Dioxygen Reactivity of Copper(I) Complexes with Tetradentate Tripodal Ligands Having Aliphatic Nitrogen Donors: Synthesis, Structures, and Properties of Peroxo and Superoxo Complexes

Article abstract of DOI:10.1246/bcsj.77.59

Oxygenation of copper(I) with tetradentate tripodal ligands (L) comprised of a tris(aminoethyl)amine (tren) skeleton having sterically bulky substituent(s) on the terminal nitrogens has been investigated, where L = tris(N-benzylaminoethyl)amine (L^H,Bn), tris(N-benzyl-N-methylaminoethyl)amine (L^Me,Bn), or tris(N,N-dimethylaminoethyl)amine (L^Me,Me). All the copper(I) complexes reacted with dioxygen at low temperatures to produce superoxocopper(II) and/ or trans-(μ-1,2-peroxo)-dicopper(II) complexes depending on the steric bulkiness of the terminal nitrogens and the reaction conditions. The reaction of a copper(I) complex [Cu(L^H,Bn)] ⁺ at -90 °C in acetone resulted in the formation of a superoxo complex [Cu(L^H,Bn)(O₂)]⁺ as a less stable species and a peroxo complex [{Cu(L^H,Bn)}2(O₂)] ²⁺ as a stable species. The structures of [Cu(L ^H,Bn)]ClO₄ and [{Cu(L_H,Bn)}₂(O ₂)](BPh₄)₂·8(CH₃) ₂CO were determined by X-ray crystallography. [{Cu(L ^H,Bn)}2(O₂)]²⁺ has a trans-(μ -1,2-peroxo)-dicopper(II) core with a trigonal bipyramidal structure. The O-O bond distance is 1.450(5) A with an intermetallic Cu...Cu separation of 4.476(2) A. The resonance Raman spectrum of [{Cu(L ^H,Bn)}2(O₂)]²⁺ measured at -90 °C in acetone-d₆ showed a broad ν(O-O) band at 837-834 cm^-1 (788 cm^-1 for an ¹⁸O labeled sample) and two ν(Cu-O) bands at 556 and 539 cm^-1, suggesting the presence of two peroxo species in solution. [Cu(L^Me,Bn)]⁺ also produced both superoxo and trans-μ-1,2-peroxo species, [Cu(L^Me,Bn)(O ₂)]⁺ and [{Cu(L^Me,Bn)}₂(O ₂)]²⁺. At a lower concentration of [Cu(L ^Me,Bn)]⁺ (～0.24 mM) and higher dioxygen concentration (P(O₂) = ～1 atm), the superoxo species is predominantly formed, whereas at a higher concentration of [Cu(L^Me,Bn)]⁺ (～1 mM) and lower dioxygen concentration (P(O₂) = ～0.02 atm) the formation of the peroxo species is observed. The resonance Raman spectrum of [Cu(L^Me,Bn)(O₂)]⁺ (～1 mM) in acetone-d₆ at ～-95 °C exhibited a μ(O-O) band at 1120 cm^-1 (1059 cm^-1 for an ¹⁸O labeled sample) and that of [{Cu(L^Me,Bn)}₂(O₂)]²⁺ (～3 mM) in acetone-d₆ at ～-90 °C showed two ν(O-O) bands at 812 and 797 cm^-1 (767 and 753 cm^-1 for an 180 labeled sample), respectively. A similar observation was also made for [{Cu(L^Me,Me)}₂(O₂)]²⁺. Relationships between the energies of the LMCT and d-d transitions and those of the ν(O-O) and ν(Cu-O) stretching vibrations and the steric constraints in the Cu(II)-(O₂^2-)-Cu(II) core are discussed.

Full text of DOI:10.1246/bcsj.77.59

Ó 2004 The Chemical Society of Japan
Bull. Chem. Soc. Jpn., 77, 59–72 (2004)
59
Headline Articles
Dioxygen Reactivity of Copper(I) Complexes with Tetradentate Tripodal
Ligands Having Aliphatic Nitrogen Donors: Synthesis, Structures,
and Properties of Peroxo and Superoxo Complexes
Kazuya Komiyama, Hideki Furutachi, Shigenori Nagatomo,1 Akifumi Hashimoto,
ꢀ
Hideki Hayashi, Shuhei Fujinami, Masatatsu Suzuki, and Teizo Kitagawa1
Department of Chemistry, Faculty of Science, Kanazawa University, Kakuma-machi, Kanazawa 920-1192
1Center for Integrative Bioscience, Okazaki National Research Institutes, Myodaiji, Okazaki 444-8585
Received June 12, 2003; E-mail: suzuki@cacheibm.s.kanazawa-u.ac.jp
Oxygenation of copper(I) with tetradentate tripodal ligands (L) comprised of a tris(aminoethyl)amine (tren) skeleton
having sterically bulky substituent(s) on the terminal nitrogens has been investigated, where L = tris(N-benzylamino-
ethyl)amine (LH,Bn), tris(N-benzyl-N-methylaminoethyl)amine (LMe,Bn), or tris(N,N-dimethylaminoethyl)amine
(LMe,Me). All the copper(I) complexes reacted with dioxygen at low temperatures to produce superoxocopper(II) and/
or trans-(ꢀ-1,2-peroxo)-dicopper(II) complexes depending on the steric bulkiness of the terminal nitrogens and the reac-
ꢂ
tion conditions. The reaction of a copper(I) complex [Cu(LH,Bn)]þ at ꢁ90 C in acetone resulted in the formation of a
superoxo complex [Cu(LH,Bn)(O2)]þ as a less stable species and a peroxo complex [{Cu(LH,Bn)}2(O2)]2þ as a stable spe-
cies. The structures of [Cu(LH,Bn)]ClO4 and [{Cu(LH,Bn)}2(O2)](BPh4)2 8(CH3)2CO were determined by X-ray crystal-
ꢃ
lography. [{Cu(LH,Bn)}2(O2)]2þ has a trans-(ꢀ-1,2-peroxo)-dicopper(II) core with a trigonal bipyramidal structure. The
ꢀ
ꢀ
O–O bond distance is 1.450(5) A with an intermetallic Cu Cu separation of 4.476(2) A. The resonance Raman spectrum
ꢃꢃꢃ
of [{Cu(LH,Bn)}2(O2)]2þ measured at ꢁ90 ꢂC in acetone-d6 showed a broad ꢁ(O–O) band at 837–834 cmꢁ1 (788 cmꢁ1 for
an 18O labeled sample) and two ꢁ(Cu–O) bands at 556 and 539 cmꢁ1, suggesting the presence of two peroxo species in
solution. [Cu(LMe,Bn)]þ also produced both superoxo and trans-ꢀ-1,2-peroxo species, [Cu(LMe,Bn)(O2)]þ and
[{Cu(LMe,Bn)}2(O2)]2þ. At a lower concentration of [Cu(LMe,Bn)]þ (ꢄ0:24 mM) and higher dioxygen concentration
(P(O2) = ꢄ1 atm), the superoxo species is predominantly formed, whereas at a higher concentration of [Cu(LMe,Bn)]þ
(ꢄ1 mM) and lower dioxygen concentration (P(O2) = ꢄ0:02 atm) the formation of the peroxo species is observed.
The resonance Raman spectrum of [Cu(LMe,Bn)(O2)]þ (ꢄ1 mM) in acetone-d6 at ꢄꢁ95 C exhibited a ꢁ(O–O) band
ꢂ
at 1120 cmꢁ1 (1059 cmꢁ1 for an 18O labeled sample) and that of [{Cu(LMe,Bn)}2(O2)]2þ (ꢄ3 mM) in acetone-d6 at
ꢄꢁ90 ꢂC showed two ꢁ(O–O) bands at 812 and 797 cmꢁ1 (767 and 753 cmꢁ1 for an 18O labeled sample), respectively.
A similar observation was also made for [{Cu(LMe,Me)}2(O2)]2þ. Relationships between the energies of the LMCT and
d–d transitions and those of the ꢁ(O–O) and ꢁ(Cu–O) stretching vibrations and the steric constraints in the Cu(II)–
(O22ꢁ)–Cu(II) core are discussed.
Synthetic copper–dioxygen complexes are of great interest
as structural and/or functional models for O2 transport proteins
such as hemocyanin and dioxygen activating copper proteins
such as tyrosinase, methane monooxygenase, etc. Various types
of Cun–O2 complexes have been developed.1–4 Some of the
complexes, such as trans-ꢀ-1,2-peroxo-dicopper(II),5 ꢀ-
ꢂ2:ꢂ2-peroxo-dicopper(II),6,7 ꢀ4-ꢂ2-peroxo-tetracopper(II),8
hydrogenperoxocopper(II),9 (side-on-superoxo)copper(II),10
and (side-on-peroxo like)copper(III) complex11 have been
structurally characterized. In addition, bend type ꢀ-ꢂ2:ꢂ2-
peroxo,12 terminal binding or unsymmetric bridging peroxo,13
and ꢀ-1,1-hydrogenperoxo-dicopper(II) complexes14 have also
been proposed. Formation of bis(ꢀ-oxo)dicopper(III) com-
plexes by the O–O bond scission of the peroxo ligand has also
been reported.15 For the superoxo copper(II) complex, an end-
on coordination mode is postulated for the complexes having
tetradentate tripodal ligands.16 Thus, the stereochemistry of
the copper–dioxygen complexes is highly flexible depending
on the stereochemical and/or electronic effects of the ligands.
It is therefore interesting to investigate how the stereochemical
and electronic features of supporting ligands influence the for-
mation, structure, and reactivity of copper–dioxygen com-
plexes.
Dioxygen reactivity of a wide variety of copper(I) com-
plexes having various tetradentate tripodal ligands has been
investigated.5,17–31 Detailed kinetic and thermodynamic studies
of copper(I) complexes of tmpa32 and analogous tetradentate
tripodal ligands revealed the stepwise formation of a superoxo
Published on the web January 13, 2004; DOI 10.1246/bcsj.77.59

