Reactions of copper(II)-phenol systems with O2: Models for TPQ biosynthesis in copper amine oxidases

Article abstract of DOI:10.1021/ic101832c

Copper(II) complexes supported by a series of phenol-containing bis(pyridin-2-ylmethyl)amine N₃ ligands (denoted as L^oH, L^mH, and L^pH) have been synthesized, and their O ₂ reactivity has been examined in detail to gain mechanistic insights into the biosynthesis of the TPQ cofactor (2,4,5- trihydroxyphenylalaninequinone, TOPA quinone) in copper-containing amine oxidases. The copper(II) complex of L^oH (ortho-phenol derivative) involves a direct phenolate to copper(II) coordination and exhibits almost no reactivity toward O₂ at 60 °C in CH₃OH. On the other hand, the copper(II) complex of L^mH (meta-phenol derivative), which does not involve direct coordinative interaction between the phenol moiety and the copper(II) ion, reacts with O₂ in the presence of triethylamine as a base to give a methoxy-substituted para-quinone derivative under the same conditions. The product structure has been established by detailed nuclear magnetic resonance (NMR), infrared (IR) spectroscopy, and electrospray ionization-mass spectroscopy (ESI-MS) (including ¹⁸O-labeling experiment) analyses. Density functional theory predicts that the reaction involves (i) intramolecular electron transfer from the deprotonated phenol (phenolate) to copper(II) to generate a copper(I)-phenoxyl radical; (ii) the addition of O₂ to this intermediate, resulting in an end-on copper(II) superoxide; (iii) electrophilic substitution of the phenolic radical to give a copper(II)-alkylperoxo intermediate; (iv) O-O bond cleavage concomitant with a proton migration, giving a para-quinone derivative; and (v) Michael addition of methoxide from copper(II) to the para-quinone ring and subsequent O₂ oxidation. This reaction sequence is similar to that proposed for the biosynthetic pathway leading to the TPQ cofactor in the enzymatic system. The generated para-quinone derivative can act as a turnover catalyst for aerobic oxidation of benzylamine to N-benzylidene benzylamine. Another type of copper(II)-phenol complex with an L^pH ligand (para-phenol derivative) also reacts with O₂ under the same experimental conditions. However, the product of this reaction is a keto-alcohol derivative, the structure of which is qualitatively different from that of the cofactor. These results unambiguously demonstrate that the steric relationship between the phenol moiety and the supported copper(II) ion is decisive in the conversion of active-site tyrosine residues to the TPQ cofactor.

Full text of DOI:10.1021/ic101832c

Inorg. Chem. 2011, 50, 1633–1647 1633
DOI: 10.1021/ic101832c
Reactions of Copper(II)-Phenol Systems with O2: Models for TPQ Biosynthesis in
Copper Amine Oxidases
Kae Tabuchi,† Mehmed Z. Ertem,‡ Hideki Sugimoto,† Atsushi Kunishita,† Tetsuro Tano,† Nobutaka Fujieda,†
Christopher J. Cramer,*,‡ and Shinobu Itoh*,†
†Department of Material and Life Science, Division of Advanced Science and Biotechnology, Graduate School of
Engineering, Osaka University, 2-1 Yamada-oka, Suita, Osaka 565-0871, Japan, and ‡Department of Chemistry
and Supercomputing Institute, University of Minnesota, 207 Pleasant St. SE, Minneapolis, Minnesota 55455,
United States
Received September 7, 2010
Copper(II) complexes supported by a series of phenol-containing bis(pyridin-2-ylmethyl)amine N3 ligands (denoted as
LoH, LmH, and LpH) have been synthesized, and their O2 reactivity has been examined in detail to gain mechanistic
insights into the biosynthesis of the TPQ cofactor (2,4,5-trihydroxyphenylalaninequinone, TOPA quinone) in copper-
containing amine oxidases. The copper(II) complex of LoH (ortho-phenol derivative) involves a direct phenolate to
copper(II) coordination and exhibits almost no reactivity toward O2 at 60 ꢀC in CH3OH. On the other hand, the copper(II)
complex of LmH (meta-phenol derivative), which does not involve direct coordinative interaction between the phenol
moiety and the copper(II) ion, reacts with O2 in the presence of triethylamine as a base to give a methoxy-substituted
para-quinone derivative under the same conditions. The product structure has been established by detailed nuclear
magnetic resonance (NMR), infrared (IR) spectroscopy, and electrospray ionization-mass spectroscopy (ESI-MS)
(including 18O-labeling experiment) analyses. Density functional theory predicts that the reaction involves (i)
intramolecular electron transfer from the deprotonated phenol (phenolate) to copper(II) to generate a copper(I)-
phenoxyl radical; (ii) the addition of O2 to this intermediate, resulting in an end-on copper(II) superoxide; (iii) electrophilic
substitution of the phenolic radical to give a copper(II)-alkylperoxo intermediate; (iv) O-O bond cleavage concomitant
with a proton migration, giving a para-quinone derivative; and (v) Michael addition of methoxide from copper(II) to the
para-quinone ring and subsequent O2 oxidation. This reaction sequence is similar to that proposed for the biosynthetic
pathway leading to the TPQ cofactor in the enzymatic system. The generated para-quinone derivative can act as a
turnover catalyst for aerobic oxidation of benzylamine to N-benzylidene benzylamine. Another type of copper(II)-phenol
complex with an LpH ligand (para-phenol derivative) also reacts with O2 under the same experimental conditions.
However, the product ofthis reactionis a keto-alcohol derivative, the structure ofwhich is qualitatively differentfromthatof
the cofactor. These results unambiguously demonstrate that the steric relationship between the phenol moiety and the
supported copper(II) ion is decisive in the conversion of active-site tyrosine residues to the TPQ cofactor.
Introduction
addition of the copper ion and O2 to the apo-enzymes, i.e.,
without the intermediacy of any external enzymes (so-called
“self-processing”).
Redox reactions between copper(II) ions and the phenol
groups of tyrosines in enzyme active sites are involved in the
biogenetic production of novel organic cofactors such as
TPQ (2,4,5-trihydroxyphenylalaninequinone, TOPA quinone)
by copper-containing amine oxidases, LTQ (lysine tyrosyl-
quinone) by lysyl oxidase, and Tyr-Cys by galactose oxidase
(see Figure 1).1 All of these cofactors are derived from
encoded tyrosine residues (with their phenol groups) placed
near the mononuclear copper reaction sites; the post-transla-
tional modification reactions proceed spontaneously with the
As an example, the TPQ cofactor of copper-containing
amine oxidases has been demonstrated to be generated by
self-processing of the active site tyrosine (Tyr), a potential
mechanism for which is illustrated in Scheme 1.1a,b,e,g,h,j,2 In
particular, interaction of the phenolate oxygen (deprotonated
form) of Tyr with the copper(II) ion in the holo-enzyme A
may induce electron transfer from the phenolate to copper-
(II), generating a phenoxyl radical-copper(I) intermediate
(B), to which molecular oxygen (O2) may then be added. In
this process, the O2 moiety is formally reduced by two elec-
trons;one from copper(I) and the other from the phenoxyl
*Author to whom correspondence should be addressed. E-mail:
shinobu@mls.eng.osaka-u.ac.jp (S.I.), cramer@umn.edu (C.J.C.).
r
2011 American Chemical Society
Published on Web 02/01/2011
pubs.acs.org/IC

