Synthesis and reactivity of (μ-η2:η2-peroxo)dicopper(II) complexes with dinucleating ligands: Hydroxylation of xylyl linker with a NIH shift

Article abstract of DOI:10.1016/j.jorganchem.2006.05.068

New hexadentate dinucleating ligands having a xylyl linker, X-L-R, were synthesized, where X-L-R = 1,3-bis[bis(6-methyl-2-pyridylmethyl)aminomethyl]-2,4,6-trimethybenzene (Me₂-L-Me) and 1,3-bis[bis(6-methyl-2-pyridylmethyl)aminomethyl]-2-fluoro

Full text of DOI:10.1016/j.jorganchem.2006.05.068

Journal of Organometallic Chemistry 692 (2007) 111–121
www.elsevier.com/locate/jorganchem
Synthesis and reactivity of (l-g2:g2-peroxo)dicopper(II) complexes with
dinucleating ligands: Hydroxylation of xylyl linker with a NIH shift
a
a
b
b
Takahiro Matsumoto , Hideki Furutachi , Shigenori Nagatomo , Takehiko Tosha ,
a
b
a,
*
Shuhei Fujinami , Teizo Kitagawa , Masatatsu Suzuki
a
Department of Chemistry, Graduate School of Natural Science and Technology, Kanazawa University, Kakuma-machi, Kanazawa 920-1192, Japan
b
Center for Integrative Bioscience, Okazaki National Research Institutes, Myodaiji, Okazaki 444-8585, Japan
Received 16 March 2006; received in revised form 3 May 2006; accepted 12 May 2006
Available online 9 September 2006
Abstract
New hexadentate dinucleating ligands having a xylyl linker, X–L–R, were synthesized, where X–L–R = 1,3-bis[bis(6-methyl-2-pyr-
idylmethyl)aminomethyl]-2,4,6-trimethybenzene (Me2–L–Me) and 1,3-bis[bis(6-methyl-2-pyridylmethyl)aminomethyl]-2-fluorobenzene
(H–L–F). They form dinuclear copper(I) complexes, [Cu2(X–L–R)]2+ (Me2–L–Me (1) and H–L–F (2)). The copper(I) complexes in ace-
tone at ꢀ78 ꢀC react with O2 to produce intra- and intermolecular (l-g2:g2-peroxo)dicopper(II) species depending on the concentrations
of the complexes: both complexes generate intramolecular (l-g2:g2-peroxo)dicopper(II) species [Cu2(O2)(X–L–R)]2+ (1-O2 and 2-O2) at
the concentrations below ꢁ5 mM, whereas at ꢁ60 mM, both complexes produce intermolecular (l-g2:g2-peroxo)dicopper(II) species,
which were confirmed by the electronic and resonance Raman spectroscopies. The electronic spectrum of 1-O2 in acetone at concentra-
tions below ꢁ5 mM showed an absorption band at (kmax = 442 nm, e = 5600 Mꢀ1 cmꢀ1) assignable to the pꢂ(O–O)-to-Cu(II) (ðdx2ꢀy2
þ
dx2ꢀy2 Þ component) LMCT transition in addition to an intense band attributable to the pꢂr(O–O)-to-Cu(II) (ðrdx2ꢀy2 ꢀ dx2ꢀy2 Þ component)
LMCT transition (kmax = 359 nm, e = 21000 Mꢀ1 cmꢀ1), indicating that the (l-g2:g2-peroxo)Cu(II)2 core of 1-O2 takes a butterfly struc-
ture. Decomposition of 1-O2 resulted in hydroxylation of the 2-position of the xylyl linker with 1,2-methyl migration (NIH shift), sug-
gesting that the hydroxylation reaction proceeds via a cationic intermediate as proposed for closely related (l-g2:g2-peroxo)Cu(II)2
complexes having a xylyl linker. Kinetic study of the decomposition (hydroxylation of the xylyl linker) of 1-O2 suggests that a stereo-
chemical effect of the methyl group in the 2-position of the xylyl linker has a significant influence on a transition state for decomposition
(hydroxylation of the xylyl linker).
ꢁ 2006 Elsevier B.V. All rights reserved.
Keywords: Dinuclear copper complexes; Arene hydroxylation; Oxygenation; Peroxo dicopper complexes
1. Introduction
hydroxylation of the xylyl linkers of the supporting ligands
[3–5]. The detailed kinetic studies of the decompositions of
[Cu2(O2)(X–XYL–H)]2+ (X = NO2, H, and t-Bu, see
Scheme 1) have revealed that Hammett q value is ꢀ2.1,
suggesting that the hydroxylation mediated by [Cu2(O2)-
(X–XYL–H)]2+ proceeds via an electrophilic aromatic sub-
stitution mechanism [6]. In addition to such intramolecular
reactions, intermolecular arene hydroxylation reactions
have also been demonstrated where phenolates are hydrox-
ylated by the (l-g2:g2-peroxo)Cu(II)2 complexes [8–11].
Itoh and co-workers have shown that the phenolate
hydroxylation by a (l-g2:g2-peroxo)copper(II) complex
The Cun/O2-mediated arene hydroxylation is of current
interest for understanding the reaction mechanisms of
dioxygen activating copper proteins in biological systems
such as tyrosinase and utilizing metal complexes as oxida-
tion catalysts [1–15]. Karlin and co-workers have first dem-
onstrated that the (l-g2:g2-peroxo)Cu(II)2 complexes,
[Cu2(O2)(X–XYL–R)]2+
,
are capable of performing
*
Corresponding author. Fax: +81 76 264 5742.
E-mail address: suzuki@cacheibm.s.kanazawa-u.ac.jp (M. Suzuki).
0022-328X/$ - see front matter ꢁ 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2006.05.068

112
T. Matsumoto et al. / Journal of Organometallic Chemistry 692 (2007) 111–121
Scheme 1.
well mimics that performed by tyrosinase [9,12]. Recently,
we have also found that a (l-g2:g2-peroxo)Cu(II)2 having
a xylyl linker [Cu2(O2)(H–L–H)]2+ (3-O2) is capable of per-
forming not only intramolecular hydroxylation of the xylyl
linker of the dinucleating ligand H–L–H, but also intermo-
lecular epoxidation of styrene and hydroxylation of THF
by hydrogen atom abstraction [13].
It has been shown that a bis(l-oxo)Cu(III)2 species is
also capable of hydroxylation of phenyl group of a sup-
porting ligand via an electrophilic aromatic substitution
mechanism [14]. Recently, Stack and co-workers also have
reported intermolecular hydroxylation reaction of pheno-
lates by a bis(l-oxo)Cu(III)2 species, where coordination
of a phenolate to a (l-g2:g2-peroxo)Cu(II)2 species causes
the O–O bond cleavage to generate a bis(l-oxo)Cu(III)2
species which is responsible for the hydroxylation of a phe-
nolate via an electrophilic aromatic substitution mecha-
nism [15]. Thus, these model studies have provided
chemical insights into the arene hydroxylation reactions
performed by both (l-g2:g2-peroxo)Cu(II)2 and bis(l-
oxo)Cu(III)2 species.
ing that the hydroxylation proceeds via an electrophilic
pathway involving carbocationic intermediate [4].
a
However, in the above oxidation reaction, no (l-g2:g2-per-
oxo)Cu(II)2 species has been identified. So, it is impor-
tant to definitively confirm the presence of a (l-g2:g2-
peroxo)Cu(II)2 species as a reaction intermediate and, if
present, to know the nature of (l-g2:g2-peroxo)Cu(II)2
species, since the methyl group in the 2-position of a
xylyl linker could have significant influence on the stereo-
chemistry and electronic nature of (l-g2:g2-peroxo)Cu(II)2
species. Recently, we found that the (l-g2:g2-peroxo)-
Cu(II)2 species [Cu2(O2)(H–L–H)]2+ (3-O2) is 100 times
stable compared to [Cu2(O2)(H–XYL–H)]2+ (6-O2), which
allowed us to observe 3-O2 by benchtop UV–Vis
monitoring.