60
Bull. Chem. Soc. Jpn., 77, No. 1 (2004)
HEADLINE ARTICLES
by the addition of aqueous NaOH. An oily layer was extracted with
R2
R1
three 50 cm3 portions of chloroform. The chloroform extracts were
combined and dried over Na2SO4 and filtered. The filtrate was
evaporated under a reduced pressure to give a desired ligand as a
yellow oil, which was used for preparation of complexes without
further purification. Yield: 13.6 g (80%). FAB–MS, m=z ¼ 456
[M ꢁ 2]þ. 1H NMR (CDCl3, 400 MHz) ꢃ 2.14 (9H, s, NCH3),
2.41 (6H, t, N(CH2CH2N)3), 2.62 (6H, t, N(CH2CH2N)3), 3.46
(6H, s, NCH2Ph), 6.82 (15H, m, PhH).
R1
R2
N
LH,Bn : R1 = H, R2 = CH2Ph
LMe,Bn : R1 = CH3, R2 = CH2Ph
LMe,Me : R1 = R2 = CH3
N
N
N
R1
R2
[CuI(L)]+
(L = LH,Bn
(L = LMe,Bn
(L = LMe,Me
[CuII(L)(O2)]+ [{CuII(L)}2(O2)]2+
(L = LH,Bn (L = LH,Bn
(L = LMe,Bn (L = LMe,Bn
(L = LMe,Me (L = LMe,Me
[CuII(Cl)(L)]+
(L = LH,Bn
(L = LMe,Bn
(L = LMe,Me
Tris(N,N-dimethylaminoethyl)amine (LMe,Me):
This was
)
)
)
)
synthesized according to the literature method.34
)
)
)
)
Synthesis of Complexes. All the manipulations for preparation
of the copper(I) complexes were carried out under Ar or N2 atmos-
phere using standard Schlenk techniques. (Caution: All the per-
chlorate salts are potentially explosive and should be handled with
care).
[Cu(LH,Bn)]ClO4: [Cu(CH3CN)4]ClO4 (666 mg, 2.04 mmol)
was dissolved into an acetone solution (10 cm3) of LH,Bn (858
mg, 2.06 mmol) under N2 by warming to afford a colorless solu-
tion, to which was added diethyl ether. The resulting solution
was allowed to stand for a few hours to give white crystals suitable
for X-ray crystallography. Yield: 0.83 g (70%). Anal. Calcd for
C27H36N4CuClO4: C, 55.95; H, 6.26; N, 9.67%. Found: C,
55.87; H, 6.26; N, 9.69%. IR (KBr, cmꢁ1) 1446, 1091 (ClO4),
)
)
)
)
Scheme 1. Tetradentate tripodal ligands and their complexes.
complex [Cu(tmpa)(O2)]þ (Cu:O2 = 1:1) and a ꢀ-peroxo com-
plex [{Cu(tmpa)}2(O2)]2þ (Cu:O2 = 2:1).17 The relative stabil-
ity of those species has been shown to be dependent on the ster-
eochemical nature of the tripodal ligands. Recently, Schindler
et al. have reported similar observations for the copper(I) com-
plexes with LH,Bn and LMe,Me ligands given in Scheme 1. How-
ever, the studies are mainly focused on the kinetics of oxygen-
ation of the copper(I) complexes.24–26 In order to gain further
understanding of the properties of copper–dioxygen com-
plexes, structural and spectroscopic studies of copper(I) and
the resulting dioxygen-copper(II) complexes are needed.
1
ꢂ
748, 698, 623 (ClO4). H NMR (CD3COCD3, 400 MHz, 20 C)
ꢃ 2.83 (6H, t, N(CH2CH2N)3), 2.90 (6H, t, N(CH2CH2N)3), 3.72
(6H, s, NCH2Ph), 7.18–7.31 (15H, m, PhH).
[Cu(LH,Bn)]BPh4: [Cu(CH3CN)4]ClO4 (167 mg, 0.517 mmol)
was dissolved into a methanol (10 cm3)–acetone (2 cm3) mixture
containing LH,Bn (267 mg, 0.640 mmol) under N2, to which was
added a methanol solution (5 cm3) of NaBPh4 (365 mg, 1.07
mmol). The resulting solution was allowed to stand for a few hours
to give white crystals. Yield: 333 mg (80.6%). Calcd for
C51H56N4BCu: C, 76.63; H, 7.06; N, 7.01%. Found: C, 76.60; H,
6.94; N 7.06%. IR (KBr, cmꢁ1) 1579 (BPh4), 1452, 728, 692, 611.
[Cu(LMe,Bn)]ClO4: To LMe,Bn (945 mg, 2.06 mmol) in acetone
(10 cm3) was added [Cu(CH3CN)4]ClO4 (666 mg, 2.04 mmol) un-
der N2 to give a pale yellow solution, to which a small amount of
diethyl ether was added to produce white crystals. Yield: 1.03 g
(81%). Anal. Calcd for C30H42N4CuClO4: C, 57.96; H, 6.81; N,
k1
[CuI(L)]þ þ O2 ꢀ [CuII(L)(O2)]þ
ð1Þ
ð2Þ
k
ꢁ1
k2
[CuII(L)(O2)]þ þ [CuI(L)]þ ꢀ [fCuII(L)g2(O2)]2þ
k
ꢁ2
In this study, in order to explore how the nature of the ali-
phatic nitrogen donors and the alkyl substituents of the support-
ing ligands derived from a tren framework influences the for-
mation, structure, and properties of copper–dioxygen com-
plexes, we have investigated the oxygenation of a series of
Cu(I) complexes with tri- or hexa-alkylsubstituted tren deriva-
tives: LH,Bn, LMe,Bn, and LMe,Me in Scheme 1, the crystal struc-
ture of a trans-ꢀ-1,2-peroxo-dicopper(II) complex [{Cu-
(LH,Bn)}2(O2)]2þ, and the spectroscopic properties (UV–vis
and resonance Raman) of copper–dioxygen complexes. The
crystal structures and some physicochemical properties of the
copper(II) complexes, [Cu(Cl)(L)]þ, are also reported.
9.01%. Found: C, 57.91; H, 6.74; N, 9.26%. IR (KBr, cmꢁ1
1471, 1455, 1089 (ClO4), 752, 704, 623 (ClO4).
)
[Cu(LMe,Bn)]CF3SO3: This was prepared by the same method
for [Cu(LMe,Bn)]ClO4 except for using [Cu(CH3CN)4]CF3SO3.
Yield: 0.865 g (63%). Calcd for C31H42N4CuSO3F3: C, 55.47;
H, 6.31; N, 8.35%. Found: C, 55.16; H, 6.25; N, 8.31%. IR
(KBr, cmꢁ1) 1471, 1456, 1277 (CF3SO3), 1151 (CF3SO3), 1030
Experimental
1
Materials. Acetonitrile and acetone were dried over Molecular
Sieves 4A and distilled before use. All other reagents and solvents
were commercially available and used without further purification.
(CF3SO3)ꢂ, 750, 702, 636 (CF3SO3). H NMR (CD3COCD3, 400
MHz, 20 C) ꢃ 2.52 (9H, s, NCH3), 2.94 (6H, t, N(CH2CH2N)3),
3.09 (6H, t, N(CH2CH2N)3), 3.76 (6H, s, NCH2Ph), 7.21–7.47
(15H, m, PhH).
Syntheses of Ligands.
Tris(N-benzylaminoethyl)amine
(LH,Bn): This was synthesized according to the literature meth-
od.33
[{Cu(LH,Bn)} (O )](BPh ) 8(CH ) CO:
Complex [Cu-
2Á
2
2
4
3 2
(LH,Bn)]BPh4 (100 mg) was dissolved in a small amount of acetone
at ꢁ78 ꢂC under N2 and then O2 gas was introduced into the solu-
tion to produce a deep violet color. Diethyl ether was poured gently
onto the deep violet solution to make a bilayered solution of diethyl
ether and acetone, which was allowed to stand for two days to af-
ford violet crystals suitable for X-ray crystallography.
Tris(N-benzyl-N-methylaminoethyl)amine (LMe,Bn): LH,Bn
(15.4 g, 37 mmol) was dissolved in formic acid (71.3 g, 1.55
mol), to which 37% formaldehyde (17.34 g, 214 mmol) was added,
ꢂ
followed by refluxing for three days at 100 C. Concentrated hy-
drochloric acid was added to make the solution acidic (pH = ca.
1), and the mixture was then evaporated to dryness under reduced
pressure. The residue was dissolved in water (50 cm3) and the
aqueous solution was washed with three 50 cm3 portions of diethyl
ether. The resulting aqueous solution was made basic (pH ca. 12)
[Cu(Cl)(LH,Bn)]ClO4: To an ethanol solution (20 cm3) of
LH,Bn (458 mg, 1.10 mmol) was added dropwise an ethanol solu-
tion (12 cm3) of Cu(ClO4)2 6H2O (370 mg, 1.0 mmol) and LiCl
(80 mg, 1.9 mmol) to give a blue powder, which was dissolved
ꢃ

K. Komiyama et al.
Bull. Chem. Soc. Jpn., 77, No. 1 (2004)
61
by heating to give a blue solution. The resulting solution was al-
lowed to stand overnight to afford blue crystals suitable for X-
ray crystallography. X-ray crystallography of [Cu(Cl)(LH,Bn)]ClO4
is given in supplementary materials.35 Yield: 530 mg (86%). Anal.
Calcd for C27H36N4CuCl2O4: C, 52.73; H, 5.90; N, 9.11%. Found:
C, 52.59; H, 5.98; N, 9.22%. IR (KBr, cmꢁ1) 1454, 1090 (ClO4),
737, 698, 625 (ClO4). UV–vis (ꢄmax/nm ("/Mꢁ1 cmꢁ1 (1 M =
1 mol dmꢁ3)) in acetonitrile) 710 (181), 904 (347).
working electrode, a platinum coil auxiliary electrode, and a satu-
rated calomel electrode (SCE) as a reference electrode. Acetoni-
trile was used as the solvent, and n-tetrabutylammonium perchlo-
rate as the supporting electrolyte. The E1=2 (ꢁE) value of ferro-
cene/ferricinium (FeIII/FeII) with this set up was 395 mV (92
mV). GC–MS measurements were obtained using a Shimadzu
GCMS-QP5050A equipped with a mass spectral detector.
Resonance Raman scattering was measured with a liquid nitro-
gen cooled CCD detector (Model LN/CCD-1340 ꢅ 400PB,
Princeton Instruments) attached to a 1 m-single polychromator
(Model MC-100DG, Ritsu Oyo Kogaku). The 413.1 nm line of a
Krþ laser (Model 2060 Spectra Physics), 514.5 nm line of an
Arþ laser (Model 2017 Spectra Physics), and a dye laser (607
nm) (Model 375B Spectra Physics) with rhodamine-6G dye pump-
ed by an Arþ laser were used as the exciting sources. The laser
powers used for the 413.1, 514.5, and 607 nm excitations were
9.4, 30, and 50 mW, respectively, at the sample points. All meas-
urements were carried out with a spinning cell (1000 rpm) kept at
ꢄꢁ45 to ꢄꢁ95 ꢂC. Raman shifts were calibrated with indene and
the accuracy of the peak positions of the Raman bands was ꢆ1
cmꢁ1. The spectra of the oxygenated samples of [Cu(LMe,Me)]þ
were measured by using the samples prepared by mixing stoichio-
metric amounts of [Cu(CH3CN)4]ClO4 and LMe,Me in acetone or
acetone-d6.
[Cu(Cl)(LH,Bn)]ClO CH Cl 0.25H O: The copper(I) com-
4Á
2Á
2
2
plex [Cu(LH,Bn)]ClO4 (120 mg, 0.21 mmol) was dissolved in di-
chloromethane (8 cm3) under Ar to give a blue-green solution at
ambient temperature. After the reaction mixture was stirred for 1
h, a small amount of grayish green powder was precipitated, which
was removed by filtration under an air atmosphere and the resulting
solution was allowed to stand overnight to give blue crystals suit-
able for X-ray crystallography. The elemental analysis revealed the
loss of CH2Cl2 after exposure of the sample under air, although
X-ray crystallography showed the above formula (vide infra).
Yield: 30 mg (23%). IR and UV–vis spectra were identical to
those of [Cu(Cl)(LH,Bn)]ClO4 described above. Anal. Calcd for
C27H38CuN4Cl2O5: C, 51.23; H, 6.05; N, 8.85%. Found: C,
51.35; H, 5.68; N, 8.99%.
[Cu(Cl)(LMe,Bn)]ClO 0.5H O: To an ethanol solution (20
4Á
2
cm3) of LMe,Bn (505 mg, 1.1 mmol) was added dropwise an ethanol
solution (12 cm3) containing Cu(ClO4)2 6H2O (370 mg, 1.0
X-ray Crystallography. X-ray diffraction studies were made
on a Rigaku/MSC Mercury diffractometer and a Rigaku RAXIS-
IV imaging plate area detector with graphite monochromated
ꢃ
mmol) and LiCl (80 mg, 1.9 mmol) to afford a blue-green solution,
from which green crystals were obtained. Recrystallization from a
methanol/diethyl ether mixture gave green crystals suitable for X-
ray crystallography. Yield: 390 mg (59%). Anal. Calcd for
C30H43N4CuCl2O4:5: C, 54.09; H, 6.51; N, 8.41%. Found: C,
53.75; H, 6.13; N, 8.30%. IR (KBr, cmꢁ1) 1475, 1454, 1092
ꢁ10
ꢀ
ꢀ
Mo-Kꢅ radiation (ꢄ ¼ 0:71070 A (1A = 1 ꢅ 10
m)). The
structures were solved by direct methods (SHELXS-86 or SIR-
92)37 and expanded using a Fourier technique.38 The structures
were refined by a full-matrix least-squares method by using the
teXsan crystallographic software package (Molecular Structure
Corporation).39 Hydrogen atoms were placed at calculated posi-
tions. They were included, but not refined, in the final least-squares
cycles.
(ClO4), 739, 704, 625 (ClO4). UV–vis (ꢄmax/nm ("/Mꢁ1 cmꢁ1
)
in acetonitrile) 744 (289), 934 (705). This complex was also isolat-
ed as 2.5 hydrate from [Cu(LMe,Bn)]ClO4 in dichloromethane sim-
ilar to [Cu(Cl)(LH,Bn)]ClO4 CH2Cl2 0.25H2O. Anal. Calcd for
ꢃ
ꢃ
C30H47N4CuCl2O6:5: C, 51.32; H, 6.75; N, 7.98%. Found: C,
51.49; H, 6.65; N, 8.05%. IR and UV–vis spectra were identical
Crystallographic data are summarized in Table 1 and have been
deposited at the CCDC, 12 Union Road, Cambridge CB2 1EZ, UK
and copies can be obtained on request, free of charge, by quoting
the publication citation and deposition numbers CCDC 218701–
218705. Experimental details for X-ray crystallography, tables of
final atomic coordinates, thermal parameters, and full bond dis-
tances and angles are given in the supplementary materials.35
[Cu(LH,Bn)]ClO4 (I): A single crystal with dimensions of
0:55 ꢅ 0:08 ꢅ 0:08 mm was picked up on a hand-made cold cop-
per plate mounted inside a liquid N2 Dewar vessel and mounted on
a glass fiber at ca. ꢁ80 ꢂC. There are two independent complex cat-
ions and two perchlorate anions in an asymmetric unit. All non-
hydrogen atoms were refined anisotropically.
to [Cu(Cl)(LMe,Bn)]ClO4 0.5H2O.
ꢃ
[Cu(Cl)(tren)]ClO4: A methanol solution (5 cm3) of tren (366
mg, 2.5 mmol) was added dropwise to an aqueous solution (10
cm3) of Cu(ClO4)2 6H2O (459 mg, 1.24 mmol) and CuCl2 2H2O
ꢃ
ꢃ
(211 mg, 1.24 mmol) to afford blue crystals. Yield: 500 mg (58%).
Anal. Calcd for C6H18N4CuCl2O4: C, 20.91; H, 5.26; N, 16.25%.
Found: C, 20.90; H, 5.36; N, 16.12%. IR (KBr, cmꢁ1) 1594 (NH2),
1473, 1089 (ClO4), 627 (ClO4). UV–vis (ꢄmax/nm ("/Mꢁ1 cmꢁ1
in acetonitrile) 740 (187), 932 (440).
)
[Cu(LMe,Me)]ClO4,24 [Cu(Cl)(LMe,Me)]ClO4,24 [Cu(tmpa)-
(CH3CN)]ClO4,5b and [Cu(Cl)(tmpa)]ClO4:36 These were syn-
thesized according to literature methods.
[{Cu(LH,Bn)} (O )](BPh ) 8(CH ) CO (II): A dark-violet
2Á
2
2
4
3 2
Physical Measurements. Electronic spectra were measured on
a Hitachi U-3400 spectrophotometer, and for low temperature
measurements, an Otsuka Denshi optical glass fiber attachment
with the corrected light path lengths of 1.01 and 0.279 cm was
used. Low temperature electronic spectra were also obtained on
a Shimadzu MultiSpec-1500 diode array spectrophotometer using
a 1 cm or 1 mm quartz cell with a Unisoku thermostated cell holder
designed for low-temperature experiments. Infrared spectra were
obtained by the KBr-disk method with a HORIBA FT-300B spec-
trophotometer. 1H NMR spectra were measured with a JEOL JNM-
LA400 spectrometer using tetramethylsilane (TMS) as the internal
standard. Cyclic voltammograms were obtained with a BAS CV-
27 using a three-electrode configuration, including a glassy carbon
crystal with dimensions of 0:50 ꢅ 0:20 ꢅ 0:20 mm was mounted
on a glass fiber at ca. ꢁ80 ꢂC as described above. The asymmetric
unit consists of a half of a complex cation, a tetraphenylborate
anion, and four acetone molecules. All non-hydrogen atoms were
refined with anisotropic displacement parameters.
[Cu(Cl)(LH,Bn)]ClO CH Cl 0.25H O (III): A block crys-
4Á
2Á
2
2
tal (0:35 ꢅ 0:25 ꢅ 0:25 mm) was mounted on a glass fiber at ca.
ꢂ
ꢁ80 C as described above. The asymmetric unit consists of two
independent complex cations and perchlorate anions. The perchlo-
rate anions and dichloromethane molecules were solved by disor-
dered models. All non-hydrogen atoms, except for those of disor-
dered perchlorate anions and dichloromethane molecules, were re-
fined with anisotropic displacement parameters.

