A double arene hydroxylation mediated by dicopper(II)-hydroperoxide species

Article abstract of DOI:10.1021/ja0280776

The dicopper(II) complex [Cu₂(L)]⁴⁺ (L =α,α′-bis{bis[2-(1′-methyl-2′- benzimidazolyl)ethyl]amino}-m-xylene) reacts with hydrogen peroxide to give the dicopper(II) -hydroquinone complex in which the xylyl ring of the ligand has undergone a double hydroxylation reaction at ring positions 2 and 5. The dihydroxylated ligand 2,6-bis({bis[2-(3-methyl-1 H-benzimidazo[-2-yl)ethyl]amino}methyl)benzene-1,4-diol was isolated by decomposition of the product complex. The incorporation of two oxygen atoms from H₂O₂ into the ligand was confirmed by isotope labeling studies using H₂¹⁸O₂. The pathway of the unusual double hydroxylation was investigated by preparing the two isomeric phenolic derivatives of L, namely 3,5-bis({bis[2-(1-methyl- 1H-benzimidazol-2-yl)ethyl]amino}methyl) phenol (6) and 2,6-bis({bis[2-(1-methyl-1H-benzimidazol-2-yl)- ethyl]amino}methyl)phenol (7), carrying the hydroxyl group in one of the two positions where L is hydroxylated. The dicopper(II) complexes prepared with the new ligands 6 and 7 and containing bridging μ-phenoxo moieties are inactive in the hydroxylation. Though, the dicopper(II) complex 3 derived from 6 and containing a protonated phenol is rapidly hydroxylated by H202 and represents the first product formed in the hydroxylation of [Cu₂(L)]⁴⁺. Kinetic studies performed on the reactions of [Cu₂(L)]⁴⁺ and 3 with H₂O₂ show that the second hydroxylation is faster than the first one at room temperature (0.13 ± 0.05 s^-1 vs 5.0(±0.1) X 10^-3 s^-1) and both are intramolecular processes. However, the two reactions exhibit different activation parameters (ΔH? = 39.1 ± 0.9 kJ mol^-1 and ΔS? = -115.7 ± 2.4 J K^-1 mol-1 for the first hydroxylation; ΔH? = 77.8 ± 1.6 kJ mol^-1 and ΔS? = -14.0 ± 0.4 J K^-1 mol^-1 for the second hydroxylation). By studying the reaction between [Cu₂(L)]⁴⁺. and H₂O₂ at low temperature, we were able to characterize the intermediate η¹:η¹-hydroperoxodicopper(II) adduct active in the first hydroxylation step, [Cu₂(L)(OOH)]³⁺ [λ_max = 342 (ε 12 000), 444 (ε 1200), and 610 nm (ε 800 M^-1cm^-1); broad EPR signal in frozen solution indicative of magnetically coupled Cu(II) centers].

Full text of DOI:10.1021/ja0280776

Published on Web 03/14/2003
A Double Arene Hydroxylation Mediated by
Dicopper(II)-Hydroperoxide Species
†
†
†
†
Giuseppe Battaini, Enrico Monzani, Angelo Perotti, Cristina Para,
Luigi Casella,*, Laura Santagostini, Michele Gullotti, Ren e´ e Dillinger,
†
‡
‡
§
§
§
Christian N a¨ ther, and Felix Tuczek
Contribution from the Dipartimento di Chimica Generale, UniVersit a` di PaVia,
Via Taramelli 12, 27100 PaVia, Italy, Dipartimento C.I.M.A., UniVersit a` di Milano,
Centro CNR, Via Venezian 21, 20133 Milano, Italy, and Institut f u¨ r Anorganische Chemie,
UniVersit a¨ t Kiel, Otto-Hahn-Platz 6/7, 24098 Kiel, Germany
Received August 9, 2002; E-mail: luigi.casella@unipv.it
(L)]4 (L ) R,R′-bis{bis[2-(1′-methyl-2′-benzimidazolyl)ethyl]amino}-
+
Abstract: The dicopper(II) complex [Cu
2
m-xylene) reacts with hydrogen peroxide to give the dicopper(II)-hydroquinone complex in which the xylyl
ring of the ligand has undergone a double hydroxylation reaction at ring positions 2 and 5. The dihydroxylated
ligand 2,6-bis({bis[2-(3-methyl-1H-benzimidazol-2-yl)ethyl]amino}methyl)benzene-1,4-diol was isolated by
decomposition of the product complex. The incorporation of two oxygen atoms from H
2
O
2
into the ligand
. The pathway of the unusual double hydroxylation
was investigated by preparing the two isomeric phenolic derivatives of L, namely 3,5-bis({bis[2-(1-methyl-
H-benzimidazol-2-yl)ethyl]amino}methyl)phenol (6) and 2,6-bis({bis[2-(1-methyl-1H-benzimidazol-2-yl)-
18
was confirmed by isotope labeling studies using H
2 2
O
1
ethyl]amino}methyl)phenol (7), carrying the hydroxyl group in one of the two positions where L is
hydroxylated. The dicopper(II) complexes prepared with the new ligands 6 and 7 and containing bridging
µ-phenoxo moieties are inactive in the hydroxylation. Though, the dicopper(II) complex 3 derived from 6
and containing a protonated phenol is rapidly hydroxylated by H
in the hydroxylation of [Cu (L)] . Kinetic studies performed on the reactions of [Cu (L)] and 3 with H O
2 2 2 2
show that the second hydroxylation is faster than the first one at room temperature (0.13 ( 0.05 s vs
2
O
2
and represents the first product formed
4+
4+
-
1
-
3
-1
5.0((0.1) × 10
s ) and both are intramolecular processes. However, the two reactions exhibit different
q
-1
q
-1
-1
activation parameters (∆H ) 39.1 ( 0.9 kJ mol and ∆S ) -115.7 ( 2.4 J K mol for the first
hydroxylation; ∆H ) 77.8 ( 1.6 kJ mol and ∆S ) -14.0 ( 0.4 J K mol for the second hydroxylation).
By studying the reaction between [Cu
q
-1
q
-1
-1
(L)]4 and H
+
at low temperature, we were able to characterize
2
2
O
2
1
1
3+
2
the intermediate η :η -hydroperoxodicopper(II) adduct active in the first hydroxylation step, [Cu (L)(OOH)]
[λ
max ) 342 (ꢀ 12 000), 444 (ꢀ 1200), and 610 nm (ꢀ 800 M-1cm ); broad EPR signal in frozen solution
-1
indicative of magnetically coupled Cu(II) centers].
ligand4 or an exogenous substrate with insertion of an oxygen
atom. In other dinuclear copper systems performing ligand
hydroxylation reactions, the putative copper-dioxygen inter-
mediate could not be characterized due to its low stability even
-7
8
Introduction
Synthetic copper complexes performing various types of
oxygen activation processes have become increasingly important
in the past decade, in relation with the enzymatic reactions
1
9
2
at low temperature.
catalyzed by copper enzymes, particularly tyrosinase, and the
The key step in the monophenolase activity of tyrosinase is
the reaction of the η :η -peroxodicopper(II) complex with the
phenolic substrate to produce an o-quinone. Different pathways
potential applications in catalytic transformations of organic
2
2
3
substrates. In general, in the best characterized systems, the
adduct formed by dioxygen with a couple of copper atoms, in
the form of either a µ-η :η -peroxodicopper(II) or a bis(µ-oxo)-
dicopper(III) complex, performs C-H bond cleavage of the
2
2
(
4) (a) Karlin, K. D.; Hayes, J. C.; Gultneh, Y.; Cruse, R. W.; McKown, J.
W.; Hutchinson, J. P.; Zubieta, J. J. Am. Chem. Soc. 1984, 106, 2121-
2
128. (b) Karlin, K. D.; Nasir, M. S.; Cohen, B. I.; Cruse, R. W.; Kaderli,
†
Universit a` di Pavia.
Universit a` di Milano.
Universit a¨ t Kiel.
S.; Zuberb u¨ hler, A. D. J. Am. Chem. Soc. 1994, 116, 1324-1336. (c)
Pidcock, E.; Obias, H. V.; Zhang, C. X.; Karlin, K. D.; Solomon, E. I. J.
Am. Chem. Soc. 1998, 120, 7841-7847. (d) Becker, M.; Schindler, S.;
Karlin, K. D.; Kaden, T. A.; Kaderli, S.; Palanch e´ , T.; Zuberb u¨ hler, A. D.
Acc. Chem. Res. 1997, 30, 139-147.
‡
§
(
1) Karlin, K. D.; Zuberb u¨ hler, A. D. In Bioinorganic Catalysis, 2nd ed.;
Reedijk, J., Bouwman, E., Eds.; Dekker: New York, 1999; pp 469-534.
2) (a) Solomon, E. I.; Sundaram, U. M.; Machonkin, T. E. Chem. ReV. 1996,
(
(5) (a) Mahapatra, S.; Halfen, J. A.; Tolman, W. B. J. Am. Chem. Soc. 1996,
118, 11575-11586. (b) Mahapatra, S.; Kaderli, S.; Llobet, A.; Neuhold,
Y.-M.; Palanch e´ , T.; Halfen, J. A.; Young, V. G.; Kaden, T. A.; Que, L.,
Jr.; Zuberb u¨ hler, A. D.; Tolman, W. B. Inorg. Chem. 1997, 36, 6343-
6356. (c) Holland, P. L.; Rodgers, K. R.; Tolman, W. B. Angew. Chem.,
Int. Ed. 1999, 38, 1139-1142. (d) Halfen, J. A.; Young, V. G.; Tolman,
W. B. Inorg. Chem. 1998, 37, 2102-2103.
9
6, 2563-2605. (b) McGuirl, M. A.; Dooley, D. M. Curr. Opin. Chem.
Biol. 1999, 3, 138-144. (c) Decker, H.; Dillinger, R.; Tuczek, F. Angew.
Chem., Int. Ed. 2000, 39, 1591-1595. (d) Holland, P. L.; Tolman, W. B.
Coord. Chem. ReV. 1999, 190-192, 855-869.
(3) (a) Foote, C. S.; Valentine, J. S.; Greenberg, A.; Liebman, J. F., Eds. ActiVe
Oxygen in Chemistry; Chapman & Hall: London, 1995.
10.1021/ja0280776 CCC: $25.00 © 2003 American Chemical Society
J. AM. CHEM. SOC. 2003, 125, 4185-4198
9
4185