In order to investigate physicochemical properties of a
(l-g2:g2-peroxo)Cu(II)2 species which has a methyl or a
fluorine in the 2-position of xylyl linker, we prepared new
dinucleating ligands, Me2–L–Me and H–L–F shown in
Scheme 1, and their dinuclear copper(I) complexes,
[Cu2(Me2–L–Me)]2+ (1) and [Cu2(H–L–F)]2+ (2). The
complexes produce (l-g2:g2-peroxo)Cu(II)2 species,
[Cu2(O2)(Me2–L–Me)]2+ (1-O2) and [Cu2(O2)(H–L–F)]2+
(2-O2), which were characterized by resonance Raman
and UV–Vis spectroscopies. Decomposition of 1-O2
resulted in hydroxylation of the xylyl linker of Me2–L–
Me with a 1,2-migration of the methyl group (NIH shift)
Reaction of dinuclear copper(I) complexes [Cu2(Me2–
XYL–Me)]2+ and [Cu2(H–XYL–Me)]2+ with O2 has been
shown to cause the hydroxylation at the 2-position of the
xylyl linker with a 1,2-migration of the methyl group
(NIH shift) involving the elimination of a bis(pyridylethyl)
aminomethyl group from the supporting ligand, support-

T. Matsumoto et al. / Journal of Organometallic Chemistry 692 (2007) 111–121
113
to produce dinuclear [Cu2(Me3–L–O)2]2+ (4). Kinetic study
of the decompositions of 1-O2 and 2-O2 is also described.
m(C–O) bands at 2096 and 2085 cmꢀ1 for 1-CO, 2089 cmꢀ1
for 2-CO, and 2092 cmꢀ1 for 3-CO, which are comparable
to those of [Cu(Me2-pybn)(CO)]2+ (5-CO, 2100 and
2088 cmꢀ1) that has a half part of the dinucleating ligand
H–L–H (see Scheme 1) [16] and 7-CO (2079 cmꢀ1) [7], indi-
cating that the electron donor properties of these mono-
and dinucleating ligands are similar [7,16,17]. The coordi-
nated CO ligands in [Cu2(X–L–R)(CO)2]2+ can be removed
by refluxing the methanol solutions under N2 to afford
[Cu2(X–L–R)]2+ (1, 2, and 3) as air-sensitive yellow pow-
ders.
2. Results and discussion
2.1. Synthesis and characterization of copper(I) complexes
Copper(I) complexes, [Cu2(X–L–R)(CH3CN)2]2+ (X–L–
R = Me2–L–Me (1-CH3CN) and H–L–F (2-CH3CN)),
were prepared as yellow–green powders by the reaction
of [Cu(CH3CN)4]+ with the dinucleating ligands in acetoni-
trile/ethanol/water mixture under N2 as in the case of
[Cu2(H–L–H)(CH3CN)2]2+ (3-CH3CN) [13]. Solid samples
are relatively stable against O2 and can be handled in air.
Crystal structure of 1a-CH3CN shows that each Cu(I) site
has a tetrahedral structure with two pyridyl nitrogens, one
tertiary amine nitrogen, and an additional CH3CN as
˚
shown in Fig. 1. The Cuꢃ ꢃ ꢃCu distance is 8.995 A, which
is comparable to that of [Cu2(H–XYL–H)]2+ (6)
˚
(8.940 A) [3] and shorter than that of [Cu2(m-
XYLiPr4)(CO)2]2+ (7-CO, 10.061 A) [7]. H–XYL–H forms
˚
2.2. Spectroscopic characterization of
(l-g2:g2-peroxo)Cu(II)2 complexes
a tri-coordinate complex, whereas m-XYLiPr4 and X–L–R
form four-coordinate complexes with an additional
CH3CN. However, dioxygen reactivities of 7-CH3CN and
[Cu2(X–L–R)(CH3CN)2]2+ significantly differ; oxygenation
of 1-CH3CN, 2-CH3CN, and 3-CH3CN is very slow and
not fully completed probably due to strong coordination
of CH3CN, whereas 7-CH3CN is readily oxygenated [7].
Thus, for full oxygenation, we prepared [Cu2(X–L–R)]2+
(1, 2, and 3) from [Cu2(X–L–R)(CH3CN)2]2+ through
The copper(I) complexes 1, 2, and 3 in acetone at ꢀ78 ꢀC
react with O2 to produce the brown (l-g2:g2-peroxo)dicop-
per(II) complexes, [Cu2(O2)(X–L–R)]2+ (1-O2 (Me2–L–Me),
2-O2 (H–L–F), and 3-O2 (H–L–H)), which were character-
ized by resonance Raman and electronic spectroscopies. It
has been shown that the oxygenation of the dicopper(I) com-
plexes having a xylyl linker such as [Cu2(NO2–XYL–H)]2+
(NO2-6) and [Cu2(m-XYLiPr4)]2+ (7) generate both intramo-
lecular (l-g2:g2-peroxo)Cu(II)2 species and intermolecular
(l-g2:g2-peroxo)Cu(II)2 or bis(l-oxo)Cu(III)2 species such
as a dimer of dimer depending on the concentrations of the
complexes and the nature of the side arms [5,7]. NO2-6 gen-
erates an intramolecular (l-g2:g2-peroxo)Cu(II)2 species at
the concentrations bellow 4 mM, whereas 7 generates an
intermolecular bis(l-oxo)Cu(III)2 species even at the con-
centration of 0.1 mM. Thus the formation of intramolecular
and intermolecular species are highly dependent on not only
the concentration of complex, but also the nature of the side
arm of the dinucleating ligand. Similar observation was also
made for the present complexes.
[Cu2(X–L–R)(CO)2]2+
.
Reaction
of
[Cu2(X–L–
R)(CH3CN)2]2+ in ethanol with CO gave corresponding
carbonyl complexes, [Cu2(X–L–R)(CO)2]2+ (1-CO, 2-CO,
and 3-CO). The IR spectra of the complexes showed strong
The resonance Raman spectra of 1-O2 and 3-O2 in ace-
tone (ꢁ5 mM) measured with a 406.7 nm laser excitation
show isotope-sensitive bands at 752 cmꢀ1 16–18D = 42)
(
(Fig. 2) and 744 cmꢀ1 16–18D = 40) [13], respectively, which
(
are typical of the m(O–O) of the (l-g2:g2-peroxo)Cu(II)2
complexes and no m(Cu–O) band in the 550–630 cmꢀ1
region attributable to the bis(l-oxo)Cu(III)2 species is
observed [18,19]. However, the spectrum of 2-O2 is quite
different from those of 1-O2 and 3-O2; the spectrum shows
Fig. 1. ORTEP view (50% probability) of the cation of [Cu2(Me2–L–
Me)(CH3CN)2](PF6)2 Æ 0.5CH3CN Æ 0.25H2O (1a-CH3CN). Hydrogen
˚
atoms are omitted for clarity. Selected bond distances (A) and angles
(ꢀ): Cu1 ꢃ ꢃ ꢃ Cu2 8.995(1), Cu1–N1 2.233(5), Cu1–N2 2.069(5), Cu1–N3
2.044(5), Cu1–N7 1.906(5), Cu2–N4 2.251(4), Cu2–N5 2.007(5), Cu2–N6
2.063(5), Cu2–N8 1.893(5), N1–Cu1–N2 77.8(2), N1–Cu1–N3 80.2(2),
N1–Cu1–N7 141.5(2), N2–Cu1–N3 110.0(2), N2–Cu1–N7 113.8(2), N3–
Cu1–N7 124.0(2), N4–Cu2–N5 80.9(2), N4–Cu2–N6 81.4(2), N4–Cu2–N8
125.8(2), N5–Cu2–N6 113.4(2), N5–Cu2–N8 127.1(2), N6–Cu2–N8
115.2(2).
two isotope sensitive bands at 744 cmꢀ1
726 cmꢀ1
(
16–18D = 38) and
(
16–18D = 38) in the m(O–O) region, and some iso-
tope sensitive features in the 550–600 cmꢀ1 region (the
m(Cu–O) region). The two bands at 744 and 726 cmꢀ1 sug-
gest the presence of two types of (l-g2:g2-peroxo)Cu(II)2

114
T. Matsumoto et al. / Journal of Organometallic Chemistry 692 (2007) 111–121
Fig. 3. Resonance Raman spectra of 1-O2 (red), 2-O2 (blue), and 3-O2
(black) (ꢁ60 mM) in acetone at ꢀ90 ꢀC measured with a 514.5 nm laser
excitation. Asterisks are the solvent bands. (For interpretation of the
references in colour in this figure legend, the reader is referred to the web
version of this article.)