62
Bull. Chem. Soc. Jpn., 77, No. 1 (2004)
HEADLINE ARTICLES
Table 1. Crystallographic Data for [Cu(LH,Bn)](ClO4) (I), [{Cu(LH,Bn)}2(O2)](BPh4)2 8(CH3)2CO (II), [Cu(Cl)(LH,Bn)]ClO4
ꢃ
ꢃ
CH2Cl2 0.25H2O (III), and [Cu(Cl)(LMe,Bn)](ClO4) 1.5CH3OH 0.5H2O (V)
ꢃ
ꢃ
ꢃ
I
II
III
V
Formula
C27H36N4O4ClCu
ꢁ150
C126H160N8O10B2Cu2
ꢁ116
C28H38:5N4O4:25Cl4Cu
ꢁ116
C31:5H49N4O6Cl2Cu
ꢁ150
Temp/ꢂC
MW
Cryst system
579.60
monoclinic
Pn
2095.41
monoclinic
P21=c
704.49
monoclinic
C2=c
714.21
triclinic
ꢂ
P1
Space group
ꢀ
a/A
ꢀ
11.602(1)
16.431(1)
14.661(2)
90
103.438(6)
90
2718.5(5)
4
55.0
1216.0
1.416
9.41
15.302(4)
18.924(4)
21.335(8)
90
108.61(2)
90
5855(2)
2
51.5
2240.0
1.188
4.24
24879
7004 (I ꢇ 3ꢉðIÞ)
668
1.24
1.17; ꢁ0:71
0.058
0.087
26.652(5)
18.226(2)
27.444(4)
90
102.20(1)
90
13029(3)
16
51.5
5848.0
1.436
10.38
15.808(1)
16.229(2)
16.225(1)
83.451(7)
61.912(4)
79.578(7)
3609.8(5)
4
55.0
1508.0
1.314
7.98
31465
12422 (I ꢇ 3ꢉðIÞ)
778
1.52
1.60; ꢁ0:77
0.049
0.084
b/A
ꢀ
c/A
ꢅ/deg
ꢆ/deg
ꢇ/deg
ꢀ 3
V/A
Z
2ꢈmax
Fð000Þ
Dcalcd/g cm3
Abs coeff/cmꢁ1
No. of reflcns collcd
No. of indpt reflcns
No. of refined params
GOF
26297
21187
8532 (I ꢇ 3ꢉðIÞ)
667
1.52
1.02; ꢁ0:81
0.059
0.082
8698 (I ꢇ 3ꢉðIÞ)
736
1.66
1.10; ꢁ1:39
0.072
0.107
ꢀ ꢁ3
Largest peak; hole/e A
RaÞ
Rw
bÞ
2
2
1=2
2
a) R ¼ ꢃ½jFoj ꢁ jFcjꢈ=ꢃjFoj. b) Rw ¼ ½ꢃwðjFoj ꢁ jFcjÞ =ꢃwjFoj ꢈ ; w ¼ 1=½ꢉ2ðFoÞ þ p2jFoj =4ꢈ (p ¼ 0:073 for I;
p ¼ 0:092 for II; p ¼ 0:096 for III; p ¼ 0:089 for V).
ꢂ
Table 2. Selected Bond Distances (A) and Angles ( ) of [Cu(LH,Bn)](ClO4) (I), [{Cu(LH,Bn)}2(O2)]-
ꢀ
(BPh4)2 8(CH3)2CO (II), [Cu(Cl)(LH,Bn)]ClO4 CH2Cl2 0.25H2O (III), and [Cu(Cl)(LMe,Bn)](ClO4)
1.5CH3OH 0.5H2O (V)
ꢃ
ꢃ
ꢃ
ꢃ
ꢃ
I
II
III
V
ꢀ
Bond Distances (A)
1.868(3)
2.104(4)
Cu1–Cl1 (or O1)
Cu1–N1
Cu1–N2
Cu1–N3
Cu1–N4
2.261(2)
2.026(4)
2.109(5)
2.127(5)
2.160(4)
2.2193(8)
2.029(3)
2.183(3)
2.165(2)
2.193(3)
2.235(4)
2.064(4)
2.084(4)
2.050(5)
2.142(3)
2.083(3)
2.149(3)
1.450(5)
4.476(2)
ꢀ
Cu1–Cu1
O1–O1
ꢀ
Cu2–Cl2
Cu2–N5
Cu2–N6
Cu2–N7
Cu2–N8
2.247(2)
2.046(4)
2.140(5)
2.134(5)
2.110(5)
2.2392(9)
2.022(3)
2.185(3)
2.188(3)
2.184(3)
2.232(4)
2.077(5)
2.081(5)
2.087(4)
Bond Angles (ꢂ)
111.5(3)
125.4(1)
ꢀ
Cu1–O1–O1
N2–Cu1–N3
N2–Cu1–N4
N3–Cu1–N4
N6–Cu2–N7
N6–Cu2–N8
N7–Cu2–N8
119.0(2)
126.8(2)
111.4(2)
115.5(2)
123.8(2)
118.2(2)
122.9(2)
118.4(2)
115.9(2)
115.1(2)
115.8(2)
125.8(2)
119.14(10)
119.03(9)
119.40(10)
120.5(1)
116.1(1)
120.8(1)
113.8(1)
116.4(1)
[Cu(Cl)(LH,Bn)]ClO4 (IV): A block crystal (0:80 ꢅ 0:10 ꢅ
ꢁ120 ꢂC. An asymmetric unit consists of a discrete complex cation
and one perchlorate anion. All non-hydrogen atoms were refined
anisotropically. The crystal structure of the complex cation is given
0:10 mm) was picked up on the cold coꢂpper plate described above,
and mounted on a glass fiber at ꢁ80 C. Data were collected at

K. Komiyama et al.
Bull. Chem. Soc. Jpn., 77, No. 1 (2004)
63
0.25H2O (III), and [Cu(Cl)(LMe,Bn)]ClO 1.5CH OH
C16
4Á
Á
C42
3
C17
C43
0.5H2O (V). X-ray crystallography of [Cu(LH,Bn)]ClO4 (I) re-
vealed that an asymmetric unit contains two independent mole-
cules (A and B) which are conformational isomers as shown in
Fig. 1. The selected bond distances and angles are given in
Table 2. The structures of complex cations (A) and (B) are
trigonal pyramidal, similar to that of [Cu(LMe,Me)]þ.24a The
structural difference in these two isomers is in the orientation
of three benzyl groups. The three benzyl groups of molecule
(A) have an equatorial orientation with respect to a gauche con-
formation of the 1,2-diaminoethane chelate ring, whereas one
of the benzyl groups in molecule (B) adopts an axial orientation
with respect to a gauche conformation of the 1,2-diaminoethane
chelate ring and the others an equatorial orientation. The three
benzyl groups for both structures form picket fences around the
Cu atoms. There is no significant difference in the bond dis-
tances and angles around the copper atoms in the two isomers.
The average Cu–Neq bond distances of the present type com-
C15
C7
C52
C25
C44
C18
C41
C35
C26
C51
C14
C45
C40
C34
C13
C53
C24
C8
C9
C27
C50
C6
C39
C12
C54
C23
C36
C33
C22
C5
C4
C49
C48
N7
N3
C31
C32
C21
N4
Cu2
C3
Cu1
N8
N2
N6
C38
C11
C30
C37
C10
C2
C29
N1
C20
N5
C47
C46
C19
C1
C28
A
B
Fig. 1. ORTEP views (50% probability) of two isomers (A
and B) of [Cu(LH,Bn)]ClO4 (I). Hydrogen atoms are omit-
ted for clarity.
in Fig. S1 in the supplementary materials.35
ꢀ
[Cu(Cl)(LMe,Bn)]ClO 1.5CH OH 0.5H O (V):
A single
plexes having aliphatic nitrogen donors (2.074 A for I and
4Á
Á
3
2
2.122 A for [Cu(LMe,Me)]þ24a) are significantly longer than
ꢀ
crystal with dimensions of 0:60 ꢅ 0:50 ꢅ 0:40 mm was mounted
ꢂ
those of the complexes having aromatic nitrogen donors ([Cu-
19
on a glass fiber at ca. ꢁ80 C as described above. There are two
41
(tmqa)]þ (2.01 and 2.007 A), [Cu(Ph3tren)]þ (2.01 A),
ꢀ
15c
ꢀ
independent complex cations and perchlorate anions in an asym-
metric unit, respectively. The structures of the two cations are al-
most mirror images each other. A methanol molecule is present,
but disordered. All non-hydrogen atoms of the complex cations
and perchlorate anions were refined anisotropically.
[Cu(Me2-tpa)]þ (2.008 A), [Cu(Me3-tpa)]þ (2.01 A). The
long bond distances of I and [Cu(LMe,Me)]þ are partly due to
the steric requirement of the bulky substituents and partly due
to the hardness of aliphatic nitrogen donors, which bind more
weakly to Cu(I) ion than the aromatic nitrogen donors.
42
ꢀ
ꢀ
Results
Figure 2 shows the crystal structure of a peroxo complex cat-
ion of [{Cu(LH,Bn)}2(O2)](BPh4)2 8(CH3)2CO (II), which con-
The tripodal ligands, LH,Bn and LMe,Bn, formed the copper(I)
complexes [Cu(LH,Bn)]þ and [Cu(LMe,Bn)]þ, which can be read-
ily isolated as white crystals by the reaction of the correspond-
ing ligand with [Cu(CH3CN)4]þ in acetone under N2. Schindler
et al. reported that the isolation of [Cu(LH,Bn)]þ was unsuccess-
ful due to serious disproportionation reaction to copper metal
and copper(II) complex.26 However, it was found that a freshly
prepared ligand did not cause disproportionation. Use of the li-
gand under CO2 gas resulted in serious disproportionation, sug-
gesting that a carbonate salt causes disproportionation. The
copper(I) complexes [Cu(LH,Bn)]þ and [Cu(LMe,Bn)]þ are reac-
tive with dichloromethane. Dissolving the complexes into di-
chloromethane under Ar at ambient temperature gave blue-
green solutions within a few minutes. As mentioned in the
experimental section, [Cu(LH,Bn)]þ and [Cu(LMe,Bn)]þ afforded
the corresponding chlorocopper(II) complexes [Cu(Cl)-
(LH,Bn)]ClO4 CH2Cl2 0.25H2O and [Cu(Cl)(LMe,Bn)]ClO4
ꢃ
sists of a centrosymmetric Cu(ꢀ-1,2-O2)Cu core with LH,Bn
ligands. The structure is quite similar to that of [{Cu-
(tmpa)}2(O2)]2þ
.
The selected bond distances and angles
5a
are given in Table 2. Unlike I, the six benzyl groups spread
out to accommodate the oxygen atoms of the peroxo ligand
and to avoid steric interaction between benzyl groups (Fig.
2a). Each copper ion adopts a five coordinate structure. The
structural index parameter (ꢊ ¼ ðꢆ ꢁ ꢅÞ=60, ꢆ > ꢅ) for five-
coordinate complexes introduced by Addison et al.43 is 0.81
comparable to that of [{Cu(tmpa)}2(O2)]2þ (ꢊ ¼ 0:86), where
those for an ideal square pyramid and trigonal bipyramid are
ꢊ ¼ 0 and ꢊ ¼ 1, respectively. Thus each copper ion has a
ꢀ
slightly distorted trigonal bipyramidal structure. The O1–O1
ꢀ
A
distance is 1.450(5)
comparable to that of
[{Cu(tmpa)}2(O2)]2þ (1.432(6) A) and those of the peroxo
ꢀ
complexes of various transition metal ions.44 The Cu1–O1
ꢃ
ꢃ
ꢃ
ꢀ
and Cu1–N1 bond distances are 1.868(3) and 2.104(4) A, re-
2.5H2O from the dichloromethane solution. In addition, GC–
MS analysis of a blue-green dichloromethane solution pro-
duced by dissolving [Cu(LH,Bn)]þ under N2 atmosphere re-
vealed the formation of 1,2-dichloroethane, which is likely to
be derived from the coupling of dichloromethane, as found
for the reaction of [Cu(tmpa)(CH3CN)]þ with organic halides
reported by Karlin et al.40 They also observed that dichloro-
methane readily reacts with [Cu(tmpa)(CH3CN)]þ to afford
[Cu(Cl)(tmpa)]þ in a high yield. The copper(I) complexes
[Cu(LH,Bn)]þ and [Cu(LMe,Bn)]þ are very reactive with O2 in
solution state, but relatively stable against O2 in the solid state
and can be handled in air.
spectively, which are also comparable to those of
[{Cu(tmpa)}2(O2)]2þ (1.852(5) and 2.104(6) A, respectively).
ꢀ
ꢀ
However, the average Cu–Neq bond distance (2.125 A) is sub-
stantially longer than that of [{Cu(tmpa)}2(O2)]2þ (2.069 A),
ꢀ
which seems to be due to unfavorable steric interaction between
the methylene hydrogens of the benzyl groups and the bridging
ꢀ
oxygens of the peroxo ligand. The Cu1–O1–O1 angle of
111.5(3)ꢂ is slightly larger than that of [{Cu(tmpa)}2(O2)]2þ
ꢀ
(107.7(2)ꢂ), resulting in slight expansion of the Cu1 Cu1 sep-
ꢀ
aration (4.476(2) A) relative to that of [{Cu(tmpa)}2(O2)]
ꢃꢃꢃ
2þ
ꢀ
(4.359(1) A). The space-filling model of II shown in Fig. 2b re-
veals that the Cu(II)(ꢀ-1,2-O2)Cu(II) core is almost covered by
a hydrophobic cavity formed by the six benzyl groups of two
Structures of [Cu(LH,Bn)]ClO4 (I), [{Cu(LH,Bn)}2(O2)]-
(BPh ) 8(CH ) CO (II), [Cu(Cl)(LH,Bn)]ClO CH Cl
4 2Á
4Á
2Á
3 2
2