1634 Inorganic Chemistry, Vol. 50, No. 5, 2011
Tabuchi et al.
Figure 1. Organic cofactors of amine oxidase (TPQ), lysyl oxidase (LTQ), and galactose oxidase (Tyr-Cys) post-translationally derived from an active-site
tyrosine.
radical moiety;to produce a copper(II)-alkylperoxo-type
intermediate (C). Subsequent heterolytic cleavage of the
O-O bond in C can occur to give the copper(II)-hydroxo
o-quinone species D, and following Michael addition of the
generated hydroxide to the 5-position of the o-quinone inter-
mediate D0 gives the reduced TPQ cofactor TPQred. Aerobic
oxidation of TPQred is the last step required to produce
TPQox. Similar mechanisms may be inferred for the biosynth-
esis of the LTQ cofactor of lysyl oxidase, where Michael
addition of a nearby lysine residue occurs instead of the
addition of hydroxide to the o-quinone intermediate. To date,
the hydroxide addition step of TPQ biosynthesis (D0 f TPQ)
has been investigated in model systems,3 but little has been
explored for the initial aerobic oxidation process of Tyr
catalyzed by the copper(II) ion (Tyr f D).
In this study, we examine the reactions of mononuclear
copper(II) complexes supported by a series of bis(pyridin-2-
ylmethyl)amine ligands that have a phenol group attached to
the 6-position of one of the pyridine donor groups (LoH,
LmH, and LpH; see Chart 1) when exposed to molecular
oxygen. The ligands are designed to mimic the reaction center
of the apo-form of copper-containing amine oxidases (see
Figure 2), where three histidine imidazole ligands and a
tyrosine residue (the precursor of TPQ cofactor) are present.
To examine the effect on the biosynthetic process of different
Tyr phenol geometries, relative to the copper ion, three types
of phenol-containing ligands;LoH, LmH, and LpH;have
been designed. In the case of LoH, direct coordination of the
phenolate oxygen to the bound metal ion will occur, whereas
such a coordinative interaction is unlikely to be involved in
the LmH and LpH ligand systems. On the basis of product
analyses and theoretical calculations, mechanistic aspects of
the post-translational modification of Tyr to the TPQ co-
factor are discussed.
Experimental Section
General. Reagents and solvents used in this study, except the
ligands and complexes, were commercial products of the highest
available purity and were further purified by standard methods,
if necessary.5 N-(6-Phenylpyridin-2-yl)methylbenzylamine (1),
6-(3-methoxyphenyl)pyridin-2-ylcarboxyaldehyde (2m), and
6-(4-methoxyphenyl)pyridin-2-ylcarboxyaldehyde (2p) were
prepared according to reported methods.6 [CuII(Lo)](ClO4)
(CuIILo) was prepared as described in a previous study.7 Fourier
transform infrared (FT-IR) spectra were recorded using a
Shimadzu Model FTIR-8200PC system. Mass spectra were
recorded using a JEOL Model JMS-700T Tandem MS station
or a JEOL Model JMS-700 apparatus. NMR spectra were
recorded using various equipment (JEOL Model LMN-
ECP300WB, JEOL Model ECP400, JEOL Model ECS400, or
Varian Model UNITY INOVA 600 MHz). ELectron spin
resonance (ESR) measurements were performed on a Bruker
Model E-500 spectrometer at -150 ꢀC, using a variable-tem-
perature cell holder or a JEOL Model JES-FA100 apparatus.
Anaerobic reactions were carried out after bubbling argon gas
gently through a thin syringe needle for ∼30 min to a solution in
a reaction vessel or in an ultraviolet-visible light (UV-vis) cell,
which was tightly closed by a silicon rubber cap.
(1) (a) Matsuzaki, R.; Suzuki, S.; Yamaguchi, K.; Fukui, T.; Tanizawa,
K. Biochemistry 1995, 34, 4524–4530. (b) Tanizawa, K. J. Biochem. 1995, 118,
671–678. (c) Klinman, J. P. J. Biol. Chem. 1996, 271, 27189–27192. (d) McGuirl,
M. A.; Dooley, D. M. Curr. Opin. Chem. Biol. 1999, 3, 138–144. (e) Dooley,
D. M. J. Biol. Inorg. Chem. 1999, 4, 1–11. (f) Rogers, M. S.; Baron, A. J.;
McPherson, M. J.; Knowles, P. F.; Dooley, D. M. J. Am. Chem. Soc. 2000, 122,
990–991. (g) Okeley, N. M.; van der Donk, W. A. Chem. Biol. 2000, 7, R159–
R171. (h) Halcrow, M. A. Angew. Chem., Int. Ed. 2001, 40, 346–349. (i) Xie,
L. L.; van der Donk, W. A. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 12863–
12865. (j) Kim, M.; Okajima, T.; Kishishita, S.; Yoshimura, M.; Kawamori, A.;
Tanizawa, K.; Yamaguchi, H. Nat. Struct. Biol. 2002, 9, 591–596. (k) Mure, M.;
Mills, S. A.; Klinman, J. P. Biochemistry 2002, 41, 9269–9278. (l) Firbank, S. J.;
Rogers, M.; Hurtado-Guerrero, R.; Dooley, D. M.; Halcrow, M. A.; Phillips,
S. E. V.; Knowles, P. F.; McPherson, M. J. Biochem. Soc. Trans. 2003, 31, 506–
509. (m) Rogers, M. S.; Dooley, D. M. Curr. Opin. Chem. Biol. 2003, 7, 189–
196. (n) Whittaker, M. M.; Whittaker, J. W. J. Biol. Chem. 2003, 278, 22090–
22101. (o) Whittaker, J. W. Arch. Biochem. Biophys. 2005, 433, 227–239. (p)
Dominy, J. E.; Hwang, J.; Guo, S.; Hirschberger, L. L.; Zhang, S.; Stipanuk, M. H.
J. Biol. Chem. 2008, 283, 12188–12201. (q) Rogers, M. S.; Hurtado-Guerrero,
R.; Firbank, S. J.; Halcrow, M. A.; Dooley, D. M.; Phillips, S. E. V.; Knowles, P. F.;
McPherson, M. J. Biochemistry 2008, 47, 10428–10439.
(2) (a) Schwartz, B.; Dove, J. E.; Klinman, J. P. Biochemistry 2000, 39,
3699–3707. (b) Dove, J. E.; Schwartz, B.; Williams, N. K.; Klinman, J. P.
Biochemistry 2000, 39, 3690–3698. (c) Samuels, N. M.; Klinman, J. P. J. Biol.
Chem. 2006, 281, 21114–21118. (d) DuBois, J. L.; Klinman, J. P. Biochemistry
2006, 45, 3178–3188. (e) Moore, R. H.; Spies, M. A.; Culpepper, M. B.;
Murakawa, T.; Hirota, S.; Okajima, T.; Tanizawa, K.; Mure, M. J. Am. Chem.
Soc. 2007, 129, 11524–11534.
(3) (a) Mure, M.; Tanizawa, K. Biosci., Biotechnol., Biochem. 1997, 61,
410–417. (b) Mandal, S.; Lee, Y.; Purdy, M. M.; Sayre, L. M. J. Am. Chem. Soc.
2000, 122, 3574–3584.
X-ray Structure Determination. Each single crystal was
mounted on a glass fiber. X-ray diffraction (XRD) data were
collected by a Rigaku Model RAXIS-RAPID imaging plate
two-dimensional area detector, using graphite-monochromated
˚
Mo KR radiation (λ = 0.71069 A) to 2θmax of 55ꢀ. All of the
crystallographic calculations were performed using the Crystal
(4) Wilce, M. C. J.; Dooley, D. M.; Freeman, H. C.; Guss, J. M.; Matsunami,
H.; McIntire, W. S.; Ruggiero, C. E.; Tanizawa, K.; Yamaguchi, H. Biochemistry
1997, 36, 16116–16133.
(5) Perrin, D. D.; Armarego, W. L. F.; Perrin, D. R. Purification of
Laboratory Chemicals, 4th ed.; Pergamon Press: Elmsford, NY, 1996.
(6) (a) He, Z.; Craig, D. C.; Colbran, S. B. J. Chem. Soc., Dalton Trans.
2002, 4224–4235. (b) Chuang, C.-l.; Lim, K.; Chen, Q.; Zubieta, J.; Canary, J. W.
Inorg. Chem. 1995, 34, 2562–2568. (c) Jensen, M. P.; Lange, S. J.; Mehn, M. P.;
Que, E. L.; Que, L. J. Am. Chem. Soc. 2003, 125, 2113–2128. (d) He, Z.;
Colbran, S. B.; Craig, D. C. Chem.;Eur. J. 2003, 9, 116–129. (e) Tabuchi, K.;
Sugimoto, H.; Kunishita, A.; Fujieda, N.; Itoh, S. Inorg. Chem. 2010, 49, 6820–
6822.
(7) Kunishita, A.; Teraoka, J.; Scanlon, J. D.; Matsumoto, T.; Suzuki,
M.; Cramer, C. J.; Itoh, S. J. Am. Chem. Soc. 2007, 129, 7248–7249.

Article
Inorganic Chemistry, Vol. 50, No. 5, 2011 1635
Scheme 1
(200 mL) with cooling, using a water-ice bath. The resulting
mixture was stirred at room temperature for 3 days and then
acidified by adding concentrated hydrochloric acid to quench the
remaining sodium cyanoborohydride. The pH level of the solution
was brought to 14 by adding concentrated sodium carbonate, and
the resulting solution was evaporated under reduced pressure to
remove methanol. The mixture was extracted with CH2Cl2 (50 mL
ꢀ 3), and the combined organic layer was dried over Na2SO4.
After removal of Na2SO4 by filtration, evaporation of the solvent
afforded a yellowish oily material. Column chromatography on
silica gel (hexane and ethyl acetate eluent) gave the titled compound
3m in 86% yield (2.1 g, 4.45 mmol). 1H NMR (400 MHz, CDCl3): δ
3.83 (2 H, s, -CH2-), 3.89 (3 H, s, -CH3), 3.98 (2 H, s, -CH2-),
3.99 (2H, s, -CH2-), 6.96 (1 H, ddd, J = 8.2 Hz, 2.6 and 0.9 Hz),
7.23-7.28 (1 H, m), 7.33-7.42 (4 H, m), 7.44-7.53 (4 H, m),
7.56-7.64 (6 H, m), 7.74 (2 H, td, J = 7.9 and 1.6 Hz), 8.02
(2 H, d, J = 7.3 Hz); HRMS (FABþ): m/z = 472.2384. Calcd for
(C32H30N3O)þ 472.2389.
Figure 2. Active site of apo-form of copper-containing amine oxidase
generated from PDB data (1AVK) using PyMol (DeLano Scientific LLC).4
Chart 1
N-[(6-(3-Hydroxylphenyl)pyridin-2-ylmethyl)-N-(6-phenyl-
pyridin-2-ylmethyl)benzylamine (LmH). BBr3 (5 mL of a 1.0 M
CH2Cl2 solution) was added to compound 3m (1.5 g, 3.18 mmol)
dissolved in CH2Cl2 (50 mL) in a three-neck round-bottom
flask under a nitrogen atmosphere. The mixture was refluxed for
30 h under a nitrogen atmosphere. The final reaction mixture
was then poured into methanol (50 mL), evaporated to dryness
on a rotary evaporator, and three such cycles of addition of
methanol and evaporation of solvent were performed to elim-
inate any volatile compounds. Finally, the yellowish oil was
suspended in CH2Cl2 (30 mL), which was dissolved in methanol
(20 mL) and water (20 mL). The pH of the solution was
increased to 7, and at this pH level, a pale yellow compound
was extracted with CH2Cl2 (50 mL ꢀ 3). The combined organic
layer was dried over Na2SO4. After removal of Na2SO4 by
filtration, evaporation of the solvent afforded a yellowish oily
product. Column chromatography on silica gel (hexane and
ethyl acetate eluent) gave the titled compound LmH in 63% yield
(0.92 g, 2.01 mmol,). 1H NMR (300 MHz, CDCl3): δ 3.77 (2 H, s,
-CH2-), 3.94 (2 H, s, -CH2-), 3.99 (2 H, s, -CH2-), 6.80
(1 H, ddd, J = 8.1 Hz, 2.7 Hz, and 0.9 Hz), 7.22-7.24 (1 H, m),
7.27-7.33 (2 H, m), 7.37-7.45 (7 H, m), 7.48-7.57 (5 H, m),
7.67-7.72 (2 H, m), 7.67-7.72 (2 H, m), 7.96 (2 H, d, J = 7.2
Hz); HRMS (FABþ): m/z = 458.2235. Calcd for (C31H27N3O)þ
458.2232.
Structure software package of the Molecular Structure Cor-
poration [Crystal Structure: Crystal Structure Analysis Package
version 3.8.1, Molecular Structure Corp. and Rigaku Corp.
(2005)]. The structures were solved using the SIR92 or SHELX97
programs and refined with CRYSTALS. All non-hydrogen
atoms and hydrogen atoms were refined anisotropically and
isotropically, respectively. Atomic coordinates, thermal param-
eters, and intramolecular bond distances and angles are provided
in the Supporting Information (CIF file in a text format).
Ligand Synthesis. N-[6-(3-Methoxyphenyl)pyridin-2-ylmethyl]-
N-(6-phenylpyridin-2-ylmethyl)benzylamine (3m). Sodium cyano-
borohydride (0.39 g, 6.21 mmol) was added slowly to a mixture
of 6-(3-methoxyphenyl)pyridin-2-ylcarboxyaldehyde (2m) (1.19 g,
5.24 mmol), N-(6-phenylpyridin-2-yl)methylbenzylamine (1) (1.42 g,
5.18 mmol), and acetic acid (0.33 mg, 5.49 mmol) in methanol

