Zirconium(IV)- and hafnium(IV)-containing polyoxometalates as oxidation precatalysts: Homogeneous catalytic epoxidation of cyclooctene by hydrogen peroxide

Article abstract of DOI:10.1016/j.molcata.2014.07.020

The homogeneous epoxidation of cis-cyclooctene with H₂O ₂, catalyzed by various Zr/Hf-containing Keggin sandwich polyoxometalates (POMs), i.e., the 1:2 complexes [M(α-PW ₁₁O₃₉)₂]^10- (M = Zr, 1 and M = Hf, 2), the 4:2 complexes [[{M(H₂O)}₂{M(H₂O) ₂}₂(μ-OH)₃(μ-OH)₂](α-1, 2-PW₁₀O₃₇)₂]^7- (M = Zr, 3 and M = Hf, 4), the 2:2 complexes [{M(μ-OH)(H₂O)}₂(α- PW₁₁O₃₉)₂]^8- (M = Zr, 5 and M = Hf, 6), and the 3:2 complex [Zr₃(μ₂-OH)₃(A- α-PW₉O₃₄)₂]^9- (7), was examined. At least two different reaction types exist: (1) one brought about by the highly catalytically active Venturello complex [PO₄{WO(O ₂)₂}₄]^3-, which was generated by reactions of the Zr/Hf-containing POMs with hydrogen peroxide, and (2) the reaction with active species formed on Zr/Hf clusters in the POMs. In the first type, the original sandwich structures are not maintained during the reactions (as shown in 1, 2, 3, and 4), but in the second type, the sandwich structures are maintained (as seen in 5 and 6). 7 was inactive. The reaction, (5 + cis-cyclooctene) + H₂O₂, was explicitly influenced by changing the reaction time of 5 and olefin, before addition of H ₂O₂, indicating a significant interaction between 5 and olefin, such as coordination of olefin to the POM. Thus, for homogeneous olefin epoxidation by hydrogen peroxide catalyzed by di-Zr/Hf clusters, a new mechanism is proposed through olefin-coordinated species to di-Zr/Hf centers.

Full text of DOI:10.1016/j.molcata.2014.07.020

Journal of Molecular Catalysis A: Chemical 394 (2014) 224–231
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Editor’s Choice paper
Zirconium(IV)- and hafnium(IV)-containing polyoxometalates as
oxidation precatalysts: Homogeneous catalytic epoxidation of
cyclooctene by hydrogen peroxide
Hiroki Aoto, Keisuke Matsui, Yoshitaka Sakai, Teppei Kuchizi, Hiromi Sekiya,
∗
Hironori Osada, Takuya Yoshida, Satoshi Matsunaga, Kenji Nomiya
Department of Chemistry, Faculty of Science, Kanagawa University, 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:
Received 8 May 2014
Received in revised form 8 July 2014
Accepted 12 July 2014
Available online 19 July 2014
The homogeneous epoxidation of cis-cyclooctene with H O , catalyzed by various Zr/Hf-containing
2
2
10−
Keggin sandwich polyoxometalates (POMs), i.e., the 1:2 complexes [M(␣-PW11O39)2]
(M = Zr, 1 and
7−
M = Hf, 2), the 4:2 complexes [[{M(H2O)}2{M(H2O)2}2(␮-OH)3(␮-OH)2](␣-1,2-PW10O37)2] (M = Zr, 3
8−
and M = Hf, 4), the 2:2 complexes [{M(␮-OH)(H2O)}2(␣-PW11O39)2] (M = Zr, 5 and M = Hf, 6), and the 3:2
9
−
complex [Zr3(␮2-OH)3(A-␣-PW9O34)2] (7), was examined. At least two different reaction types exist:
3−
(1) one brought about by the highly catalytically active Venturello complex [PO4{WO(O2)2}4] , which
Keywords:
was generated by reactions of the Zr/Hf-containing POMs with hydrogen peroxide, and (2) the reaction
with active species formed on Zr/Hf clusters in the POMs. In the first type, the original sandwich structures
are not maintained during the reactions (as shown in 1, 2, 3, and 4), but in the second type, the sandwich
structures are maintained (as seen in 5 and 6). 7 was inactive. The reaction, (5 + cis-cyclooctene) + H2O2,
was explicitly influenced by changing the reaction time of 5 and olefin, before addition of H2O2, indicat-
ing a significant interaction between 5 and olefin, such as coordination of olefin to the POM. Thus, for
homogeneous olefin epoxidation by hydrogen peroxide catalyzed by di-Zr/Hf clusters, a new mechanism
is proposed through olefin-coordinated species to di-Zr/Hf centers.
Epoxidation
Hafnium(IV)
Zirconium(IV)
Hydrogen peroxide
Polyoxometalates
©
2014 Elsevier B.V. All rights reserved.
POM [␣-PW11TiO ]5 [20–23] and its conjugated acid [␣-
−
40
1
. Introduction
4−
P(TiOH)W11O39
substituted POM [(␣-PTiW11O39)2O]
Keggin 1,5-dititanium(IV)-substituted POM [␣-1,5-PW10Ti O40]
]
[21,23]; the dimeric Keggin monotitanium(IV)-
8−
In academic and industrial fields, the selective catalytic epoxida-
[23], and the monomeric
7−
tion of alkenes with an environmentally friendly oxidant, aqueous
hydrogen peroxide, is an interesting objective [1–5]. Transition-
metal-substituted polyoxometalates (POMs) have attracted con-
siderable attention as oxidation catalysts because they are resistant
to oxidative degradation. Furthermore, the “active sites” of their
transition metals and countercations can undergo extensive syn-
thetic modifications. The most commonly studied types of reaction
are H O -based epoxidation and sulfoxidation catalyzed by metal-
2
[22,24,25].
With regard to the catalytic activities of the dimeric mono-,
di-, and tri-titanium(IV)-substituted ␣-Keggin polyoxotungstates
8
−
10−
[(␣-PTiW11O39)2O] (Ti1), [(␣-1,2-PW Ti O ) O ]
2
(Ti2), and
(Ti3) in alkene epoxidation with
H O , it has been reported that the structure around the tita-
10
2
38
2
12−
[(␣-1,2,3-PTi W O ) O ]
3
9
37
2
3
2
2
2
2
nium centers has a strong influence on the catalytic activities [26].
The formation of catalytically active peroxo- and hydroperoxo-
titanium(IV) species has been confirmed [26–29]. Previous
studies of H O -based oxidation catalyzed by Ti-substituted
substituted POMs [6–18].