64
Bull. Chem. Soc. Jpn., 77, No. 1 (2004)
HEADLINE ARTICLES
N1*
N3*
Cu1*
N2*
N4*
O1*
Cu
O
N
C24
C23
C25
C26
O1
C21
N4
C17
C9
C8
C22
C7
C12
C27
Cu
Cu1
C18
C13
N2
C3
C2
C4
C16
C5
C6
N3
N
C20
C14
C15
N1
C11
C19
C1
C10
(a)
(b)
Fig. 2. ORTEP view (50% probability) of the complex cation of [{Cu(LH,Bn)}2(O2)](BPh4)2 8(CH3)2CO (II) (a) and a space filling
view (b). Hydrogen atoms are omitted for the ORTEP view.
ꢃ
C42
C16
C43
C17
C41
C44
C15
C18
C40
C45
C39
Cl1
C14
C13
C12
Cl2
N3
C32
C36
N7
C33
C35
C30
C31
Cu2
C3
N6
C11
N4
C9
C38
Cu1
N2
C21
C48
C8
C4
C34
N8
C27
C22
C49
C26
C10
C37
C7
C50
C51
N5
C29
N1
C2
C28
C1
C5
C54
C23
C20
C47
C6
C46
C19
C53
C24
C25
C52
A'
B'
Fig. 3. ORTEP views (50% probability) of complex cations (A0 and B0) of [Cu(Cl)(LH,Bn)]ClO4 CH2Cl2 0.25H2O (III). Hydrogen
atoms are omitted for clarity.
ꢃ
ꢃ
LH,Bn ligands, which seems to be responsible for protecting the
unstable Cu(II)(ꢀ-1,2-O2)Cu(II) core against some deleterious
decay reactions.
C28
C27
C26
C29
C30
X-ray crystallography of the chlorocopper(II) complex
[Cu(Cl)(LH,Bn)]ClO4 CH2Cl2 0.25H2O (III) showed that an
C25
ꢃ
ꢃ
C24
Cl1
asymmetric unit contains two independent molecules (A0 and
B0) which are also the conformational isomers with respect to
the orientations of the benzyl groups as found for I in Fig. 3.
The selected bond distances and angles are given in Table 2.
In the course of this study, Schindler reported a crystal structure
of a chloride salt ([Cu(Cl)(LH,Bn)]Cl) which has only one
isomer with a symmetric orientation of benzyl groups.26 The
structures of two isomers in III are slightly distorted trigonal
bipyramidal with the ClN4 donor set. All the benzyl groups
spread out to accommodate a chloride ion as in II. There is
no significant difference in the metric parameters around the
Cu atoms of two isomers. The crystal structure of [Cu(Cl)-
(LMe,Bn)]ClO4 1.5CH3OH 0.5H2O (V) also showed that there
N4
C23
C3
C6
C22
C21
Cu1
N2
C14
C7
C5
C15
C20
C17
C4
C2
N3
C19
C18
C8
N1
C13
C12
C1
C16
C10
C11
C9
Fig. 4. ORTEP view (50% probability) of one of complex
cations of [Cu(Cl)(LMe,Bn)]ClO4 1.5CH3OH 0.5H2O (V).
Hydrogen atoms are omitted for clarity.
ꢃ
ꢃ
Cu–Neq bond distances are 2.06 A for [Cu(Cl)(tmpa)] ,36 2.10
þ
ꢀ
ꢃ
ꢃ
þ
45
ꢀ
ꢀ
ꢀ
are two independent complexes in an asymmetric unit. Howev-
er, they are almost mirror images. One of the complex cations
of V is given in Fig. 4, and the other one is given in the supple-
mentary materials (Fig. S2).
A for [Cu(NCS)(tren)] , 2.130 A for III, 2.183 A for V, and
2.186(2) A for [Cu(Cl)(LMe,Me)]þ.24a Unlike the equatorial
ꢀ
ꢀ
bond distances, the Cu–Nax bond distances (2.036 A for III
and 2.026 A for V) are comparable to those of [Cu(Cl)-
þ
(LMe,Me)]þ (2.040(6) A) and [Cu(Cl)(tmpa)] (2.050(6) A),
ꢀ ꢀ
ꢀ
which are significantly shorter than that of II (2.104(4) A).
Electronic Spectra and Electrochemistry. The electronic
ꢀ
For the copper(II) complexes, there is a tendency that the
average Cu–Neq bond distance becomes longer as the steric
bulkiness of the terminal nitrogen donors increases: the average

K. Komiyama et al.
Bull. Chem. Soc. Jpn., 77, No. 1 (2004)
65
Table 3. Electrochemical and Electronic Spectral Data for
the Cu(I) and Cu(II) Complexes in Acetonitrile
which is significantly negative compared to that of [Cu(Cl)-
(LMe,Bn)]þ (ꢁ210 mV vs SCE), even though crystal structural
and electronic spectral features of [Cu(Cl)(LMe,Me)]þ are quite
similar to those of [Cu(Cl)(LMe,Bn)]þ as mentioned already. The
origin of such a large negative shift for [Cu(Cl)(LMe,Me)]þ is not
clear at present. The E1=2 values of [Cu(tmpa)(CH3CN)]þ and
[Cu(Cl)(tmpa)]þ are ꢁ10 and ꢁ370 mV vs SCE, respectively,
which are between those of [Cu(LH,Bn)]þ and [Cu(LMe,Bn)]þ,
and also between those of [Cu(Cl)(LH,Bn)]þ and [Cu(Cl)-
(LMe,Bn)]þ. Thus, the introduction of benzyl and/or methyl
groups on the terminal nitrogens of the tren skeleton has a sig-
nificant influence on the redox properties of the copper(I) and
chlorocopper(II) complexes.
E1=2 (ꢁEp)
mV vs SCE
Spectral data
max/nm ("/Mꢁ1 cmꢁ1
)
ꢄ
½Cu(LH,Bn)]þ
½Cu(LMe,Bn)]þ
ꢁ140 (120)
þ110 (145)
½Cu(tmpa)(CH3CN)]þ ꢁ10aÞ (85)aÞ
½Cu(Cl)(LH,Bn)]þ
½Cu(Cl)(LMe,Bn)]þ
½Cu(Cl)(LMe,Me)]þ
½Cu(Cl)(tmpa)]þ
½Cu(Cl)(tren)]þ
ꢁ380 (200) 710 (181), 904 (347)
ꢁ210 (155) 744 (289), 934 (705)
ꢁ430 (145) 740 (187), 932 (440)bÞ
ꢁ370cÞ (175) 725sh (90), 955 (210)dÞ
640sh (73), 786 (120)
a) The E1=2 value measured in CH3CN is not reported in
Ref. 36b, but ꢁ610 mV vs Ag/AgNO3 in DMF is reported.
b) The absorption spectrum in CH3CN is not reported in
Ref. 24a, but ꢄmax ¼ 876 nm (" ¼ 248 Mꢁ1 cmꢁ1) in water
is reported. c) The E1=2 value measured in CH3CN is not re-
ported in Ref. 36b, but ꢁ790 mV vs Ag/AgNO3 in DMF is re-
ported. d) Ref. 36b. sh: shoulder.
Dioxygen Reactivity of Copper(I) Complexes. Electronic
Spectra: The copper(I) complexes [Cu(LH,Bn)]þ and [Cu-
(LMe,Bn)]þ in acetone are very reactive toward dioxygen as re-
ported for [Cu(LH,Bn)]þ by Schindler et al.25,26 At room temper-
ature, the complexes underwent instantaneous irreversible oxi-
dation to give the blue copper(II) species, whereas below ꢁ80
ꢂC they form dioxygen complexes, which were characterized
by X-ray crystallography, electronic, and/or resonance Raman
spectroscopy.
spectral data of the chlorocopper(II) complexes [Cu(Cl)-
(LH,Bn)]þ, [Cu(Cl)(LMe,Bn)]þ, [Cu(Cl)(LMe,Me)]þ, [Cu(Cl)-
(tmpa)]þ, and [Cu(Cl)(tren)]þ in acetonitrile are given in
Table 3. The electronic spectra of all the chlorocopper(II) com-
plexes exhibit two d–d absorption bands in the 600–1000 nm
region, which are characteristic of those of the trigonal bipyra-
midal copper(II) complexes. A comparison of the absorption
maxima of the chloro complexes reveals that the order of the
ligand field strength of a series of the tripodal ligands is tren
(786 nm) > LH,Bn (904 nm) > LMe,Me (932 nm) ꢄ LMe,Bn
(934 nm) > tmpa (955 nm).36b This order indicates that intro-
duction of sterically bulky substituent(s) into the terminal nitro-
gens of the tren skeleton weakens the ligand field strength and
the ligand field strength of pyridyl nitrogens is weaker than that
of aliphatic nitrogens. It is noted that there is a tendency that the
molar extinction coefficients ("/Mꢁ1 cmꢁ1) of [Cu(Cl)-
(LH,Bn)]þ, [Cu(Cl)(LMe,Bn)]þ, and [Cu(Cl)(LMe,Me)]þ are sig-
nificantly larger than those of the other trigonal bipyramidal
complexes as seen in Table 3. Especially that of [Cu(Cl)-
(LMe,Bn)]þ is ꢄ700 Mꢁ1 cmꢁ1 at 934 nm, which is exceptional-
ly large for d–d transitions. The origin of such a large molar
extinction coefficient, however, is not known at present.
Buꢂbbling of O2 into an acetone solution of [Cu(LH,Bn)]þ at
ꢁ80 C caused a color change from colorless to deep violet
to produce a peroxo species [{Cu(LH,Bn)}2(O2)]2þ. The elec-
tronic spectrum of an acetone solution of [{Cu(LH,Bn)}2(O2)]2þ
(0.123 mM, all the concentrations for the spectral measure-
ments are given as the copper concentration) at ꢁ80 ꢂC exhibits
an intense absorption band at 518 nm (" ¼ 14900 Mꢁ1 cmꢁ1),
two moderately intense bands at ꢄ435 nm (shoulder, " ¼
ꢄ2800 Mꢁ1 cmꢁ1) and ꢄ590 nm (" ¼ ꢄ7700 Mꢁ1 cmꢁ1),
and a weaker band at 976 nm (" ¼ 300 Mꢁ1 cmꢁ1, 4.04 mM)
as shown in Fig. 5 and Table 4, where the " values are given
based on a dimer [{Cu(LH,Bn)}2(O2)]2þ. This spectral feature
is quite similar to that of [{Cu(tmpa)}2(O2)]2þ as seen in
5a
Table 4, indicating that the violet species is a trans-(ꢀ-1,2-per-
oxo)-dicopper(II) complex, [{Cu(LH,Bn)}2(O2)]2þ, as con-
firmed by X-ray crystallography. Schindler et al. reported
(a)
Cyclic voltammograms (CV) of the copper(I) complexes and
the chlorocopper(II) complexes measured in acetonitrile
showed quasi-reversible waves corresponding to Cu(II)/Cu(I)
redox couple. The E1=2 values of [Cu(LH,Bn)]þ and [Cu(Cl)-
(LH,Bn)]þ are ꢁ140 and ꢁ380 mV vs SCE, respectively, and
those of [Cu(LMe,Bn)]þ and [Cu(Cl)(LMe,Bn)]þ are þ110 and
ꢁ210 mV vs SCE, respectively (Table 3). Thus, N-benzyl-N-
methylamino groups cause a significant positive shift of E1=2
values, indicating that the sterically bulky tertiary amine nitro-
gen stabilizes the copper(I) oxidation state and/or destabilizes
the copper(II) oxidation state compared to the secondary amine
nitrogen. These positive shifts of E1=2 values for [Cu(LMe,Bn)]þ
and [Cu(Cl)(LMe,Bn)]þ may be partly attributable to the weaker
electron donor ability of LMe,Bn relative to that of LH,Bn. This is
in line with the weaker ligand field strength of the LMe,Bn ligand
relative to that of the LH,Bn ligand. It should be noted, however,
that the E1=2 value of [Cu(Cl)(LMe,Me)]þ is ꢁ430 mV vs SCE
(c)
(b)
(b)
Fig. 5. Electronic spectra of [Cu(LH,Bn)]þ in acetone under
P(O2) = ꢄ1 atm; (a) at ꢁ80 ꢂC (0.123 mM), (b) at ꢁ90 ꢂC
ꢂ
(0.117 mM) and (c) at ꢁ80 C (4.04 mM). " values are
given based on a dimer.