A R T I C L E S
Battaini et al.
can be hypothesized for this reaction, and the currently accepted
view suggests that either the side-on bound peroxo group directly
attacks the substrate or a bis(µ-oxo) species resulting from
Scheme 1
2
c,d
cleavage of the O-O bond is the oxygenating species.
A
third pathway, involving a hydroperoxodicopper(II) species, is
possible even though this implies previous protonation of the
bound peroxide moiety. However, so far, hydroperoxodicopper-
(II) complexes have not been shown to be capable of supporting
phenol hydroxylation reactions.
An interesting aspect of the aromatic ligand hydroxylation
mediated by the peroxo complex derived from the reaction of
+
[
Cu2(XYL-H)]2 with dioxygen, reported by Karlin et al. as
4
the first example of tyrosinase model, is that the reaction is
not observed when the pyridine rings in the chelating arms of
the m-xylyl-containing ligand are replaced by other nitrogen
9c,10
heterocycles such as imidazoles or benzimidazoles.
For the
2+
dicopper(I) complex [Cu2(L)] , 1, where the ligand L contains
N-methylbenzimidazole instead of pyridine rings,1 this feature
has a positive effect from a practical point of view, since the
dioxygen adduct of this complex is capable of hydroxylating
an exogenous phenol, thereby exhibiting true tyrosinase-like
activity.8 However, the origin of this markedly different
reactivity behavior between apparently strictly related peroxo-
dicopper(II) systems is still unclear.
0b
ligand hydroxylation according to an intermolecular mechanism
which is different from the intramolecular pathway involved in
the reaction between the corresponding dicopper(I) complex and
4
c
dioxygen. In this paper, we report synthetic, mechanistic, and
spectroscopic studies on a novel reaction between 1 and
hydrogen peroxide, which produces a stable complex with the
ligand containing a hydroquinone moiety, 4, resulting from a
double hydroxylation of the xylyl aromatic ring (Scheme 1, B
) 1-methylbenzimidazole). This extremely facile reaction occurs
in two successive steps. The first one is slower (at room
temperature) and proceeds through a dicopper(II)-hydroper-
oxide complex, 2, which evolves to the monohydroxylated
complex 3, where the aromatic position meta to the xylylene-
diamino substituents has been hydroxylated. The second step
is faster and involves binding of hydrogen peroxide to 3 and a
second hydroxylation to give the hydroquinone complex 4, from
which the dihydroxylated ligand 5 can be isolated. This
somewhat unexpected pathway was probed by preparing the
two isomeric phenolic derivatives of the ligand L, namely 6
and 7 (Scheme 2), complex 3, and the two µ-phenoxodicopper-
(II) complexes derived from 6 and 7, that is, 8 and 9. The latter
µ-phenoxo complexes are inactive in the hydroxylation, whereas
3 reacts rapidly with H2O2 to give 4. The intermediate
In Karlin’s system, the xylyl ligand hydroxylation, to produce
a µ-hydroxo,µ-phenoxodicopper(II) species, can be achieved
also by reacting the dinuclear Cu(II) complex, [Cu2(XYL-
4
+
11
H)] , and hydrogen peroxide. In the latter case, although the
1
1
intermediate active species could not be identified, a η :η -
hydroperoxide complex has been proposed.11 This performs the
(
6) (a) Itoh, S.; Nakao, H.; Berreau, L. M.; Kondo, T.; Komatsu, M.; Fukuzumi,
S. J. Am. Chem. Soc. 1998, 120, 2890-2899. (b) Itoh, S.; Taki, M.; Nakao,
H.; Holland, P. L.; Tolman, W. B.; Que, L., Jr.; Fukuzumi, S. Angew.
Chem., Int. Ed. 2000, 39, 398-400. (c) Allen, W. E.; Sorrell, T. N. Inorg.
Chem. 1997, 36, 1732-1734.
(
7) (a) Mahadevan, V.; Hou, Z.; Cole, A. P.; Root, D. E.; Lal, T. K.; Solomon,
E. I.; Stack, T. D. P. J. Am. Chem. Soc. 1997, 119, 11996-11997. (b)
Mahadevan, V.; DuBois, J. L.; Hedman, B.; Hodgson, K. O.; Stack, T. D.
P. J. Am. Chem. Soc. 1999, 121, 5583-5584. (c) Mahadevan, V.; Henson,
M. J.; Solomon, E. I.; Stack, T. D. P. J. Am. Chem. Soc. 2000, 122, 10249-
1
0250.
(
8) Santagostini, L.; Gullotti, M.; Monzani, E.; Casella, L.; Dillinger, R.;
Tuczek, F. Chem. Eur. J. 2000, 6, 519-522.
3
+
hydroperoxo complex [Cu2(L)(OOH)] was characterized at
(
9) (a) Casella, L.; Gullotti, M.; Pallanza, G.; Rigoni, L. J. Am. Chem. Soc.
1
988, 110, 4221-4227. (b) Gelling, O. J.; van Bolhuis, F.; Meetsma, A.;
low temperature. This is the first dinuclear copper(II)-hydro-
Feringa, B. L. J. Chem. Soc., Chem. Commun. 1988, 552-553. (c) Sorrell,
T. N.; Vankai, V. A.; Garrity, M. L. Inorg. Chem. 1991, 30, 207-210. (d)
Alzuet, G.; Casella, L.; Villa, M. L.; Carugo, O.; Gullotti, M. J. Chem.
Soc., Dalton Trans. 1997, 4789-4794. (e) Ghosh, D.; Mukherjee, R. Inorg.
Chem. 1998, 37, 6597-6605. (f) Monzani, E.; Battaini, G.; Perotti, A.;
Casella, L.; Gullotti, M.; Santagostini, L.; Nardin, G.; Randaccio, L.;
Geremia, S.; Zanello, P.; Opromolla, G. Inorg. Chem. 1999, 38, 5359-
-
peroxo complex in which the HOO moiety is the only bridging
ligand between the copper(II) centers.
Results and Discussion
5
369.
Aromatic Hydroxylation: Characterization of the Prod-
ucts. The reaction between 1 and hydrogen peroxide produces
(
10) (a) Sorrell, T. N.; Garrity, M. L. Inorg. Chem. 1991, 30, 210-215. (b)
Casella, L.; Gullotti, M.; Radaelli, R.; Di Gennaro, P. J. Chem. Soc., Chem.
Commun. 1991, 1611-1612. (c) Casella, L.; Monzani, E.; Gullotti, M.;
Cavagnino, D.; Cerina, G.; Santagostini, L.; Ugo, R. Inorg. Chem. 1996,
(11) Cruse, R. W.; Kaderli, S.; Meyer, C. J.; Zuberb u¨ hler, A. D.; Karlin, K. D.
3
5, 7516-7525.
J. Am. Chem. Soc. 1988, 110, 5020-5024.
4186 J. AM. CHEM. SOC.
9
VOL. 125, NO. 14, 2003

Hydroxylation by Dicopper(II)−Hydroperoxide Species
A R T I C L E S
Scheme 2
Scheme 3 a
a species characterized by an intense absorption band near 345
-
1
-1
nm (ꢀ ≈ 9000 M
cm ). The reaction occurs with a
4+
stoichiometry of 2 H2O2:1 [Cu2(L)]
(see Figure 1S in
Supporting Information) and proceeds rapidly to completion
when carried out on a preparative scale using a moderate excess
of hydrogen peroxide. Chromatographic separation of the
product mixture resulting from the reaction carried out in
acetonitrile solution enabled us to isolate two products: a major
green fraction consisting of the µ-(4-hydroxy)-phenolate (µ-
hydroquinone) complex 4 and a minor, light brown fraction
consisting of the µ-hydroxo,µ-hydroquinone complex 10 (B )
a
(
i) N-bromosuccinimide; (ii) bis[2-(1-methyl-1H-benzimidazole-2-
yl)ethyl]amine.
a somewhat unstable, photosensitive oily material which
undergoes a slow oxidative degradation process in air at room
temperature. The origin of the photosensitivity of 5 is quite
certainly due to its moderate absorption near 340 nm (ꢀ ≈ 3000
-
1
-1
M
cm ), which shifts to higher energy by the addition of
1
-methylbenzimidazole), which is probably formed spontane-
acid and reveals that 5 assumes a zwitterionic structure. The
ously from 4 during workup.
4+
same ligand hydroxylation of [Cu2(L)] can be observed using
tert-butyl hydroperoxide instead of H2O2, although in this case
the reaction proceeds at a slower rate and, therefore, a much
larger excess of oxidant must be used in order to observe
significant conversion of 1 to 4.
Since the hydroxylation occurs on the two carbon positions
2
and 5 of L, it could be anticipated that one of the phenoxo-
dicopper(II) complexes 8 or 9 were intermediates of the reaction.
The two putative intermediates were prepared for this purpose
from the new phenolic ligands 6 and 7 (prepared according to
Scheme 3). To our surprise, neither 8 nor 9 are able to support
the ligand hydroxylation reaction by hydrogen peroxide observed
with 1. Complex 8 exhibits an extremely slow reaction with
H2O2, but it is likely that this reaction is actually undergone by
a small fraction of complex 3, with unbound phenol, present in
equilibrium in solution. In fact, compound 3 reacts rapidly with
hydrogen peroxide to produce 4. Attempts to prepare derivatives
of complex 9 with either a protonated or O-acetylated phenol
group were unsuccessful because in any conditions only the
bridged µ-phenoxo complex 9 spontaneously forms.
The bis(hydroxylated) ligand 5 can be obtained from either
4
or 10 by removal of the copper ions using a base extraction
4
a
procedure. After flash chromatographic purification, it is
possible to isolate compound 5 in more than 80% yield. The
1
13
H and C NMR characterization data of this compound confirm
that the ligand L has undergone two oxygen atom insertions at
positions 2 and 5 of the xylyl aromatic ring (see Figure 2S in
Supporting Information and Experimental Section). The incor-
poration of two oxygen atoms from H2O2 into the ligand was
unequivocally confirmed by mass spectrometric analysis of 5,
which shows a cluster of peaks centered at m/z 801.3. When
the reaction between 1 and hydrogen peroxide was carried out
By following the reaction between 1 and H2O2 in acetonitrile
by IR spectroscopy, some significant changes can be observed
-1
in the 1100-1350 cm -range (Figure 1). In particular, a new
18
using isotope labeled H2 O2 (90% enrichment), the dihydroxyl-
ated ligand isolated after decomposition exhibited an MS
spectrum where the cluster of molecular peaks had shifted to
-1
band appears at 1304 cm concomitant with the transformation
-1
of 1 to 4. Similar IR bands near 1300 cm , assigned to the
12
phenolate ν(C-O) mode, have been observed in the spectra
+
+
+
+
m/z 805.5 (ratio of M /(M + 2)/M /(M + 4) ) 28:75:100,
1
8
more than 80% O incorporated in each step) (see Figure 3S
(
12) Pyrz, J. W.; Karlin, K. D.; Sorrell, T. N.; Vogel, G. C.; Que, L., Jr. Inorg.
in Supporting Information). The free hydroquinoid ligand 5 is
Chem. 1984, 23, 4581-4584.
J. AM. CHEM. SOC.
9
VOL. 125, NO. 14, 2003 4187