Fig. 2. Resonance Raman spectra of 1-O2 (red) and 2-O2 (blue) (ꢁ5 mM)
in acetone at ꢀ90 ꢀC measured with
a 406.7 nm laser excitation.
Asterisks are the solvent bands. (For interpretation of the references in
colour in this figure legend, the reader is referred to the web version of this
article.)
Cu(II)2 species of [Cu2(O2)(NO2–XYL–H)]2+ (NO2–6–O2)
has also been reported, although the shift (14 cmꢀ1 (from
747 to 733 cmꢀ1)) is significantly smaller than those of
the present complexes [5]. Unlike 1-O2 and 3-O2, 2-O2
showed only a single band at 714 cmꢀ1 attributable to an
intermolecular (l-g2:g2-peroxo)Cu(II)2 species. It has been
reported that [Cu2(H–XYL–F)]2+ generates only an inter-
molecular (l-g2:g2-peroxo)Cu(II)2 species even at 4 mM
[5]. Thus, introduction of the fluoro group into the xylyl
linker tends to produce intermolecular (l-g2:g2-peroxo)-
Cu(II)2 species, although the origin of this tendency is
not known at present.
It is noted that the dilution of the 60 mM solution of 3-
O2 in acetone to 10 mM causes no significant change in the
relative intensities of the m(O–O) bands at 715 and
744 cmꢀ1 for 30 min at ꢀ90 ꢀC as shown in Fig. 4, implying
that the conversion between intramolecular and intermo-
lecular (l-g2:g2-peroxo) species is slow under the condi-
tions. A similar behavior was also observed in the
electronic spectra of 1-O2, 2-O2, and 3-O2 (vide infra).
The electronic spectra of 1-O2, 2-O2, and 3-O2 in acetone
species, since 18O2 sample also showed the two correspond-
ing bands. The features in the 550–600 cmꢀ1 region sug-
gests a possibility of the presence of a bis(l-oxo)Cu(III)2
species (vide infra) [18,19]. It has been shown that the
bands in the 550–600 cmꢀ1 region attributable to the
bis(l-oxo)Cu(III)2 species are unusually strongly enhanced
by ꢁ400 nm laser excitation. Thus, the concentration of the
bis(l-oxo)Cu(III)2 species, if present, seems to be low.
The resonance Raman spectra of the concentrated ace-
tone solutions (ꢁ60 mM) of 1-O2, 2-O2, and 3-O2 measured
with a 514.5 nm laser excitation showed new bands at 719,
714, and 715 cmꢀ1, respectively, as shown in Fig. 3. For 1-
O2 and 3-O2, the spectra also exhibit the bands at 752 and
745 cmꢀ1 observed for the 5 mM samples, indicating that
the bands at 719 and 714 cmꢀ1, and the bands at 752 and
745 cmꢀ1 are attributable to those of the intermolecular
and intramolecular (l-g2:g2-peroxo)Cu(II)2 species,
respectively. It is noted that the m(O–O) of [Cu2(O2)(Me2-
pybn)2]2+ (5-O2) is observed at 714 cmꢀ1 [16], which is
almost the same as those of the present intermolecular
(l-g2:g2-peroxo)Cu(II)2 species. Thus, the xylyl bridges
of 1-O2 and 3-O2 have a substantial influence on the
m(O–O) frequency. Such low m(O–O) frequency shifts in
the intermolecular (l-g2:g2-peroxo)Cu(II)2 species may
be partly attributable to the decrease of a donation from
pꢂv orbital of the peroxo ligand to 4px orbital of the cop-
per(II) compared to those of the intramolecular (l-g2:g2-
peroxo)Cu(II)2 species which have a butterfly structure
(vide infra) [20]. A similar lower frequency shift of the
m(O–O) band of the intermolecular (l-g2:g2-peroxo)-
at ꢀ80 ꢀC showed intense pꢂr(O–O)-to-Cu(II) (ðdx2ꢀy2
ꢀ
dx2ꢀy2 Þ component) LMCT bands (kmax/nm (e/Mꢀ1 cmꢀ1))
at 359 (21000), 353 (26000), and 351 (32000) [13], respec-
tively, which are characteristic of those of the (l-g2:g2-per-
oxo)Cu(II)2 complexes (Fig. 5a). The intensity of the
LMCT transition decreases as the transition energy
decreases, which seem to be closely related to the bulkiness
of the substituents (F and CH3) in the bridging xylyl link-
ers. The electronic spectra of 1-O2 and 2-O2 also showed
additional bands at 442 nm (e = 5600 Mꢀ1 cmꢀ1) and 433
nm (e = 3400 Mꢀ1 cmꢀ1), respectively. Gaussian analysis

T. Matsumoto et al. / Journal of Organometallic Chemistry 692 (2007) 111–121
115
Scheme 2.
have a butterfly structure. The intensity and transition
energy of the forbidden band has been shown to be depen-
dent on the bending of the Cu2O2 plane; as the angle /
shown in Scheme 2 becomes smaller, the intensity increases
and the transition energy lowers. Thus, the order of the
angles / is 1-O2 < 3-O2. This is also supported by the trend
observed for the m(O–O) frequencies (752 cmꢀ1 for 1-O2
and 744 cmꢀ1 for 3-O2). Solomon et al. reported that the
m(O–O) is dependent on the magnitude of the angle / of
the butterfly core: a smaller angle / causes a high frequency
shift of the m(O–O), since the bending increases a donation
from pꢂv orbital of the peroxo ligand to 4px orbital of the
copper(II) [20]. Although the electronic spectrum of 2-O2
shows a similar additional absorption band at 433 nm
(e = 3400 Mꢀ1 cmꢀ1) as noted earlier, there is a possibility
of the presence of bis(l-oxo)Cu(III)2 species which also has
a LMCT band in this region. Further study is needed for
fuller understanding of the spectroscopic properties of 2-
O2.
Fig. 4. Resonance Raman spectra of 3-O2 at (a) ꢁ60 mM (red), (b)
ꢁ30 mM (blue), and (c) ꢁ10 mM (black) in acetone at ꢀ90 ꢀC measured
with a 514.5 nm laser excitation, where the samples of (b) and (c) were
prepared by dilution of a 60 mM solution (a) (see text). (For interpretation
of the references in colour in this figure legend, the reader is referred to the
web version of this article.)
of the electronic spectrum of 3-O2 revealed that 3-O2 also
has an additional band at ꢁ390 nm. Similar spectral fea-
tures have also been reported for [Cu2(O2)(NnPY2)]2+
(n = 3–5) [20]. The additional absorption bands for 1-O2
and 3-O2 can be assigned to the pꢂr(O–O)-to-Cu(II)
(ðdx2ꢀy2 þ dx2ꢀy2 Þ component) LMCT transition based on
the assignment made by Solomon et al. [20]. This transition
has been shown to be forbidden for (l-g2:g2-peroxo)-
Cu(II)2 complex in a planar structure (C2h), whereas the
transition become allowed by a symmetry lowering (C2v).
Thus, the electronic spectra suggest that 1-O2 and 3-O2 also
It is noted that the intensities of the absorption bands
observed at ꢁ390–442 nm for the acetone solutions
(ꢁ0.5 mM) of 1-O2, 2-O2, and 3-O2 prepared by dilution
of the 60 mM solutions decrease as shown in Fig. 5b, sug-
gesting that the intermolecular (l-g2:g2-peroxo)Cu(II)2
species have a planar structure.
Fig. 5. (a) Electronic spectra of 1-O2 (0.254 mM, red), 2-O2 (0.265 mM, blue), and 3-O2 (0.255 mM, black) in acetone at ꢀ90 ꢀC. Inset: Gaussian analysis
of the spectrum of 3-O2. (b) Electronic spectra of 1-O2 (0.5 mM, red), 2-O2 (0.5 mM, blue), and 3-O2 (0.5 mM, black) in acetone diluted from 60 mM
acetone solutions at ꢀ90 ꢀC. Decrease of the intensities of the ꢁ350 nm bands appear to be due to some decomposition during dilution processes. (For
interpretation of the references in colour in this figure legend, the reader is referred to the web version of this article.)