66
Bull. Chem. Soc. Jpn., 77, No. 1 (2004)
HEADLINE ARTICLES
Table 4. Spectroscopic Data for Peroxodicopper(II) and Superoxocopper(II) Complexes
16{18
16{18
Complexes
UV–vis bands
max/nm ("/Mꢁ1 cmꢁ1
ꢁ(O–O) (
ꢁ)
ꢁ(Cu–O) (
ꢁ)
Ref.
ꢁ1
ꢁ1
~
~
ꢄ
)
ꢁ/cm
ꢁ/cm
Peroxo complexes
[{Cu(LH,Bn)}2(O2)]2þ
ꢄ435sh (ꢄ2800), 518 (14900), ꢄ590sh 839, 829 (51–41)
(ꢄ7700), 976 (300) (d–d band)
556 (28), 539 (27)
531 (26)
This work, 26aÞ
This work
[{Cu(LMe,Bn)}2(O2)]2þ
[{Cu(LMe,Me)}2(O2)]2þ
[{Cu(tmpa)}2(O2)]2þ
567 (>10000), ꢄ618sh (>8700),
ꢄ1088 (>220)
812 (45), 797 (44)
ꢄ460sh (ꢄ1500), 553 (ꢄ11600),
ꢄ615sh (ꢄ8600), 1044 (ꢄ240)
822 (47), 808 (44sh) 535 (25)
This work, 24bÞ
5, 46, 47
ꢄ440sh (2000), 525 (11500), ꢄ590sh 827 (44)
561 (26)
(7600), 1035 (160)
565 (17900)
550 (10200), 600 (9700)
525 (4010), 625 (2290)
415sh (2333), 505 (10500),
620 (5400)
[{Cu(2L)}2(O2)]2þ
807–746 (753–704)cÞ ꢄ550–522 (533–507)cÞ
29
21a
22
[{Cu(Lpy)}2(O2)]2þ
[Cu2(MEPY22PZ)(O2)]2þ
[Cu2(bpman)(O2)]2þ
822 (51)
844 (46)
831 (44)
530 (24)
558 (26)
561
30
[{Cu(L1)}2(O2)]2þ
[Cu2(L2)(O2)]2þ
520, 630
495 (7900), 623 (5400)
840
827
23
23
Superoxo complexes
[Cu(LH,Bn)(O2)]þ
[Cu(LMe,Bn)(O2)]þ
412
416 (ꢄ5400), 591 (ꢄ1640), 737
(ꢄ2340)
This work, 26dÞ
This work
1120 (61)
1122
474 (20)
[Cu(LMe,Me)(O2)]þ
[Cu(tmpa)(O2)]þ
[Cu(2L)(O2)]2þ
414
410 (4000), 747 (1000)
416 (4600), 654 (1800)
This work, 24eÞ
17, 19
29
10b
1120 (62)
1043 (59)
450 (8), 422 (5)
[Cu(HB(3-Ad-5-iPrpz)3)(O2)] 383, 452, 699, 980
sh: shoulder. a) ꢄmax (LMCT) = 506 nm is reported in Ref. 26. b) ꢄmax (LMCT) = 552 nm and ꢁ(O–O) = 825–801 cmꢁ1 are reported in
Ref. 24b. c) Corresponding relation between the bands may not be clear. d) ꢄmax (LMCT) = 406 nm is reported in Ref. 26. e) ꢄmax
(LMCT) = 412 nm (" ¼ 4800) and ꢁ(O–O) = 1122 cmꢁ1 are reported in Ref. 24b.
similar absorption spectra of peroxo and superoxo species
(ꢄmax ¼ 506 for peroxo species and ꢄmax ¼ 406 nm for a su-
peroxo species in propionitrile (vide infra)) using time-resolved
electronic spectroscopy.25,26 The intense absorption bands at
518 and ꢄ590 nm can be assigned to the spin-allowed
is only partially formed. A similar spectrum has already been
found upon oxygenation of [Cu(tmpa)(CH3CN)]þ, which ex-
hibits a transient absorption band at ꢄ410 nm observable only
under stopped-flow conditions.17 Karlin et al. assigned this
transient species to a superoxo species [Cu(tmpa)(O2)]þ
(ꢄmax ¼ 410 nm (" ¼ 4000 Mꢁ1 cmꢁ1) and 747 nm (" ¼
1000 Mꢁ1 cmꢁ1)) and established the stepwise equilibriums
given in Eqs. 1 and 2 described above. Bubbling of Ar gas
caused the conversion of [Cu(LH,Bn)(O2)]þ to [{Cu(LH,Bn)}2-
(O2)]2þ, which was completed within a few minutes, indicating
that a low dioxygen concentration stabilizes the peroxo species,
as can be expected from the stepwise equilibriums Eqs. 1 and 2.
The oxygenation is also affected by the concentration of
[Cu(LH,Bn)]þ. The electronic spectrum of a more concentrated
acetone solution (1.05 mM) at ꢁ90 ꢂC under P(O2) = ꢄ1 atm
revealed that the formation of the superoxo species is highly
suppressed (not shown in Fig. 5), indicating that higher concen-
tration of [Cu(LH,Bn)]þ also leads to the formation of the peroxo
species as expected from the equilibriums.
ꢀ
ꢀ
O2(ꢋꢉ ) ! CuII(dz2 ) and O2(ꢋꢁ ) ! CuII(dz2 ) charge transfer
(LMCT) transitions, respectively, and the band at ꢄ435 nm to
ꢀ
II
2
the spin-forbidden O2(ꢋꢁ ) ! Cu (dz ) LMCT transition ac-
cording to the assignment made for [{Cu(tmpa)}2(O2)]2þ by
Solomon et al.46 The weaker absorption band at 976 nm of
[{Cu(LH,Bn)}2(O2)]2þ can be attributed to the d–d transitions,
which is red-shifted compared to that of the chloro complex
[Cu(Cl)(LH,Bn)]þ (904 nm). Similar observations were made
for the peroxo species [{Cu(LMe,Bn)}2(O2)]2þ, [{Cu(LMe,Me)}2-
(O2)]2þ, and [{Cu(tmpa)}2(O2)]2þ relative to the correspond-
ing chloro complexes. Such red shifts of the d–d transitions
of the peroxo complexes seem to be partly due to the stronger
ꢋ donor ability of the coordinated peroxo ligands. [{Cu-
(LH,Bn)}2(O2)]2þ is stable at ꢁ80 ꢂC for several hours. Deoxy-
genation was not effected by bubbling of Ar at ꢁ80 ꢂC.
The reaction of [Cu(LMe,Bn)]þ with O2 in acetone at ꢁ90 ꢂC
resulted in an immediate color change from pale yellow to
green. The electronic spectrum of the green species under these
conditions (0.238 mM and P(O2) = ꢄ1 atm) exhibited three
absorption bands at 416 nm (" ¼ ꢄ5400 Mꢁ1 cmꢁ1), ꢄ591
nm (" ¼ ꢄ1640 Mꢁ1 cmꢁ1), and ꢄ737 nm (" ¼ ꢄ2340
It is noted that the reaction temperature has a significant in-
fluence on the oxygenation behavior of [Cu(LH,Bn)]þ. The elec-
tronic spectrum of an acetone solution (0.117 mM) at ꢁ90 ꢂC
under P(O2) = ꢄ1 atm showed an absorption band at 412 nm
together with those of the peroxo species [{Cu(LH,Bn)}2(O2)]2þ
(Fig. 5b) as already observed by Schindler et al. (ꢄmax ¼ 406
nm at ꢁ70 ꢂC in propionitrile) under stopped-flow condi-
tions.25,26 Under the present reaction conditions, this species
M
ꢁ1 cmꢁ1) as shown in Fig. 6a, where " values are given based
on a monomer. The formation of a superoxo species [Cu-
(LMe,Bn)(O2)]þ was confirmed by resonance Raman spectrosco-