Several H O -based oxidation reactions catalyzed by
2
2
titanium(IV)-substituted POMs have been extensively investigated
2
2
[
19], e.g., the monomeric Keggin monotitanium(IV)-substituted
POMs have also indicated that the protons in countercations
and/or reaction solutions are crucial for the catalytic activi-
ties and the formation of active hydroperoxo–titanium and/or
protonated peroxo species [23,26,30–35]. In alkene epoxidation
∗ Corresponding author at: Department of Chemistry, Faculty of Science, Kana-
with H O2 catalyzed by Ti-substituted POMs, homolytic O
O
2
gawa University, Hiratsuka, Kanagawa 259-1293, Japan. Tel.: +81 463 59 4111;
fax: +81 463 58 9684.
cleavage of the Ti–hydroperoxo species, followed by oxygen
transfer to the alkene [radical mechanism], has been proposed
E-mail address: nomiya@kanagawa-u.ac.jp (K. Nomiya).
http://dx.doi.org/10.1016/j.molcata.2014.07.020
1
381-1169/© 2014 Elsevier B.V. All rights reserved.

H. Aoto et al. / Journal of Molecular Catalysis A: Chemical 394 (2014) 224–231
225
[
33]. For example, it is generally accepted that in oxida-
Venturello complex (Bu N) [PO {WO(O ) } ] was synthesized
4
3
4
2 2 4
tion catalyzed by dititanium-containing 19-tungstodiarsenate(III),
[
using a modified version of the reported method [38] and char-
8−
31
183
Ti (OH) As W19O67(H O)] , the Ti–hydroperoxo groups are the
acterized by elemental analysis, P and
W NMR spectroscopy
2
2
2
2
active oxygen-donating intermediates for alkene epoxidation [32].
In contrast, only a few examples of H O -based oxidation
(see Supporting Information). (Et NH ) [{␣-PW11O39Zr(␮-
2
2 8
OH)(H O) } ]·6H O (EtN-5) and the lithium salt (Li-5) were
2
2
2
2
2
2
catalyzed by Zr-containing POMs have been reported, by Khold-
eeva’s group, e.g., oxidation catalyzed by 2:2 complexes of Zr in
dimeric Keggin POMs such as (Bu N)7H[{PW11O39Zr(␮-OH)} ],
prepared using a modified version of the reported method [47],
and identified by CHN analysis, FTIR spectroscopy, TG/DTA, and
3
1
P NMR spectroscopy. The side-on peroxo dinulcear Zr complex
4
2
2
(
Bu N) [{PW
O
Zr(␮-OH)} ], and (Bu N) [{PW
O
Zr} (␮-
(Et NH ) [{Zr(␮-␩ -O )} (␣-PW11O39) ]·11H O (EtN-8) was
4
8
11 39
2
4
9
11 39
2
2
2
10
2
2
2
2
OH)(␮-O)] [36]. Kholdeeva’s group proposed that the active
synthesized by a reaction of EtN-5 with 30% aqueous H O using
2 2
species for cyclohexene oxidation was an unstable Zr–peroxo
a modified version of the reported method [37], and characterized
by elemental analysis, TG/DTA, FTIR, P NMR and x-ray crystallog-
3
1
31
species ( P NMR at ı −12.3 ppm) generated by ligand exchange
of the coordinating water ligands of the monomeric POM,
raphy (also see Supporting Information). Other precatalysts were
prepared according to the reported methods or modified versions
of these [47–52], and identified by elemental analysis, FTIR spec-
troscopy, TG/DTA, and 31P NMR spectroscopy. The precatalysts and
their abbreviations are shown in Table 1. The abbreviation consists
of a combination of the countercation and polyoxoanion (EtN-
and BuN- represent Et NH2 and Bu N countercations, respec-
(
Bu N)3+n[PW11O39Zr(OH)n(H O)3(2)-n] (n = 0 and 1), with the per-
4 2
oxo (H O ) ligand. They also suggested that the acid proton
2
2
is crucial for catalysis by Zr-containing POMs. With respect to
catalysis by Zr-containing POMs, Kortz’s group reported the stoi-
chiometric oxidation of l-methionine by the side-on peroxo species
2
(
i.e., ␮-␩ -peroxo-containing Zr2 species) of the dimeric Keggin
2
4
POM K12[Zr (O ) (␣-SiW11O39) ]·25H O [37].
tively). The polyoxoanion moieties are denoted as complexes,
by a combination of the number of metal centers and the two
lacunary POMs, i.e., 1:2 complexes of one metal center (Zr/Hf) and
two monolacunary POMs (EtN-1 and EtN-2) [49], 4:2 complexes
of four metal centers (Zr/Hf) and two dilacunary POMs (BuN-3
and BuN-4) [51,52], 3:2 complexes of three Zr centers and two
trilacunary POMs (BuN-7) [50], and several types of 2:2 complexes
such as (i) 2:2 complexes consisting of two hydrated metal centers
(Zr/Hf) and two monolacunary POMs with a P heteroatom (BuN-5,
BuN-6, and Li-5) [47], and (ii) 2:2 complexes consisting of a Zr2
center with two side-on peroxo groups and two monolacunary
POMs with a P heteroatom (EtN-8). Instrumentation/analytical
procedures are described in Supporting Information.
2
2
2
2
2
In this work, we focused on epoxidation of cis-cyclooctene
with aqueous hydrogen peroxide catalyzed by various struc-
turally
characterized
Zr/Hf-containing
Keggin
sandwich
10−
POMs, i.e., [M(␣-PW11O39)2]
(M = Zr,
[{M(H O)} {M(H O) } (␮-OH) (␮-OH) ](␣-1,2-PW10O37) ]
1 and M = Hf, 2),
7−
[
(
(
2
2
2
2
2
3
2
2
8−
M = Zr,
3
and M = Hf, 4), [{M(␮-OH)(H O)} (␣-PW11O ) ]
2
2
39 2
9
−
M = Zr, 5 and M = Hf, 6), and [Zr (␮ -OH) (A-␣-PW O ) ] (7).
3
2
3
9
34 2
In the reactions with 5 and 6, the original sandwich structures were
kept after the reactions and the active species were formed on the
Zr/Hf centers, whereas in the reactions with 1–4, the reaction was
brought about by the highly active Venturello complex [38–46],
which was generated during the reaction. No catalytic activity was
observed for 7 and the side-on peroxo species with heteroatom
2
10−
(8), the latter of which was
P [{Zr(␮-␩ -O )} (␣-PW11O39) ]
2
2
2
synthesized from a reaction between 5 and aqueous H O [see
2
2
2.2. Catalytic reactions
Supporting Information].