116
T. Matsumoto et al. / Journal of Organometallic Chemistry 692 (2007) 111–121
2.3. Self-decompositions
change from brown through green to dark-brown. The
electronic spectrum of the final dark-brown solution
showed an absorption band at 460 nm, reminiscent of for-
mation of a Cu(II) phenolate complex. We have succeeded
in the isolation and crystallization of the dark-brown spe-
cies. Crystal structure revealed that the complex has a
dimer structure having Me3–L–Oꢀ ligands, indicating that
Me2–L–Me ligand is hydroxylated at the 2-position of the
xylyl linker with a 1,2-methyl migration (NIH shift) accom-
panied by the elimination of a bis(6-methyl-2-pyridylm-
ethyl)aminomethyl sidearm as shown in Fig. 6 and
Scheme 3. As noted earlier, a similar observation has
already been reported for the oxidation of [Cu2(H–XYL–
Me)]2+ and [Cu2(Me2–XYL–Me)]2+ with dioxygen [4].
Ligand recovery experiment gave Me3–L–OH (yield =
ꢁ77% based on 1-O2) and bis(6-methyl-2-pyridylmethyl)-
amine (Me2-bpa) (yield could not be determined from
NMR). Isotope labeling experiment by 18O2 revealed that
the oxygen atom of Me3–L–OH comes from 18O2. Occur-
rence of 1,2-methyl migration indicates that the hydroxyl-
ation reaction proceeds via an electrophilic aromatic
substitution mechanism involving a carbocationic interme-
diate [4]. We have also studied decomposition of 2-O2 in
acetone at room temperature. Although ligand recovery
experiments showed an original H–L–F and some modified
ligands which were detected by NMR and ESI-TOF/MS, it
was difficult to identify those modified ligands.
Self decomposition of [Cu2(O2)(H–L–H)]2+ (3-O2) in
acetone (< ꢁ 5 mM) causes almost quantitative hydroxyl-
ation of the xylyl linker of H–L–H via an electrophilic aro-
matic substitution reaction [13]. However, decomposition
of 3-O2 in a concentrated acetone solution (ꢁ60 mM) at
ꢀ60 ꢀC which contains intra- and intermolecular (l-
g2:g2-peroxo) species gave a hydroxylated ligand H–L–
OH (ꢁ44% yield based on 3-O2) together with H–L–H
(ꢁ38%), a ligand-based alcohol (ꢁ7%), and a ligand-based
aldehyde (ꢁ9%). In the latter two modified ligands, one of
the methyl groups is oxidized to alcohol and aldehyde,
respectively, which were identified by NMR and ESI-
TOF/MS. Such a low oxidation yield of the xylyl linker
(ꢁ44%) suggests that the amount of intermolecular species
is more than 50% under the above conditions. The oxida-
tion of the 6-methyl group is in line with the observation
for the decomposition of [Cu2(O2)(Me2-pybn)2]2+ (5-O2),
which also gave a ligand-based alcohol and a ligand-based
aldehyde [21]. Thus intermolecular (l-g2:g2-peroxo) spe-
cies is capable of oxidation of the 6-methyl group of the
supporting ligand.
Decomposition of [Cu2(O2)(Me2–L–Me)]2+ (1-O2,
0.754 mM) in acetone at ꢀ50 ꢀC showed a successive color
2.4. Kinetic study of decompositions of 1-O2, 2-O2, and 3-O2
Decompositions of [Cu2(O2)(Me2–L–Me)]2+ (1-O2) and
[Cu2(O2)(H–L–F)]2+ (2-O2) in acetone (0.1–1.0 mM) obey
Fig. 6. ORTEP views (50% probability) of the cationic portion of
[Cu2(Me3–L–O)2](PF6)2 Æ C2H5OH Æ H2O (4). Hydrogen atoms are omitted
˚
for clarity. Selected bond distances (A) and angles (ꢀ): Cu1 ꢃ ꢃ ꢃ Cu1*
3.103(1), Cu1–O1 1.952(3), Cu1–O1* 1.989(3), Cu1–N1 2.133(4), Cu1–N2
2.001(4), Cu1–N3 2.091(4), O1–Cu1–O1* 76.1(1), O1–Cu1–N1 91.4(1),
O1–Cu1–N2 169.3(1), O1–Cu1–N3 91.7(1), O1*–Cu1–N1 137.1(1), O1*–
Cu1–N2 102.8(1), O1*–Cu1–N3 137.4(1), N1–Cu1–N2 96.3(1), N1–Cu1–
N3 82.7(1), N2–Cu1–N3 81.9(1).
Scheme 3. A possible oxidation pathway of [Cu2(O2)(Me2–L–Me)]2+ (1-
O2) with 1,2-methyl migration (NIH shift).

T. Matsumoto et al. / Journal of Organometallic Chemistry 692 (2007) 111–121
117
(q = ꢀ1.84 at ꢀ55 ꢀC) [11], and [Cu2(p–R–PhO)
(O)2(DBED)2]2+ (q = ꢀ2.2 at ꢀ120 ꢀC) [15]. The rate con-
stant observed for 1-O2 is in line with the electron donating
effect of the methyl groups, which appears to be partly
responsible to the decomposition rates of the complex.
However, steric effect of the methyl group in the 2-position
of the xylyl linker may also contribute to the decomposi-
tion rate (vide infra).
Fig. 8 showed Eyring plots for the decompositions of 1-
O2 and 2-O2. The activation parameters of 2-O2
(DHꢀ = 62 1 kJ molꢀ1 and DSꢀ = ꢀ10 6 J molꢀ1 Kꢀ1
)
are comparable to those of 3-O2 (DHꢀ = 63 1 kJ molꢀ1
and DSꢀ = ꢀ11 3 J molꢀ1 Kꢀ1) as shown in Table 1.
Unlike 2-O2, the high reactivity of 1-O2 relative to 3-O2
is attributable to favorable enthalpy effect (a decrease of
the activation enthalpy change, DHꢀ = 53 1 kJ molꢀ1),
although entropy effect is unfavorable (a decrease of the
activation entropy change, DSꢀ = ꢀ44 6 J molꢀ1 Kꢀ1).
These enthalpy and entropy effects are not in line with a
trend attributed to the electronic effect of p-substituents
X (X = MeO, t-Bu, H, and NO2) observed for a series of
[Cu2(O2)(X–L–H)]2+ as seen in Table 1; as the electron
donating nature of the substituent X increases, the activa-
tion enthalpy changes decrease, but the activation entropy
changes increase, suggesting that stereochemical effect of
the methyl group in the 2-position of the xylyl linker has
Fig. 7. Representative spectral change for decomposition of 1-O2 (10 s
interval) in acetone at ꢀ50 ꢀC. Experimental conditions: [1-
O2] = 0.737 mM. Inset: Time course of the absorbance change at
359 nm and the first-order kinetic plot.
first-order kinetics (Fig. 7). Comparison of the first order
rate constants of 1-O2 (1.04 · 10ꢀ2 sꢀ1), 2-O2 (3.25 ·
10ꢀ3 sꢀ1), and 3-O2 (2.12 · 10ꢀ3 sꢀ1) at ꢀ50 ꢀC shown in
Table 1 reveals that introduction of fluorine into the xylyl
linker has no significant influence on the decomposition
rate, but introduction of three methyl groups accelerates
the decomposition rate by ꢁ5 times. Very recently, we have
reported that decomposition rate constants (hydroxylation
rate constants of the xylyl linker) of a series of [Cu2(O2)(X–
L–H)]2+ (X–L–H = H–L–H, MeO–L–H, t-Bu–L–H, and
NO2–L–H) are highly dependent on the electron donating
ability of the substituent X and a Hammett q value is
ꢀ1.9 at ꢀ50 ꢀC, indicating that the aromatic hydroxylation
mediated by [Cu2(O2)(X–L–H)]2+ proceeds via an electro-
philic aromatic substitution mechanism. Similar observa-
tions have also been made for various complexes such as
[Cu2(O2)(X–XYL–H)]2+ (q = ꢀ2.1 at ꢀ80 ꢀC) [6b],
[Cu2(O2)(Lpy2Bz)2]2+ (q = ꢀ1.8) [12], [Cu2(O2)(MeL66)]2+
a
significant influence on
a
transition state for
decomposition.
It should be noted that, although a Hammett q value
(ꢀ2.1 at ꢀ80 ꢀC) of a series of [Cu2(O2)(X–XYL–H)]2+ is
quite similar to that (ꢀ1.9 at ꢀ50 ꢀC) of [Cu2(O2)(X–L–
H)]2+ complexes, the enthalpy and entropy effects are quite
different; for a series of [Cu2(O2)(X–XYL–H)]2+, as the
electron donating nature of the substituent X increases,
both activation enthalpy and entropy changes decrease.