1636 Inorganic Chemistry, Vol. 50, No. 5, 2011
Tabuchi et al.
N-[6-(4-Methoxyphenyl)pyridin-2-ylmethyl]-N-(6-phenylpyri-
din-2-ylmethyl)benzylamine (3p). This compound was prepared
in a similar manner for the synthesis of 3m, using 6-(4-methox-
yphenyl)pyridin-2-ylcarboxyaldehyde (2p), instead of 2m (70%
(hexane and ethyl acetate eluent) converted some of compound
4 to compound 5H. Thus, the isolated yield of 4 was 6% and that
of 5H was 25%, based on the starting copper(II) complex
CuIILmH after the SiO2 column chromatography.
1
4. 1H NMR (400 MHz, CDCl3): δ 3.71 (3 H, s, -CH3), 3.77
(2 H, s, -CH2-), 3.91(2H, s, -CH2-), 3.91 (2 H, s, -CH2-), 6.74
(1 H, d, J = 10.0 Hz), 6.77 (1 H. d, J = 10.4 Hz), 7.16 (1 H, dd,
J = 7.4 and 1.6 Hz), 7.22-7.25 (1 H, m), 7.33 (2 H, t, J = 7.4
Hz), 7.39 (1 H, t, J = 7.2 Hz), 7.44-7.47 (4 H, m), 7.51 (1 H, d,
J = 7.2 Hz), 7.59 (1 H, d, J = 7.6 Hz), 7.70-7.77 (3 H, m), 7.99
(2 H, d, J = 6.8 Hz); FT-IR (NaCl) 1672 and 1727 cm-1 (CdO);
isolated yield). H NMR (400 MHz, CDCl3): δ 3.82 (2 H, s,
-CH2-), 3.86 (3 H, s, -OCH3), 3.95 (2 H, s, -CH2-), 3.97
(2 H, s, -CH2-), 6.98 (2 H, d, J = 8.8 Hz), 7.23-7.27 (1 H. m),
7.34 (2 H, t, J = 7.4 Hz), 7.41 (1 H, d, J = 7.2 Hz), 7.44-7.54
(6 H, m), 7.59 (2 H, t, J = 7.8 Hz), 7.68-7.76 (2 H, m), 7.97 (2 H,
d, J = 8.8 Hz), 8.01 (2 H, d, J = 6.8 Hz); HRMS (FABþ): m/z =
472.2385. Calcd for (C32H30N3O)þ 472.2389.
þ
N-[(6-(3-Hydroxylphenyl)pyridin-2-ylmethyl)-N-(6-phenyl-
pyridin-2-ylmethyl)benzylamine (LpH). This compound was pre-
pared in a similar manner for the synthesis of LmH, using 3p
instead of 3m (41% isolated yield). 1H NMR (400 MHz, CDCl3):
δ 3.79 (2 H, s, -CH2-), 3.95 (2 H, s, -CH2-), 3.97 (2 H, s,
-CH2-), 6.75 (2 H, d, J = 9.2), 7.23 (1 H, t, J = 7.4), 7.32 (2 H,
t, J = 7.6 Hz), 7.52-7.39 (1 H, m), 7.42-7.48 (5 H, m),
7.52-7.59 (3 H, m), 7.66-7.40 (2 H, m), 7.78 (2 H, d, J =
9.2 Hz), 7.96 (2 H, d, J = 6.8 Hz); HRMS (FABþ): m/z =
458.2220. Calcd for (C31H27N3O)þ 458.2232.
HRMS (FABþ) m/z = 502.2159; Calcd for C32H28N3O3
502.2131.
5H. 1H NMR (400 MHz, CDCl3): 3.89 (2 H, s, H23), 3.96 (2
H, s, H1), 4.03 (2 H, s, H13), 6.71 (1 H, d, J = 10.2 Hz, H10), 6.83
(1 H. d, J = 10.2 Hz, H9), 7.00 (1 H, d, J = 7.6 Hz, H3),
7.22-7.26 (1 H, m, H27), 7.31 (2 H, t, J = 7.6 Hz, H26), 7.39 (1 H,
t, J = 7.2 Hz, H22), 7.43-7.47 (2 H, m, H21), 7.52 (2 H, d, J = 7.2
Hz, H25), 7.56 (1 H, d, J = 7.6 Hz, H17), 7.61 (1 H, d, J = 7.2 Hz,
H15), 7.71 (1 H, t, J = 7.6 Hz, H16), 7.77 (2 H, t, J = 8.2 Hz, H4),
7.94 (2 H, d, J = 7.2 Hz, H20), 9.04 (1 H, d, J = 9.2 Hz, H5); 13
C
Synthesis of Copper(II) Complexes. [CuII(LmH)(CH3CN)]-
NMR (100 MHz, CDCl3): 55.26 (C23), 60.02 (C1), 61.12 (C13),
104.28 (C7), 117.99 (C3), 119.05 (C17), 121.81 (C5), 122.21 (C15),
126.81 (C20), 127.68 (C27), 128.54 (C26), 128.70 (C21), 128.93
(C22), 129.38 (C25), 132.74 (C9), 137.14 (C24), 137.37 (C16),
139.15 (C19), 141.90 (C4), 143.56 (C10), 148.71 (C2), 152.88
(C6), 156.52 (C18), 157.45 (C14), 173.90 (C12), 183.52 (C8),
184.38 (C11); FT-IR (NaCl, cm-1) 1686 and 1731 (CdO), 347þ4
(OH); HRMS (FABþ) m/z = 488.1990; Calcd for C32H28N3O3
488.1974.
[CuII(6)](ClO4) (CuII6). CuIILpH (50 mg, 0.066 mmol) was
dissolved in methanol (30 mL) and Et3N (2 equiv) was added to
the solution. The resulting solution was stirred for 3 days at 60
ꢀC, and then the solvent was reduced under reduced pressure to
give a brown residue. The addition of ether (100 mL) to the
residue gave a brown powder, which precipitated by allowing
the mixture to stand for several minutes. The supernatant was
then removed by decantation, and the remaining brown solid
was washed with ether three times and dried to give complex
CuII6 in 80%. FT-IR (KBr, cm-1): 1192 and 621 (ClO4-), 1665
(CdO); MS (ESIþ) m/z = 535.1; Calcd for C31H26N3O2Cu
535.1.
(ClO4)2 (CuIILmH). CuII(ClO4)2 6H2O (74 mg, 0.20 mmol) was
3
added to an acetonitrile solution (5 mL) of ligand LmH (92 mg,
0.20 mmol) and stirred for 15 min at room temperature. The
addition of ether (100 mL) to the solution gave a green powder
that was precipitated by allowing the mixture to stand for 30
min. The supernatant was then removed by decantation, and the
remained green solid was washed with ether three times and
dried to give complex CuIILmH in 75%. Single crystals of
CuIILmH were obtained by vapor diffusion of ether into an
acetone:CH3CN (5:1) solution of the complex. FT-IR (KBr,
cm-1): 1114 and 623 (ClO4-), 3374 (OH); HRMS (FABþ) m/z
= 520.1445; Calcd for C31H27N3OCu 520.1450; Anal. Calcd
for: [CuII(Lm)(CH3CN)](ClO4)2 H2O: C, 50.87; H, 4.14; N,
3
7.19. Found: C, 50.71; H, 3.92; N, 7.00.
[CuII(LpH)(CH3CN)](ClO4)2 (CuIILpH). This compound was
synthesized by following the same procedure as that described
for the preparation of CuIILmH, using LpH (92 mg, 0.20 mmol)
instead of LmH as a green powder in a 82% yield. Single crystals
of CuIILpH were obtained by vapor diffusion of ether into an
acetone:CH3CN (5:1) solution of the complex. FT-IR (KBr,
cm-1): 1094 and 626 (ClO4-), 3371 (OH); HRMS (FABþ) m/z
= 520.1474; Calcd for C31H27N3Ocu 520.1450; Anal. Calcd for
[CuII(LpH)(CH3CN)](ClO4)2 H2O 1/2CH3CN: C, 50.67; H,
Isolation and Structure Determination of Modified Ligand 6H.
After the reaction described above, the reaction mixture was
dissolved in CH2Cl2 (20 mL) and 25% NH3 aqueous solution
(10 mL) was added to the solution. The aqueous solution was
then extracted with CH2Cl2 (30 mL ꢀ 3), and the combined
organic layer was dried over Na2SO4. After the removal of
Na2SO4 by filtration, evaporation of the solvent gave a brown
3
3
4.05; N, 6.46. Found: C, 50.87; H, 4.24; N, 6.66.
[CuII(4)(OCH3)](ClO4) (CuII4). CuIILmH (50 mg, 0.066
mmol) was dissolved in methanol (30 mL), and then Et3N
(2 equiv) was added to the solution. The resulting solution was
stirred for 3 days at room temperature. After the reaction, the
solvent was removed under reduced pressure to give a brown
residue, to which ether (100 mL) was added. Brown powder was
gradually precipitated by allowing the mixture to stand for
several minutes. The supernatant was then removed by decanta-
tion, and the remaining brown solid was washed with ether three
times and dried to give complex CuII4 in 94%. FT-IR (KBr,
cm-1): 1108 and 622 (ClO4-); HRMS (FABþ) m/z = 595.1514;
Calcd for C33H30N3O4Cu 595.1532.
1
material. The H NMR spectrum of the resulting brown ma-
terial indicated that compound 6H was produced in 56% (NMR
yield based on CuIILpH using CH2Br2 as an internal standard).
1H NMR (400 MHz, CDCl3): δ 3.79 (2 H, s, H21), 3.93 (2 H, s,
H11), 3.94 (2 H, s, H1), 6.26 (2 H, d, J = 8.8 Hz, H8), 6.48 (1 H, s,
OH26), 6.72 (2 H. d, J = 8.8 Hz, H9), 7.09 (1 H, d, J = 7.6 Hz,
H5), 7.26-7.31 (1 H, m, H25), 7.36 (2 H, t, J = 8.4 Hz, H24),
7.40-7.42 (1 H, m, H20), 7.46 (2 H, d, J = 7.2 Hz, H23),
7.49-7.54 (3 H, m, H13 and H19), 7.59-7.64 (2 H, m, H3 and
H15), 7.68 (1 H, t, J = 7.6 Hz, H4), 7.74 (1 H, t, J = 7.8 Hz, H14),
8.00 (2 H, d, J = 8.4 Hz, H18); 13C NMR (100 MHz, CDCl3):
58.82 (C21), 59.30 (C1), 60.19 (C11), 70.78 (C7), 118.73 (C5
and C15), 120.95 (C13), 122.65 (C3), 126.87 (C18), 127.19
(C25), 127.85 (C8), 128.38 (C24), 128.67 (C19 and C23), 128.87
(C20), 137.13 (C14), 138.25 (C4), 138.88 (C22), 139.32 (C17),
150.81 (C9), 154.19 (C6), 156.65 (C16), 159.14 (C2 or C12),
159.25 (C2 or C12), 185.74 (C10). FT-IR (NaCl, cm-1) 1688
and 1733 (CdO), 3465 (OH); HRMS (FABþ) m/z = 488.1990;
Isolation and Structure Determination of Modified Ligands 4
and 5H. After the reaction described above, the mixture was
dissolved in CH2Cl2 (20 mL) and a 25% NH3 aqueous solution
(10 mL) was added to the solution. The aqueous solution was
then extracted with CH2Cl2 (30 mL ꢀ 3), and the combined
organic layer was dried over Na2SO4. After removal of Na2SO4
by filtration, evaporation of the solvent gave a brown material.
1
The H NMR spectrum of the resulting brown material indi-
cated that compound 4 was produced in 51% (NMR yield based
on CuIILmH, using CH2Br2 as an internal standard). A SiO2
column chromatographic treatment of the brown material
þ
Calcd for C32H28N3O3 488.1974. FT-IR (NaCl, cm-1) 1671