In this paper, we report full details of homogeneous cis-
cyclooctene epoxidation with hydrogen peroxide catalyzed by
Zr/Hf-containing Keggin sandwich POMs (1–8), and propose that
a significant interaction between olefin and POMs (5 and 6), i.e.,
an olefin coordination process is present before attack of hydrogen
peroxide.
Homogeneous reactions of cis-cyclooctene, catalyzed by several
POMs as Et NH , Bu N, or Li salts (Table 1), were carried out in
2
2
4
round-bottomed flasks. cis-Cyclooctene (1.0 mL, 7.70 mmol), 30%
aqueous H O2 (1.0 mL, 12.72 mmol), and a POM (0.01–0.02 mmol)
2
in a mixed solvent (30 mL) consisting of CH CN and CH Cl (15/15
3
2
2
v/v), a mixed solvent (31 mL) consisting of CH CN and water (30/1
3
v/v), or a mixed solvent (33 mL) consisting of CH CN and water
3
◦
2
. Experimental
(30/3 v/v) were mixed at 25 C under air. The reaction solution was
sampled after 0.0, 0.5, 0.75, 1.0, 1.5, 2.0, 2.25, 2.5, 3.0, 4.0, 5.0, 6.0,
and 24 h and analyzed using Shimadzu GC-17AAT and Shimadzu
GC-2010 Plus gas chromatographs (TCD) and a DB-FFAP capillary
column (0.53 mm × 15 m). The catalytic activities of POMs 1–8,
evaluated as turnover numbers (TON) and/or turnover frequen-
cies per second (TOF), were compared with those of the Venturello
complex and several previously reported Ti-containing Keggin
POMs.
2
.1. Materials
The following reactants were used as received: 30% aque-
ous H O (Wako), Bu NCl (TCI), and CD CN and D O (Isotec).
2
2
4
3
2
H [PW12O40]·3H O was prepared by a traditional method and
3
2
identified using FTIR spectroscopy, thermogravimetry (TG) and
differential thermal analysis (DTA), and 31P NMR spectroscopy.
Table 1
Compositions, formulas, and abbreviations of precatalysts.
M: lacunary POM ratio
Formula
Abbreviation
Ref.
1
4
2
:2 complex
:2 complex
:2 complex
(Et2NH2)10[Zr(␣-PW11O39)2]·7H2O
EtN-1
EtN-2
BuN-3
BuN-4
BuN-5
BuN-6
Li-5
EtN-5
EtN-8
BuN-7
[49]
[49]
[51,52]
[51,52]
[47]
[47]
[47]
[47]
This work
[50]
(
Et2NH2)10[Hf(␣-PW11O39)2]·8H2O
(Bu4N)7[[{Zr(H2O)}2{Zr(H2O)2}2(␮-OH)3(␮3-OH)2](␣-1,2-PW10O37)2]·3H2O
Bu4N)7[[{Hf(H2O)}2{Hf(H2O)2}2(␮-OH)3(␮3-OH)2](␣-1,2-PW10O37)2]·2H2O
(Bu4N)7H[{Zr(␮-OH)(H2O)}2(␣-PW11O39)2]·3H2O
Bu4N)7H[{Hf(␮-OH)(H2O)}2(␣-PW11O39)2]·3H2O
(
(
Li8[{Zr(␮-OH)(H2O)}2(␣-PW11O39)2]·20H2O
(
(
Et2NH2)8[{Zr(␮-OH)(H2O)}2(␣-PW11O39)2]·6H2O
2
Et2NH2)10[{Zr(␮-␩ -O2)}2(␣-PW11O39)2]·11H2O
3
:2 complex
(Bu4N)6.5H2.5[Zr3(␮-OH)3(A-␣-PW9O34)2]

2
26
H. Aoto et al. / Journal of Molecular Catalysis A: Chemical 394 (2014) 224–231
Fig. 3. TON–time curve of cis-cyclooctene epoxidation with 30% aqueous H2O2
catalyzed by POMs (Li-5, BuN-5, BuN-6, BuN-7, and EtN-8). Reaction conditions:
catalysts 0.02 mmol, substrate 7.70 mmol, 30% H2O2 aq. 12.72 mmol, under air at
◦
25
C.
The conversions of cis-cyclooctene after 24 h-reactions were
Fig. 1. TON–time curve of cis-cyclooctene epoxidation with 30% aqueous H2O2
catalyzed by POMs (EtN-1 and EtN-2). Reaction conditions: catalysts 0.02 mmol,
substrate 7.70 mmol, 30% H2O2 aq. 12.72 mmol, under air at 25 C.
more than 83–95% for both EtN-1 and EtN-2, and more than 77–95%
for both BuN-3 and BuN-4. BuN-4 shows a rapid increase in activity
after an induction period of more than 6 h and the final TON was
almost the same as that of BuN-3 after 24 h-reaction. The TONs after
◦
2
4 h-reactions for EtN-1 and EtN-2 (Fig. 1) and those for BuN-3 and
3
. Results and discussion
BuN-4 (Fig. 2) reached more than 300.
The reactions by EtN-1, EtN-2, BuN-3 and BuN-4 (Figs. 1 and 2)
are quite different from those by BuN-5 and BuN-6 (Fig. 3). In
the solutions of the 24 h-reactions of EtN-1 and EtN-2 (6.35
3
.1. Zr/Hf-containing POMs as precatalysts for cis-cyclooctene
epoxidation by H O : Catalysis by 1–4 and Venturello complex
2
2
and 6.27 ␮mol, respectively) and H O2 (0.636 mmol), formation of
the Venturello complex was confirmed by P NMR as signals at
2
Homogeneous epoxidation of cis-cyclooctene with hydrogen
peroxide catalyzed by various Zr/Hf-containing POMs was exam-
ined under the conditions of POM (0.02 mmol), cis-cyclooctene
31
3
.57 ppm (from EtN-1) and 3.54 ppm (from EtN-2). In fact, the
highly active reaction by the Venturello complex has been well
known [38–46]. In Figs. 1 and 2, it should be noted that the TONs
after 6 h-reactions for EtN-1, EtN-2, and BuN-3 were about two
times of the TON after 6 h-reaction for the separately prepared,
fresh Venturello complex, indicating that the POMs (EtN-1, EtN-2,
and BuN-3) generate 2 mol of the Venturello complex during the
reactions.