Table 1
Kinetic parameters for arene hydroxylation by (l-g2:g2-peroxo)Cu(II)2
complexes having xylyl linker
Ligand
DHꢀ
DSꢀ
k1 (sꢀ1 a
)
Reference
(kJ molꢀ1
)
(J molꢀ1 Kꢀ1
)
Me2–L–Me
H–L–F
H–L–H
MeO–L–H
t–Bu–L–H
NO2–L–H
t-Bu–XYL–H
H–XYL–H
NO2–XYL–H
m-XYLiPr4
53
62
63
59
62
65
41
50
55
1
1
1
1
1
1
2
1
1
ꢀ44
ꢀ10
ꢀ11
8
6
6
1
4
4
3
8
2
2
1.04 · 10ꢀ2 This Work
3.25 · 10ꢀ3 This Work
2.12 · 10ꢀ3 [13]
1.35 · 10ꢀ1 [13]
1.79 · 10ꢀ2 [13]
1.43 · 10ꢀ4 [13]
1.07 · 10ꢀ4 [6]
1
ꢀ26
ꢀ59
ꢀ35
ꢀ32
1.49 · 10ꢀ1 [6]
1.34 · 10ꢀ2 [6]
50.1 0.2
ꢀ50.4 0.9
1.99 · 10ꢀ2 [7]
a
At ꢀ50 ꢀC.
Fig. 8. Eyring plots for the decompositions of 1-O2 and 2-O2 in acetone.

118
T. Matsumoto et al. / Journal of Organometallic Chemistry 692 (2007) 111–121
Thus, it appears likely that nature of the transition states
for decomposition (hydroxylation) of [Cu2(O2)(X–L–
120.0 (py), 60.2 (CH2), 52.8 (CH2), 24.3 (pyCH3), 20.4
(xylCH3), 16.1 (xylCH3). ESI-TOF/MS (acetonitrile solu-
tion containing a small amount of formic acid): m/z (rela-
tive intensity) = 300.19 (100) [M + 2H]2+, 599.39 (3)
[M + H]+. Anal. Calc. for C39H46N6: C, 78.22; H, 7.74;
N, 14.03. Found: C, 78.04; H, 7.84; N, 13.90%.
H)]2+ is different from that of [Cu2(O2)(X–XYL–H)]2+
.
In summary, we have synthesized dicopper(I) complexes
with dinucleating ligands having a xylyl linker, [Cu2(X–L–
R)]2+ (1 and 2), which generate (l-g2:g2-peroxo)Cu(II)2
species, [Cu2(O2)(X–L–R)]2+ (1-O2 and 2-O2), by the reac-
tion with O2 in acetone at low temperatures. Resonance
Raman and UV–Vis spectroscopies revealed that at low
concentrations (less than 5 mM), 1 and 2 produce intramo-
lecular (l-g2:g2-peroxo)Cu(II)2 species, whereas at
ꢁ60 mM concentration, they form intermolecular (l-
g2:g2-peroxo)Cu(II)2 species. Decomposition of 1-O2
resulted in hydroxylation of the 2-position of the xylyl lin-
ker with a NIH shift of the methyl group, suggesting that
the hydroxylation proceeds via an electrophilic aromatic
1,3-Bis[bis(6-methyl-2-pyridylmethyl)aminomethyl]-2-
fluorobenzene Æ 0.5H2O (H–L–F). To a mixture of bis(6-
methyl-2-pyridylmethyl)amine (5.45 g, 24.0 mmol) and
1,3-bis(bromomethyl)-2-fluorobenzene (3.39 g, 12.0 mmol)
in chloroform (100 mL) was added an aqueous solution
(30 mL) of Na2CO3 Æ H2O (2.98 g, 24.0 mmol). The reac-
tion mixture was vigorously stirred at room temperature
for three days. The mixture was extracted with chloroform
(3 · 30 cm3) and the combined extracts were dried over
Na2SO4 and evaporated under reduced pressure to give
an orange oil, which was dissolved into hot hexane
(200 mL) and insoluble material was removed by decanta-
tion. The resulting hexane solution was allowed to stand at
substitution mechanism involving
intermediate.
a
carbocationic
1
3. Experimental
ꢀ80 ꢀC to give a white powder. Yield: 6.53 g (93%). H
NMR ((CD3)2CO, 400 MHz): d (ppm) = 7.59 (4H, t,
pyH), 7.54 (2H, t, xylH), 7.43 (4H, d, pyH), 7.13 (1H, t,
xylH), 7.05 (4H, d, pyH), 3.76 (12H, s, NCH2py +
NCH2xyl), 2.44 (12H, s, pyCH3). 13C{1H} NMR (CDCl3,
100.4 MHz): d (ppm) = 160.0 (d, xylC-F), 159.2 (py),
157.4 (py), 136.6 (py), 129.8 (xyl), 125.7 (xyl), 123.4 (xyl),
121.3 (py), 119.3 (py), 60.3 (CH2), 51.6 (CH2), 24.4
(pyCH3). ESI-TOF/MS (acetonitrile solution containing a
small amount of formic acid): m/z (relative inten-
sity) = 288.17 (100) [M + 2H]2+, 575.34 (4) [M + H]+.
Anal. Calc. for C36H40N6O0.5F1: C, 74.07; H, 6.91; N,
14.40. Found: C, 74.06; H, 6.84; N, 14.48%.
3.1. General
Acetone was dried over Molecular Sieves 4A and
distilled under N2 before use. Bis(6-methyl-2-pyridyl-
methyl)amine (Me2-bpa) [22], 1,3-bis(bromomethyl)-2-flu-
orobenzene [4b], 1,3-bis[bis(6-methyl-2-pyridylmethyl)ami-
nomethyl]benzene (H–L–H) [13] and [Cu2(H–L–H)](PF6)2
[13] were synthesized according to the literature methods.
All other reagents and solvents were commercially avail-
able and used without further purification.
3.2. Syntheses of ligands
3.3. Syntheses of complexes
1,3-Bis[bis(6-methyl-2-pyridylmethyl)aminomethyl]-2,4,
6-trimethylbenzene (Me2–L–Me). To a mixture of bis(6-
methyl-2-pyridylmethyl)amine (4.72 g, 20.8 mmol) and
2,4-bis(chloromethyl)-1,3,5-trimethylbenzene (2.25 g, 10.4
mmol) in chloroform (100 mL) was added an aqueous solu-
tion (30 mL) of Na2CO3 Æ H2O (2.58 g, 20.8 mmol). Reac-
tion mixture was vigorously stirred at room temperature
for 3 days. The mixture was extracted with chloroform
(3 · 30 cm3) and the combined extracts were dried over
Na2SO4 and evaporated under reduced pressure to give
an orange oil, which was dissolved into hot hexane
(200 mL) and insoluble material was removed by decanta-
tion. The resulting hexane solution was allowed to stand at
ꢀ80 ꢀC to give a yellow oil, which was isolated and dis-
solved into 20 mL of ethanol. The solution was allowed
to stand for a week to yield a yellow powder. Yield:
Copper(I) complexes were prepared under N2 atmo-
sphere using standard Schlenk techniques. Caution: All
the perchlorate salts are potentially explosive and should be
handled with care.
[Cu2(Me2–L–Me)(CH3CN)2](PF6)2 Æ H2O (1-CH3CN).
Using the same procedure as described for the synthesis
of [Cu2(H–L–H)(CH3CN)2](PF6)2 [13], an ethanol suspen-
sion (10 mL) containing [Cu(CH3CN)4]ClO4 (0.327 g,
1.00 mmol), Me2–L–Me (0.308 g, 0.514 mmol), and
NH4PF6 (0.330 g, 2.03 mmol) was heated to give a yellow
solution, which was allowed to stand for a day to afford a
yellow–green powder. Yield: 0.425 g (76 %). 1H NMR
((CD3)2CO, 300 MHz): d (ppm) = 7.83 (4H, t, pyH), 7.43
(4H, d, pyH), 7.33 (4H, d, pyH), 6.93 (1H, s, xylH), 4.06
(4H, s, NCH2xyl), 3.99 (8H, s, NCH2py), 2.77 (12H, s,
pyCH3), 2.43 (3H, s, xylCH3), 2.32 (6H, s, xylCH3), 2.23
(6H, s, CH3CN). FTIR (KBr, cmꢀ1): 1604 (C@C, aro-
matic), 1577 (C@C, aromatic), 847 ðPFꢀ6 Þ, 557 ðPF6ꢀÞ.
