2006) 16496;
(c) L. Ackermann, P. Nova`k, R. Vicente, N. Hofmann, Angew. Chem., Int. Ed. 48
(2009) 6045;
H2O
R1
R2
R2
R2
R
R
1 = Alkyl group
2 = Electron-withdrawing group
Scheme 3. Radical generation mechanism.
(d) M. Tobisu, I. Hyodo, N. Chatani, J. Am. Chem. Soc. 131 (2009) 12070.
[3] (a) Y. Ikeda, T. Nakamura, H. Yorimitsu, K.J. Oshima, Am. Chem. Soc. 124
(2002) 6514;
(b) T. Fujioka, T. Nakamura, H. Yorimitsu, K. Oshima, Org. Lett. 4 (2002) 2257;
(c) H. Ohmiya, T. Tsuji, H. Yorimitsu, K. Oshima, Chem. Eur. J. 10 (2004) 5640.
[4] P. Shukla, Y.-C. Hsu, C.-H.J. Cheng, Org. Chem. 71 (2006) 655.
[5] (a) S. Aizawa, A. Majumder, Y. Yokoyama, M. Tamai, D. Maeda, A. Kitamura,
Organometallics 28 (2009) 6067;
1.984–2.005 Å for [Pd(p2S2)2](BF4)2 [5b] and 2.005–2.011 Å for
[Pd3Cl6(pp3S3)2] [15]) are definitely elongated compared with
those in the free phosphine sulfide, e.g. pp3S4 (1.957–1.958 Å)
[16], while such a difference in bond distances is not clearly ob-
served for phosphine oxides (1.449–1.518 Å for the bound P@O
in 6 and 1.473–1.487 Å for the pendant P@O in [Pd(4-ClC6H4S)(pp3-
O)]+ and [Pd(4-ClC6H4S)2(pp3O2)] [17]). The present comparison of
the catalytic activities also indicates the efficiency of phosphine
sulfides in the case of the Co(I)/Co(III) catalytic cycle. On the other
hand, the phosphine oxide p2O2 deactivated the Co catalyst, prob-
(b) S. Aizawa, M. Kondo, R. Miyatake, M. Tamai, Inorg. Chim. Acta 360 (2007)
2809.
[6] S. Aizawa, T. Iida, S. Funahashi, Inorg. Chem. 35 (1996) 5163.
}
[7] D. Fernández, M.I. G.-Seijo, T. Kégl, G. Petocz, L. Kollár, M.E. G.-Fernández,
Inorg. Chem. 41 (2002) 4435.
[8] (a) C.A. Ghilardi, S. Midollini, L. Sacconi, Inorg. Chem. 14 (1975) 1790;
(b) W.H. Hohman, D.J. Kountz, D.W. Meek, Inorg. Chem. 25 (1986) 616.
[9] A. Altomare, G. Cascarano, C. Giacovazzo, A. Guagliardi, M.C. Burla, G. Polidori,
M. Camalli, J. Appl. Crystallogr. 27 (1994) 435.
[10] G.M. Sheldrick, SHELXL97, Program for Crystal Structure Refinement, University
of Göttingen, Germany, 1997.
[11] (a) S. Aizawa, T. Hase, T.J. Wada, Organomet. Chem. 692 (2007) 813;
(b) S. Aizawa, T. Kawamoto, Y. Asai, C.J. Ishimura, Organomet. Chem. 695
(2010) 1253.
[12] (a) J. Tsuji, Palladium Reagents and Catalysts, Wiley, Chichester, 1995;
(b) R.F. Heck, Palladium Reagents in Organic Synthesis, VCH, Weinheim, 1996;
(c) A.J. Suzuki, Organomet. Chem. 576 (1999) 147;
ably due to the poor
p accepting ability resulting in destabilization
of the Co(I) active species.
Two different reaction mechanisms for the catalytic cycle were
proposed recently [4]. One is an oxidative addition driven mecha-
nism (Scheme 2) in which the oxidative addition of a Co(I) species
with an alkyl halide gives the alkyl Co(III) intermediate, followed
by coordination of an alkene. The alkene is then inserted into the
Co-alkyl bond, protonation gives the final reductive coupling prod-
uct and reduction of the Co(III) species by zinc powder regenerates
the active Co(I) species. The other is a radical generation mecha-
nism (Scheme 3) in which electron transfer from a low-valent
Co(I) species to an alkyl halide generates an alkyl radical, followed
by radical addition to the acrylate, trapping by the Co(II) species to
give the Co(III) intermediate. Subsequent protonation gives the fi-
nal product, and reduction of Co(III), similar to the oxidative addi-
tion driven mechanism, regenerates the Co(I) species. It is obvious
that the active catalyst is the Co(I) complex, generated by reduc-
tion of the Co(II) species with zinc powder, because the coupling
product, n-butyl-4-methylhexanoate, was stoichiometrically
formed by the reaction of the Co(I) complex 7 without zinc powder.
Deactivation of complex 4 with the tetradentate pp3 compared to
(d) I.P. Beletskaya, A.V. Cheprakov, Chem. Rev. 100 (2000) 3009.
[13] (a) A. Streitwieser Jr., A. Rajca, R.S. McDowell, R.J. Glaser, Am. Chem. Soc. 109
(1987) 4184;
(b) D.G. Gilheany, Chem. Rev. 94 (1994) 1339;
(c) A.E. Reed, P.R.J. Schleyer, Am. Chem. Soc. 112 (1990) 1434;
(d) M.W. Schmidt, S. Yabushita, M.S.J. Gordon, Phys. Chem. 88 (1984) 382;
(e) P.A. Schultz, R.P.J. Messmer, Am. Chem. Soc. 115 (1993) 10925;
(f) P.A. Schultz, R.P.J. Messmer, Am. Chem. Soc. 115 (1993) 10938;
(g) J.A. Dobado, H. Martínez-García, J.M. Molina, M.R.J. Sundberg, Am. Chem.
Soc. 120 (1998) 8461.
[14] T.S. Labana, R. Verma, A. Singh, M. Shikha, A. Castineiras, Polyhedron 21 (2002)
205.
[15] S. Aizawa, T. Kawamoto, S. Nishigaki, A.J. Sasaki, Organomet. Chem. 696 (2011)
2471.
[16] B. Deb, P.P. Sarmah, K. Saikia, A.L. Fuller, R.A.M. Randall, A.M.Z. Slawin, J.D.
Woollins, D.K.J. Dutta, Organomet. Chem. 696 (2011) 3279.
[17] S. Aizawa, T. Kawamoto, K. Saito, Inorg. Chim. Acta 357 (2004) 2191.