A R T I C L E S
Battaini et al.
Scheme 4
zimidazole bands at higher energy and the 345-nm band at lower
energy. The origin of this 345-nm band is mostly due to a πfπ*
transition of the hydroquinone chromophore, since a similar
absorption is present in the ligand, albeit with lower intensity.
However, also weaker π(benzimidazole)fCu(II) LMCT transi-
tions occur in the range between 300 and 400 nm, as is shown
4+
14
by the electronic spectrum of the starting [Cu2(L)] complex.
Figure 1. IR spectra recorded in acetonitrile solution of (A) [Cu2(L)][ClO4]4
(
[
20 mM); (B) [Cu2(L)][ClO4]4 (20 mM) + H2O2 (3-fold excess); and (C)
It is likely that hydroquinone-to-copper LMCT bands contribute
to both the intense 345-nm band and the weaker absorption band
near 420 nm. The latter band is usually observed in the range
1
8
Cu2(L)][ClO4]4 (20mM) + H2 O2 (3-fold excess). Spectra B and C were
recorded after a 2-h reaction time.
Table 1. UV-visible Spectral Data for Copper(II) Complexes in
Acetonitrile Solution
between 400 and 470 nm, with variable intensity, for µ-phen-
13,15
oxodicopper(II) complexes,
and occurs also in the spectra
IL λmax, nm
LMCT λmax, nm
d−d λmax, nm
of 8 and 9, but not in that of 3, the near-UV spectrum of which
resembles that of 1 (Table 1). Both the µ-hydroquinone and
µ-phenoxo complexes exhibit d-d bands in the 650-700-nm
range, indicative of an essentially square-pyramidal geometry
for the Cu(II) centers. It is worth noting that the Cu(II)
coordination geometry in the starting complex 1 is essentially
-1
(ꢀ, M cm-1
-1
(ꢀ, M cm-1
)
-1
complex
(ꢀ, M cm-1
273 (32 200)
80 (30 800)
270 (30 600)
78 (29 800)
273 (32 000)
80 (30 400)
272 (27 300)
80 (27 000)
272 (27 600)
80 (25 300)
273 (30 500)
80 (29 100)
)
)
1
310 sh (4000)
410 sh (600)
318 sh (4600)
360 sh (2000)
345 (8500)
420 sh (1400)
314 (7200)
410 sh (3200)
344 (4000)
675 (450)
850 (500)
660 (300)
2
3
4
8
9
16
2
700 (250)
698 (700)
668 (250)
670 (220)
2
trigonal bipyramidal, but on conversion to the bis(µ-hydroxo)
2+
derivative [Cu2(L)(OH)2] , the Cu(II) coordination sphere also
2
14
changes to square pyramidal. Upon addition of an excess base
like triethylamine to solutions of 4 and 10, the characteristic
band at 344 nm quickly disappears and the solutions turn dark
brown. As represented in Scheme 4, deprotonation of the
hydroxyl group in 4 destabilizes the hydroquinone-Cu(II)
moiety, giving rise to a fast oxidative decomposition.
2
424 (2300)
320 sh (9600)
345 (9200)
10
2
420 sh (1500)
of related µ-phenoxodicopper(II) complexes and are present also
The NMR spectra enable further characterization of the
dicopper(II) complexes. As shown in Figure 2, the spectra of 1
and 3 are similar and contain broad, paramagnetically shifted
signals in both the low field (between 15 and 30 ppm) and high
field (between -5 and -30 ppm) regions. Other signals may
be too broad to be detectable. The spectra of the µ-phenoxo
-
1
in the spectra of compounds 8 and 9. In fact, the 1304-cm
band undergoes a shift to 1294 cm-1 when H2 O2 is used in
18
the reaction with 1, with a shift comparable to that reported for
2+ 12,13
the formation of [Cu2(XYL-O)(OH)] .
Both the bis(hydroxylated) complexes 4 and 10 exhibit a
-
1
-1
prominent absorption band at 345 nm (ꢀ ≈ 9000 M cm )
(
14) Casella, L.; Carugo, O.; Gullotti, M.; Garofani, S.; Zanello, P. Inorg. Chem.
and a weaker, broad absorption band centered near 420 nm (ꢀ
1993, 32, 2056-2067.
-
1
-1
≈
1000 M cm ) (Table 1). The spectrum of 10 contains an
(15) (a) Sorrell, T. N. Tetrahedron 1989, 45, 3-68. (b) Berends, H. P.; Stephan,
D. W. Inorg. Chem. 1987, 26, 749-754. (c) Holz, R. C.; Brink, J. M.;
Godena, F. T.; O’Connor, C. J. Inorg. Chem. 1994, 33, 6086-6092. (d)
Holz, R. C.; Bradshaw, J. M.; Bennett, B. Inorg. Chem. 1998, 37, 1219-
1225. (e) Torelli, S.; Belle, C.; Gautier-Luneau, I.; Pierre, J. L.; Saint-
Aman, E.; Latour, J. M.; Le Pape, L.; Luneau, D. Inorg. Chem. 2000, 39,
additional, ill-defined absorption in the 300-330-nm range,
likely due to hydroxo-Cu(II) ligand-to-metal charge-transfer
transition (LMCT),14 which overlaps with the strong intraben-
3526-3536.
(
13) Ling, J.; Farooq, A.; Karlin, K. D.; Loehr, T. M.; Sanders-Loehr, J. Inorg.
(16) Hathaway, B. J. In ComprehensiVe Coordination Chemistry; Wilkinson,
Chem. 1992, 31, 2552-2556.
G., Ed.; Pergamon Press: New York, 1987; Vol. 5, pp 533-774.
4188 J. AM. CHEM. SOC.
9
VOL. 125, NO. 14, 2003

Hydroxylation by Dicopper(II)−Hydroperoxide Species
A R T I C L E S
J. AM. CHEM. SOC.
9
VOL. 125, NO. 14, 2003 4189

A R T I C L E S
Battaini et al.
Figure 2. Proton NMR spectra of 1 mM CD3CN solutions of (a) complex 1; (b) complex 3; (c) complex 9; (d) complex 8; (e) complex 11; and (f) complex
0.
1
complexes 8 and 9 contain several relatively sharp resonances
at very low field (down to about 80 ppm), while those of the
dibridged complexes 10 and 11 contain much sharper signals
in the diamagnetic region (Figure 2), indicative of a strong
coupling between the metal centers. While a detailed analysis
of these spectra and, more generally, the magnetic properties
of the complexes are beyond the scope of the present investiga-
tion, it is clear that the NMR behavior confirms the presence
or absence of single (µ-phenoxo) or double (µ-hydroxo,µ-
phenoxo) bridges between the Cu(II) centers of the present
dinuclear complexes.
Figure 3. Molecular structure of the unit of complex 11 in the crystal
structure of 11[CLO4]2‚CH3CN‚(C2H5)2.
(
Figure 3 and Table 2). The coordination polyhedron of the
Crystal Structure of Complex 11. Single crystals of
compound 11[CLO4]2‚CH3CN‚(C2H5)2O were obtained in the
attempt to prepare the dicopper(II) complex of the O-acetylated
ligand 7. However, structure determination of the crystals
isolated shows that the acetyl ligand was hydrolyzed on complex
formation and a second hydroxo bridge formed during crystal-
lization. The compound crystallizes in the triclinic centrosym-
metric space group P 1h with two formula units in the unit cell
and all atoms located in general positions. Each copper(II) center
is five-coordinated by the phenolate oxygen atom, three nitrogen
atoms of the ligand, and the oxygen atom of the hydroxyl group
copper(II) centers can be described as distorted tetragonal
pyramids sharing common edges each with one nitrogen atom
on the top of the pyramid. The Cu-N distances to these nitrogen
atoms are significantly longer compared to those in the basal
plane of the pyramids (Table 2). Both copper(II) centers are
connected via the phenolate oxygen atom of the ligand and one
µ-hydroxo bridge forming a nearly planar four-membered
heteroring. The Cu-Cu distance amounts to 3.11(1) Å. The
structure clearly shows that a bridging hydroperoxo group
replacing the bridged hydroxo group would not be suitably
4190 J. AM. CHEM. SOC.
9
VOL. 125, NO. 14, 2003