ESI-TOF/MS (acetonitlie under N2 atmosphere): m/z (rela-
tive intensity) = 362.12 (100) [Cu2(Me2–L–Me)]2+. Anal.
Calc. for C43H54N8O1Cu2F12P2: C, 46.28; H, 4.88; N,
10.04. Found: C, 46.03; H, 4.73; N, 9.96%. Recrystallization
4.46 g (72 %). 1H NMR (CDCl3, 400 MHz):
d
(ppm) = 7.46 (t, 4H, pyH), 7.21 (d, 4H, pyH), 6.95 (d,
4H, pyH), 6.71 (s, 1H, xylH), 3.70 (s, 12H,
NCH2py + NCH2xyl), 2.50 (s, 12H, pyCH3), 2.29 (s, 3H,
xylCH3), 2.22 (s, 6H, xylCH3). 13C{1H} NMR (CDCl3,
100.4 MHz): d (ppm) = 159.4 (py), 157.0 (py), 138.9 (xyl),
137.0 (xyl), 136.2 (py), 132.8 (xyl), 130.0 (xyl), 121.1 (py),

T. Matsumoto et al. / Journal of Organometallic Chemistry 692 (2007) 111–121
119
of 1-CH3CN from EtOH/H2O/CH3CN mixture gave single
364.09 (12) [Cu2(H–L–F)(CO)]2+
.
Anal. Calc. for
crystals, [Cu2(Me2–L–Me)(CH3CN)2](PF6)2 Æ 0.5CH3CN Æ
0.25H2O (1a-CH3CN), suitable for X-ray crystallography.
[Cu2(Me2–L–Me)(CO)2](PF6)2 (1-CO). This was synthe-
sized as a white powder in the same manner as described
for synthesis of [Cu2(H–L–H)(CO)2](PF6)2 [13] using
[Cu2(Me2–L–Me)(CH3CN)2](PF6)2 Æ H2O (0.425 g, 0.381
mmol). Yield: 0.376 g (92 %). 1H NMR ((CD3)2CO,
400 MHz): d (ppm) = 7.94 (4H, t, pyH), 7.52 (4H, d,
pyH), 7.46 (4H, d, pyH), 7.11 (1H, s, xylH), 3.75–4.32
(12H, m, NCH2py + NCH2xyl), 2.81 (12H, s, pyCH3),
2.59 (3H, s, xylCH3), 2.33 (6H, s, xylCH3). FTIR (KBr,
cmꢀ1): 2096 (C„O), 2085 (C„O), 1606 (C@C, aromatic),
1577 (C@C, aromatic), 837 ðPFꢀ6 Þ, 557 ðPF6ꢀÞ. ESI-TOF/
MS (acetone under N2 atmosphere): m/z (relative inten-
C38H39N6O2Cu2F13O2P2: C, 43.55; H, 3.75; N, 8.02.
Found: C, 43.87; H, 3.77; N, 8.10%.
[Cu2(H–L–F)](PF6)2 (2). This was prepared by the same
method as described for synthesis of [Cu2(H–L–H)](PF6)2
[13] except [Cu2(H–L–F)(CO)2](PF6)2 was used. ESI-
TOF/MS (acetone under N2 atmosphere): m/z (relative
intensity) = 350.09 (100) [Cu2(H–L–F)]2+. FTIR (KBr,
cmꢀ1): 1610 (C@C, aromatic), 1576 (C@C, aromatic),
843 ðPFꢀ6 Þ, 559 ðPFꢀ6 Þ.
[Cu2(Me3–L–O)2](PF6)2 Æ C2H5OH Æ H2O (4). [Cu2(Me2–
L–Me)](PF6)2 (0.0250 mmol) prepared from [Cu2(Me2–L–
Me)(CO)2](PF6)2 (0.0268 g, 0.0250 mmol) was dissolved
in acetone (10 mL). The solution was then exposed to O2
at room temperature to give a dark-brown solution, to
which was added 5 mL of ethanol. The solution was
allowed to stand for a week to yield dark brown crystals
suitable for X-ray analysis. Yield: 0.0200 g (65 %). FTIR
(KBr, cmꢀ1): 1608 (C@C, aromatic), 1577 (C@C, aro-
matic), 837 ðPFꢀ6 Þ, 557 ðPFꢀ6 Þ. UV–Vis (acetonitrile,
0.132 mM) (kmax/nm (e/Mꢀ1 cmꢀ1)): 460 (6300), 330
(4200, sh). Anal. Calc. for C50H64N6Cu2F12O4P2: C,
48.82; H, 5.24; N, 6.83. Found: C, 48.73; H, 5.33; N,
6.69%. ESI-TOF/MS (acetonitrile): m/z (relative inten-
sity) = 362.12 (100) [Cu2(Me2–L–Me)]2+
, 376.12 (16)
[Cu2(Me2–L–Me)(CO)]2+. Anal. Calc. for C41H46N6O2-
Cu2F12P2: C, 45.94; H, 4.33; N, 7.84. Found: C, 46.15;
H, 4.34; N, 7.91%.
[Cu2(Me2–L–Me)](PF6)2 (1). This was prepared by the
same method as described for synthesis of [Cu2(H–L–
H)](PF6)2 [13] except [Cu2(Me2–L–Me)(CO)2](PF6)2 was
used. ESI-TOF/MS (acetone under N2 atmosphere): m/z
(relative intensity) = 362.12 (100) [Cu2(Me2–L–Me)]2+
.
FTIR (KBr, cmꢀ1): 1610 (C@C, aromatic), 1576 (C@C,
aromatic), 839 ðPFꢀ6 Þ, 557 ðPF6ꢀÞ.
sity) = 437.15 (100) [Cu2(Me3–L–O)2]2+
.
[Cu2(H–L–F)(CH3CN)2](PF6)2 (2-CH3CN). The com-
plex was prepared by the method similar to the synthesis
of [Cu2(H–L–H)(CH3CN)2](PF6)2 [13]. An ethanol suspen-
sion (10 mL) containing [Cu(CH3CN)4]ClO4 (0.327 g,
1.00 mmol), H–L–F Æ 0.5H2O (0.292 g, 0.500 mmol), and
NH4PF6 (0.325 g, 1.99 mmol) was heated to give a yellow
solution, which was allowed to stand for a day to afford
a yellow–green powder. Yield: 0.489 g (91%). 1H NMR
((CD3)2CO, 400 MHz): d (ppm) = 7.84 (4H, t, pyH), 7.56
(2H, t, xylH), 7.43 (4H, d, pyH), 7.33 (4H, d, pyH), 7.22
(1H, t, xylH), 3.82–4.17 (12H, m, NCH2py + NCH2xyl),
2.79 (12H, s, pyCH3), 2.34 (6H, s, CH3CN). FTIR (KBr,
cmꢀ1): 1604 (C@C, aromatic), 1577 (C@C, aromatic),
850 ðPFꢀ6 Þ, 557 ðPF6ꢀÞ. ESI-TOF/MS (acetone under N2
atmosphere): m/z (relative intensity) = 350.09 (100)
3.4. Physical measurements
The electronic spectra were measured with a Shimadzu
diode array spectrometer Multispec-1500 with a Unisoku
thermostated cell holder designed for low-temperature
experiments.
Resonance Raman spectra were obtained with a liquid
nitrogen cooled CCD detector (LN/CCD-1100-PB, Roper
Scientific) attached to a 1 m-single polychromator (Model
MC-100DG, Ritsu Oyo Kogaku). The 406.7 nm line of a
Kr+ laser (Model 2060 Spectra Physics) and 514.5 nm line
of an Ar+ laser (GLC 3200, NEC) were used as the exciting
sources. The laser powers used for the 406.7 and 514.5 nm
excitations were 2 and 10 mW, respectively, at the sample
points. All measurements were carried out with a spinning
cell (1000 rpm) keeping at ꢀ80 to ꢀ90 ꢀC. Raman shifts
were calibrated with indene and the accuracy of the peak
[Cu2(H–L–F)]2+, 370.61 (34) [Cu2(H–L–F)(CH3CN)]2+
,
391.12 (2) [Cu2(H–L–F)(CH3CN)2]2+. Anal. Calc. for
C40H45N8Cu2F13P2: C, 44.74; H, 4.22; N, 10.43. Found:
C, 44.62; H, 4.24; N, 10.55%.
positions of the Raman bands was 1 cmꢀ1
.