(
7.70 mmol) and 30% aqueous H O2 (12.72 mmol) in a mixed sol-
2
◦
vent (30 mL) consisting of CH CN:CH Cl (15/15 v/v), at 25 C in air.
3
2
2
The POMs used were 1:2 complexes (EtN-1 and EtN-2), 4:2 com-
plexes (BuN-3 and BuN-4), and the freshly prepared Venturello
complex, (Bu N) [PO {WO(O ) } ]. The time course versus TON
4
3
4
2 2 4
plots of epoxide formation are shown in Figs. 1 and 2.
Thus, the catalysis by EtN-1, EtN-2, BuN-3, and BuN-4 was actu-
ally due to the Venturello complex, which was derived by direct
reaction with hydrogen peroxide.
3.2. Homogeneous epoxidation of cis-cyclooctene with hydrogen
peroxide catalyzed by POMs 5–8
The 2:2 complexes BuN-5 and BuN-6, and the 3:2 complex
BuN-7 (POMs 0.02 mmol) were used in a mixed solvent (30 mL)
consisting of CH CN/CH Cl (15/15 v/v) as homogeneous catalysts
3
2
2
for epoxidation of cis-cyclooctene. The 2:2 complex of Zr as a Li salt
Li-5, 0.01 mmol) and the separately prepared, side-on peroxo–Zr2
(
species EtN-8 (0.02 mmol), both of which were water soluble, but
insoluble in organic solvents, were also used as homogeneous cat-
alysts in a mixed solvent (31 mL) consisting of CH CN/water (30/1
3
v/v). The TON–time course curves are shown in Fig. 3.
In the reactions by 5 and 6 in the presence of 30% aqueous
H O , a color change was not observed, i.e., the original colorless,
2
2
clear, homogeneous solution was maintained during the course
of the reaction. This is in contrast to the oxidation reactions with
3
0% aqueous H O2 catalyzed by mono-, di-, and tri-titanium(IV)-
2
Fig. 2. TON–time curve of cis-cyclooctene epoxidation with 30% aqueous H2O2 cat-
alyzed by POMs (freshly prepared Venturello complex, BuN-3, and BuN-4). Reaction
conditions: catalysts 0.02 mmol, substrate 7.70 mmol, 30% H2O2 aq. 12.72 mmol,
substituted Keggin POMs (Ti1, Ti2, and Ti3 as Bu N salts), in
which the colors of all the POMs changed from colorless to yel-
low or orange as a result of formation of peroxotitanium(IV)
4
◦
under air at 25 C.

H. Aoto et al. / Journal of Molecular Catalysis A: Chemical 394 (2014) 224–231
227
Table 2
Epoxidation of olefins (cis-cyclooctene, cyclohexene, 1-octene, and styrene) with 30% aqueous H2O2 catalyzed by BuN-5 and Li-5.
Precatalyst
Substrate
Epoxide
TONc
Selectivity (%)c
Other products
–
2-hydroxycyclohexanone
heptanal
BuN-5a
cis-cyclooctene
cyclohexene
cyclooctene oxide
trans-cyclohexane-1,2-diol
1,2-epoxyoctane
166
101
23
99
37
–
1-octene
styrene
cis-cyclooctene
cyclohexene
styrene oxide
7
132
38
–
benzaldehyde
–
2-hydroxycyclohexanone
heptanal
Li-5b
cyclooctene oxide
trans-cyclohexane-1,2-diol
1,2-epoxyoctane
99
62
–
1-octene
23
a
Solvent 15:15 (v/v) CH2Cl2/CH3CN, 30 mL.
Solvent 30:1 (v/v) CH3CN/H2O, 31 mL.
After 24 h.
b
c
Reaction conditions: catalyst 0.02 mmol, substrate cis-cyclooctene 7.70 mmol, cyclohexene 9.86 mmol, 1-octene 6.37 mmol, or styrene 8.70 mmol, 30% H2O2 aq. 12.72 mmol,
◦
under air, at 25 C.
Table 3
Epoxidation of mixed olefins (cis-cyclooctene and cyclohexene) with 30% aqueous H2O2 catalyzed by BuN-5 and Li-5.
Precatalysts
Mixed substrate (mmol)
Total conversion (%)c
Products (mmol)c
TONc
Selectivity (%)c
BuN-5a
cis-cyclooctene (1.54)
cyclohexene (4.62)
cis-cyclooctene (4.62)
cyclohexene (1.54)
92
68
61
58
cyclooctene oxide (1.4)
trans-cyclohexane-1,2-diol (1.2)
cyclooctene oxide (2.8)
71
58
138
11
99
37
97
25
trans-cyclohexane-1,2-diol (0.2)
Li-5b
cis-cyclooctene (1.54)
cyclohexene (4.62)
cis-cyclooctene (4.62)
cyclohexene (1.54)
53
23
51
23
cyclooctene oxide (0.8)
trans-cyclohexane-1,2-diol (0.5)
cyclooctene oxide (2.4)
41
31
118
9
99
58
99
51
trans-cyclohexane-1,2-diol (0.2)
◦
Reaction conditions: catalyst 0.02 mmol, mixed substrate cis-cyclooctene and cyclohexene, 30% H2O2 aq. 12.72 mmol, under air, at 25 C.
a
Solvent 15:15 (v/v) CH2Cl2/CH3CN, 30 mL.
Solvent 30:1 (v/v) CH3CN/H2O, 31 mL.
After 24 h.
b
c
and/or hydroperoxotitanium(IV) intermediates [26,27]. The cat-
alytic activities of BuN-5 and BuN-6 at 25 C were monitored by
3.3. Homogeneous epoxidation of olefins with hydrogen peroxide
catalyzed by BuN-5 and Li-5: selection of cis-cyclooctene as a test
substrate
◦
GC during 24 h-reactions. Cyclooctene oxide was the major prod-
uct (selectivity > 99%), and no induction period was observed for
these reactions. The ratio of the activity with BuN-5 to that of the
Venturello complex was about 38% after 6 h and 49% after 24 h.
Also, the ratio of the activity with BuN-6 to that of the Venturello
complex was about 29% after 6 h and 45% after 24 h.