Hydroxylation by Dicopper(II)−Hydroperoxide Species
A R T I C L E S
Table 2. Selected Bond Lengths (Å) and Angles (deg) for
Complex 11[CLO4]2·CH3CN·(C2H5)2
and number of components from those exhibited by mono-
nuclear, terminally bound Cu-OOH complexes.
17-21
Attempts to measure characteristic vibrations of this inter-
mediate by resonance Raman spectroscopy using an excitation
wavelength of 350 nm failed; that is, after addition of an excess
4
+
of H2O2 to a 10 mM solution of [Cu2(L)] in acetone/DMF or
acetone/DMSO, no additional peaks could be detected. The
subsequent addition of several equivalents of NEt3 caused the
solution to turn dark olive-green which is correlated with a shift
of the CT band from 342 to 360 nm and the appearance of a
shoulder at 420 nm. This optical spectrum is tentatively ascribed
to the doubly bridged µ-hydroxo µ-hydroperoxo species. No
Raman signals of this intermediate were detectable either, using
excitation wavelengths of 350 or 413 nm. A possible reason
for these findings might be the kinetic lability or, more likely,
conformational flexibility of the hydroperoxo adducts, which
is so important for the double hydroxylation, causes severe
broadening of the Raman signals.
When the low-temperature solution of 2 is warmed to about
-30 °C, the absorption bands of the hydroperoxo intermediate
decrease quickly in intensity and are replaced by weaker spectral
features (see Figure 4S in Supporting Information). A weaker
and asymmetric band appears in the near-UV region, with a
maximum near 370 nm in the difference spectrum (∆ꢀ ≈ 2000
-
1
-1
M cm ). We identify this species as an hydroxo adduct of
arranged for an intramolecular attack to the xylyl C-H bond
para to the phenoxo group in the second hydroxylation step,
but of course it may react intermolecularly with a second
molecule of the complex.
the monohydroxylated complex 3 on the basis of the near-UV
feature in the difference spectrum, since hydroxo-to-Cu(II)
9f,14
LMCT transitions typically occur in the same range.
When
the low-temperature solution of this intermediate (with excess
hydrogen peroxide present) is further warmed, some dramatic
spectral changes occur. The band at 370 nm progressively
decreases in intensity and around -20 °C is rapidly replaced
by the stronger absorption band near 345 nm. The final spectrum
corresponds to that previously described for complex 4, contain-
ing the hydroquinone ligand 5. Therefore, in the low-temperature
experiment, we can clearly detect the intermediate and product
of the first hydroxylation step. The second step involves the
reaction between hydrogen peroxide and the intermediate
hydroxo adduct of 3, which is likely to give a second
µ-hydroperoxo adduct. But this intermediate is too reactive and
cannot be detected. Also, compound 3 reacts rapidly with
hydrogen peroxide in the same conditions to give the bis-
Low-Temperature Spectroscopy: Characterization of the
Intermediates. When a large excess of hydrogen peroxide
4
+
(
about 100-fold) is added to a solution of [Cu2(L)] in
acetonitrile at -40 °C, a moderately intense band at 342 nm
-
1
-1
(
≈
∆ꢀ ≈ 8600 M cm ) and two weaker bands at 444 nm (∆ꢀ
-
1
-1
-1
-1
850 M cm ) and 610 nm (∆ꢀ ≈ 400 M cm ) develop
in the difference spectra (see Figure 4S in Supporting Informa-
tion). The reaction is complete in 0.5 h, probably because
hydrogen peroxide is partially frozen and undissolved in the
solution. The UV bands do not decrease in intensity even upon
applying several vacuum/argon cycles at low temperature. The
spectral features and the irreversibility of the reaction exclude
2
2
that this intermediate species can be the same η :η -peroxodi-
copper(II) adduct that is formed by the reaction between [Cu2-
(hydroxylated) compound 4, as represented in Scheme 1, and
2+
2+
(
L)] and O2, that is, [Cu2(L)(O2)] (12), which is charac-
terized by λmax at 362 (ꢀ ≈ 15 000), 455 (ꢀ ≈ 2000), and 550
nm (ꢀ ≈ 900 M cm ).
even following the reaction at low temperature (down to -40
°
C), no intermediate could be observed.
-
1
-1 8
The hydroperoxo intermediate 2 was characterized also by
EPR spectroscopy. As shown in Figure 4, the EPR spectrum of
complex 1 in a solvent mixture of MeOH-DMF at -150 °C
exhibits a well resolved signal for noninteracting Cu(II) centers.
-
4
-1
The EPR parameters, g| ) 2.555, A| ) 172 × 10 cm , g⊥ )
.05, with a partially resolved superhyperfine structure due to
2
(
17) (a) Wada, A.; Harata, M.; Hasegawa, K.; Jitsukawa, K.; Masuda, H.; Mukai,
M.; Kitagawa, T.; Einaga, H. Angew. Chem., Int. Ed. 1998, 37, 798-799.
(
b) Chen, P.; Fujisawa, K.; Solomon, E. I. J. Am. Chem. Soc. 2000, 122,
1
0177-10193.
18) Solomon, E. I.; Tuczek, F.; Root, D. E.; Brown, C. A. Chem. ReV. 1994,
4, 827-856.
(19) Root, P. E.; Mahroof-Tahir, M.; Karlin, K. D.; Solomon, E. I. Inorg. Chem.
(
9
We thus propose that the bands that appear in the spectrum
1998, 37, 4838-4849.
are associated with LMCT transitions of a dicopper(II)-
(20) Karlin, K. D.; Ghosh, P.; Cruse, R. W.; Farooq, A.; Gultneh, Y.; Jacobson,
1
1
hydroperoxide species with the η :η structure 2. The bridging
structure of this hydroperoxide complex is suggested by the
pattern of LMCT bands, which is different for both the position
R. R.; Blackburn, N. J.; Strange, R. W.; Zubieta, J. J. Am. Chem. Soc.
1988, 110, 6769-6780.
(21) Mahroof-Tahir, M.; Murthy, N. N.; Karlin, K. D.; Blackburn, N. J.; Shaikm,
S. N.; Zubieta, J. Inorg. Chem. 1992, 31, 3001-3003.
J. AM. CHEM. SOC.
9
VOL. 125, NO. 14, 2003 4191

A R T I C L E S
Battaini et al.
steps could be followed for L, and their cumulative log â values
were determined as follows: 8.04 (0.2), 14.94 (0.2), 19.86 (0.3),
2
3.98 (0.1), 25.85 (0.3). The ligand easily and simultaneously
incorporates two copper(II) ions according to the equilibrium:
2
+
4+
2
Cu + L T [Cu (L)]
log â120 ) 19.89 (1)
2
Successively, the complex containing two copper ions, each
coordinated to three nitrogen atoms of the ligand and two water
14
molecules, undergoes a stepwise deprotonation of coordinated
water molecules, as is shown in the species distribution diagram
of Figure 5, according to the equilibria:
4
+
3+
+
[
Cu (L)(H O) ] T [Cu (L)(H O) (OH)] + H (2)
2 2 4 2 2 3
with a pKa1 of 7.66 ( 0.02, and
3
+
2+
+
[
Cu (L)(H O) (OH)] T [Cu (L)(H O) (OH) ] + H
2 2 3 2 2 2 2
(3)
Figure 4. Frozen solution EPR spectra in MeOH-DMF (5:3, v/v), at -150
with a pKa2 of 7.84 ( 0.02. Both pKa values are rather low and
consistent with the formation of a bis(µ-hydroxo)dicopper(II)
species. It can be noted that the first and the second deproto-
nation have similar pKa values even though the charge decreases
after the first step. This behavior is not surprising. As we
previously noted, the monohydroxo species that forms after
the first deprotonation step preorganizes the complex and makes
easier the formation of the dibridged complex containing the
ring closed Cu2(OH)2 core.
°C, of complex 1 (a) and after adding excess H2O2 at -80 °C, in the presence
of NEt3, to obtain complex 2 (b).
N
coupling with three nitrogen donors (A | ) 11.5 G), are indi-
cative of the presence of five-coordinated Cu(II) centers in a
square-pyramidal geometry. Upon addition of excess H2O2 at
low temperature, the EPR spectrum of the resulting hydroper-
oxide adduct 2 changes to a fairly broad and unresolved signal
at g ≈ 2.12, consistent with a weakly antiferromagnetically
coupled, dinuclear Cu(II) center containing a single bridging
ligand.
9
f
Kinetics. Kinetic studies were performed to characterize both
steps of the double hydroxylation reaction undergone by [Cu2-
4+
Ligand Protonation and Stability Constants of Copper-
(L)] in the presence of hydrogen peroxide. When the reaction
is studied by adding aqueous hydrogen peroxide to an aceto-
nitrile solution of the complex, the development of the near-
UV band near 350 nm due to 4 is preceded by a rapid decrease
in intensity of the broad shoulder near 310 nm (kobs ) 3.3 ×
4+
(
II)-L Complexes. Since the reaction between [Cu2(L)] and
hydrogen peroxide occurs in a mixed aqueous-organic solvent,
it is important to characterize the proton dissociation equilibria
that metal-bound water molecules can undergo in the reaction
conditions. Thus, the protonation and complex forming equilibria
-
2
-1
10 s ) (see Figure 1S in Supporting Information). This is
due to an equilibration process of the species upon addition of
the aqueous reagent, as it is shown by the fact that the process
2
+
with Cu of the ligand L were investigated potentiometrically
in a mixture of acetonitrile-water, 4:1 (v/v). Five protonation
Figure 5. Species distribution in the 2 Cu/L system as a function of pH in acetonitrile/water solution (80:20, v/v): (- -) [Cu2(L)(OH2)2]4+; (s) [Cu2-
3
+
2+
(
L)(OH)] ; (‚‚‚) [Cu2(L)(OH)2] . The dot profile reproduced on the bottom right of the diagram shows the dependence of the rate of hydroxylation of
4+
complex [Cu2(L)] (in ∆Abs350/s units) on the solution pH. The concentration of the complex was 50 µM, while that of H2O2 was 0.2 mM in all experiments.
The reactions were studied in a 4:1 mixture of MeCN/aqueous 5 mM phosphate buffer at 25 ( 0.1 °C; the pH of the solution was controlled by Gran’s
method. The species formed below pH 4 are protonated forms of L and free Cu2 ion.
+
4192 J. AM. CHEM. SOC.
9
VOL. 125, NO. 14, 2003