1H and 13C{1H} NMR spectra were measured with
JEOL JNM-LM300 or -LM400 using tetramethylsilane
(TMS) for 1H NMR and d1-chloroform for 13C{1H}
NMR as the internal standards. Infrared spectra were
obtained by KBr disk method with a HORIBA FT-200
spectrophotometer.
[Cu2(H–L–F)(CO)2](PF6)2 (2-CO). This was synthesized
as a white powder in the same manner as described for syn-
thesis of [Cu2(H–L–H)(CO)2](PF6)2 [13] using [Cu2(H–L–
F)(CH3CN)2](PF6)2 (0.489 g, 0.455 mmol). Yield: 0.447 g
1
(94%). H NMR ((CD3)2CO, 400 MHz): d (ppm) = 7.90
(4H, t, pyH), 7.67 (2H, t, xylH), 7.49 (4H, d, pyH), 7.39
(4H, d, pyH), 7.26 (1H, t, xylH), 4.43 (4H, d, NCH2py),
4.39 (4H, s, NCH2xyl), 4.00 (4H, d, NCH2py), 2.82–2.85
(12H, m, pyCH3). FTIR (KBr, cmꢀ1): 2089 (C„O), 1608
(C@C, aromatic), 1577 (C@C, aromatic), 839 ðPFꢀ6 Þ, 557
ðPFꢀ6 Þ. ESI-TOF/MS (acetone under N2 atmosphere): m/
3.5. Isolation and identification of modified ligands by
thermal decomposition of [Cu2(O)2(H–L–H)](PF6)2
(3-O2) under O2
Ligand recovery experiment was performed as follows.
z
(relative intensity) = 350.09 (100) [Cu2(H–L–F)]2+
,
Typically,
3
(0.0605 mmol) prepared from 3-CO

120
T. Matsumoto et al. / Journal of Organometallic Chemistry 692 (2007) 111–121
(0.0623 g, 0.0605 mmol) was dissolved into 1 mL of ace-
tone and the resulting solution (60.5 mM) was exposed to
O2 at ꢀ78 ꢀC and allowed to stand at ꢀ60 ꢀC overnight
to complete the decomposition. Evaporation of acetone
under reduced pressure yielded a dark green oil. To this
oil was added CHCl3 (20 mL) and concentrated aqueous
ammonia (20 mL) with stirring. Organic compounds were
extracted with CHCl3 (3 · 20 mL). The combined extracts
were dried over Na2SO4 and chloroform was removed by
evaporation under a reduced pressure to give an oil. H–
L–OH, H–L–H and the modified ligands in which one of
6-methyl group of pyridyl group is oxidized to alcohol
and aldehyde which were identified by 1H NMR and
3.7. Kinetic measurements
The decompositions of [Cu2(O2)(Me2–L–Me)]2+ (1-O2)
and [Cu2(O2)(L–F)]2+ (2-O2) were followed in 1 mm path
length UV–Vis cell that was held in a Unisoku thermos-
tared cell holder designed for low temperature experiments.
Acetone solutions of 1-O2 (0.2–1.0 mM) and 2-O2 (0.1–
0.5 mM) were kept at the desired temperatures for
30 min, O2 gas was bubbled for 5–10 s. Rate constants
for decompositions of 1-O2 and 2-O2 were determined by
monitoring decrease of the absorbance at 359 and
353 nm, respectively.
1
ESI-TOF/MS, and their amounts were determined by H
3.8. 18O2 labeling experiments
NMR by addition of 1 mL of 2,6-dimethyl-1,4-benzoqui-
none CDCl3 solution (0.0256 M) as an internal standard.
Yield of H–L–H, H–L–OH, ligand-based alcohol, and
ligand-based aldehyde were 38, 44, 7, and 9% based on 3-
CO, respectively. H–L–OH: 1H NMR (CDCl3,
300 MHz): d (ppm) = 7.49 (4H, t, pyH), 7.33 (4H, d,
pyH), 7.20 (2H, d, xylH), 6.97 (4H, d, pyH), 6.74 (1H, t,
xylH), 3.83 (8H, s, NCH2py), 3.79 (4H, s, NCH2xyl),
2.52 (12H, s, pyCH3). ESI-TOF/MS (acetonitrile solution
containing a small amount of formic acid): m/z (relative
Hydroxylation of ligand was also performed using 18O2.
The procedures were the same as those of 16O2 experiments
except for using 18O2. Me3–L–18OH: ESI-TOF/MS (aceto-
nitrile solution containing a small amount of formic acid):
m/z (relative intensity) = 378.25 (100) [Me3–L–18OH +
H]+. Me3–L–16OH: ESI-TOF/MS (acetonitrile solution
containing a small amount of formic acid): m/z (relative
intensity) = 376.24 (100) [Me3–L–16OH + H]+.
intensity) = 287.17 (100) [M + 2H]2+
[M + H]+.
,
573.33 (24),
Table 2
Crystallographic data for [Cu2(Me2–L–Me)(CH3CN)2](PF6)2 Æ 0.5CH3C-
N Æ 0.25H2O (1a-CH3CN), and [Cu2(Me3–L–O)2](PF6)2 Æ C2H5OH Æ H2O
(4)
1a-CH3CN
4
3.6. Isolation and identification of modified ligands by
thermal decomposition of [Cu2(O)2(Me2–L–Me)](PF6)2
(1-O2) under O2
Empirical formula
Formula weight
Crystal system
C
44H54O0.25N8.5Cu2P2F12 C50H64O4N6Cu2P2F12
1122.99
Monoclinic
P21/c (#14)
1230.11
Triclinic
P1 (#2)
ꢀ
Space group
Ligand recovery experiment was performed as follows.
Unit cell dimension
˚
a (A)
23.111(2)
8.6299(9)
27.442(2)
90
112.972(2)
90
5039.1(6)
4
60.9
9.446(2)
12.033(2)
13.149(2)
73.65(1)
65.24(1)
86.27(1)
1299.9(4)
1
Typically,
1
(0.0188 mmol) prepared from 1-CO
˚
b (A)
(0.0202 g, 0.0188 mmol) was dissolved into 25 mL of ace-
tone and the resulting solution (0.754 mM) was exposed
to O2 at ꢀ50 ꢀC and allowed to stand one hour to complete
the decomposition. Evaporation of acetone under reduced
pressure yielded a dark brown oil. To this oil was added
CHCl3 (20 mL) and concentrated aqueous ammonia
(20 mL) containing excess Na4EDTA with vigorously stir-
ring. Organic compounds were extracted with CHCl3
(3 · 20 mL). The combined extracts were dried over
Na2SO4 and chloroform was removed by evaporation
under a reduced pressure to give an oil. The reaction prod-
˚
c (A)
a (ꢀ)
b (ꢀ)
c (ꢀ)
3
˚
V (A )
Z
2hmax
F(000)
55.0
634.00
1.571
2302.00
1.480
9.93
Dcalc (g cmꢀ3
)
l (Mo Ka) (cmꢀ1
Temperature (ꢀC)
)
9.74
ꢀ155
ꢀ150
Number of measured 29443
13456
1
reflections
Number of unique
reflections
ucts were identified by H NMR and ESI-TOF/MS, and
6534 (I P 3.00r(I))
4461 (I P 3.00r(I))
1
their amounts were determined by H NMR by addition
of 1 mL of 2,6-dimethyl-1,4-benzoquinone CDCl3 solution
(0.0190 M) as an internal standard. Yield of Me3–L–OH
was 77 % based on 1-CO. Me3–L–OH: H NMR (CDCl3,
Number of variables 620
347
1.69
1.85 and ꢀ0.85
Goodness-of-fit
Maximum and
minimum of
1.45
1.38 and ꢀ0.62
1
400 MHz): d (ppm) = 7.53 (2H, t, pyH), 7.20 (2H, d, pyH),
7.01 (2H, d, pyH), 6.46 (1H, s, xylH), 3.81 (6H, s,
NCH2py + NCH2xyl), 2.56 (6H, s, pyCH3), 2.21 (3H, s,
xylCH3), 2.18 (3H, s, xylCH3), 2.15 (3H, s, xylCH3). ESI-
TOF/MS (acetonitrile solution containing a small amount
of formic acid): m/z (relative intensity) = 376.24 (100)
[M + H]+.
residual elecꢀtr3on
Ra
wRb
˚
density (e A
)
0.056
0.088
0.068
0.100
P
P
a
R = [|Fo| ꢀ |Fc|]/ |Fo| (I P 3.0r(I)).