The epoxidations of olefins (cis-cyclooctene, cyclohexene, 1-
octene, and styrene) with 30% aqueous H O catalyzed by BuN-5
2
2
◦
and Li-5 were examined at 25 C under air in a mixed solvent
(30 mL) consisting of CH CN and CH Cl (15/15 v/v) and a mixed
3
2
2
In comparing Li-5 and BuN-5 as precatalysts for cyclooctene
epoxidation, it should be noted that the catalytic activity was
BuN-5 > Li-5 (see Tables 2 and 3, Section 3.5. (4)), whereas the cat-
alytic stability was Li-5 > BuN-5. Powder samples of BuN-5 and
BuN-6, recovered after 24 h-reactions by evaporating the reac-
tion solutions and reprecipitating with excess diethyl ether, were
characterized using FTIR and 31P NMR spectroscopies. The orig-
inal sandwich structures were maintained during and after the
solvent (31 mL) consisting of CH CN and water (30/1 v/v), respec-
tively.
3
Cyclohexene has been known as a useful test substrate that
allows one to distinguish between homolytic and heterolytic
oxidation mechanism [35]. It has been said that formation of
allylic oxidation products, 2-cyclohexene-1-ol, 2-cyclohexene-1-
one, cyclohexene oxide is an indication of a homolytic oxidation
by hydrogen peroxide (radical mechanism), whereas the selec-
tive formation of trans-cyclohexane-1,2-diol (major product) and
cyclohexene oxide (minor product) suggests a heterolytic mecha-
nism (ionic mechanism) [35]; distribution of products containing
cyclohexene oxide depends on the reaction mechanism. Our results
of cyclohexene oxidation suggest the ionic mechanism, i.e., het-
erolytic mechanism (Table 2), in which 2-hydroxycyclohexanone
as other products is derived by further oxidation of trans-
cyclohexane-1,2-diol.
reactions. However, the recovered sample of the Bu N salt was
4
3−
contaminated with saturated Keggin POM species [PW12O40
as minor species, which was confirmed by 31P NMR signal at -
4.40 ppm in CD CN, whereas that of the Li salt was not. Repeated
]
1
3
reuse of the Bu N salt resulted in increased amounts of the
4
3−
saturated Keggin [PW12O40
]
species. Thus, with regard to cat-
alyst stability, it appears that the Li salt is superior to the
Bu N salt.
4
BuN-7 was inactive for epoxidation of cis-cyclooctene. The
side-on coordinated peroxo–Zr2 species (EtN-8) did not show any
activity (Fig. 3); the TON after 24 h was 0.8 without addition of tri-
fluoromethanesulfonic acid (TFSA) and 17.3/19.0 with addition of
TFSA 1:2/1:4, respectively. Kortz et al. reported that the side-on per-
oxo species of a dimeric Zr complex with Si heteroatom was active
in the stoichiometric oxidation of l-methionine, but the reaction
was not catalytic [37]. In the POM with P heteroatom, the side-
on peroxo species (EtN-8) is probably formed during the course
of reaction from the active, end-on hydroperoxo species Zr–OOH,
which will be formed at the early stage of the reaction of POM 5
In this work, we selected cis-cyclooctene as a test substrate for
further studying the reaction in detail rather than cyclohexene,
because cis-cyclooctene gave the best results in the viewpoints of
the sole product of oxidation, i.e., highest selectivity for the epoxide,
and the highest TON of the epoxide (Table 2).
3.4. Results from changing reaction times between Li-5 and
cis-cyclooctene before addition of H O and between Li-5 and
2
2
H O before addition of cis-cyclooctene (Control experiments #1)
2
2
Control experiments (Fig. 4) consisting of the reac-
with aqueous H O (Fig. S3).
2
2
tions (I) (Li-5 + cis-cyclooctene) + H O2 (Entries 1–3) and (II)
2

2
28
H. Aoto et al. / Journal of Molecular Catalysis A: Chemical 394 (2014) 224–231
Fig. 4. Results of control experiments consisting of two types of reaction: (I) (Li-5 + olefin) + H2O2 and (II) (Li-5 + H2O2) + olefin, which were examined by changing the
addition times of the second additive (olefin or H2O2) to Li-5. Reaction conditions for (I) and (II): catalysts 0.02 mmol, substrate 7.70 mmol, 30% H2O2 aq. 12.72 mmol, under
◦
air at 25 C. Reaction product in (I) and (II): cyclooctene oxide.
(
Li-5 + H O ) + cis-cyclooctene (Entries 4–6) were carefully car-
on 24 h-reactions after addition of cis-cyclooctene (Entries 5 and
6, respectively) were much higher than that (TON 11.9) of Entry 4.
TONs by changing reaction times of catalyst + H O , i.e., 24 h and
2
2
ried out by changing the reaction times of the second additive
cis-cyclooctene in (I) or H O2 in (II)) to Li-5. GC analysis was
(
2
2
2
performed during 24 h-reactions after addition of H O2 in (I) or
10 min indicate that decomposition of H O2 does not occur. Entry
2
2
cis-cyclooctene in (II). In the reaction (I), the highest activity (TON
4 also shows that the lowest activity results from the formation of
side-on peroxo species (POM 8) during the prolonged mixing time
(see Fig. S3). Entries 5 and 6 are probably due to the active, end-
on coordinated hydroperoxo species, which are alive at the early
stage.
2
05.8) was observed for Entry 3, in which hydrogen peroxide
was added on 10 min after Li-5 and cis-cyclooctene were mixed,
and the final solution was analyzed after 24 h-reaction. As the
reaction time of Li-5 and cis-cyclooctene increased, the activity
decreased (Entries 1 and 2). This indicates that there is a significant
interaction between the POM and cis-cyclooctene before addition
of H O . An olefin coordination as such an interaction is proposed.
Since the activities in (I) are higher than those in (II), in
the case of that Li-5, olefin and H O2 are simultaneously coex-
2
istent, the process based on (I), i.e., olefin coordination and
subsequent nucleophilic attack of free and/or coordinated hydro-
gen peroxide, will be proposed as plausible reaction scheme
(Fig. 5).
2
2
Based on theoretical and experimental studies containing x-ray
IV
0
crystallography, several Zr ion (d )–olefin bonding complexes
have been reported in the literature [53–57]. The longest reaction
time (one week) between Li-5 and cis-cyclooctene resulted in
the lowest activity (Entry 1). The reason may be attributable to
polymerization such as ring-opening metathesis of cis-cyclooctene
coordinated to Li-5 [58]. The coordinated olefin can undergo the
nucleophilic attack of free and/or coordinated hydrogen peroxide.
Thus, the reaction (I) is due to the ionic oxidation mechanism by
hydrogen peroxide rather than the radical oxidation mechanism.