Article
Inorganic Chemistry, Vol. 50, No. 5, 2011 1637
(CdO), 3300 (OH); HRMS (FABþ) m/z = 474.2175; Calcd for
C32H28N3O2þ 474.2182.
work, spin-purified doublet energies were computed relative to
the quartet state through invocation of the standard Heisenberg-
Dirac Hamiltonian14 for two centers A and B, which has
eigenvalues of
Catalytic Oxidation of Benzylamine with CuII4. Benzylamine
(77 mg, 0.72 mmol) was added to an acetonitrile solution (5 mL)
of CuII4 (5 mg, 7.2 mmol) and the solution was stirred for 24 h at
room temperature. After the reaction, the solvent was removed
under reduced pressure to give a brown residue, which was then
dissolved in CH2Cl2 (10 mL) and treated with a 25% NH3
aqueous solution (5 mL). The aqueous solution was then extrac-
tedwithCH2Cl2 (15mLꢀ 3), andthe combinedorganiclayerwas
dried over Na2SO4. After the removal of Na2SO4 by filtration,
evaporation of the solvent gave a brown material, in which
ES ¼ - JABSðS þ 1Þ
ð1Þ
where JAB is the coupling constant between the two centers, S
takes on whole or half-integer values (ranging from |SA - SB| to
|SA þ SB|), and -SA(SA þ 1) - SB(SB þ 1) has been moved into
the zero of energy. For the case where center A is a phenoxyl
radical (SA = 0.5) and center B is a triplet-coupled Cu(II)-
superoxide (SB = 1.0), S = 0.5 for the doublet state and S = 1.5
for the quartet state, such that
1
formation of N-benzylidene benzylamine was confirmed by H
NMR. Yieldofthe imine productwas determined as 840%, based
on CuII4, using CHCl2CHCl2 as an internal reference; turnover
number of the catalyst is 8.4. 1H NMR (300 MHz, CDCl3): δ 4.83
(2 H, s, CH2), 7.26-7.43 (8 H, m), 8.40 (1 H, s, CH); MS (FABþ)
m/z = 196.2; Calcd for C32H25N3O3þ 196.1.
DE ¼ QE þ 3JAB
We compute JAB using the Yamaguchi equation:15
ð2Þ
Computational Methods. All geometries were fully optimized
at the M06-L level of density functional theory (DFT),8
using the Stuttgart [8s7p6d | 6s5p3d] ECP10MWB contracted
pseudopotential basis set on Cu9 and the 6-31G(d) basis set10 on all
other atoms. In addition, three uncontracted f functions, having
exponents of 5.100, 1.275, and 0.320, were placed on Cu. The
grid=ultrafine option (in Gaussian 0911) was chosen for integral
evaluation and an automatically generated density-fitting basis set
was used within the resolution-of-the-identity approximation for
the evaluation of Coulomb integrals. The nature of all stationary
points was verified by analytic computation of vibrational fre-
quencies, which were also used for the computation of zero-point
vibrational energies, molecular partition functions (with all fre-
quencies below 50 cm-1 replaced by 50 cm-1 when computing free
energies), and for determining the reactants and products asso-
ciated with each transition-state structure (by following the normal
modes associated with imaginary frequencies). Partition functions
were used in the computation of 298 K thermal contributions to
free energy, employing the usual ideal-gas, rigid-rotator, harmonic
oscillator approximation.12 Solvation effects associated with
methanol as a solvent were taken into account using the SMD
continuum solvation model.13
HSE - BSE
JAB ¼ - HS
ð3Þ
ÆS2æ - BSÆS2æ
where HSE is the energy of the single-determinantal high-spin
state at the low-spin geometry, BSE is the energy of the broken-
symmetry low-spin state, and ÆS2æ is the expectation value of the
total spin operator applied to the appropriate high-spin (HS) or
broken-symmetry (BS) determinant.
To predict the NMR 1H chemical shifts of possible products,
structures were fully optimized at the M06-2X/6-31þG(d,p)
level of theory,16 taking into account solvation effects using
the integral equation formalism17 of the polarized continuum
model18 (IEFPCM) for chloroform as a solvent. The NMR
shifts were calculated at the B3LYP/6-311þG(2d,p) level19 with
IEFPCM solvation. The molecular cavities for these calcula-
tions were constructed as a sum of atom-centered spheres, using
the radii of Bondi.20
Results and Discussion
Synthesis. Ligands LmH and LpH were prepared by
reductive coupling between compound 1 and compound
2p or 2m, using NaBH3CN as the reductant, followed by
deprotection of the methoxy group using BBr3 (see
Scheme 2).
The addition of triplet molecular oxygen to the doublet Cu(II)
complexes may, in principle, generate either doublet or quartet
products. Many of the doublet spin states were characterized by
weak coupling between two different spin centers (the phenoxyl
radical and a Cu(II) superoxide) and, therefore, are not well
described by a single determinant nor appropriately treated by
the standard Kohn-Sham density functional formalism. In this
The copper(II) complexes CuIILmH and CuIILpH were
prepared by treating the corresponding ligands with CuII-
(ClO4)2 6H2O in acetonitrile, and CuIILo was synthesized
3
as described in our previous study.7 Figure 3 shows the
crystal structures of CuIILmH and CuIILpH. The crystal-
lographic data of the copper(II) complexes, together with
selected bond lengths and angles, are summarized in
Tables 1 and 2, respectively. The crystal structure
of CuIILo has already been reported in our previous
(8) Zhao, Y.; Truhlar, D. G. J. Chem. Phys. 2006, 125, 194101–194118.
(9) Dolg, M.; Wedig, U.; Stoll, H.; Preuss, H. J. Chem. Phys. 1987, 86,
866–872.
(10) Hehre, W. J.; Radom, L.; Schleyer, P. v. R.; Pople, J. A. Ab Initio
Molecular Orbital Theory; Wiley: New York, 1986.
(11) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson,
G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.;
Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.;
Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.;
Nakai, H.; Vreven, T.; Montgomery, J. A.; , Jr., Peralta, J. E.; Ogliaro, F.;
Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.;
Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.;
Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.;
Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts,
R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.;
Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth,
G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas,
O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09,
Revision A.02; Gaussian, Inc.: Wallingford, CT, 2009.
(14) (a) Heisenberg, W. Z. Physik 1928, 49, 619. (b) Cramer, C. J.; Truhlar,
D. G. Phys. Chem. Chem. Phys. 2009, 11, 10757–10816.
(15) (a) Yamaguchi, K.; Jensen, F.; Dorigo, A.; Houk, K. N. Chem. Phys.
Lett. 1988, 149, 537–542. (b) Soda, T.; Kitagawa, Y.; Onishi, T.; Takano, Y.;
Shigeta, Y.; Nagao, H.; Yoshioka, Y.; Yamaguchi, K. Chem. Phys. Lett. 2000,
319, 223–230.
(16) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215–241.
(17) Cances, E.; Mennucci, B.; Tomasi, J. J. Chem. Phys. 1997, 107, 3032–
ꢀ
3041.
(18) Miertus, S.; Scrocco, E.; Tomasi, J. Chem. Phys. 1981, 55, 117–129.
(19) (a) Becke, A. D. Phys. Rev. A 1988, 38, 3098–3100. (b) Lee, C.; Yang,
W.; Parr, R. G. Phys. Rev. B 1988, 37, 785–789. (c) Becke, A. D. J. Chem. Phys.
1993, 98, 5648–5652. (d) Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch,
M. J. J. Phys. Chem. 1994, 98, 11623–11627.
(12) Cramer, C. J. Essentials of Computational Chemistry: Theories and
Models, 2nd ed.; John Wiley & Sons: Chichester, U.K., 2004.
(13) Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. J. Phys. Chem. B
2009, 113, 6378–6396.
(20) Bondi, A. J. Phys. Chem. 1964, 68, 441–451.