Hydroxylation by Dicopper(II)−Hydroperoxide Species
A R T I C L E S
Figure 6. Dependence of the initial rates of hydroxylation of 1 (0) and 3 (4) by hydrogen peroxide as a function of the complex concentration (A) and
hydrogen peroxide concentration (B). The insets show the plots referring to 1 on expanded scales.
can be reproduced by simply adding water to an acetonitrile
solution of the complex. To avoid these problems, the kinetic
experiments were performed in a solvent mixture of acetonitrile
and aqueous phosphate buffer at pH 7.5, in a 4:1 ratio (v/v), at
tion exhibited a saturation behavior (Figure 6B). On the basis
of these results, for both hydroxylation steps, the following
simplified mechanism can be proposed:
-
+
25 °C. The first step was analyzed by studying the reaction of
H O h HO + H
(4)
(5)
2
2
2
1
, while the second step was studied employing complex 3.
K
k
3+
4
+
-
Since the first hydroxylation reaction is slower than the second
one, the intermediate compound 3 (Scheme 1) is not ac-
cumulated at room temperature. However, monitoring the
development of the absorption band at 345 nm associated with
compound 4 and considering the initial rate of the reaction (r0),
we could analyze the mechanism of the rate-determining, first
hydroxylation step. On the other hand, by studying the formation
of 4 from 3 and hydrogen peroxide, we could obtain the rate of
the second hydroxylation.
[Cu (L′)] + HO2 y\ z [Cu (L′)OOH] 98 4
2
2
where L′ ) L in the case of 1, and L′ ) 6 in the case of 3. The
kinetic parameters could be obtained by fitting the rate data to
the equation corresponding to the previous preequilibrium
scheme:
k‚K‚[Complex][H O ]
2
2
r0 )
(6)
1 + K‚[H O ]
2
2
The rate dependence of the hydroxylation reactions was
studied varying both the concentration of the complex and that
of hydrogen peroxide. The dependence of the rate versus
complex concentration was found to be linear for both 1 and 3
The following values of the kinetic parameters were obtained
at 25 °C:
First hydroxylation:
(Figure 6A), indicating that both hydroxylation reactions are
k1 ) (5.0 ( 0.1) × 10 s-1 and K ) 9500 ( 400 M
-
3
-1
intramolecular processes. This result accounts for the high
1
1
8
percentage of the O incorporation found in 5 when labeled
hydrogen peroxide was used. The rate dependence of the
hydroxylations of both 1 and 3 on hydrogen peroxide concentra-
Second hydroxylation:
-
1
-1
k2 ) (0.13 ( 0.05) s and K ) 7800 ( 300 M
2
J. AM. CHEM. SOC.
9
VOL. 125, NO. 14, 2003 4193

A R T I C L E S
Battaini et al.
The slopes of the r0 versus [1] and r0 versus [3] plots are in
perfect agreement with the extrapolated values of k and K for
both steps. Thus, the second hydroxylation is about 25 times
faster than the first one, which can be explained considering
that the ligand 6 is considerably electron richer than L. The K1
and K2 constants can be considered as binding constants of
hydrogen peroxide to 1 and 3, respectively. The similarity
between the two values indicates that the peroxide binds to
species with the same ionic charge and, likely, also with the
same binding mode in the two hydroxylation steps.
Scheme 5
The rates of hydroxylation undergone by 1 and 3 display
somewhat different temperature dependences, as shown by the
Eyring plots included in the Supporting Information (Figure 5S).
The activation parameters obtained for the two hydroxylation
steps are the following:
First hydroxylation:
q
-1
q
∆
H ) 39.1 ( 0.9 kJ mol ; ∆S ) -115.7 (
.4 J K mol-1
-
1
2
Second hydroxylation:
q
-1
q
∆
H ) 77.8 ( 1.6 kJ mol ; ∆S ) -14.0 (
-
1
hydroxylation, we can explain the pH dependence of the rate
according to the equilibria reported in Scheme 5, which shows
that different hydroperoxo species can be formed in different
conditions. In the acid pH range, binding of hydrogen peroxide
0
.4 J K mol-1
q
The ∆S values indicate that a more highly constrained transition
state is required for the hydroxylation of 1 than for the
hydroxylation of 3. Though, in the more constrained excited
state configuration, an easier and probably concerted bond-
making and bond-breaking process accompanies the oxygen
transfer reaction from the bound hydroperoxide to the ligand
4
+
to [Cu (L)] is unfavored because strong competition by
2
protons prevents formation of the adduct with the hydroperoxide
anion. Dicopper(II) hydroperoxo adducts can be formed in
neutral and in basic conditions, but in the latter case, the
simultaneous presence of a hydroxo bridge locks in the bound
hydroperoxide, making impossible an intramolecular attack of
this residue to the xylyl C-H bond of the ligand. It is only
near neutral pH that the species containing a single hydroperoxo
ligand exists in appreciable amount. This species has the
conformational mobility and, possibly, also the electrophilicity
to allow an intramolecular hydroxylation to occur. The pH
dependence of the rate was not investigated for 3 because this
compound readily gives rise to the unreactive µ-phenoxo
complex in a neutral medium. Neither doubly bridged µ-hy-
droxo,µ-hydroperoxo nor µ-phenoxo,µ-hydroperoxo complexes
are able to support the aromatic hydroxylation.
q
aromatic ring, as shown by the ∆H data. The smaller activation
entropy in the second hydroxylation is coupled to a larger
activation enthalpy, and this results in a reduced difference in
the rates of the two hydroxylations as the temperature is lowered.
It is interesting to note that the activation parameters found for
the ligand hydroxylation undergone by [Cu2(XYL-H)] in the
presence of O2 (∆H ) 50 kJ mol ; ∆S ) -35 J K mol )
lie somewhat between those found here for the hydroxylations
of 1 and 3 in the presence of H2O2, while smaller ∆H and
larger negative ∆S values were found for the benzylic ligand
hydroxylation undergone by the Cu(I) complex and O2 reported
by Itoh et al.6a But, in the latter case, the association of two
mononuclear units in a dimeric peroxo complex was in-
volved.
2+
q
-1
q
-1
-1 1
q
q
Mechanistic Considerations. On the basis of the results
For the first hydroxylation reaction, the pH dependence of
the rate was also studied. The exact pH of the mixed solvent of
acetonitrile-aqueous buffer was obtained by calibrating the
described previously, it is possible to propose a mechanism for
4+
the double aromatic hydroxylation undergone by [Cu (L)] in
2
the presence of hydrogen peroxide. In Scheme 6, the first
22
electrodes according to Gran’s method. In these experiments,
the reaction rates were studied using a hydrogen peroxide
concentration that gives the highest rate of hydroxylation of 1
hydroxylation step is illustrated, and it is assumed that the
second hydroxylation step proceeds with a similar intramolecular
mechanism. The monohydroxo derivative of the dinuclear
-
4
at pH 7.5 and 25 °C (2.5 × 10 M). Interestingly, as reported
in Figure 5, the rate dependence on pH follows an asymmetric
bell shaped curve. The maximum rate is observed in the pH
range around 8, corresponding to the conditions where the
monohydroxo species [Cu2(L)(OH)]3 is maximized, according
to the species distribution curve shown in the same figure.
Taking into account the deprotonation of hydrogen peroxide,
1
1
complex 13 reacts with hydrogen peroxide to form the η :η -
hydroperoxodicopper(II) complex 14, which performs a re-
giospecific electrophilic attack on the aromatic nucleus of the
ligand. The electrophilic nature of the reaction is shown by the
fact that the second hydroxylation, occurring on the electron
rich phenol nucleus of 3, is faster than the first one on 1. A key
step is obviously the cleavage of the O-O bond of the
hydroperoxo intermediate. A heterolytic cleavage, with a
concerted attack on the aromatic nucleus, is preferred since this
can better account for the regiospecificity of the reaction.
+
which is favored on increasing the pH, and that the [Cu2(L)-
+
(
OH)]3 species is apparently much more reactive in the
(22) Rossotti, F. J. C.; Rossotti, H. J. Chem. Educ. 1965, 42, 375-378.
4
194 J. AM. CHEM. SOC. VOL. 125, NO. 14, 2003
9

Hydroxylation by Dicopper(II)−Hydroperoxide Species
A R T I C L E S
Scheme 6
Homolysis of the hydroperoxide bond would produce a hydroxyl
radical that could easily attack the aromatic nucleus in more
than a single position, particularly in the second step, where
the phenol nucleus of 3 has accessible ortho positions. Inser-
tion of the OH group into the aromatic ring produces 15,
which, upon aromatization, yields the hydroxylated complex
1
6.
Stereochemical aspects are extremely important in the present
double hydroxylation reaction. Using a frontier molecular orbital
description, Solomon hypothesized two possible pathways for
although, in 17 and 18, C5 is farther away from the â-oxygen
atom of the hydroperoxo group than C2, might be attributed to
2
2
the arene hydroxylation performed by the η :η -peroxodicopper-
the fact that the p -orbital component of the arene HOMO at
π
2+ 4c
(
II) complex [Cu2(XYL-H)(O2)] . These depend on the
C5 has positive overlap with the p orbital at the â-oxygen atom,
σ
configuration of the N3CuOOCuN3 coordination units, contain-
ing approximately square-pyramidal copper centers, which
involves either an eclipsed disposition of the two tridentate
nitrogen donors and cis axial ligands, or a staggered disposition
of the same donors and trans axial ligands. In the first case, the
Cu2O2 core is noncoplanar with the arene ring and the peroxide
σ* orbital can interact favorably with the HOMO of the arene.
In the other case, the peroxide O-O atoms and the arene C2
and C5 atoms lie approximately on the same line but the plane
whereas the p -orbital component of the arene HOMO at C2 is
close to the nodal plane of the σ* orbital, which reduces overlap.
π
In addition, electron transfer from the arene HOMO into the
hydroperoxo σ* orbital acts to increase the O-O bond length
and moves the â-oxygen atom toward C5, thus promoting a
concerted O-O cleavage and aromatic hydroxylation at this
position, in agreement with the activation data. Finally, both
hydroxylations may require slight rotations of the arene ring
around a C1-C3 axis in order to increase orbital overlap
(counterclockwise for hydroxylation at C5 and clockwise for
hydroxylation at C2; 19).
of the arene ring is tilted with respect to the Cu2O2 plane, as it
is in [Cu2(XYL-O)(OH)]2
+ 4a
and in 11 (vide supra), and in
these conditions, the peroxide π*σ component of the LUMO
can overlap with the arene HOMO. The second pathway is
favored for the arene hydroxylation undergone by [Cu2(XYL-
2+
4c
H)(O2)] , which occurs at position 2 of the xylyl ring. In
contrast, the hydroxylations proceed via hydroperoxo intermedi-
ates in the present case, and the first hydroxylation occurs at
C5. As shown by the simple MM+ steric interaction mini-
mization performed on 2 using the program Hyper Chem
(Hyper Cube, release 5.01), when the two square-pyramidal
copper(II) centers assume the eclipsed configuration, with cis
axial ligands, the xylyl aromatic nucleus lies just below the OH
group of the bound hydroperoxo moiety, as shown by structures
1
7 and 18.
In this disposition, a direct overlap between the arene HOMO
and the peroxide σ* component is possible (19), and taking into
account that the C2 and C5 atoms give larger orbital contribu-
tions to the HOMO,4 we can explain the regiochemistry of the
two hydroxylations. The observation that, in the first step,
hydroxylation at C5 is favored over hydroxylation at C2,
c
1
1
In contrast to the µ-η :η -hydroperoxodicopper(II) complex,
2
2
the corresponding µ-η :η -peroxodicopper(II) complex 12 was
J. AM. CHEM. SOC.
9
VOL. 125, NO. 14, 2003 4195