In contrast, the similar control experiments were also carried
out using a dimeric trititanium(IV)-substituted Keggin POM (Ti3)
that has been known as epoxidation catalysis via radical mech-
anism [26]. The reaction times between Ti3 and cis-cyclooctene
were changed (1 h and one week), followed by adding H O , and
3.5. Catalytic epoxidation of mixed substrates by Li-5 and BuN-5
(
Control experiments #2)
Epoxidations of mixed substrates (cis-cyclooctene and cyclo-
hexene in molar ratios of 1:3 and 3:1) with aqueous H O catalyzed
2
2
by BuN-5 and Li-5 were examined for 24 h. The results are shown
in Table 3.
2
2
(1) In cyclohexene oxidation, the products were cyclohexene epox-
ide as the minor product and trans-cyclohexane-1,2-diol as the
major product. trans-Cyclohexane-1,2-diol was derived from
cyclohexene epoxide in an acidic environment, based on disso-
ciation of the coordinated water molecules in the Zr centers.
The selectivity for cyclohexene epoxide is therefore actually
the same as that for trans-cyclohexane-1,2-diol. The selectivity
analysing the products after 24 h-reactions. The TONs of epoxide
produced during different reaction times (1 h and one week) were
almost unchanged (∼150). These results are quite different from
the results using Li-5.
On the other hand, in (II), the reaction times between Li-5 and
H O were changed. TONs (130.7 and 121.3) of epoxide produced
2
2

H. Aoto et al. / Journal of Molecular Catalysis A: Chemical 394 (2014) 224–231
229
Fig. 5. Proposed reaction scheme.
after 24 h-reaction for trans-cyclohexane-1,2-diol was evalu-
“homogeneous biphasic” system, whereas that consisting
ated as
of BuN-5 in CH CN and CH Cl becomes a “homogeneous
3
2
2
monophasic” system.
[trans − cyclohexane − 1, 2 − diol]/([cyclohexene]t=0
3
.6. 31P NMR spectra of POM 5 in the presence of mixed or single
substrate (cis-cyclooctene and/or cyclohexene) and aqueous H O
−
[cyclohexene] ).
t=24
2
2
(
(
(
2) Despite the different starting molar ratios of the mixed sub-
strates, cis-cyclooctene was oxidized more than cyclohexene
by either BuN-5 or Li-5, showing strong substrate depend-
ence of these precatalysts. This fact suggests that coordination
ability of cis-cyclooctene to Zr centers is higher than that of
cyclohexene. The coordination of cyclohexene and subsequent
epoxidation will begin after the cis-cyclooctene has been almost
consumed.
Solution 31P NMR spectra of 5 are strongly dependent on sol-
vents and countercations; BuN-5 in CD CN shows a signal at
3
−2.00 ppm, EtN-5 in D O at −3.59 ppm, Li-5 in D O at −3.53 ppm
2
2
together with minor peaks at −4.47 and −4.58 ppm, and Li-5 in
3
1
CD CN at −2.07 ppm (Table S2). The P NMR spectrum of Li-5
3
in CD CN in the presence of aqueous H O showed a signal at
3
2
2
−2.80 ppm as a broad peak after 1 h-mixing and at −2.57 ppm after
216 h-mixing. The latter signal was in good agreement with that
3) The selectivity for trans-cyclohexane-1,2-diol in oxidation with
BuN-5 is lower than that with Li-5. The present reaction
is significantly influenced by the solvent system. BuN-5 in
CH CN/CH Cl (15/15 v/v, 30 mL) is much more active than Li-
of the separately prepared, side-on peroxo species EtN-8 in CD CN
3
(−2.55 ppm). Thus, the broad signal at −2.80 ppm is attributable
to the end-on hydroperoxo species. The 31P NMR spectrum of
Li-5 in CD CN on 1 h after excess cis-cyclooctene or cyclohexene
3
2
2
3
5
in CH CN/H O (30/1 v/v, 31 mL). Thus, in catalysis by BuN-5,
was added showed signals at −2.11 ppm or −2.14 ppm, respec-
tively; the signals on 168 h after the addition were observed at
−2.14 ppm for cis-cyclooctene and at −2.16 (main) and −2.60 ppm
(minor) for cyclohexene. The signals at −2.11 ppm and −2.14 ppm
are assignable to the olefin-coordinated species of cis-cyclooctene
and cyclohexene, respectively. The 31P NMR signal at −2.08 ppm
3
2
further oxidation products of trans-cyclohexane-1,2-diol such
as 2-hydroxycyclohexanone can be also formed, resulting in
lower selectivity.
4) Difference between catalytic activities of BuN-5 and Li-5:
although the order of the catalyst stability is Li-5 > BuN-5,
that of the catalytic activity is BuN-5 > Li-5. Both reactions
proceed in homogeneous systems. The solubilities of BuN-5
and Li-5 are different in the constituent solvents: Li-5 is sol-
uble in H O, but insoluble in CH CN, whereas BuN-5 is soluble
for L-5 (2 mmol) in CD CN solution containing olefin mixtures of
3
cis-cyclooctene (154 mmol) and cyclohexene (462 mmol) was the
same as the signal for olefins mixed in a different ratio, i.e., cis-
cyclooctene (462 mmol) and cyclohexene (154 mmol), even after
1 h and/or longer (Table S2). These facts are consistent with that
the coordination ability of cis-cyclooctene is higher than that of
cyclohexene. Thus, it is suggested that under the co-existence of cis-
cyclooctene, neither cyclohexene coordinated species nor mixed
coordinated species are formed. These results are also consistent
2
3
in CH CN, but insoluble in CH Cl . The substrates are solu-
3
2
2
ble in both CH CN and CH Cl . Since the present reaction is
3
2
2
significantly influenced by the solvent systems, the different
activities are attributable to the different reaction systems;
the system consisting of Li-5 in H O and CH CN becomes a
2
3

2
30
H. Aoto et al. / Journal of Molecular Catalysis A: Chemical 394 (2014) 224–231
with those for epoxidation of mixed substrates with different molar
ratios (see Section 3.5).
References
[
[
1] R.A. Sheldon, J.K. Kochi, Metal-Catalyzed Oxidations of Organic Compounds,
Academic Press, New York, 1981 (Chapter 3).
2] G.W. Parshall, S.D. Ittel, Homogeneous Catalysis: The Applications and Chem-
istry of Catalysis by Soluble Transition Metal Complexes, 2nd ed, Wiley, New
York, 1992, pp. 151.
3
.7. Comparison of present work with catalysis by Ti-containing
POMs and Zr-containing POMs reported in the literature
(
1) The 2:2 complexes (5 and 6) and the 1:2 complexes
[3] K.S. Suslick, in: C.L. Hill (Ed.), Activation and Functionalization of Alkenes,
Wiley, New York, 1989, p. 219.