1638 Inorganic Chemistry, Vol. 50, No. 5, 2011
Tabuchi et al.
Scheme 2
Table 1. Summary of the X-ray Crystallographic Data of CuIILmH and CuIILpH
CuIILmH
CuIILpH
formula
formula weight
C33H30N4Cl2O9Cu
761.07
C33H30N4Cl2O9Cu
761.07
crystal I
space group
a
b
c
monoclinic
C2/c (#15)
28.313(11) A
monoclinic
C2/c (#15)
23.3974(11) A
16.2969(7) A
˚
17.6996(9) A
˚
˚
˚
˚
8.669(2) A
29.689(11) A
˚
β
V
112.927(18)ꢀ
6711(4) A
99.6154(5)ꢀ
6654.1(5) A
3
3
˚
˚
Z
F(000)
Dcalcd
T
crystal size
8
3128.00
8
3128.00
1.506 g/cm3
153 K
1.591 g/cm3
113 K
0.40 mm ꢀ 0.10 mm
ꢀ 0.10 mm
8.705 cm-1
55.0ꢀ
0.40 mm ꢀ 0.20 mm
ꢀ 0.20 mm
8.780 cm-1
54.9ꢀ
μ(Mo KR)
2θmax
no. of reflns
30720
31950
measd
no. of reflns
obsd
7649
([I > 2.00σ(I)])
479
0.0470
0.1264
0.912
7572
(all reflections)
498
0.0706
0.2443
0.977
no. of variables
R1a
wR2b
goodness of fit, GOF
P
P
P
P
a R1 = ||Fo| - |Fc||/ |Fo|. b wR2 = [ (w (Fo2 - Fc2)2/ Fo2)2]1/2
.
Both CuIILmH and CuIILpH exhibit monomeric struc-
tures that have distorted square pyramidal structures,
where the equatorial positions are occupied by the three
nitrogen atoms N(1), N(2), N(3) from the ligand and
an additional nitrogen atom from the external ligand
CH3CN; one of the oxygen atoms of the ClO4 counter-
anion occupies the axial position. The τ values of
CuIILmH and CuIILpH are 0.127 and 0.065, respec-
tively.21 Apparently, there is no direct interaction be-
tween the phenol group and the copper(II) ion in either
complex, in contrast to the case of CuIILo. In both
CuIILmH and CuIILpH, the phenol oxygen atoms are
Figure 3. ORTEP drawings of (A) CuIILmH and (B) CuIILpH showing
50% probability thermal ellipsoids. Hydrogen atoms are omitted for
clarity.
Communication, and it has direct coordination of the
deprotonated phenolic oxygen to copper(II), as indicated
in Chart 2.7
(21) Addison, A. W.; Rao, T. N.; Reedijk, J.; van Rijn, J.; Verschoor,
G. C. J. Chem. Soc., Dalton Trans. 1984, 1349–1356.

Article
Inorganic Chemistry, Vol. 50, No. 5, 2011 1639
Table 2. Selected Bond Lengths and Bond Angles of CuIILmH and CuIILpH
Compound CuIILmH
Table 3. ESR Parameters of CuIILo, CuIILmH, and CuIILpHa
ESR Parameters
g^
complex
g//
A//
Bond Lengths
˚
CuIILo
2.231
2.429
2.429
2.056
2.093
2.093
174 G
107 G
107 G
˚
2.0267(16) A
Cu(1)-N(1)
Cu(1)-N(3)
Cu(1)-O(2)
2.0230(17) A Cu(1)-N(2)
CuIILmH
CuIILpH
˚
˚
1.962(2) A
2.008(2) A
Cu(1)-N(4)
˚
2.491(2) A
a In CH3OH at 77 K; microwave frequency = 9.03 GHz, modulation
frequency = 100 kHz, modulation amplitude = 3 G, microwave power
= 1.00 mW.
Bond Angles
N(1)-Cu(1)-N(2) 161.86(9)ꢀ
N(1)-Cu(1)-N(4) 97.26(8)ꢀ
N(2)-Cu(1)-N(4) 99.98(8)ꢀ
N(1)-Cu(1)-N(3) 81.36(8)ꢀ
N(2)-Cu(1)-N(3) 82.71(8)ꢀ
N(3)-Cu(1)-N(4) 169.38(7)ꢀ
Compound CuIILpH
Bond Lengths
˚
˚
Cu(1)-N(1)
Cu(1)-N(3)
Cu(1)-O(3)
2.020(4) A
Cu(1)-N(2)
2.021(4) A
˚
˚
1.956(4) A
2.008(3) A
Cu(1)-N(4)
˚
2.510(3) A
Bond Angles
N(1)-Cu(1)-N(2) 164.13(17)ꢀ N(1)-Cu(1)-N(3) 82.14(17)ꢀ
N(1)-Cu(1)-N(4) 95.90(17)ꢀ
N(2)-Cu(1)-N(4) 99.64(16)ꢀ
N(2)-Cu(1)-N(3) 83.31(16)ꢀ
N(3)-Cu(1)-N(4) 168.02(16)ꢀ
Chart 2
Figure 4. Ultraviolet-visible light (UV-vis) spectrum of CuIILo (0.5
mM) under O2 atmosphere in methanol at 60 ꢀC in a UV cell with a path
length of 1.0 cm.
Reactivity of Copper(II) Complexes toward O2. To
examine the redox interaction between the copper(II)
ion and the phenolate moiety of the ligand, the reactions
of the copper(II) complexes and O2 were studied in
methanol in the presence of Et3N as a base; in the case
of CuIILo, Et3N was not added to the reaction solution,
since, in that instance, the phenol proton of the ligand is
already deprotonated in the starting complex.
disordered, being found partially at both the meta- and
para-positions of the phenyl groups of the ligand side
arms.
O2-Reactivity of CuIILo. The copper(II) complex of
L
o- (CuIILo) exhibits an intense phenolate to copper(II)
LMCT band at 380 nm (ε = 3650 M-1 cm-1), together
with a shoulder near 470 nm and a weak d-d band at 619
nm (ε = 140 M-1 cm-1) in methanol under anaerobic
conditions.22 Introduction of O2 into the solution hardly
affected the spectrum at 60 ꢀC for several hours, demon-
strating that complex CuIILo was stable under an O2
atmosphere (Figure 4). In the ESI-MS of the final reac-
tion solution, there was a set of peaks at 519.4, corre-
sponding to the starting material [CuII(Lo)]þ (calcd value:
519.1) after exposure to O2. These results clearly indicate
that this complex, with direct coordination of phenolate
to copper(II), exhibits almost no reactivity toward mo-
lecular oxygen under the experimental conditions.
ESR spectra of the copper(II) complexes have been
measured in CH3OH at 77 K, as shown in Figure S1 in the
Supporting Information, and the ESR parameters of each
complex are summarized in Table 3. Complex CuIILo
exhibits a typical spectrum of copper(II) with a tetragonal
geometry (see Figure S1A in the Supporting In-
formation). Double integration of the ESR spectrum of
CuIILo indicated that more than 95% spin persisted in
solution, demonstrating a monomeric structure of CuIILo
in solution. Notably, complexes CuIILmH and CuIILpH
showed effectively identical ESR spectra with a relatively
small A// value (107 G, as seen in Figures S1B and S1C in
the Supporting Information). Thus, it can be said that the
OH groups at the meta- and para-positions of the phenyl
substituent in CuIILmH and CuIILpH do not interact with
the copper(II) ion in solution, in agreement with what is
seen in the crystal structures. The relatively small A//
value of the complexes CuIILmH and CuIILpH suggests
that the geometry of the complexes changes from dis-
torted square pyramidal in the crystal to distorted trigo-
nal bipyramidal in solution. Spin quantification of
CuIILmH and CuIILpH also demonstrated that both
complexes exist in a monomeric form in solution, again
as in the case of the crystals.
O2-Reactivity of CuIILmH. In contrast to the above o-
phenol system CuIILo, which does not show any reactivity
toward O2, treatment of CuIILmH in methanol with 2
equiv of Et3N at 60 ꢀC under an O2 atmosphere gave a
spectral change, as shown in Figure 5, where an intense
absorption band at 410 nm and a weak broad band at
550-700 nm gradually appeared ((8.7 ( 0.9) ꢀ 10-5 s-1).
(22) (a) Jazdzewski, B. A.; Tolman, W. B. Coord. Chem. Rev. 2000,
200-202, 633–685. (b) Itoh, S.; Taki, M.; Fukuzumi, S. Coord. Chem. Rev.
2000, 198, 3–20.