A R T I C L E S
Battaini et al.
8
found to be unable to support the hydroxylation of L. This
245 mmol) and traces of dibenzoyl peroxide. The refluxing flask was
exposed to a 150-W tungsten lamp. After 1.5 h, the mixture was cooled
to room temperature, filtered, and washed several times with basic
situation is thus different from that present in Karlin’s [Cu2-
XYL-H)(O2)]2 complex, likely because of slight differences
+
4c
(
water. The organic layer was dried using anhydrous MgSO
removal of the solvent, a light yellow oil was obtained. The crude
product was dissolved in a mixture of CCl (20 mL) and hexane (100
4
, and upon
in the orientation of the pyridine rings in this complex with
respect to the benzimidazole rings of 12. These differences are
determined by the bite angles of the heterocyclic donors and
thus depend on the size of the heterocyclic rings, which differ
in the two cases. As a result, both with eclipsed and staggered
configurations of the chelating arms, the oxygen atoms of the
side-on bound peroxide moiety are too far from the xylyl
aromatic ring to promote an electrophilic attack, in contrast to
Karlin’s system.4 In addition, the orientation of the xylyl
aromatic ring may be different in the two systems because it
may be controlled by ring stacking interactions with the
heterocyclic rings.
In conclusion, the present investigation describes a double
arene hydroxylation mediated by hydroperoxodicopper(II) com-
plexes bearing potential relevance for the mechanism of copper
monooxygenases and, in particular, tyrosinase. Although the
mechanisms proposed for tyrosinase generally involve a peroxo
4
mL), and the solution was left for 1 day at 0 °C. Recrystallization of
the white crystalline material that formed, in MeOH at -15 °C, yielded
1
the purified product (yield 10%). H NMR (CDCl
3
): δ 7.42 (d, 2H,
phenyl-H), 7.25 (m, 2H, phenyl-H), 4.4 (s, 4H, CH
2
-Br), 2.48 (s, 3H,
CH CO).
3
Acetic Acid 3,5-Dimethylphenyl Ester. 3,5-Dimethylphenol (15 g,
0.123 mmol) was slowly added to a mixture of acetic anhydride and
pyridine 1:1 (v/v) kept at 0 °C. Then, the resulting yellow solution
was allowed to stir for 24 h at room temperature. Removal of the solvent
c
followed by distillation under vacuum afforded the ester (yield 60%)
1
(
bp 72 °C at 0.1 mmHg). H NMR (CDCl
3
): δ 6.92 (s, 1H, phenyl-
), 2.32 (s, 3H, CH CO).
H), 6.73 (s, 2H, phenyl-H), 2.38 (s, 6H, CH
3
3
Acetic Acid 3,5-Bis(bromomethyl)phenyl Ester. Bromination of
acetic acid 3,5-dimethylphenyl ester (12 g, 73 mmol) was carried out
by refluxing it in CCl
4
(500 mL) for 1.75 h with N-bromosuccinimide
(26 g, 146 mmol) in the presence of a trace dibenzoyl peroxide. The
2
intermediate, a rearrangement of the bound peroxide to
hydroperoxide, with concomitant proton transfer, has been
suggested by inhibition studies of the monophenolase reaction
refluxing mixture was also irradiated with a 150-W tungsten lamp. Then,
the reaction mixture was cooled and washed with basic water, and the
organic layer was dried over MgSO . Upon rotary evaporation, the
4
23
and, more recently, by transient-phase kinetics of the diphenolase
solvent was removed yielding an orange oil. This was dissolved in a
reaction.24
mixture of CCl (20 mL) and hexane (100 mL), and the solution was
4
cooled for 1 day at 0 °C. The yellow crystalline solid that formed was
Experimental Section
recrystallized with hot methanol giving pure acetic acid 3,5-bis-
1
(bromomethyl)phenyl ester (yield 34%). H NMR (CDCl
3
): δ 7.31 (s,
All reagents were purchased from commercial sources and used as
received unless otherwise noted. DMF and acetonitrile were purified
1
3
H, phenyl-H), 7.13 (s, 2H, phenyl-H), 4.48 (s, 4H, CH
H, CH CO).
,5-Bis({bis[2-(1-methyl-1H-benzimidazol-2-yl)ethyl]amino}-
methyl)phenol (6). This compound was obtained by reacting bis[2-
1-methyl-1H-benzimidazol-2-yl)ethyl]amine (330 mg, 1 mmol) with
,5-bis(bromomethyl)phenyl ester (161 mg, 0.46 mmol) at reflux
2
-Br), 2.33 (s,
3
9
f
as described previously. The ligand L-66 and the corresponding
dinuclear complex [Cu (L)][ClO were prepared according to the
published procedures. Bis[2-(1-methyl-1H-benzimidazol-2-yl)ethyl]-
3
2
4 4
]
10b
(
3
2
5
amine was prepared as described previously. Infrared spectra were
recorded on a Perkin-Elmer Spectrum BX FT/IR instrument. Elemental
analyses were obtained from the microanalytical laboratory of the
Chemistry Department in Milano. NMR spectra were recorded on a
Bruker Avance 400 spectrometer. EPR spectra were measured in frozen
solutions using a Varian E-109 spectrometer operating at X-band
frequencies. Optical spectra were measured on HP 8452A and 8453
diode array spectrophotometers. Electrospray ionization MS spectra
were acquired using a Finnigan MAT system equipped with an ion
trap detector. The solution was introduced into the electrospray source
temperature in dry DMF (30 mL), under argon, and in the presence of
anhydrous sodium carbonate (110 mg, 1 mmol), as a base, for 24 h.
Then, a NaHCO -saturated aqueous solution (10 mL) was added. The
3
resulting brown mixture was stirred for 24 h. The solvent was removed
by rotary evaporation, and the crude orange oil was chromatographed
on silica gel, with MeOH/CH
2
Cl
2
8:2 (v/v) as eluent, to afford 6 as a
OD): δ 7.53 (m, 4H,
white powder (yield 26%). 1H NMR (CD
3
benzimidazole-H), 7.25 (m, 4H, benzimidazole-H), 7.18 (m, 8H,
benzimidazole-H), 6.62 (s, 2H, phenyl-H), 6.41 (s, 1H, phenyl-H), 3.53
-
1
at 10 µL min using a syringe pump instrument. The ESI source was
operated at 3.5 kV, the capillary temperature was set at 180 °C, and its
voltage was at 10 V; the experiments were performed in positive ion
mode.
13
(
s, 12H, N-CH
3
), 3.47 (s, 4H, CH
2
-phenyl), 2.85 (m, 16H, CH
2
).
C
NMR (CD
(
3
OD): δ 157.5 (C), 154.5(C), 142.5 (C), 140.9 (C), 136.2
C), 122.4 (CH), 122.2 (CH), 121.5 (CH), 117.8 (CH), 113.6 (CH),
1
09. 8 (CH), 58.4 (CH
2
2 3 2
), 51.8 (CH ), 29.2 (CH ), 25.3 (CH ). UV-vis
Acetic Acid 2,6-Dimethylphenyl Ester. 2,6-Dimethylphenol (20 g,
MeOH, λmax, nm (ꢀ, M-1 cm )]: 252 (24 300), 276 (26 000), 282
-1
[
(
1
1
64 mmol) was added to a mixture of acetic anhydride and pyridine
:1 (v/v) (40 mL) in an ice bath. The pale yellow solution was stirred
-1
23 900), 310 (1450). IR (Nujol mull, cm ): 3549 m, 3470 m, 3412
s, 1763 w, 1736 w, 1638 m, 1616 m, 1596 m, 1509 w, 1332 w, 1289
w, 1237 w, 1210 m, 1152 w, 1128 w, 1095 m, 1007 m, 969 w, 874 w,
overnight at room temperature. The solvent was then removed by rotary
evaporation. The remaining oil was washed with basic water and
distilled under vacuum (bp 87 °C at 0.05 mmHg) (yield 75%). H NMR
8
47 w, 802 w, 766 w, 743 s, 722 m.
1
2
,6-Bis({bis[2-(1-methyl-1H-benzimidazol-2-yl)ethyl]amino}-
(
CDCl
3
3
): δ 7.1 (m, 3H, phenyl-H), 2.43 (s, 3H, CH CO), 2.15 (s, 6H,
methyl)phenol (7). A mixture of acetic acid 2,6-bis(bromomethyl)-
phenyl ester (0.97 g, 2.8 mmol), bis[2-(1-methyl-1H-benzimidazol-2-
yl)ethyl]amine (2 g, 6 mmol), anhydrous sodium carbonate (0.65 g, 6
mmol), and dry DMF (60 mL) was refluxed under argon for 51 h. The
crude oily product obtained upon evaporation to dryness under vacuum
of the mixture was found to be a mixture of 7 and acetic acid 2,6-bis-
CH ).
3
Acetic Acid 2,6-Bis(bromomethyl)phenyl Ester. Bromination of
acetic acid 2,6-dimethylphenyl ester (20 g, 122 mmol) was carried out
by refluxing it in CCl (500 mL) with N-bromosuccinimide (43.5 g,
4
(
23) Conrad, J. S.; Dawso, S. R.; Hubbard, E. R.; Meyers, T. E.; Strothkamp,
({bis[2-(1-methyl-1H-benzimidazol-2-yl)ethyl]amino}methyl)phenyl es-
K. G. Biochemistry 1994, 33, 5739-5744.
ter, as determined by TLC (MeOH/CH Cl , 8/2, v/v) and NMR. The
2
2
(24) Rodriguez-Lopez, J. N.; Fenoll, L. G.; Garcia-Ruiz, P. A.; Varon, R.; Tudela,
J.; Thorneley, R. N. F.; Garcia-Canovas, F. Biochemistry 2000, 39, 10497-
crude mixture was stirred for 10 h at room temperature in a solution
of MeOH (10 mL) and sodium bicarbonate-saturated water (10 mL).
Then, upon reduction of the volume of the solution by rotary
1
0506.
(25) Casella, L.; Gullotti, M.; Redaelli, R.; Di Gennaro, P. J. Chem. Soc., Chem.
Commun. 1991, 1611-1612.
4196 J. AM. CHEM. SOC.
9
VOL. 125, NO. 14, 2003