P
P
b
wR = [ w(|Fo| ꢀ |Fc|)2/ w|Fo|2]1/2
;
w = 1/[r2(Fo) + p2|Fo|2/4] (p =
0.079 for 1a-CH3CN and p = 0.098 for 4).

T. Matsumoto et al. / Journal of Organometallic Chemistry 692 (2007) 111–121
121
(b) K.D. Karlin, J.C. Hayes, Y. Gultneh, R.W. Cruse, J.W. McKown,
J.P. Hutchinson, J. Zubieta, J. Am. Chem. Soc. 106 (1984) 2121.
[4] (a) K.D. Karlin, B.I. Cohen, R.R. Jacobson, J. Zubieta, J. Am.
Chem. Soc. 109 (1987) 6194;
3.9. X-ray crystallography. General procedures
Data collections were carried out on a Rigaku/MSC
Mercury diffractometer with graphite monochromated
(b) M.S. Nasir, B.I. Cohen, K.D. Karlin, J. Am. Chem. Soc. 114
(1992) 2482.
˚
Mo Ka radiation (k = 0.71070 A). All the structures were
[5] E. Pidcock, H.V. Obias, C.X. Zhang, K.D. Karlin, E.I. Solomon, J.
Am. Chem. Soc. 120 (1998) 7841.
[6] (a) R.W. Cruse, S. Kaderli, K.D. Karlin, A.D. Zuberbuhler, J. Am.
¨
solved and refined by the teXsan crystallographic software
package [23] Lorentz and polarization corrections were
applied to the data. Empirical absorption corrections
(multi-scan) were also applied. The structures were solved
by direct methods using SIR92 [24] for 1a-CH3CN and
SHELXS86 [25] for 4, and expanded using DIRDIF-94 [26].
The space groups were confirmed by the systematic
absences, and the successful solution and refinements of
the structures. Crystallographic data are summarized in
Table 2. Crystallographic data have been deposited with
the Cambridge Crystallographic Data Center as supple-
mental publication no. CCDC 600563-600564. Copies of
this information may be obtained free of charge from
The Director, CCDC, 12 Union Road, Cambridge, CB2
1EZ, UK (fax: +44-1223-336-033; e-mail: deposit@ccdc.
cam.ac.uk or www: http://www.ccdc.cam.ac.uk).
Chem. Soc. 110 (1988) 6882;
(b) K.D. Karlin, M.S. Nasir, B.I. Cohen, R.W. Cruse, S. Kaderli,
A.D. Zuberbuhler, J. Am. Chem. Soc. 116 (1994) 1324.
¨
´
[7] S. Mahapatra, S. Kaderli, A. Llobet, Y.-M. Neuhold, T. Palanche,
J.A. Halfen, V.G. Young Jr., T.A. Kaden, L. Que Jr., A.D.
Zuberbu¨hler, W.B. Tolman, Inorg. Chem. 36 (1997) 6343.
[8] L. Santagostini, M. Gullotti, E. Monzani, L. Casella, R. Dillinger,
F. Tuczek, Chem. Eur. J. 6 (2000) 519.
[9] S. Itoh, H. Kumei, M. Taki, S. Nagatomo, T. Kitagawa, S.
Fukuzumi, J. Am. Chem. Soc. 123 (2001) 6708.
[10] L.M. Mirica, M. Vance, D.J. Rudd, B. Hedman, K.O. Hodgson, E.I.
Solomon, T.D.P. Stack, J. Am. Chem. Soc. 124 (2002) 9332.
[11] S. Palavicini, A. Granata, E. Monzani, L. Casella, J. Am. Chem. Soc.
127 (2005) 18031.
[12] S. Yamazaki, S. Itoh, J. Am. Chem. Soc. 125 (2003) 13034.
[13] T. Matsumoto, H. Furutachi, M. Kobino, M. Tomii, S. Nagatomo,
T. Tosha, T. Osako, S. Fujinami, S. Itoh, T. Kitagawa, M. Suzuki,
J. Am. Chem. Soc. 128 (2006) 3874.
Acknowledgement
[14] P.L. Holland, K.R. Rodgers, W.B. Tolman, Angew. Chem., Int. Ed.
38 (1999) 1139.
[15] L.M. Mirica, M. Vance, D.J. Rudd, B. Hedman, K.O. Hodgson, E.I.
Solomon, T.D.P. Stack, Science 308 (2005) 1890.
[16] T. Osako, S. Terada, T. Tosha, S. Nagatomo, H. Furutachi, S.
Fujinami, T. Kitagawa, M. Suzuki, S. Itoh, J. Chem. Soc. Dalton
(2005) 3514.
This work was supported by Grants-in-Aid for Scientific
Research (KAKENHI) on Priority Areas ‘‘Dynamic Com-
plex’’ from the Ministry of Education, Science, and Cul-
ture, Japan.
[17] S. Mahapatra, J.A. Halfen, E.C. Wilkinson, G. Pan, X. Wang, V.G.
Young Jr., C.J. Cramer, L. Que Jr., W.B. Tolman, J. Am. Chem.
Soc. 118 (1996) 11555.
Appendix A. Supplementary data
[18] M.J. Henson, P. Mukherjee, D.E. Root, T.D.P. Stack, E.I. Solomon,
J. Am. Chem. Soc. 121 (1999) 10332.
[19] P.L. Holland, C.J. Cramer, E.C. Wilkinson, S. Mahapatra, K.R.
Rodgers, S. Itoh, M. Taki, S. Fukuzumi, L. Que Jr., W.B. Tolman,
J. Am. Chem. Soc. 122 (2000) 792.
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.jorganchem.
2006.05.068.
[20] E. Pidcock, H.V. Obias, M. Abe, H.-C. Liang, K.D. Karlin, E.I.
Solomon, J. Am. Chem. Soc. 121 (1999) 1299.
References
[21] S. Terada, M. Suzuki, et al. Unpublished data.
[22] H. Nagano, N. Komeda, M. Mukaida, M. Suzuki, K. Tanaka,
Inorg. Chem. 35 (1996) 6809.
[23] teXsan: Crystal Structure Analysis Package, Molecular Structure
Corporation: The Woodlands, TX, 1985 and 1992.
[24] SIR-92: A. Altomare, G. Cascarano, C. Giacovazzo, A. Guagliardi,
M.C. Burla, G. Polidori, M. Camalli, J. Appl. Crystallogr. 27
(1994) 435.
[25] G.M. Sheldrick, in: G.M. Sheldrick, C. Kruger, R. Goddard (Eds.),
Crystallographic Computing 3, Oxford University Press, 1985, pp.
175–189.
[1] (a) K.D. Karlin, A.D. Zuberbuhler, in: J. Reedijk, E. Bouwman
¨
(Eds.), Bioinorganic Catalysis, 2nd ed., Marcel Dekker, New York,
1999, pp. 469–534;
(b) E.I. Solomon, P. Chen, M. Metz, S.-K. Lee, A.E. Palmer, Angew.
Chem., Int. Ed. 40 (2001) 4570;
(c) S. Itoh, S. Fukuzumi, Bull. Chem. Soc. Jpn. 75 (2002) 2081;
(d) E.A. Lewis, W.B. Tolman, Chem. Rev. 104 (2004) 1047;
(e) L.M. Mirica, X. Ottenwaelder, T.D.P. Stack, Chem. Rev. 104
(2004) 1013;
(f) L.Q. Hatcher, K.D. Karlin, J. Biol. Inorg. Chem. 9 (2004) 669.
[2] Y. Matoba, T. Kumagai, A. Yamamoto, H. Yoshitsu, M. Sugiyama,
J. Biol. Chem. 281 (2006) 8981.
[3] (a) K.D. Karlin, Y. Gultneh, J.P. Hutchinson, J. Zubieta, J. Am.
Chem. Soc. 104 (1982) 5240;
[26] P.T. Beurskens, G. Admiraal, G. Beurskens, W.P. Bosman, R. de
Gelder, R. Israel, J.M.M. Smits, The DIRDIF-94 program system,
Technical Report of the Crystallography Laboratory, University of
Nijmegen, The Netherlands, 1994.