1640 Inorganic Chemistry, Vol. 50, No. 5, 2011
Tabuchi et al.
proton Hc on the pyridine nucleus (see Scheme 3 and
Figure S3 in the Supporting Information). This confirms
the close proximity of the methoxy proton (-OCH3) and
the pyridine nucleus, and it rules out an alternative
structure 40, in which the -OCH3 group and the olefinic
proton Hb are sufficiently close that an NOE should be
observed between them. In addition, irradiation of Hc at δ
= 7.17 gave a NOE to the aromatic proton Hd on the
pyridine nucleus at δ = 7.73 and the methoxy protons at δ
= 3.71, but showed no NOE to the olefinic protons at δ
= 6.75. This observation rules out structure 400, in which
Hc and olefinic proton Ha should be close enough in some
conformations to exhibit a NOE. Finally, the chemical
shifts of Ha and Hb were estimated using ChemBioDraw
Ultra (version 12.0); expert-system predictions are δ =
6.87 and 6.87 for 4, 5.89 and 8.09 for 40, and 7.23 and 5.89
for 400. Obviously, the best match to the experimental data
(δ = 6.75) is for 4. Thus, the HRMS, IR, and 1H NMR
data, including NOE, support the p-quinone structure of
4. The isolated compound 4 exhibited an absorption
spectrum similar to the final spectrum shown in Figure 5,
exhibiting an intense absorption band at ∼410 nm and a
weak broad band at 550-700 nm assignable to π-π* and
n-π* transitions of the p-quinone moiety, respectively.
The absorption spectrum resembles those of the TPQ
cofactors,1-3 further supporting formation of the p-qui-
none derivative 4.
Figure 5. Spectral change of the aerobic reaction of CuIILmH (0.5 mM)
in methanol containing Et3N (1.0 mM) at 60 ꢀC in a UV cell with a path
length of 1.0 cm. Inset: The time course of the absorption change at 410 nm.
Such a spectral change was not observed in the absence of
O2 (anaerobic conditions), clearly demonstrating that O2
is required for the formation of the 410-nm chromophore.
Also note that no distinct intermediate was observed
during the course of the reaction, as seen in Figure 5.
This may indicate that, in the presence of 2 equiv of Et3N,
the initial deprotonation step from the phenol moiety of
CuIILmH is the rate-limiting step in the overall reaction
pathway, as discussed below (cf. Scheme 4).
The ESI-MS of the final reaction mixture shown
in Figure 6 exhibits a set of peaks starting at 595.1
(Figure 6A), the mass number and isotope distribution
pattern of which are identical to those of the simulation
spectrum for a copper(II) complex containing modified
ligand 4 (see Chart 3) and methoxide (OCH3) (Figure 6C).
Isotope labeling experiments using 18O2 (purity 95%,
Cambridge Isotope Laboratories, Inc.) gave the ESI-
MS result shown in Figure 6B, where the main peak
appeared at 597.0 (a two-mass-unit shift). This clearly
demonstrates that one of the oxygen atoms incorporated
into 4 originates from molecular oxygen. Computer simu-
lation of the spectrum shown in Figure 6C (Figure 6D)
indicates that the content of 18O is 69%.
A second isolated product from chromatography, 5H,
exhibited m/z = 488.1990 in the HRMS analysis (calcd
for [5H þ H]þ, 488.1896), and two carbonyl CdO stretch-
ing vibration peaks at 1686 and 1731 cm-1, together with
a broad intense absorption band at 3474 cm-1 due to the
1
OH group in the IR spectrum. In the H NMR shown
in Figure S4 in the Supporting Information, the methoxy
protons found in the spectrum of compound 4 are absent
in 5H. However, the two doublet peaks at δ = 6.71 and
6.83 indicate that the two olefinic protons on the modified
ligand still remain. Furthermore, two carbonyl carbons
are detected at 183.5 and 184.3 ppm in the 13C NMR
spectrum shown in Figure S5 in the Supporting Informa-
tion. On the basis of these spectral data, together with the
detailed 2D NMR analyses (HSQC, HH-COSY, and
HMBC) shown in Figure S6 in the Supporting Informa-
tion, the structure of compound 5H has been secured and
To confirm the structure of the modified ligand, the
organic products were isolated by an ordinary workup of
the final reaction mixture using aqueous NH3 solution
(demetallation) and subsequent chromatography on sili-
ca (see the Experimental Section). Two organic products,
4 and 5H, were isolated and identified as indicated below.
Compound 4 exhibited m/z = 502.2159 in the HRMS
analysis (calcd for [4 þ H]þ, 502.2123), and two carbonyl
CdO stretching vibration peaks at 1672 and 1727 cm-1 in
the IR spectrum. In the 1H NMR spectra of 4, shown in
Figure S2 in the Supporting Information, there is a sharp
singlet at δ = 3.71 that has a peak area corresponding to
the three protons of the methoxy group (-OCH3). An-
1
all the H and 13C NMR peaks have been completely
assigned, as indicated in the insets of Figures S4 and S5 in
the Supporting Information. Since compound 5H was not
contained in the crude reaction mixture prior to SiO2
column chromatography (as judged by 1H NMR), com-
pound 5H must be generated from compound 4 during
chromatography.23
A possible mechanism for the formation of CuII4 from
CuIILmH is presented in Scheme 4. Treatment of CuII4
with triethylamine (2 equiv) in methanol may induce
deprotonation of methanol and the phenolic proton of
the ligand to give, after coordination of methoxide,
1
other characteristic in the H NMR is the existence of a
doublet of doublets in an ab quartet pattern centered at
δ = 6.75, which clearly indicates the existence of two
olefinic protons strongly coupled to each other (see
Figure S2 in the Supporting Information). Further struc-
tural information for 4 was obtained by NOE measure-
ments, as indicated in Scheme 3. Irradiation of the
methoxide peak at δ = 3.71 provided an NOE (1.3%)
only on a peak at δ = 7.17, which is assigned as aromatic
(23) The NMR data of compound 5H are slightly different from those of
essentially the same compound (L2H) reported in our previous Commu-
nication (ref 6e). Detailed analysis of the FAB-MS data of the present
compound (5H) indicated that there is a small peak at 504.2, which
corresponds to a molecular mass of [5H þ H2O]. Thus, in the present
compound, 5H may bind one molecule of H2O in the ligand cavity, which
influences the NMR chemical shifts of the compound.

Article
Inorganic Chemistry, Vol. 50, No. 5, 2011 1641
Figure 6. ESI-MS spectra of (A) [CuII(4)(OCH3)]þ (CuII4) generated using 16O2 and (B) [CuII(4)(OCH3)]þ (CuII4) generated using 18O2. Simulated
spectra of these products are shown in panels C and D, respectively.
Scheme 3
Scheme 4
intermediate A. Either concomitantly or subsequently,
intramolecular electron transfer from the generated phe-
nolate moiety to copper(II) ion may occur to generate a
copper(I)-phenoxyl radical species B, to which molecular
oxygen may be added to generate an alkylperoxo-type
intermediate C, where the peroxide group bridges the
para-position of the aromatic substituent and the copper
ion. Heterolytic cleavage of the O-O bond in C generates
p-quinone intermediate D. In subsequent steps, Michael
addition of methoxide and aerobic oxidation of the
reduced quinone occur, to deliver the final product CuII4.
To gain further insight into the precise mechanistic
pathway and the nature of the various potential inter-
mediates, density functional theory (DFT) calculations

1642 Inorganic Chemistry, Vol. 50, No. 5, 2011
Tabuchi et al.
Chart 3
to B is beyond the scope of the present study, because the
complex dynamics of electron transfer and tight ion-pair
dissociation in a polar solvent are both involved and
could likely be addressed only by explicit quantum me-
chanical simulations. Nevertheless, it is apparent that
both of the two diabatic states represented by structures
A and B in Scheme 4 are energetically accessible under the
experimental conditions.
It may be that deprotonation of the phenol moiety of
CuIILmH to give A is the rate-determining step in the
presence of a small amount of the base (2 equiv of Et3N)
in the overall reaction pathway, as mentioned previously.
If the deprotonation step were fast enough to generate a
sufficient amount of phenolate A, a copper(I)-phenoxyl
radical species (intermediate B) should be detected in the
initial stage of the reaction. However, the formation of
such an organic radical species could not be detected
by ESR and UV-vis. When a large excess of Et3N
(<100 equiv) was added to enhance the deprotonation
step, a complicated mixture of oligomeric products was
obtained. This may be due to the C-C coupling reactions
of the phenoxyl radical intermediate B generated in
solution. Thus, to suppress such an undesirable side
reaction (C-C coupling reaction), the addition of a
limited amount of Et3N (2 equiv) was necessary (see
Figure 5).
Next, we examined the coordination of molecular
oxygen to the copper atom. We successfully located four
low- to moderate-energy conformational isomers for each
of the doublet and quartet electronic states that may be
generated by the addition of triplet molecular oxygen to
doublet B. The lowest-energy such structure, quartet ii, is
shown in Figure 7 (see also Figure 8 for a traditional
reaction coordinate diagram, but bearing in mind that
each intermediate and TS structure is more properly
viewed as a population of related but different stereo-
structures, so that a single reaction coordinate curve, as
opposed to a family of many such curves, is somewhat of a
simplification). Structure ii is predicted to have a free
energy of 2.9 kcal/mol, relative to infinitely separated iB
and molecular O2; the other quartet conformers have free
energies ranging from 3.3 kcal/mol to 8.0 kcal/mol. The
doublet conformers are structurally quite similar to the
quartet conformers and are predicted to have free en-
ergies that are ∼4.8-9.6 kcal/mol, relative to infinitely
separated iB and molecular O2. Within the accuracy of
our DFT model, the two electronic states may be con-
sidered to be degenerate, which is consistent with very
weak coupling between what spin density calculations
indicate to be a phenoxyl radical and a triplet end-on
Cu(II)-superoxide complex. The triplet end-on Cu(II)-
superoxide character of the CuO2 fragment is consistent
with expectations, given the nature of the supporting
ligands coordinated to the metal atom.24
were undertaken (see the above Computational Methods
section for full details). To begin, the uncharged mole-
cule with a deprotonated phenol and a methoxide ligand
coordinated to the copper atom was chosen. Because of
the distorted tetragonal coordination of the copper
atom and the asymmetry of the phenol group, several
conformational isomers were considered and four dis-
tinct structures were found as local minima, differing in
(a) the absolute configuration of the benzyl-substituted
nitrogen atom (which will readily invert at experimental
temperatures) and (b) the rotational orientation of the
phenol group; these conformers had relative free ener-
gies of 0.0, 0.2, 6.8, and 7.8 kcal/mol (all free energies
presented in this discussion include the effects of con-
tinuum methanol solvation). While all structures and
energies are provided in the Supporting Information, in
the interests of brevity, here, we will focus our discussion
on the theoretical results for the lowest-energy species;
the lowest-energy conformer iB may be found in Figure 7.
[For discussion of the microscopic mechanism, we have
adopted lowercase roman numerals for the individual
intermediates and transition-state (TS) structures. In
cases where those structures correspond to species in
Scheme 4, we add, as a subscript, the capital letter
corresponding to the labeling in Scheme 4. Thus, for
example, iB indicates the correspondence between opti-
mized intermediate i and structure B.]
The electronic structure of intermediate iB, which has a
doublet spin state, is best described as a Cu(I) complex
with the unpaired spin density located primarily on the
phenoxyl fragment; in particular, the spin densities on the
O, C(2), C(4), and C(6) positions of the phenoxyl ring
(taking C(3) as the position of substitution of the attached
pyridine) were computed to be 0.22, 0.17, 0.18, and 0.14 a.
u., respectively. That is, DFT predicts the ground electro-
nic state of this complex to correspond to B in Scheme 4.
In this respect, structure A in Scheme 4 is formally an
excited electronic state, relative to B. To better under-
stand the nature of the complex immediately after the
deprotonation step, and to further illuminate the relation-
ship between A and B (Scheme 4), complex i was also
optimized in the presence of the ammonium ion generated
directly from deprotonation of the phenol (including
methanol continuum solvation). Population analysis
clearly shows that, in the presence of the ammonium
ion, the ground-state electronic structure of the tight ion
pair is best represented by A in Scheme 4;the unpaired
spin is essentially entirely found in a Cu d orbital.
TDDFT calculations indicate that the lowest-energy
electronic transition in the tight ion pair is indeed that
corresponding to intramolecular electron transfer to gen-
erate B. Computation of the rate of reaction for A going
Following characterization of the stable oxygen ad-
ducts, we proceeded to optimize TS structures for the
addition of the distal oxygen of the O2 fragment to the
phenoxyl ring. We found 12 such structures, differing in
conformation and the position of attack on the phenoxyl
(24) (a) Cramer, C. J.; Tolman, W. B. Acc. Chem. Res. 2007, 40, 601–608.
(b) Cramer, C. J.; Gour, J. R.; Kinal, A.; Wtoch, M.; Piecuch, P.; Moughal Shahi,
A. R.; Gagliardi, L. J. Phys. Chem. A 2008, 112, 3754–3767.