Hydroxylation by Dicopper(II)−Hydroperoxide Species
A R T I C L E S
evaporation, a light yellow solid precipitated. This was purified by
chromatography on silica gel, by eluting with a gradient of MeOH/
CH Cl , which was varied from 1:1 to 8:2 (v/v), yielding pure 7 (yield
2 2
reflection, cm-1) 3100 s, 2950 s, 1610 m, 1492 m, 1461 m, 1330 m,
1306 w, 1292 w, 1238 w, 1100 s, 1009 m, 929 w, 869 w, 750 m, 713
w, 622 m.
1
2
1
0%). H NMR (CDCl
2H, benzimidazole-H), 7.02 (d, 2H, phenyl-H), 6.68 (t, 1H, phenyl-
-phenyl), 3.45 (s, 12H, CH -N), 3.15 (t, 8H, CH ),
). 13C NMR (CDCl
): δ 156.1 (C), 153.8 (C), 143.0
C), 136.0 (C), 129.5 (CH) 124.5 (C), 122.7 (CH), 122.53 (CH), 119.5
CH), 118. 9 (CH), 109.8 (CH), 56.0 (CH ), 51.5 (CH ), 30.2 (CH ),
, λmax, nm (ꢀ, M- cm )]: 256 (22 300),
3
): δ 7.68 (m, 4H, benzimidazole-H), 7.20 (m,
Isolation of the Hydroxylated Ligand, 2,6-Bis({bis[2-(3-methyl-
1H-benzimidazol-2-yl)ethyl]amino}methyl)benzene-1,4-diol (5). The
(L)]4 (100 mg, 0.07 mmol) was dissolved in acetonitrile
+
H), 3.8 (s, 4H, CH
.01 (t, 8H, CH
2
3
2
complex [Cu
2
3
(
(
2
2
2
3
(25 mL), and concentrated aqueous hydrogen peroxide (0.28 mmol)
was added. The solution was allowed to react for 24 h under stirring
at room temperature. Then, it was concentrated under reduced pressure
2
2
3
1
-1
2
6.3 (CH
2
). UV-vis [CH
2
Cl
2
and chromatographed on a Sephadex LH-20 column (1.5 × 40 cm ).
-
1
78 (22 900), 284 (21 150), 336 (500). IR (Nujol mull, cm ): 3286
Both fractions which eluted were collected and combined; the solvent
volume was reduced through rotary evaporation. The resulting con-
centrated solution was added to a chilled mixture of concentrated
s, 2726 w, 1732 w, 1614 m, 1594 m, 1505 m, 1331 w, 1261 m, 1233
w, 1150 w, 1123 m, 1095 m, 1090 m, 1007 m, 1006 m, 875 w, 842 w,
8
02 w, 766 m, 742 s.
Copper Complexes 3[ClO
aqueous ammonia and CH
aqueous layer was separated and further extracted with cold CH
(50 mL, 3 times). The organic extracts were combined, washed with
water, and then dried over MgSO . Every step of the decomposition
and extraction procedure was carried out at 5 °C using aluminum foil
protection of the glassware from light exposure, to prevent any oxidative
and photochemical decomposition of 5. The final organic solution was
purified by flash chromatography on a silica gel column, eluting with
2
Cl
2
1:1 (v/v, 100 mL) under stirring. The
Cl
2
4
]
4
and 8[ClO
4
]
3
. To a stirred solution
2
of Cu(ClO
phosphate buffer pH 5.1, 30:1 (v/v) (15 mL), a solution of the ligand
(32 mg, 0.04 mmol) in MeOH (5 mL) was gently added. The solution
4
)
2
2
‚6H O (30 mg, 0.09 mmol) in methanol/aqueous 50 mM
4
6
was stirred for 30 min, and the light green precipitate that immediately
formed was filtered, washed with a small portion of tetrahydrofuran,
and dried under vacuum (yield 35%). This precipitate corresponds to
3
[ClO
4
]
4
. The light brown mother solution was allowed to stand in a
2
a solvent mixture of Et O-iPrOH (5:4, v/v) to which an increasing
refrigerator until a second precipitate formed. A light brown powder
was collected by filtration and dried under vacuum (yield 40%). This
amount of concentrated ammonia, from 1% to 15% (v/v), was added.
The fractions containing the product were identified through the
characteristic UV absorption of the benzimidazole residues (at 284,
276, and 254 nm) and collected. Upon evaporation of the solvent, a
light yellow, somewhat air-sensitive oil was obtained (yield > 80%).
product corresponds to 8[ClO
[ClO . Anal. Calcd for C48
N, 10.69. Found: C, 44.02; H, 3.95; N, 10.31. IR (KBr, diffuse
4 3
] .
3
4
]
4
52 10 4 2
H N O17Cl Cu : C, 44.01; H, 4.00;
-
1
1
H NMR (CDCl ): δ 7.69 (m, 4H, benzimidazole-H), 7.25 (m, 14H,
reflection, cm ): 3420 br, 2936 s, 2023 m, 1658 s, 1617 m, 1496 m,
485 w, 1458 s, 1414 m, 1390 w, 1332 w, 1301 w, 1260 m, 1100 br
vs, 940 m, 871 m, 801 m, 760 m, 751 s, 627 s.
[ClO . Anal. Calcd for C48 Cu
N, 11.58. Found: C, 47.25; H, 4.15; N, 11.31. IR (KBr, diffuse
3
1
2
benzimidazole-H + phenyl-H), 3.72 (s, 4H, CH -phenyl), 3.6 (s, 12H,
13
CH
54.13 (C), 153.85 (C), 151.03 (C), 141.05 (C), 136.03 (C), 130.08
CH), 122.71 (CH), 122.53 (CH), 119.36 (CH), 109.67 (CH), 58.81
3
2 2 3
-N), 3.1 (t, 8H, CH ), 3.22 (t, 8H, CH ). C NMR (CDCl ): δ
1
8
4
]
3
H
51
10
N O13Cl
3
2
: C, 47.67; H, 4.25;
(
-
1
-
1
2 2 3 2
(CH ), 53.82 (CH ), 30.20 (CH ), 26.37 (CH ). IR (liquid film, cm ):
reflection, cm ): 3410 br, 2959 w, 2013 w, 1615 m, 1596 w, 1495
m, 1483 w, 1459 m, 1415 w, 1332 m, 1308 w, 1293 w, 1262 w, 1238
w, 1150 s, 1090 br vs, 940 w, 870 w, 854 w, 803 w, 793 w, 760 s, 751
s, 627 s.
3
1
1
340 s, 3059 w, 3007 w, 2937 m, 2852 w, 1666 m, 1629 m, 1510 s,
469 s, 1442 s, 1402 m, 1330 m, 1282 m, 1238 m, 1217 m, 1151 w,
126 w, 1095 w, 1028 w, 1008 w, 879 w, 753 s, 669 m. UV-vis
-
1
-1
[MeOH, λmax, nm (ꢀ, M cm )]: 248 (31 400), 276 (26 600), 282
Complex 9[ClO
2 mL) was slowly added to a solution of Cu(ClO
.23 mmol) in dry acetonitrile (10 mL). The resulting green-yellow
4
]
3
. A sample of 7 (80 mg, 0.11 mmol) in CH
2 2
Cl
+
(25 500), 340 (3200). ESI-MS, m/z: 643.4 [MH-(CH
2
CH
2
Bz)] ,
(
0
4
)
2
‚6H O (76 mg,
2
+
+
8
01.3 [MH] , 823.5 [M + Na] .
The 18O-labeled derivative of 5 was prepared by following the same
solution was stirred under argon for 30 min at room temperature. Diethyl
ether was added until turbidity was observed. The solution was then
cooled to 0 °C for 1 h, and the green solid that precipitated was filtered
procedure described previously but using smaller amounts of the
reagents: [Cu
2
(L)]4 (0.004 mmol) and H
+
2
18
O
2
(90% O, 2% in H
18
2
O,
+
Icon Isotope, 0.016 mmol). ESI-MS, m/z: 647.5 [MH-(CH
05.5 [MH] , 827.3 [M + Na] .
Potentiometric Determinations. Potentiometric titrations of L in
2 2
CH Bz)] ,
and dried under vacuum (yield 25%). Anal. Calcd for C48
51 10 13
H N O -
+
+
8
Cl Cu : C, 47.67; H, 4.25; N, 11.58. Found: C, 47.88; H, 4.64; N,
3
2
-
1
1
2
1
1.62. IR (KBr, diffuse reflection, cm ): 3550 br, 3064 s, 2976 s,
881 m, 2021 m, 1617 m, 1596 w, 1560 w, 1533 m, 1483 m, 1459 s,
418 m, 1361 w, 1333 m, 1292 w, 1288 w, 1249 m, 1090 br vs, 1011
the absence and in the presence of copper(II) ions were performed,
with apparatus and methods described elsewhere,26 in a mixture of
acetonitrile-water 80:20 (v/v), made 0.1 M in ionic strength with
w, 982 w, 929 m, 866 w, 851 m, 760 s, 753 s, 625 s, 569 w.
Isolation of the Hydroxylated Complexes 4[ClO and 10[ClO
(L)] (100 mg, 0.07 mmol) was dissolved in
4
NaClO (50 mL) at 298 K. The electrodes were dipped for more than
]
4 3
4 2
] .
4
+
1 h in the previously mentioned solvent mixture before standardization
The complex [Cu
2
2
6
of the system, which was also made as previously described. The
HYPER-QUAD27 program was used to process the data and calculate
both the protonation and stability constants.
acetonitrile (25 mL), and a concentrated aqueous solution of hydrogen
peroxide (0.28 mmol) was added. The solution was allowed to react
for 24 h under stirring at room temperature. Then, it was concentrated
under reduced pressure and chromatographed on a Sephadex LH-20
Solution IR Measurements. The hydroxylation reaction was
investigated by IR spectroscopy in solution using a cell with a 0.05-
2
column (1.5 × 40 cm ). Two bands were observed and collected
2
cm path length and CaF optical windows (HELLMA). The reaction
separately: a major green fraction corresponding to complex 4 and a
minor light brown fraction corresponding to 10. The products were
obtained upon rotary evaporation of the solutions resulting from
chromatography.
solutions were prepared in a separate flask by adding 3 molar equiv.
18
18
hydrogen peroxide (H
O, Icon Isotope) to an acetonitrile solution of [Cu
00 µL). The solutions were allowed to react for 2 h, and then they
2 2 2 2
O , 30% in water or H O , 90% O, 2% in
4
+
H
2
2
2
(L)] (20 mM,
4
[ClO
4 3 54 11 3 2
] ‚MeCN. Anal. Calcd for C50H N O14Cl Cu : C, 47.42;
were injected into the cell, and the spectra recorded. Blank spectra were
recorded immediately after mixing of the reagents.
H, 4.30; N, 12.17. Found: C, 47.88; H, 4.63; N, 12.35. IR (KBr diffuse
-
1
reflection, cm ): 3400 br, 2050 w, 1619 m, 1501 m, 1485 m, 1461
m, 1454 m, 1418 m, 1304 w, 1291 w, 1240 w, 1100 s, 933 w, 859 w,
(
26) Monzani, E.; Quinti, L.; Perotti, A.; Casella, L.; Gullotti, M.; Randaccio,
7
48 m, 625 m.
0[ClO . Anal. Calcd for C48
N, 12.25. Found: C, 50.84; H, 4.85; N, 12.64. IR (KBr diffuse
L.; Geremia, S.; Nardin, G.; Faleschini, P.; Tabb `ı , G. Inorg. Chem. 1998,
37, 553-562.
1
]
4 2
52 10 2 2
H N O11Cl Cu : C, 50.44; H, 4.59;
(27) Sabatini, A.; Vacca, A.; Gans, P. Coord. Chem. ReV. 1982, 120, 389-
405.
J. AM. CHEM. SOC.
9
VOL. 125, NO. 14, 2003 4197