1
0−
[
M(PW11O39)2]
(M = Zr 1 and Hf 2) underwent mutual inter-
[
[
4] K.A. Jorgensen, Chem. Rev. 89 (1989) 431–458.
5] B. Meanier, Chem. Rev. 92 (1992) 1411–1456.
conversion under appropriate pH conditions [47], suggesting
that these complexes are not rigid, but flexible. However, they
showed quite different catalytic behaviors in the epoxida-
[6] N. Mizuno, K. Kamata, K. Yamaguchi, Top. Catal. 53 (2010) 876–893.
7] N. Mizuno, K. Yamaguchi, K. Kamata, Coord. Chem. Rev. 249 (2005)
944–1956.
[
1
tion of cis-cyclooctene with H O ; the 2:2 complexes formed
2
2
[8] A.C. Estrada, I.C.M.S. Santos, M.M.Q. Simoes, M.G.P.M.S. Neves, J.A.S. Cavaleiro,
A.M.V. Cavaleiro, Appl. Catal. A 392 (2011) 28–35.
active species on dinuclear Zr/Hf centers in sandwich struc-
tures, whereas the 1:2 complexes generated the highly active
Venturello complex.
[
9] N. Mizuno, K. Yamaguchi, K. Kamata, Catal. Surv. Asia. 15 (2011) 68–79.
[
[
10] N. Mizuno, K. Yamaguchi, Chem. Rec. 6 (1) (2006) 12–22.
11] N. Mizuno, S. Hikichi, K. Yamaguchi, S. Uchida, Y. Nakagawa, K. Uehara, K.
Kamata, Catal. Today 117 (2006) 32–36.
12] C. Jahier, S.S. Mal, U. Kortz, S. Nlate, Eur. J. Inorg. Chem. (2010) 1559–1566.
13] M. Carraro, N. Nsouli, H. Oelrich, A. Sartorel, A. Soraru, S.S. Mal, G. Scorrano, L.
Walder, U. Kortz, M. Bonchio, Chem. Eur. J. 17 (2011) 8371–8378.
[14] M. Craven, R. Yahya, E. Kozhevnikova, R. Boomishankar, C.M. Robertson, A.
Steiner, I. Kozhevnikov, Chem. Commun. 49 (2013) 349–351.
15] V. Mirkhani, M. Moghadam, S. Tangestaninejad, I. Mohammadpoor-Baltork, E.
Shams, N. Rasouli, Appl. Catal. A: Gen. 334 (2008) 106–111.
16] R. Hajian, S. Tangestaninejad, M. Moghadam, V. Mirkhani, A. Reza Khosropour,
C. R. Chimie 15 (2012) 975–979.
17] Z. Karimi, A.R. Mahjoub, S.M. Harati, Inorg. Chim. Acta 376 (2011) 1–9.
18] P.A. Shringarpure, A. Patel, Ind. Eng. Chem. Res. 50 (2011) 9069–9076.
[19] K. Nomiya, Y. Sakai, S. Matsunaga, Eur. J. Inorg. Chem. (2011) 179–196.
[20] O.A. Kholdeeva, G.M. Maksimov, R.I. Maksimovskaya, L.A. Kovaleva, M.A. Fedo-
tov, React. Kinet. Catal. Lett. 66 (1999) 311–317.
(
2) The active species formed by 5 and 6 are substantially differ-
ent from those formed by Ti-substituted POMs. In the latter
case, homolytic O O cleavage of the Ti–hydroperoxo species,
followed by oxygen transfer to the alkene, has been considered
[
[
[
33]. It is generally accepted that Ti–hydroperoxo groups are
[
the active oxygen-donating intermediates for the epoxidation
of alkenes, but the alkene-coordinating species have never been
considered.
[
[
[
(
3) The active species formed by 5 and 6 are also different from that
derived from the Zr -containing POMs reported by Kholdeeva’s
2
group [36], which has been proposed as an unstable Zr–peroxo
3
1
species ( P NMR, ı −2.3 ppm) formed by a ligand exchange of
[
21] O.A. Kholdeeva, R.I. Maksimovskaya, G.M. Maksimov, L.A. Kovaleva, Kinet. Catal.
42 (2001) 217–222.
the monomeric species (Bu N) [PW Zr(OH)n(H O)
O
]
3(2)-n
4
3+n
11 39
2
(
n = 0 and 1) with a peroxo (H O ) ligand.
[22] T. Yamase, E. Ishikawa, Y. Asai, S. Kanai, J. Mol. Catal. A: Chem. 114 (1996)
37–245.
2
2
2
[
23] O.A. Kholdeeva, G.M. Maksimov, R.I. Maksimovskaya, L.A. Kovaleva, M.A. Fedi-
tiv, V.A. Grigoriev, C.L. Hill, Inorg. Chem. 39 (2000) 3828–3837.
24] E. Ishikawa, T. Yamase, J. Mol. Catal. A: Chem. 142 (1999) 61–76.
25] F. Gao, T. Yamase, H. Suzuki, J. Mol. Catal. A: Chem. 180 (2002) 97–108.
26] C.N. Kato, S. Negishi, K. Yoshida, K. Hayashi, K. Nomiya, Appl. Catal. A: Gen. 292
(2005) 97–104.
4
. Conclusion
[
[
[
Two different types of cis-cyclooctene epoxidation with
hydrogen peroxide were found in catalysis by Zr/Hf-based
sandwich-structured POMs 1–8. The reactions with 1–4, which
gave very high activities, were brought about by the Venturello
complex [PO {WO(O ) } ]
[
27] K. Hayashi, C.N. Kato, A. Shinohara, Y. Sakai, K. Nomiya, J. Mol. Catal. A: Chem.
262 (2007) 30–35.
[
28] K. Hayashi, M. Takahashi, K. Nomiya, Dalton Trans. (2005) 3751–3756.
3−
[38–46], which was generated by
[29] C.N. Kato, K. Hayashi, S. Negishi, K. Nomiya, J. Mol. Catal. A: Chem. 262 (2007)
25–29.