Article
Inorganic Chemistry, Vol. 50, No. 5, 2011 1643
Figure 7. Ball-and-stick stereostructures for selected stationary points on the mechanistic pathway leading to 4, as computed at the M06-L level of theory.
For clarity, hydrogen atoms are not shown, except at the substituted aromatic position after electrophilic attack by oxygen. Hydrogen atoms are white,
carbon atoms are gray, nitrogen atoms are blue, oxygen atoms are red, and copper atoms are latte.
Figure 8. Reaction coordinate diagram for steps along the path from iB to phenol ring oxygenation. Density functional theory (DFT) energies, including
continuum solvation, are indicated in units of kcal/mol; see Figure 7 for larger structural images.

1644 Inorganic Chemistry, Vol. 50, No. 5, 2011
Tabuchi et al.
Scheme 5
ring. All TS structures were predicted to have doublet
spin states, which is consistent with the coupling of two
spins in the forming C-O bond and dominant Cu(II)
character, as illustrated for intermediate C in Scheme 4.
Generally, we found additions to C(6) to be least favored,
locating 4 structures with activation free energies, relative
to separated iB and O2, ranging from 13.0 kcal/mol to 27.6
kcal/mol, and additions to C(2) and C(4) to be roughly
equally favorable, with 4 structures in each case having
activation free energies of 9.2-17.0 kcal/mol in the case of
C(2) and 9.1-15.4 kcal/mol in the case of C(4). One such
TS structure, iii, is shown in Figure 7; addition to C(4) has
an activation free energy of 9.9 kcal/mol.
Figure 9. Spectral change for the aerobic reaction of CuIILpH (0.5 mM)
in methanol containing Et3N (1.0 mM) at 60 ꢀC in a UV cell with a path
length of 1.0 cm. Inset shows the time course of the absorption change at
380 nm.
The intermediates that result from C-O bond forma-
tion (i.e., those structures analogous to C in Scheme 4)
were computed to have free energies, relative to separated
iB and O2, ranging from -5.4 kcal/mol to 13.5 kcal/mol.
The largest free energy of activation for the reverse
reaction from any adduct (i.e., breaking the newly formed
C-O bond) is 15.3 kcal/mol, which corresponds to the
reverse reaction of intermediate ivC, which is the lowest-
energy adduct (Figure 7), passing back through structure
iii.
energy in every case for O-O bond breaking, compared
to C-O bond formation, indicates that it is the O-O
scission step that is product-determining in the phenol
oxygenation process, so that the relative stabilities of the
preceding intermediates and the ease of their formation
does not dictate the final position of oxygenation. Theory
does not address whether this step is also rate-determin-
ing, because the activation free energies of the (possibly
coupled) deprotonation and electron-transfer steps lead-
ing to iB have not been quantified.
The next step in the microscopic mechanism that we
considered was scission of the O-O bond. For 9 of the 12
intermediates, we successfully located TS structures for
this bond breakage. The TS structures fell into two
classes, differing in the magnitude of the predicted activa-
tion free energies. In the lower-energy class, the geometry
of the adduct permits a hydrogen-atom transfer from the
newly oxygenated position to the departing oxygen atom
at the same time that O-O bond breaking is taking place.
That is, the direct product of the reaction is a quinone
and a Cu(II) hydroxide (analogous to structure D in
Scheme 4). In the higher-energy class, such a concerted
hydrogen-atom transfer is geometrically impossible, and
the resulting product is instead a high-energy species
containing a copper-oxo fragment and an open-shell
monohydrogenated quinone. The TS structures that we
were unsuccessful in locating all corresponded to struc-
tures that would fall into the latter, high-energy class, and
they are presumably not readily located, because of the
high energy of the resulting products. Figure 7 illustrates
one TS structure from each class: v (leading to copper
hydroxo) and vi (leading to copper oxo).
In the case of the lower-energy class of TS structures,
bond scission for a C(2) adduct, to generate an ortho-
quinone, is predicted to have an activation free energy
(again, relative to infinitely separated iB and O2) of 15.6
kcal/mol, and an analogous TS structure for a C(4)
adduct (v), to generate a para-quinone, is predicted to
have an activation free energy of 17.7 kcal/mol. Within
the accuracy of the model, these two pathways are pre-
dicted to be about equally likely. Activation free energies
for all other TS structures for this step range from 22.5
kcal/mol to 36.1 kcal/mol. The larger activation free
The remaining steps in the proposed mechanism;
namely, Michael addition to the quinone and sub-
sequent oxidation of the product;were not assessed with
DFT. The observation of 4 as a product and not 40 is not
entirely consistent with the theoretical prediction that TS
structures leading to the precursor para- and ortho-qui-
nones, respectively, have roughly equal activation free
energies. However, mass recovery of crude 4 is only 51%
(vide supra), so this discrepancy may reflect instability of
the missing isomer to subsequent reaction or workup
conditions.
With respect to the assignment of the structure of
compound 5H, we also computed the 1H chemical shifts
for geometries of 5H and 5H00 (the latter being the
hydroxy analogue of 400 shown in Scheme 3; 5H0, the
corresponding analog of 40, was not considered, because it
is a tautomer of 5H and would be expected to readily
equilibrate). The B3LYP/6-311þG(2d,p)//M06-2X/6-
31þG(d,p) predictions for the olefinic protons in 5H are
6.63 and 6.76 ppm, which is in near-quantitative agree-
ment with the experimental values of 6.71 and 6.83. The
predicted shifts for 5H00, on the other hand, are 5.92 and
8.20 ppm. These data provide further confirmation for
the structural assignment of 5H. As noted previously,
compound 5H is likely generated from compound 4 via a
SiO2-assisted hydrolysis of the methoxy group during the
column chromatographic treatment of compound 4 (see
Scheme 5).
O2 Reactivity of CuIILpH. Complex CuIILpH also
reacted in methanol at 60 ꢀC in the presence of NEt3
(2 equiv), as shown in Figure 9. Although the time course
of the absorption change is not simple (see inset of Figure 9),