K. Komiyama et al.
Bull. Chem. Soc. Jpn., 77, No. 1 (2004)
67
(a)
(a)
ε / Cu
(c)
(b)
(b)
(a)
(b)
(c)
(d)
ε / Cu2
(b)
(b)
Fig. 7. Electronic spectra of [Cu(LMe,Me)]þ in acetone; (a)
at ꢁ90 ꢂC (1.04 mM and P(O2) = ꢄ0:02 atm),
[{Cu(LMe,Me)}2(O2)]2þ, (b) at ꢁ90 ꢂC (0.104 mM and
P(O2) = ꢄ1 atm), and (c) at ꢁ90 ꢂC (2.93 mM and
P(O2) = ꢄ0:02 atm) measured by an optical fiber with a
corrected light path length of 0.279 cm. " values are given
based on a dimer.
Fig. 6. ꢂ Electronic spectra of [Cu(LMe,Bn)]þ in acetone; (a) at
ꢂ
ꢁ90 C (0.238 mM and P(O2) = ꢄ1 atm), (b) at ꢁ90 C
after bubbling of Ar gas into the former acetone solution
for a few minutes, and (c) at ꢁ90 ꢂC (1.02 mM and
P(O2) = ꢄ0:02 atm) measured by an optical fiber with a
corrected light path length of 0.279 cm, where " values
for [{Cu(LMe,Bn)}2(O2)]2þ should be read as twice as those
given in the left hand axis, and (d) at ꢁ90 ꢂC (1.01 mM and
P(O2) = ꢄ0:02 atm).
*
(c)
18O2
*
py (vide infra). There is a possibility that the band at ꢄ591 nm
is attributable to an overlapping of the CT band of the peroxo
species present as a minor species (ꢄmax ¼ 567 and 618 nm,
vide infra). However, Gaussian analysis revealed that there is
no absorption band at 567 or 618 nm. Thus, these three bands
are undoubtedly attributable to a superoxo species [Cu-
(LMe,Bn)(O2)]þ. There is no appreciable absorption band up to
1300 nm. Assignment of those absorption bands remains to
be made. This superoxo species [Cu(LMe,Bn)(O2)]þ is stable
for at least one hour under the conditions. Bubbling of Ar into
this solution for a few minutes, however, caused a drastic spec-
tral change. As the 416 nm band decreased, new bands at 567
and 618 nm appeared (Fig. 6b). Although the absorption band
at 567 nm is significantly lower in energy than those of
[{Cu(LH,Bn)}2(O2)]2þ and [{Cu(tmpa)}2(O2)]2þ (518 and 525
nm, respectively), the spectral pattern is characteristic of those
of the peroxo species. The formation of a peroxo species was
also confirmed by resonance Raman spectroscopy (vide infra).
The result clearly indicates that the superoxo species
[Cu(LMe,Bn)(O2)]þ can be also converted to a peroxo species
[{Cu(LMe,Bn)}2(O2)]2þ at low O2 concentration as observed
for [Cu(LH,Bn)(O2)]þ. The electronic spectrum of a more con-
centrated acetone solution (1.02 mM) under P(O2) = ꢄ0:02
atm at ꢁ90 ꢂC exhibits no appreciable band at 416 nm, but in-
tense bands at 567 (" > 10000 Mꢁ1 cmꢁ1), 618 (shoulder, " >
8700 Mꢁ1 cmꢁ1), and 1088 nm (" > 220 Mꢁ1 cmꢁ1), where
the " values are given based on a dimer, indicating that the per-
oxo species is solely present under the conditions (Fig. 6c).
However, it should be noted that this species decayed within
several minutes. Thus, [{Cu(LMe,Bn)}2(O2)]2þ is thermally very
unstable compared with [Cu(LMe,Bn)(O2)]þ and [{Cu(LH,Bn)}2-
*
*
*
*
*
*
*
*
*
16O2
*
*
(b)
(a)
18O2
*
*
*
*
*
*
*
*
*
*
16O2
*
18O2
*
x 5
x 5
*
*
*
*
*
*
16O2
*
Fig. 8. Resonance Raman spectra of [{Cꢂu(LH,Bn)}2(O2)]2þ
(a) (ꢄ7:5 mM) in acetone-d6 at ꢄꢁ90 C with 514.5 nm
laser excitation, [{Cu(LMe,Me)}2(O2)]2þ (b) (ꢄ3 mM) in
acetone-d6 at ꢄꢁ90 ꢂC with 514.5 nm laser excitation,
and [{Cu(LMe,Bn)}2(O2)]2þ (c) (ꢄ3 mM) in acetone-d6 at
ꢄꢁ90 ꢂC with 607 nm laser excitation. The asterisked
bands are solvent bands.
(O2)]2þ
.
Recently, the electronic spectral changes for successive for-
mation of a superoxo species [Cu(LMe,Me)(O2)]þ and a peroxo
species [{Cu(LMe,Me)}2(O2)]2þ have been reported by
Schindler et al.24 We have also studied the oxygenation behav-
ior of [Cu(LMe,Me)]þ depending on changes in temperature, di-

68
Bull. Chem. Soc. Jpn., 77, No. 1 (2004)
HEADLINE ARTICLES
oxygen partial pressure, and concentration of the complex. The
spectral behavior is similar to that of [Cu(LMe,Bn)(O2)]þ and
[{Cu(LMe,Bn)}2(O2)]2þ as shown in Fig. 7. However, unlike
the superoxo species of LMe,Bn, the superoxo species [Cu-
(LMe,Me)(O2)]þ is very unstable even under conditions (0.104
reversibly as the temperature rises and the concentration of
the complex decreases. It is also noted that the shape and band
position of the ꢁ(O–O) band vary with temperature and concen-
tration: the band shifts from 834 to ꢂ837 cmꢁ1 with a tempera-
ture change from ꢄꢁ90 to ꢄꢁ45 C (ꢄ5 mM) as shown in
Fig. 9. Gaussian analysis of the bands revealed that the ꢁ(O–
ꢂ
mM in acetone at ꢁ90 C under P(O2) = ꢄ1 atm), whereas
the peroxo species [{Cu(LMe,Me)}2(Oꢂ2)]2þ is stable under con-
ditions (1.04 mM in acetone at ꢁ90 C under P(O2) = ꢄ0:02
atm) and no appreciable decay was observed for at least one
hour. The electronic spectrum of the peroxo species exhibited
the absorption bands (ꢄmax/nm ("/Mꢁ1 cmꢁ1 based on a dim-
er); ꢄ460 (shoulder, ꢄ1500), 553 (ꢄ11600), 615 (shoulder,
ꢄ8600), and 1044 (ꢄ240)).
O) bands consist of two components which are located at 839
and 829 cmꢁ1
48
.
Thus, there is the possibility of the presence
of two isomers, although the crystal structure of [{Cu-
(LH,Bn)}2(O2)]2þ showed only one species in the solid state.
This temperature and concentration dependency was also ob-
served for the bands at 460–440 cmꢁ1 and 760–745 cmꢁ1, al-
though they are not 18O sensitive. There is the possibility that
they are conformational isomers as observed in the crystal
structures of [Cu(LH,Bn)]þ (I) and [Cu(LH,Bn)(Cl)]þ (III). How-
ever, it should be noted that the observed spectral changes de-
pending on both temperature and concentration suggest the
presence of some complex equilibrium(s) between two species,
that may control the relative population of the isomers. Very re-
cently, similar observations have also been found for other
complexes having tetradentate tripodal ligand (2L or
L
Me,Me).24b,29 For better understanding of this puzzling phenom-
enon, further study is needed.
The resonance Raman spectrum of the peroxo species
[{Cu(LMe,Bn)}2(O2)]2þ measured in acetone-d6 at ꢄꢁ90 ꢂC
with a 607 nm laser excitation showed three isotope-sensitive
bands at 812, 797, and 531 cmꢁ1 (767, 753, and 505 cmꢁ1
Resonance Raman Spectra: The resonance Raman spec-
trum of the peroxo species [{Cu(LH,Bn)}2(O2)]2þ (ꢄ7:5–ꢄ1
mM) measured in acetone-d6 at ꢄꢁ90 ꢂC with 514.5 nm laser
excitation shows isotope-sensitive broad bands at 837–834
(vide infra), 556, and 539 cmꢁ1 (788, 528, and 512 cmꢁ1 for
an 18O labeled sample, respectively) as seen in Fig. 8a and
Table 4. The band at 837–834 cmꢁ1 can be assigned to the
ꢁ(O–O) vibration, and those at 556 and 539 cmꢁ1 to the
46,47
ꢁ(Cu–O) vibrations as found for [{Cu(tmpa)}2(O2)]2þ
.
Two ꢁ(Cu–O) vibrations suggest the presence of two peroxo
species. The relative intensities of those bands varied with tem-
ꢂ
perature (ꢄꢁ90–ꢄꢁ45 C) as shown in Fig. 9 and with the
concentration of the complex (ꢄ7:5–ꢄ1 mM) (Fig. S3 in the
supplementary materials).35 The band at 539 cmꢁ1 decreases
837
*
-90°C
(f)
*
*
*
-45°C
(b)
*
-45°C
(e)
*
834
-60°C
(d)
(c)
*
*
*
*
839
829
*
-90°C
(a)
-90°C
*
860
840
820
800
600
500
400
300
200
Raman Shift / cm-1
Raman Shift / cm-1
Fig. 9. Temperature dependency of the resonance Raman spectra of the ꢁ(O–O) region of [{Cu(LH,Bn)}2(O2)]2þ (ꢄ5 mM) in ace-
tone-d6 at ꢄꢁ90 ꢂC (a) and ꢄꢁ45 ꢂC (b), and that of the ꢁ(Cu–O) region at ꢄꢁ90 ꢂC (c), ꢄꢁ60 ꢂC (d), ꢄꢁ45 ꢂC (e), and ꢄꢁ90
ꢂC (rechilled) (f), where intensities of two ꢁ(O–O) bands (a and b) are normalized. The ꢁ(O–O) bands were fitted by Gaussian anal-
ysis with two peaks at 829 and 839 cmꢁ1 constrained to the line widths of 19 and 16 cmꢁ1. The dotted lines are the calculated spec-
tra. The asterisked bands are solvent bands.

K. Komiyama et al.
Bull. Chem. Soc. Jpn., 77, No. 1 (2004)
69
plexes. However, dioxygen reactivity of the present copper(I)
complexes is mainly dependent on the steric bulkiness of the
substituents on the terminal nitrogens, not on the redox proper-
ties. It is interesting to explore how the nature of the substitu-
ents of the tren derivatives affects the dioxygen reactivity of
their copper(I) complexes, and the structures and properties
of the resulting copper–dioxygen complexes. Schindler et al.
reported that the reaction of [Cu(tren)]þ with O2 resulted in
an instantaneous irreversible oxidation under the same condi-
tions used for the oxygenation of [Cu(LMe,Me)]þ.24 However,
the introduction of sterically bulky substituents, such as methyl
or benzyl group(s), into the terminal nitrogens significantly sta-
bilizes copper–dioxygen complexes, which are observable at
low temperatures. The space-filling model of [{Cu(LH,Bn)}2-
(O2)]2þ as shown in Fig. 2b reveals that the Cu(II)(ꢀ-O2)Cu(II)
core is almost completely covered by a hydrophobic cavity
formed by the six benzyl groups of LH,Bn ligands, which seems
to be responsible for protecting the unstable Cu(II)(ꢀ-O2)-
Cu(II) core against some deleterious decay pathways such as
substitution with solvent molecules, a disproportionation reac-
tion between dimers, etc. Such hydrophobic cavities have also
been shown to be effective for stabilization of the unstable
(ꢀ-peroxo)diiron complexes49 and high-valent bis(ꢀ-oxo)-
dinickel(III) complex.50 The methyl groups of LMe,Me also have
a similar steric effect toward the stabilization of the peroxo
species.
474
*
454
*
*
*
*
18
16
*
*
O
2
2
*
*
*
*
*
*
*
O
*
Fig. 10. Resonance Raman spectra of [Cu(LMe,Bn)(O2)]þ
with 413.1 nm laser excitation in acetone-d6 at ꢄꢁ95
ꢂC. Inset is a difference spectrum of those of 16O2 and
18O2 samples. The asterisked bands are solvent bands.
for an 18O labeled sample, respectively) as seen in Fig. 8c. The
former two bands are assigned to the ꢁ(O–O) vibrations and the
latter to the ꢁ(Cu–O) vibration. There is also the possibility of
the presence of two peroxo species similar to [{Cu(LH,Bn)}2-
(O2)]2þ, although a splitting due to Fermi-resonance can not
be ruled out.
Although recently Schindler et al. reported the Raman spec-
tra of [{Cu(LMe,Me)}2(O2)]2þ (data are given Table 4),24 we
also measured the resonance Raman spectra of [{Cu-
(LMe,Me)}2(O2)]2þ in acetone-d6 at ꢄꢁ90 ꢂC. Three isotope
sensitive bands were observed at 822, 808, and 535 cmꢁ1
(775, shoulder 764, and 510 cmꢁ1 for 18O2) as seen in Fig.
8b. Relative intensities of the ꢁ(O–O) bands for the 16O2 and
18O2 samples differ significantly, suggesting that the two
ꢁ(O–O) bands are due to Fermi-resonance. However, the origin
of two ꢁ(O–O) bands is not clear at present.
As already mentioned, the relative formation ratio of super-
oxo and peroxo complexes are highly dependent on the concen-
tration of the complex, temperature, and the partial pressure of
dioxygen and the steric bulkiness of the substituents on the ter-
minal nitrogens of tren. As expected from the stepwise equili-
briums (Eqs. 1 and 2), the superoxo complexes are favored at
lower concentrations of the complex and higher dioxygen par-
tial pressure. Complexes [Cu(LH,Bn)]þ and [Cu(LMe,Me)]þ form
both superoxo and peroxo species under the conditions ([com-
plex] = ꢄ0:1 mM, P(O2) = ꢄ1 atm, and T ¼ ꢁ90 ꢂC), where-
as [Cu(LMe,Bn)]þ produces solely a superoxo species under sim-
ilar conditions. The observed higher stability of the superoxo
species [Cu(LMe,Bn)(O2)]þ over the peroxo species [{Cu-
(LMe,Bn)}2(O2)]2þ suggests that steric constraints between the
N-benzyl-N-methyl groups is significant, which prevents dime-
rization (Eq. 2). The present results show that the steric effect of
the N-benzyl-N-methyl groups is larger than that of the N,N-di-
methyl groups of the LMe,Me ligand for dimerization. It should
be noted that [{Cu(LMe,Bn)}2(O2)]2þ is thermally unstable com-
pared to the peroxo species of LH,Bn or LMe,Me, and readily un-
derwent an irreversible oxidation within several minutes, al-
though the peroxo species of LH,Bn or LMe,Me did not show
any appreciable irreversible oxidation for at least one hour un-
der the similar conditions. These results suggest that unfavora-
ble steric interactions between the N-benzyl-N-methyl groups
of the two LMe,Bn ligands prevents not only the formation of
a dimer structure but also facilitates irreversible oxidation prob-
ably triggered by cleavage of the dimer structure, which over-
comes the protection effect of a hydrophobic cavity formed by
N-benzyl-N-methyl groups surrounding the Cu(II)(ꢀ-O2)-
The resonance Raman spectrum of the superoxo ꢂspecies
[Cu(LMe,Bn)(O2)]þ (ꢄ1 mM) in acetone-d6 at ꢄꢁ95 C with
a 413.1 nm laser excitation showed two isotope-sensitive bands
at 1120 and 474 cmꢁ1 (1059 and 454 cmꢁ1 for an 18O labeled
sample) as shown in Fig. 10. The bands at 1120 and 475 cmꢁ1
can be assigned to the ꢁ(O–O) and ꢁ(Cu–O) vibrations, respec-
tively. The ꢁ(Cu–O) band at 475 cmꢁ1 is significantly lower
than those of the peroxo complexes, suggesting a weaker bond-
ing of the superoxo ligand compared to the peroxo ligand.
Schindler et al. reported the resonance Raman spectrum of
[Cu(LMe,Me)(O2)]þ which exhibited the ꢁ(O–O) at 1122
cmꢁ1 24b although the spectrum of a 18O2 sample was not given.
,
Those ꢁ(O–O) frequencies are characteristic of those of super-
oxo complexes of various transition metal ions,44 whereas quite
different from that of [Cu(HB(3-Ad-5-iPrpz)3)(O2)] (1043
10b
cmꢁ1
as a side-on binding mode.
)
in which a superoxide coordinates to the Cu(II) ion
Discussion
The electrochemistry of the present complexes showed that
the introduction of benzyl and/or methyl groups on the termi-
nal nitrogens of the tren skeleton has a significant influence on
the redox properties of the copper(I) and chlorocopper(II) com-
Cu(II) core as observed for [{Cu(LH,Bn)}2(O2)]2þ
.
The steric nature of the terminal nitrogen of the tren deriva-
tives also has a significant influence on the electronic properties