4
2 2 4
reactions with hydrogen peroxide, whereas the reactions with 5
and 6, which gave moderate activities (about 45% those of 1–4),
proceeded via active species formed on dimeric Zr/Hf clusters in
the sandwich structure. The 2:2 complex containing side-on per-
oxo species on the Zr2 centers (8) and the 3:2 complex (7) showed
almost no activities. Two control experiments (#1 and #2) sug-
gest that the reactive species of 5 is formed as the dimeric species
containing cis-cyclooctene coordinated to Zr atom. Thus, for homo-
geneous olefin epoxidation by hydrogen peroxide catalyzed by
di-Zr/Hf clusters, a new mechanism is proposed through olefin-
coordinated species to di-Zr/Hf centers. The epoxidation of olefin
with H O catalyzed by 5 probably proceeds via nucleophilic attack
[
[
[
[
[
[
30] O.A. Kholdeeva, T.A. Trubitsina, R.I. Maksimovskaya, A.V. Golovin, W.A. Nelwert,
B.A. Kolesov, X. López, J.M. Poblet, Inorg. Chem. 43 (2004) 2284–2292.
31] O.A. Kholdeeva, T.A. Trubitsina, G.M. Maksimov, A.V. Golovin, R.I. Maksi-
movskaya, Inorg. Chem. 44 (2005) 1635–1642.
32] F. Hussain, B.S. Bassil, U. Kortz, O.A. Kholdeeva, M.N. Timofeeva, P. de Oliveira,
B. Keita, L. Nadjo, Chem. Eur. J. 13 (2007) 4733–4742.
33] N.S. Antonova, J.J. Carbo ı´ , U. Kortz, O.A. Kholdeeva, J.M. Poblet, J. Am. Chem. Soc.
132 (2010) 7488–7497.
34] Y. Matsuki, Y. Mouri, Y. Sakai, S. Matsunaga, K. Nomiya, Eur. J. Inorg. Chem.
(
2013) 1754–1761.
35] O.A. Kholdeeva, Eur. J. Inorg. Chem. (2013) 1595–1605.
[36] O.A. Kholdeeva, G.M. Maksimov, R.I. Maksimovskaya, M.P. Vanina, T.A. Tru-
bitsina, D.Y. Naumov, B.A. Kolesov, N.S. Antonova, J.J. Carbo, J.M. Poblet, Inorg.
Chem. 45 (2006) 7224–7234.
2
2
[
[
[
[
[
[
[
37] S.S. Mal, N.H. Nsouli, M. Carraro, A. Sartorel, G. Scorrano, H. Oelrich, L. Walder,
of free and/or coordinated hydrogen peroxide to the coordinated
olefin (the ionic mechanism), which is contrasted to the catalysis
by the Ti-substituted POMs via radical mechanism.
M. Bonchio, U. Kortz, Inorg. Chem. 49 (2010) 7–9.
38] C. Aubry, G. Chottard, N. Platzer, J-M. Bregeault, R. Thouvenot, F. Chauveau, C.
Huet, H. Ledon, Inorg. Chem. 30 (1991) 4409–4415.
39] L.I. Kuznetsova, N.I. Kuznetsova, R.I. Maksimovskaya, G.I. Aleshina, O.S.
Koscheeva, V.A. Utkin, Catal. Lett. 141 (2011) 1442–1450.
40] L. Liu, C. Chen, X. Hu, T. Mohamood, W. Ma, J. Lin, J. Zhao, New J. Chem. 32 (2008)
Acknowledgements
283–289.
41] D.C. Duncan, R.C. Chambers, E. Hecht, C.L. Hill, J. Am. Chem. Soc. 117 (1995)
681–691.
42] G. Roelfes, M. Lubben, R. Hage, L. Que Jr., B.L. Feringa, Chem. Eur. J. 6 (2000)
This work was supported by JSPS KAKENHI Grant number
2
2550065, and also by the Strategic Research Base Development
2152–2159.
Program for Private Universities of the Ministry of Education, Cul-
ture, Sports, Science, and Technology of Japan.
43] I.W.C.E. Arends, R.A. Sheldon, Appl. Catal. A 212 (2001) 175–187.
[44] C. Venturello, R. D’Aloisio, J. Org. Chem. 53 (1988) 1553–1557.
[
[
45] Z. Zhang, W. Zhao, B. Ma, Y. Ding, Catal. Commun. 12 (2010) 318–322.
46] R. Ishimoto, K. Kamata, N. Mizuno, Eur. J. Inorg. Chem. (2013) 1943–1950.
Appendix A. Supplementary data
[47] K. Nomiya, Y. Saku, S. Yamada, W. Takahashi, H. Sekiya, A. Shinohara, M. Ishi-
maru, Y. Sakai, Dalton Trans. (2009) 5504–5511.
[
48] H. Osada, A. Ishikawa, Y. Saku, Y. Sakai, Y. Matsuki, S. Matsunaga, K. Nomiya,
Polyhedron 52 (2013) 389–397.
Supplementary material related to this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.molcata.
[49] C.N. Kato, A. Shinohara, K. Hayashi, K. Nomiya, Inorg. Chem. 45 (2006)
108–8119.
8
2
014.07.020.

H. Aoto et al. / Journal of Molecular Catalysis A: Chemical 394 (2014) 224–231
231
[
[
[
[
50] Y. Saku, Y. Sakai, A. Shinohara, K. Hayashi, S. Yoshida, C.N. Kato, K. Yoza, K.
Nomiya, Dalton Trans. (2009) 805–813.
51] K. Nomiya, K. Ohta, Y. Sakai, T. Hosoya, A. Ohtake, A. Takakura, S. Matsunaga,
Bull. Chem. Soc. Jpn. 86 (2013) 800–812.
52] K. Nomiya, N.C. Kasuga, S. Matsunaga, Y. Sakai, T. Hasegawa, T. Kimura, Sci. J.
Kanagawa Univ. 21 (2010) 43–46.
53] H. Zhao, A. Ariafard, Z. Lin, Inorg. Chim. Acta 359 (2006)
[54] J.-F. Carpentier, Z. Wu, C.W. Lee, S. Strömberg, J.N. Christopher, R.F. Jordan, J.
Am. Chem. Soc. 122 (2000) 7750–7767.
[55] M. Dahlmann, G. Erker, K. Bergander, J. Am. Chem. Soc. 122 (2000) 7986–7998.
[56] S. Kuroda, F. Dekura, Y. Sato, M. Mori, J. Am. Chem. Soc. 123 (2001) 4139–4146.
[57] E. Negishi, F.E. Cederbaum, T. Takahashi, Tetrahedron Lett. 27 (1986)
2829–2832.
[58] H. Martinez, N. Ren, M. Matta, M. Hillmyer, Polym. Chem. (2014),
http://dx.doi.org/10.1039/C3PY01787G.
3
527–3534.