Article
Inorganic Chemistry, Vol. 50, No. 5, 2011 1645
Figure 10. ESI-MS spectra of (A) [CuII(6)]þ (CuII6) generated using 16O2 and (B) [CuII(6)]þ (CuII6) generated using 18O2. Simulated spectra of these
products are shown in panels C and D, respectively.
Chart 4
the absorption band at ∼380 nm gradually increases. In
this case as well, no spectral change was observed in the
absence of O2 (anaerobic conditions), demonstrating that
O2 is involved in the reaction as a substrate. The ESI-MS
spectrum shown in Figure 10A exhibits a set of peaks
starting at 535.1, which clearly indicates incorporation of
one oxygen atom into the product (see Figure 10C, calcd
value: 535.13). Isotope labeling experiment using 18O2
(purity 95%, Cambridge Isotope Laboratories, Inc.) gave
an ESI-MS shown in Figure 10B, where the main peak
appears at 537.1 (a two mass unit shift). This clearly
demonstrates that one of the oxygen atoms incorporated
into 6 also originates from molecular oxygen. Computer
simulation of the spectrum shown in Figure 10C
A possible mechanism for the formation of complex
CuII6 from CuIILpH is shown in Scheme 6.
Deprotonation of CuIILpH produces CuIILp (A in
Scheme 6), within which intramolecular electron transfer
occurs from the generated phenolate moiety to the
copper(II) ion to provide intermediate B. Next, molecular
oxygen adds to B to generate an alkylperoxo-type
(Figure 10D) indicates that the content of 18O is 70%.
The organic product 6H was isolated by ordinary
demetallation of the product complex using NH4OH
and subsequent SiO2 column chromatography, as in the
1
former reaction. In the H NMR spectrum shown in
Figure S7 in the Supporting Information, there are two
doublet peaks at δ = 6.26 and 6.72, which couple to each
other with a coupling constant J = 8.8 Hz. Judging from
the integral ratio of these doublets, each peak involves
two protons. Furthermore, a singlet peak at 6.48 disap-
pears when D2O is added to the 1H NMR sample,
indicating that this peak is due to an acidic (easily
exchanged) proton, such as an OH proton. In the 13C
NMR shown in Figure S8 in the Supporting Information,
there is only one peak at 185.9, which is ascribable to a
carbonyl carbon. Furthermore, in the IR spectrum, there
is an intense peak at 1671 cm-1, which is ascribable to a
CdO stretching vibration, and a broad OH absorption
peak at 3300 cm-1. On the basis of these spectral data,
together with the detailed 2D NMR analyses (HSQC,
HH-COSY, and HMBC) shown in Figure S9 in the
Supporting Information, the structure of organic product
6H has been assigned as the keto-alcohol derivative
intermediate C, where the peroxide group bridges the
C7 ipso carbon of the phenol ring and the copper ion.
Electronically, both C7 and C9 might be expected to be
available for formation of the peroxide bridge with the
copper(II) ion, but only product from the C7 adduct is
isolated. Ultimately, oxidation of the coordinated meth-
oxide by the generated peroxide may occur to give CuII6
and formaldehyde (HCHO). In this case as well, the
deprotonation step from CuIILpH may be the rate-limit-
ing process, since no distinct intermediate was detected in
the UV-vis spectrum (see Figure 9); however, the mech-
anistic details of this process have yet to be established.
Catalytic Oxidation of Benzylamine. The catalytic oxi-
dation of benzylamine was briefly examined using the
quinone complexes CuII4 in acetonitrile at room tempera-
ture (see Scheme 7). N-Benzylidene benzylamine was ob-
tained as the oxidation product in 840% yield, based on
the quinone catalyst (i.e., multiple turnovers are obser-
ved; turnover number of the catalyst is 8.4). Thus, the
quinone-copper(II) complexes generated from the phe-
nol derivatives actually act as turnover catalysts for amine
oxidation. Although the mechanistic details have yet to be
1
shown in Chart 4, and all the H and 13C NMR peaks
have been completely assigned as indicated in Figures S7
and S8 in the Supporting Information, respectively.

1646 Inorganic Chemistry, Vol. 50, No. 5, 2011
Tabuchi et al.
Scheme 6
Scheme 7
netic resonance (NMR) analyses, electrospray ionization-
mass spectroscopy (ESI-MS) (including an 18O-labeling
experiment), and infrared (IR) spectroscopy. The density
functional theory (DFT) calculations are consistent with a
mechanistic pathway that begins with an exergonic intramo-
lecular electron transfer from the deprotonated phenol
(phenolate) to copper(II) in A (shown in Scheme 4), to
generate a copper(I)-phenoxyl radical species B. Molecular
oxygen is added to B to form an end-on superoxo copper(II)
species ii, shown in Figure 7, from which C-O bond forma-
tion through electrophilic attack on the aromatic ring occurs
to give a copper(II)-alkylperoxo type intermediate C. From
this intermediate, O-O bond cleavage and proton migration
proceeds in a concerted manner to provide a para-quinone
derivative D. DFT suggests that a stepwise O-O bond
cleavage to generate a formal Cu(III)-oxo species, followed
by subsequent proton migration, is energetically much less
favorable in the reaction pathway from A to D (see Figure 8).
The final step is Michael addition of a methoxide on the
copper(II) ion to the para-quinone ring and subsequent
O2 oxidation. These reaction steps are very close to those
proposed for the biosynthesis of TPQ cofactor in the enzy-
matic system shown in Scheme 1. Thus, the present reaction
can be regarded as a good model for TPQ biosynthesis,
providing important insights into the nature of postulated
intermediates.
However, there is a critical difference between our model
reaction and the enzymatic reaction. Namely, under our
present reaction conditions, using a small amount of triethy-
lamine base (2 equiv), deprotonation from the phenol moiety
of CuIILmH is rate-limiting in regard to the overall reaction
pathway. Thus, once the phenolate is generated from the
neutral phenol by deprotonation, the subsequent reactions
(electron transfer, generating intermediate B; O2 addition to
make alkylperoxide derivative C; O-O bond cleavage to give
the p-quinone derivative D; and Michael addition of metha-
nol to form the final product) may proceed smoothly. Thus,
any intermediates were not detected in our model system (see
Figure 5). If an excess amount of the base (<100 equiv) was
added to the methanol solution of CuIILmH, a complicated
mixture of oligomeric products was obtained. This may be
examined in detail, the reaction may involve a trans-amina-
tion mechanism, as suggested for amine oxidation by the
model compounds of cofactors TPQ, PQQ (pyrroloquino-
linequinone), and TTQ (tryptophan tryptophylquinone).25
Summary
In this study, copper(II) complexes supported by a series of
redox active phenol ligands LoH, LmH, and LpH (shown in
Chart 1) have been synthesized, and their O2 reactivity has
been explored in detail, to gain mechanistic insights into the
biosynthetic processes that lead to the novel organic TPQ
cofactor (2,4,5-trihydroxyphenylalanine quinone, TOPA
quinone) in copper-containing amine oxidases.
The copper(II)-phenolate complex CuIILo (ortho-phenol
derivative) involves a strong coordination of the phenolate
oxygen to copper(II) (Cu-O distance is 1.891 A), and it is
stable under O2 in CH3OH at 60 ꢀC (essentially no change was
observed after several hours). On the other hand, the copper-
(II)-phenol complex CuIILmH (meta-phenol derivative) re-
acts with O2 in the presence of triethylamine as a base in
CH3OH at 60 ꢀC to give a methoxy-substituted para-quinone
derivative (compound 4; see Chart 3). The structure of the
product has been established through detailed nuclear mag-
7
˚
(25) (a) Mure, M.; Klinman, J. P. J. Am. Chem. Soc. 1995, 117, 8698–
8706. (b) Itoh, S.; Mure, M.; Ogino, M.; Ohshiro, Y. J. Org. Chem. 1991, 56,
6857–6865. (c) Itoh, S.; Takada, N.; Haranou, S.; Ando, T.; Komatsu, M.;
Ohshiro, Y.; Fukuzumi, S. J. Org. Chem. 1996, 61, 8967–8974.

Article
Inorganic Chemistry, Vol. 50, No. 5, 2011 1647
due to the C-C coupling reaction of the phenoxyl radical
intermediate B, which is generated in a relatively large
amount in solution in the presence of a large excess of
Et3N. Thus, the TPQ formation is only observed successfully
in the presence of a limited amount of ET3N, where the
copper(I) phenoxyl radical species exists as a very minor
component, to prevent the C-C coupling reaction. In the
enzyme active site, such a C-C coupling reaction does not
proceed at all, so that one can detect the intermediates. This is
a limitation of model studies. Nevertheless, the generated
para-quinone derivative can act as a turnover catalyst for
aerobic oxidation of benzylamine to N-benzylidene benzyl-
amine, as shown in Scheme 7. The reaction may involve a
trans-amination mechanism, as in the cases of other quinone
cofactor model compounds.25
When the position of the phenolic OH group was changed
from meta to para, relative to the pyridine donor group in
copper(II)-phenol complex CuIILpH, the oxygenation reac-
tion involvedelectrophilic substitution at the ipso-position, to
give a keto-alcohol derivative 6H (see Chart 4). The structure
of this oxygenated product is fundamentally different from
that of the TPQ cofactor. Thus, the present results strongly
suggest that the steric relation between the copper(II) ion and
the phenol moiety of the reactive tyrosine residue is critically
important for its conversion to the TPQ cofactor in the
enzyme active site. Furthermore, the poor O2 reactivity
of CuIILo may suggest that such strong coordination of
phenolate to copper(II) is not favorable for the reaction with
O2. This is consistent with the observation that the coordi-
native interaction between the copper(II) ion and the un-
modified tyrosine precursor is fairly weak in the initial stage
1j
˚
of TPQ biosynthesis: the Cu-O distance is ∼2.5 A, which is
II
o
7,26
˚
much larger than that in Cu L (1.891 A).
Acknowledgment. S.I. acknowledges the financial sup-
port by Grant-in-Aid for Science Research on Priority
Areas (No. 200360044, Synergy of Elements; No.
21108515, π-Space) from Ministry of Education, Culture,
Sports, Science and Technology, Japan and also by the
Asahi Glass Foundation and the Mitsubishi Foundation.
We also thank Dr. Kyoko Inoue of the Analytical Center
of Osaka University for her assistance in obtaining the 2D
NMR data. C.J.C. and M.Z.E. thank the U.S. National
Science Foundation for support (through Grant Nos.
CHE-0610183 and CHE-0956776).
II
m
˚
(26) The optimized structure for the Cu L H shows a distance of 4.351 A
between the oxygen atom of the phenolate and the CuII ion, which is much
longer that the distance between the copper ion and thyrosyl oxygen in the
Supporting Information Available: Further details are given in
Figures S1-S19 (PDF), Table S1 (PDF), and in a CIF file. This
material is available free of charge via the Internet at http://
pubs.acs.org.
˚
enzyme active site (2.57 A). However, through-bond electron transfer from
the phenolate ring to copper(II) could be possible in the model system,
because the phenol ring is covalently connected to the pyridine donor group
that is directly coordinated to copper(II).