70
Bull. Chem. Soc. Jpn., 77, No. 1 (2004)
HEADLINE ARTICLES
ꢀ
of the peroxo complexes. The spin-allowed O2(ꢋꢉ ) ! Cu(II)
LMCT transition energies of the peroxo complexes vary from
518 to 567 nm as seen in Table 4. Karlin et al. suggested that
the LMCT transition energy of a trans-ꢀ-1,2-peroxo-dicop-
per(II) complex having a sterically strained structure is lower
than that of a complex having a less strained structure.2,20a
The LMCT and the d–d transition energies roughly reflect the
ꢉ bonding strength between the peroxo ligand and copper(II)
ion: as the ꢉ bonding of the peroxo ligand becomes stronger,
[{Cu(LMe,Bn)}2(O2)]2þ (531 cmꢁ1). There is a rough correla-
tion between the energies of the LMCT and the d–d transitions
and those of the ꢁ(O–O) and ꢁ(Cu–O) frequencies as seen in
Fig. 11, which includes tꢁh1ose of the other trans-ꢀ-1,2-peroxo
(ꢁmax/cmꢁ1)), 807–746 (17700)
~
~
complexes (ꢁ(O–O)/cm
29
for [{Cu(2L)}2(O2)]2þ
,
822 (18180) for [{Cu(Lpy)}2-
21a
22
(O2)]2þ
,
844 (19050) for [Cu2(MEPY22PZ)(O2)]2þ
,
840 (19230) for [{Cu-
831
(19800) for [Cu2(bpman)(O2)]2þ
,
30
(L1)}2(O2)]2þ
,
and 827 (20200) for [Cu2(L2)(O2)]2þ
.
The
23
23
2
the dz orbital energy of copper(II) ion becomes higher, result-
above tendency may be qualitatively explained in terms of
the bonding strength between the peroxo ligand and copper(II)
ion: as the interaction increases, the Cu–O bond becomes
ing in a higher energy shift of the LMCT and the d–d transi-
tions. The order of the LMCT and the d–d transition energies
of the peroxo complexes is [{Cu(LH,Bn)}2(O2)]2þ (518 and
976 nm) > [{Cu(tmpa)}2(O2)]2þ (525 and 1035 nm) > [{Cu-
(LMe,Me)}2(O2)]2þ (553 and 1044 nm) > [{Cu(LMe,Bn)}2-
(O2)]2þ (567 and 1088 nm), which seems to reflect the relative
degree of steric constraint of the Cu(II)–(O22ꢁ)–Cu(II) cores
arising from the steric interactions between the substituents
on the terminal nitrogens. The order is not in line with that of
the d–d transition energies for the corresponding chloro com-
plexes ([Cu(LH,Bn)(Cl)]þ (904 nm) > [Cu(LMe,Bn)(Cl)]þ (934
nm) ꢄ [Cu(LMe,Me)(Cl)]þ (932 nm) > [Cu(tmpa)(Cl)]þ (955
nm)) mentioned already, where there is no steric constraint
due to ligand–ligand interaction.
ꢀ
stronger, and the electron density of the ꢋꢉ orbital of the per-
oxo ligand decreases, resulting in an increase of the O–O bond
order and the strength of the O–O bond.14b,47 In order to confirm
the above correlations between the LMCT and the d–d transi-
tion energies, the ꢁ(O–O) and ꢁ(Cu–O) frequencies, and the
thermal stability of the peroxo complexes against irreversible
oxidation, further studies are needed.
Formation of a superoxo copper(II) species with a tetraden-
tate tripodal ligand was confirmed by resonance Raman spec-
troscopy for [Cu(LMe,Bn)(O2)]þ, which is thermodynamically
stable under the conditions mentioned above. It should be noted
that, unlike the peroxo complexes, the superoxo complexes ex-
hibit a relatively intense charge transfer (LMCT) band at 410–
420 nm, which shows no remarkable change in the transition
energy depending on the coordination environments of the su-
peroxo complexes except for [Cu(HB(3-Ad-5-iPrpz)3)(O2)]10b
(see Table 4), which has a bidentate side-on coordination mode.
Very recently, Tolman et al. reported that [Cu(2L)(O2)]þ also
has a side-on coordination mode based on resonance Raman
spectra.29 Unlike [Cu(HB(3-Ad-5-iPrpz)3)(O2)], the electronic
The ꢁ(O–O) and ꢁ(Cu–O) frequencies of the present peroxo
complexes also vary from 839 to 797 cmꢁ1 and from 556 to 531
cmꢁ1, respectively, depending on the steric nature of the termi-
nal nitrogens of the tren derivatives. The order of the ꢁ(O–O)
frequencies is [{Cu(LH,Bn)}2(O2)]2þ (839 and 829 cmꢁ1) >
46,47
[{Cu(tmpa)}2(O2)]2þ (827 cmꢁ1
)
>
[{Cu(LMe,Me)}2-
(O2)]2þ (822 and 808 cmꢁ1) > [{Cu(LMe,Bn)}2(O2)]2þ (812
and 797 cmꢁ1) and that of the ꢁ(Cu–O) frequencies is [{Cu-
(tmpa)}2(O2)]2þ (561 cmꢁ1
)
> [{Cu(LH,Bn)}2(O2)]2þ
and Raman spectra of [Cu(2L)(O2)]þ (ꢄmax/nm ("/Mꢁ1 cmꢁ1
)
46,47
(556 and 539 cmꢁ1) > [{Cu(LMe,Me)}2(O2)]2þ (535 cmꢁ1) >
= 416 (4600) and 654 (1800), ꢁ(O–O) = 1120 cmꢁ1
(
16{18ꢌ ¼
62 cmꢁ1)) are quite similar to those of the above superoxo com-
plexes. Although there is no structural information on the su-
peroxo complexes of the present type except for [Cu(HB-
(3-Ad-5-iPrpz)3)(O2)] and [Cu(2L)(O2)]þ, an end-on terminal
coordination mode has been proposed. Thus, further inves-
tigations on the relationship between structures and spectro-
scopic properties are needed.
g
f
h
i
i
c
c
h
e
j
d
a
a
b
b
Summary
e
f
The tetradentate tripodal ligand tren can be readily modified
stereochemically by introducing various substituent(s) into the
terminal nitrogens, which allows to investigate the substituent
effect on the properties of copper–dioxygen complexes. The di-
oxygen reactivity of the copper(I) complexes (dioxygen bind-
ing mode and thermal stability of the resulting dioxygen species
toward irreversible oxidation) are mainly dependent on the ster-
ic bulkiness of the substituents on the terminal nitrogens. Al-
though the redox properties of the copper(I) and copper(II)
complexes are also substantially dependent on the substituents,
this effect seems to have a minor influence on the dioxygen re-
activity of the present copper(I) complexes. The ligands LH,Bn
and LMe,Me stabilize the trans-(ꢀ-1,2-peroxo)-dicopper(II) spe-
cies. The crystal structure of the trans-(ꢀ-1,2-peroxo)-dicop-
per(II) complex of LH,Bn showed that the benzyl groups of
the LH,Bn ligand form a cavity around the trans-(ꢀ-1,2-peroxo)-
h
h
c
a
d
b
b
~
Fig. 11. Correlations between the energies of the spin-
ꢀ
allowed O2(ꢋꢉ ) ! Cu(II) LMCT transition and the fre-
quencies of the ꢁ(O–O) ( : left hand axis) and ꢁ(Cu–O)
(
: right hand axis) vibrations of the peroxo complexes.
a: [{Cu(LMe,Bn)}2(O2)]2þ, b: [{Cu(2L)}2(O2)]2þ, c: [{Cu-
(LMe,Me)}2(O2)]2þ
, , e: [{Cu-
d: [{Cu(Lpy)}2(O2)]2þ
(tmpa)}2(O2)]2þ, f: [Cu2(MEPY22PZ)(O2)]2þ, g: [{Cu-
(L1)}2(O2)]2þ, h: [{Cu(LH,Bn)}2(O2)]2þ, i: [Cu2(bpman)-
(O2)]2þ, j: [Cu2(L2)(O2)]2þ
.