A R T I C L E S
Battaini et al.
Low-Temperature Spectroscopy. The low-temperature spectra
(0.17 mmol) in acetonitrile (8 mL). The dark green solution was stirred
for 1 h, and then small amounts of diethyl ether were added until
turbidity was observed. The mixture was allowed to stand overnight
in a refrigerator. The green precipitate was filtered off, and vapors of
diethyl ether were diffused into the filtrate. The crystallyzed material
(
-40 °C and above) recorded during the reaction of [Cu
2
(L)] 4 with
+
hydrogen peroxide were obtained with a custom designed immersible
fiber-optic quartz probe (HELLMA) fitted to a Schlenk vessel and
connected with the HP 8452A diode array spectrophotometer. The
experiment was performed using a solution of the complex in freshly
distilled acetonitrile (0.15 mM) cooled in a chlorobenzene-acetone
8:2, v/v) liquid nitrogen cryogenic bath. To a chilled (-40 °C) solution
of the complex (25 mL), 30% aqueous hydrogen peroxide was added
>100-fold molar excess), which quickly froze. Gradual warming of
the mixture by swirling outside the bath was followed spectrophoto-
metrically. When some changes became detectable in the spectra, the
mixture was promptly chilled again to -40 °C, and spectra of the
reaction mixture were recorded during 1 h. Then, the solution was
allowed to warm gently to -30 °C, maintained at this temperature for
about 1 h, and then slowly warmed again. Around -20 °C, the spectral
features started to change again, so warming was stopped and the
temperature was maintained at about -20 °C until the reaction appeared
to be complete. Then, the solution was warmed gradually again to room
temperature, but no further spectral changes were observed.
consists of light green plates of 11[CLO
4
]
2 3 2 5 2
‚CH CN‚(C H ) O. These
green plates decompose within a few minutes, presumably due to the
loss of solvent molecules incorporated in the crystals.
(
X-ray crystallographic studies were done using an imaging plate
diffraction system (IPDS) from STOE & CIE (Mo KR radiation). The
lattice parameters were obtained from 7998 reflections in the range of
24° e 2θ e 38° (a ) 13.994(2) Å, b ) 15.092(2) Å, c ) 15.314(2) Å,
(
3
R ) 77.75(1)° , â ) 78.66(1)°, γ ) 74.26(1)°, V ) 3008.4(5) Å (T )
-
3
150 K), Z ) 2, Fber ) 1.371 g cm , triclinic, space group P 1h (No. 2),
-
1
µ ) 0.86 mm , 23 088 measured reflections in the range of 3° e 2θ
e 55°, 10 896 independent reflections used for refinement (Rint
0.0621), 8958 independent reflections with F > 4σ(F )). Structure
)
o
o
solution was performed with SHELXS-94, and structure refinement
2
was done with full-matrix least squares against F using SHELXL-97
2
2
2
(726 refined parameters, w ) 1/[σ (F
o
) + (0.0863P) + 1.07P], R
1
Kinetics. The kinetics of hydroxylation of 1 and 3 by hydrogen
peroxide were studied in a magnetically stirred and thermostated optical
cell with 1-cm path length, at 25 ( 0.1 °C. The reactions were followed
through the development of the near-UV band at 344 nm characteristic
of the product. To reduce the effect of noise, the difference between
the absorbance at 344 nm and that at 820 nm, where no absorption
occurs, was monitored. The solvent employed for these measurements
was a mixture of acetonitrile and aqueous 5 mM phosphate buffer pH
for 8958 reflections with F > 4σ (F ) ) 0.0489, wR2 for all 10 896
o
o
reflections ) 0.1436, GOOF ) 1.048, residual electron density )
3
0.64/-0.76 e/Å . All non-hydrogen atoms except one carbon atom of
the diethyl ether molecule which is disordered were refined with
anisotropic displacement parameters. All hydrogen atoms were placed
in ideal positions and refined with isotropic displacement parameters
[Uiso ) 1.2Ueq(Caromatic/methylene) ) 1.5Ueq(Cmethyl/hydroxyl)] using the riding
model with the parameters C-H (aromatic) 0.95 Å, C-H (methylene)
0.99 Å, C-H (methyl) 0.98 Å, and C-H (hydroxyl) 0.85 Å. The
position of the hydrogen atom bound to O2 was located first from a
difference map. Because refinement with varying coordinates and
displacement parameters was not possible, this atom was also refined
using a riding model. The crystal contains additional solvent molecules,
presumably diethyl ether, in holes of the structure for which no
successful structure model can be found. Therefore, the data were
corrected for the disordered solvent using the “Squeeze” option in
Platon. The analysis yields a total potential solvent accessible volume
7
.5 4:1 (v/v). The dependence of the initial rate of hydroxylation as a
function of the complex concentration was studied maintaining the H O
2
2
concentration 0.2 mM for 1 and 0.35 mM for 3, while the concentration
of the complexes was varied between 20 µM and 120 µM. The
dependence of the initial rate of hydroxylation as a function of hydrogen
peroxide concentration was investigated by adding a few microliters
2 2
of concentrated H O (final concentration varied between 0.1 and 1
mM) to the solutions of the complexes, which were 50 µM for 1 and
4
1
6 µM for 3. The experiments to determine the hydroxylation rate of
as a function of pH were carried out on solutions 50 µM in the
3
of 335 Å . The electron count per cell 54 was not included in the
complex and 200 µM in H
2
O
2
, by rapidly adding small quantities of
calculated density and in F(000).
NaOH or HClO to the appropriate pH value and immediately starting
4
Acknowledgment. This work was supported by grants from
FAR (University of Pavia) and the Italian CNR. Support by
the European COST and INTAS programs is also acknowledged.
L.S. was supported by a fellowship by CIRCMSB.
the optical readings. The exact pH of the solvent mixture was
determined by calibration of the electrode through the addition of
measured quantities of standard perchloric acid solution, according to
2
2
Gran’s method. The pH interval covered by the kinetic experiments
was between 5 and 10. The initial rates were determined from the
absorbance versus time curves, typically in the initial 50-60-s reaction
time.
Supporting Information Available: Crystallographic data
including experimental details of the structure determination and
tables with bond lengths, bond angles, atomic coordinates as
well as equivalent isotropic and anisotropic displacement
parameters are provided in PDF format; also included are five
figures presenting optical spectral changes observed in the
The kinetics of the fast initial phase preceding the hydroxylation
was studied using an Applied Photophysic stopped flow system (RS-
1
00), with a dead time of 1 ms and a 1-cm path length, coupled with
a HP 8452A diode array spectrophotometer. In these experiments, a
CH CN solution of 1 (100 µM) was mixed with a 2 mM hydrogen
13
hydroxylation of complex 1, the aromatic region of the C-
3
1
peroxide solution obtained by diluting the concentrated aqueous solution
with acetonitrile. The experiment was repeated in the same conditions
using a 2 mM solution of water in acetonitrile. The cell and the reactant
solutions were thermostated at 25 ( 0.1 °C. Mixing of the solutions in
the stopped flow cuvette reduces the reactant concentrations to one-
half.
H HMQC NMR spectrum of the dihydroxylated ligand 5, the
1
8
MS spectra of 5 and its O analogue, the low-temperature
spectral changes associated with the formation and reaction of
the hydroperoxodicopper(II) intermediate 2, and the Eyring plots
related to the hydroxylation of 1 and 3. This material is available
free of charge via the Internet at http://pubs.acs.org.
Crystal Structure Determination of 11. A sample of solid Cu-
4 2 2
(ClO ) ‚6H O (0.34 mmol) was added to a solution of O-acetylated 7
JA0280776
4198 J. AM. CHEM. SOC.
9
VOL. 125, NO. 14, 2003