K. Komiyama et al.
Bull. Chem. Soc. Jpn., 77, No. 1 (2004)
71
1277 (1992).
M. Kodera, K. Katayama, Y. Tachi, K. Kano, S. Hirota, S.
Fujinami, and M. Suzuki, J. Am. Chem. Soc., 121, 11006 (1999).
a) J. Reim and B. Krebs, Angew. Chem., Int. Ed. Engl., 33,
dicopper(II) core, which seems to be responsible for protecting
the unstable Cu(II)(ꢀ-O2)Cu(II) core from unfavorable irrever-
sible oxidation. This may be also the case for the observed ther-
7
mal stability of the peroxo complex [{Cu(LMe,Me)}2(O2)]2þ
,
8
1969 (1994). b) J. Reim, R. Werner, W. Haase, and B. Krebs,
Chem.—Eur. J., 4, 289 (1998).
which has N,N-dimethyl substituents. Thus, a hydrophobic cav-
ity is very effective for stabilizing unstable and reactive spe-
cies, and probably makes its isolation possible.
9
A. Wada, M. Harata, K. Hasegawa, K. Jitsukawa, H.
Masuda, M. Mukai, T. Kitagawa, and H. Einaga, Angew. Chem.,
Int. Ed., 37, 798 (1998).
10 a) K. Fujisawa, M. Tanaka, Y. Moro-oka, and N. Kitajima,
J. Am. Chem. Soc., 116, 12079 (1994). b) P. Chen, D. E. Root, C.
Campochiaro, K. Fujisawa, and E. I. Solomon, J. Am. Chem. Soc.,
125, 466 (2002).
11 N. W. Aboelella, E. A. Lewis, A. M. Reynolds, W. W.
Brennessel, C. J. Cramer, and W. B. Tolman, J. Am. Chem. Soc.,
124, 10660 (2002).
12 a) K. D. Karlin, M. S. Haka, R. W. Cruse, G. J. Meyer, A.
Farooq, Y. Gultneh, J. C. Hayes, and J. Zubieta, J. Am. Chem. Soc.,
The relative formation ratio of the superoxo and peroxo com-
plexes can be controlled by the steric bulkiness of the substitu-
ents on the terminal nitrogens and the reaction conditions (con-
centration of complex, temperature, and dioxygen partial pres-
sure). The N-benzyl-N-methyl groups of the LMe,Bn ligand
greatly suppress the formation of the peroxo species due to sig-
nificant steric constraints arising from the N-benzyl-N-methyl
groups, and stabilize the superoxo species. Although [{Cu-
(LMe,Bn)}2(O2)]2þ can be generated under the conditions of
higher concentration (ꢄ1 mM) and lower dioxygen concentra-
tion (P(O2) = ꢄ0:02 atm) as expected from Eqs. 1 and 2, [{Cu-
(LMe,Bn)}2(O2)]2þ is very unstable and undergoes irreversible
oxidation. The results suggest that unfavorable steric inter-
actions between the N-benzyl-N-methyl groups of the two
´
110, 1196 (1988). b) K. D. Karlin, Z. Tyeklar, A. Farooq, M. S.
Haka, P. Ghosh, R. W. Cruse, Y. Gultneh, J. C. Hayes, P. J.
Toscano, and J. Zubieta, Inorg. Chem., 31, 1436 (1992). c) E.
Pidcock, H. V. Obias, M. Abe, H.-C. Liang, K. D. Karlin, and E.
I. Solomon, J. Am. Chem. Soc., 121, 1299 (1999).
13 a) K. D. Karlin, R. W. Cruse, Y. Gultneh, A. Farooq, J. C.
Hayes, and J. Zubieta, J. Am. Chem. Soc., 109, 2668 (1987). b) J. E.
Pate, R. W. Cruse, K. D. Karlin, and E. I. Solomon, J. Am. Chem.
Soc., 109, 2624 (1987).
14 a) K. D. Karlin, P. Ghosh, R. W. Cruse, A. Farooq, Y.
Gultneh, R. R. Jacobson, N. J. Blackburn, R. W. Strange, and J.
Zubieta, J. Am. Chem. Soc., 110, 6769 (1988). b) D. E. Root, M.
Mahroof-Tahir, K. D. Karlin, and E. I. Solomon, Inorg. Chem.,
37, 4838 (1998).
L
ture, but also facilitates irreversible oxidation.
Me,Bn ligands suppress not only the formation of a dimer struc-
Significant substituent effects have also been observed for
the ꢁ(O–O) and ꢁ(Cu–O) frequencies, and the LMCT and d–
d transition energies of the peroxo complexes. The resonance
Raman spectroscopy suggests that at least two trans-1,2-peroxo
species are present in relatively high concentrations at equili-
brium. There is a rough correlation between the LMCT and
the d–d transition energies and the steric constraint in the
Cu(II)(O22ꢁ)Cu(II) core as previously suggested by Karlin et
al. In addition, a rough correlation between the ꢁ(O–O),
ꢁ(Cu–O), LMCT, and d–d transition energies of the peroxo
complex is also observed. This notion seems to be useful in
elucidating the steric and/or electronic nature of the
Cu(II)(O22ꢁ)Cu(II) core depending on the supporting ligands,
although further systematic study is needed.
15 a) J. A. Halfen, S. Mahapatra, E. C. Wilkinson, S. Kaderli,
V. G. Young, Jr., L. Que, Jr., A. D. Zuberbuhler, and W. B.
¨
Tolman, Science, 271, 1397 (1996). b) V. Mahadevan, Z. Hou,
A. P. Cole, D. E. Root, T. K. Lal, E. I. Solomon, and T. D. P. Stack,
J. Am. Chem. Soc., 119, 11996 (1997). c) H. Hayashi, S. Fujinami,
S. Nagatomo, S. Ogo, M. Suzuki, A. Uehara, Y. Watanabe, and T.
Kitagawa, J. Am. Chem. Soc., 122, 2124 (2000).
16 Crystal structure of a superoxocopper(II) complex has been
reported by Harata et al. (M. Harata, K. Jitsukawa, H. Masuda, and
H. Einaga, J. Am. Chem. Soc., 116, 10817 (1994)), but the structure
has been claimed by Berreau et al. (L. M. Berreau, S. Mahapatra, J.
A. Halfen, V. G. Young, Jr., and W. B. Tolman, Inorg. Chem., 35,
6339 (1996)).
Financial support of this research by the Ministry of Educa-
tion, Culture, Sports, Science and Technology Grant-in-Aid for
Scientific Research to M. S. and T. K. is gratefully acknowl-
edged.
17 K. D. Karlin, N. Wei, B. Jung, S. Kaderli, P. Niklaus, and A.
D. Zuberbuhler, J. Am. Chem. Soc., 115, 9506 (1993).
References
¨
18 N. Wei, N. N. Murthy, Z. Tyeklar, and K. D. Karlin, Inorg.
´
1
2
N. Kitajima and Y. Moro-oka, Chem. Rev., 94, 737 (1994).
K. D. Karlin, S. Kaderli, and A. D. Zuberbuhler, Acc. Chem.
Chem., 33, 1177 (1994).
19 N. Wei, N. N. Murthy, Q. Chen, J. Zubieta, and K. D.
Karlin, Inorg. Chem., 33, 1953 (1994).
¨
Res., 30, 139 (1997).
3
4
W. B. Tolman, Acc. Chem. Res., 30, 227 (1997).
K. D. Karlin and A. D. Zuberbuhler, ‘‘Bioinorganic Catal-
´
20 a) D.-H. Lee, N. Wei, N. N. Murthy, Z. Tyeklar, K. D.
Karlin, S. Kaderli, B. Jung, and A. D. Zuberbuhler, J. Am. Chem.
¨
ysis,’’ 2nd ed, revised and expanded, ed by J. Reedijk and E.
Bouwman, Marcel Dekker, New York (1999), pp. 469–534.
¨
Soc., 117, 12498 (1995). b) K. D. Karlin, D.-H. Lee, S. Kaderli, and
A. D. Zuberbuhler, Chem. Commun., 1997, 475.
´
a) R. R. Jacobson, Z. Tyeklar, A. Farooq, K. D. Karlin, S.
Liu, and J. Zubieta, J. Am. Chem. Soc., 110, 3690 (1988). b) Z.
5
¨
21 a) J. A. Halfen, V. G. Young, Jr., and W. B. Tolman, J. Am.
Chem. Soc., 118, 10920 (1996). b) L. M. Berreau, J. A. Halfen, V.
G. Young, Jr., and W. B. Tolman, Inorg. Chim. Acta, 297, 115
(2000).
´
Tyeklar, R. R. Jacobson, N. Wei, N. N. Murthy, J. Zubieta, and
K. D. Karlin, J. Am. Chem. Soc., 115, 2677 (1993).
6
a) N. Kitajima, K. Fujisawa, Y. Moro-oka, and K. Toriumi,
J. Am. Chem. Soc., 111, 8975 (1989). b) N. Kitajima, K. Fujisawa,
C. Fujimoto, Y. Moro-oka, S. Hashimoto, T. Kitagawa, K.
Toriumi, K. Tatsumi, and A. Nakamura, J. Am. Chem. Soc., 114,
22 J. E. Bol, W. L. Driessen, R. Y. N. Ho, B. Maase, L. Que,
Jr., and J. Reedijk, Angew. Chem., Int. Ed. Engl., 36, 998 (1997).
23 H. Borzel, P. Comba, C. Katsichtis, W. Kiefer, A. Lienke,
¨

72
Bull. Chem. Soc. Jpn., 77, No. 1 (2004)
HEADLINE ARTICLES
V. Nagel, and H. Pritzkow, Chem.—Eur. J., 5, 1716 (1999).
24 a) M. Becker, F. W. Heinemann, and S. Schindler, Chem.—
Eur. J., 5, 3124 (1999). b) M. Weitzer, S. Schindler, G. Brehm,
and K. D. Karlin, Inorg. Chem., 33, 6093 (1994).
37 a) G. M Sheldrick, ‘‘SHELXS-86. A Program for Crystal
Structure Determination,’’ University of Gottingen, FRG (1986).
¨
S. Schneider, E. Hormann, B. Jung, S. Kaderli, and A. D.
¨
25 S. Schindler, Eur. J. Inorg. Chem., 2000, 2311.
26 M. Schatz, M. Becker, O. Walter, G. Liehr, and S.
Schindler, Inorg. Chim. Acta, 324, 173 (2001).
b) SIR-92: A. Altomare, G. Cascarano, C. Giacovazzo, A.
Guagliardi, M. C. Burla, G. Polidori, and M. Camalli, J. Appl.
Crystallogr., 27, 435 (1994).
38 P. T. Beurskens, G. Admiraal, G. Beurskens, W. P.
Bosman, R. de Gelder, R. Israel, and J. M. M. Smits, ‘‘The
DIRDIF-94 program system,’’ Technical Report of the Crystallog-
raphy Laboratory, University of Nijmegen, The Netherlands
(1994).
¨
Zuberbuhler, Inorg. Chem., 42, 1800 (2003).
27 M. Schatz, M. Becker, F. Thaler, F. Hampel, S. Schindler,
´
R. R. Jacobson, Z. Tyeklar, N. N. Murthy, P. Ghosh, Q. Chen, J.
Zubieta, and K. D. Karlin, Inorg. Chem., 40, 2312 (2001).
28 M. Weitzer, M. Schatz, F. Hampel, F. W. Heinemann, and
S. Schindler, J. Chem. Soc., Dalton Trans., 2002, 686.
39 teXsan: Crystal Structure Analysis Package, Molecular
Structure Corporation (1985 & 1992).
´
40 R. R. Jacobson, Z. Tyeklar, and K. D. Karlin, Inorg. Chim.
29 B. A. Jazdzewski, A. M. Reynolds, P. L. Holland, V. G.
Young, Jr., S. Kaderli, A. D. Zuberbuhler, and W. B. Tolman, J.
Biol. Inorg. Chem., 8, 381 (2003).
30 C. He, J. L. DuBois, B. Hedman, K. O. Hodgson, and S. J.
Lippard, Angew. Chem., Int. Ed., 40, 1484 (2001).
31 C. X. Zhang, S. Kaderli, M. Costas, E. Kim, Y.-M.
Acta, 181, 111 (1991).
41 E. C. Alyea, G. Ferguson, M. C. Jenning, and Z. Xu, Poly-
hedron, 9, 739 (1990).
42 H. Hayashi, K. Uozumi, S. Fujinami, S. Nagatomo, K.
Shiren, H. Furutachi, M. Suzuki, A. Uehara, and T. Kitagawa,
Chem. Lett., 2002, 416.
43 A. W. Addison, T. N. Rao, J. Reedijk, J. Rijin, and G. C.
Verschoor, J. Chem. Soc., Dalton Trans., 1984, 1349.
44 E. C. Niederhoffer, J. H. Timmons, and A. E. Martell,
Chem. Rev., 84, 137 (1984), and the references therein.
45 P. C. Jain and E. C. Lingafelter, J. Am. Chem. Soc., 89, 724
(1967).
¨
Neuhold, K. D. Karlin, and A. D. Zuberbuhler, Inorg. Chem., 42,
1807 (2003).
¨
32 Abbreviations of ligands used: tmpa = tris(2-pyridylmeth-
yl)amine; tren = tris(aminoethyl)amine; Ph3tren = tris(4-phenyl-
3-aza-3-butenyl)amine; Me2-tpa = bis(6-methyl-2-pyridylmeth-
yl)(2-pyridylmethyl)amine; Me3-tpa = tris(6-methyl-2-pyridyl-
methyl)amine; tmqa = tris(2-quinolylmethyl)amine; MEPY22PZ
´
=
[9,22-di(pyridin-2-ylmethyl)-5,13,18,26-tetramethyl-1,4,9,14,
46 M. J. Baldwin, P. K. Ross, J. E. Pate, Z. Tyeklar, K. D.
17,22,27,28,29,30-decaaza]pentacyclo[22.2.1.14;7.111;14.117;20]tri-
acontan-5,7(28),11(29),12,18,20(30)24(27),25-octaene; Lpy = 1,4-
diisopropyl-7-(2-pyridylmethyl)-1,4,7-triazacyclononane; HB(3-
Ad-5-iPrpz)3 = hydrotris(3-adamantyl-5-isopropyl-1-pyrazolyl)-
borate; 2L = 1-(2-hydroxy-3,5-di-tert-butylbenzyl)-4,7-diisoprop-
yl-1,4,7-triazacyclononane; bpman = 2,7-bis[bis(2-pyridylmeth-
yl)aminomethyl]-1,8-naphthyridine.
Karlin, and E. I. Solomon, J. Am. Chem. Soc., 113, 8671 (1991).
47 M. J. Henson, M. A. Vance, C. X. Zhang, H.-C. Liang, K.
D. Karkin, and E. I. Solomon, J. Am. Chem. Soc., 125, 5186 (2003).
48 The ꢁ(O–O) curves were fitted with two peaks at 829 and
839 cmꢁ1 constrained to the line widths of 19 and 16 cmꢁ1
.
49 a) Y. Hayashi, T. Kayatani, H. Sugimoto, M. Suzuki, K.
Inomata, A. Uehara, Y. Mizutani, T. Kitagawa, and Y. Maeda, J.
Am. Chem. Soc., 117, 11220 (1995). b) T. Ookubo, H. Sugimoto,
T. Nagayama, H. Masuda, T. Sato, K. Tanaka, Y. Maeda, H.
Okawa, Y. Hayashi, A. Uehara, and M. Suzuki, J. Am. Chem.
Soc., 118, 701 (1996).
50 K. Shiren, S. Ogo, S. Fujinami, H. Hayashi, M. Suzuki, A.
Uehara, Y. Watanabe, and Y. Moro-oka, J. Am. Chem. Soc., 122,
254 (2000).
33 A. A. Naiini, W. M. P. B. Menge, and J. G. Verkade, Inorg.
Chem., 30, 5009 (1991).
34 M. Ciampolini and N. Nardi, Inorg. Chem., 5, 41 (1966).
35 Deposited as Document No. 03203 at the Office of the Ed-
itor of Bull. Chem. Soc. Jpn.
36 a) K. D. Karlin, J. C. Hayes, S. Juen, J. P. Hutchinson, and J.
Zubieta, Inorg. Chem., 21, 4106 (1982). b) N. Wei, N. N. Murthy,