Fine-tuning the electronic properties of highly stable organometallic CuIII complexes containing monoanionic macrocyclic ligands

Article abstract of DOI:10.1002/chem.200500088

A family of highly stable organometallic Cu^III complexes with monoanionic triazamacrocyclic ligands (L′) with general formula [CuL′]⁺ have been prepared and isolated, and their structural, spectroscopic, and redox properties thorough

Full text of DOI:10.1002/chem.200500088

Fine-Tuning the Electronic Properties of Highly Stable Organometallic
CuIII Complexes Containing Monoanionic Macrocyclic Ligands
Raül Xifra,[a] Xavi Ribas,*[a] Antoni Llobet,*[b] Albert Poater,[c] Miquel Duran,[c]
Miquel Solà,*[c] T. Daniel. P. Stack,[d] Jordi Benet-Buchholz,[e] Bruno Donnadieu,[f]
[h]
JosØ Mahía,[g] andTeodor Parella
Abstract: A family of highly stable or-
ganometallic CuIII complexes with
monoanionic triazamacrocyclic ligands
(Li) with general formula [CuLi]+ have
been prepared and isolated, and their
structural, spectroscopic, and redox
properties thoroughly investigated. The
HLi ligands have been designed in
order to understand and quantify the
electronic effects exerted by electron
donor and electron-withdrawing groups
on either the aromatic ring or the cen-
tral secondary amine or on both. In the
solid state the CuIII complexes were
mainly characterized by single-crystal
X-ray diffraction analysis, whereas in
solution their structural characteriza-
tion was mainly based on 1H NMR
spectroscopy given the diamagnetic
nature of the d8 square-planar CuIII
complexes. Cyclic voltammetry togeth-
er with H NMR and UV/Vis spectros-
1
copy have allowed us to quantify the
electronic effects exerted by the ligands
on the CuIII metal center. A theoretical
analysis of this family of CuIII com-
plexes has also been undertaken by
DFT calculations to gain a deeper in-
sight into the electronic structure of
these complexes, which has in turn al-
lowed a greater understanding of the
nature of the UV/Vis transitions as
well as the molecular orbitals involved.
ꢀ
Keywords: C H
copper(iii) complexes
activation
·
density
electro-
·
·
functional calculations
chemistry · spectroscopic properties
Introduction
special interest since these are the major constituents of nat-
ural gas and petroleum.[2] However, the full potential of
TMC in profitable practical applications, for example, in cat-
alytic conversion to more valuable products, has not yet
been realized.[3]
The controlled activation of small and relatively inert mole-
cules by transition-metal complexes (TMC) under mild con-
ditions is a very active area of research that has made im-
pressive advances over the last 20 years.[1] In particular, the
By choosing the appropriate ligands and metal, a remark-
able degree of control over the activation processes can be
ꢀ
activation of C H bonds in saturated hydrocarbons is of
[a]Dr. R. Xifra, Dr. X. Ribas
[e]Dr. J. Benet-Buchholz
Departament de Química, Universitat de Girona
Campus de Montilivi, 17071, Girona (Spain)
E-mail: xavi.ribas@udg.es
Institut Català d’Investigació Química (ICIQ)
Avgda. Països Catalans, s/n, 43007 Tarragona (Spain)
[f]Dr. B. Donnadieu
[b]Prof. A. Llobet
Service de Cristallochimie, Laboratoire de Chimie de Coordination
UPR CNRS 8241, Route de Narbonne 205, 31077, Toulouse Cedex 4
(France)
Departament de Química, Universitat Autònoma de Barcelona
Bellaterra, 08193 Barcelona (Spain)
Fax : (+34)935-813-101
[g]Dr. J. Mahía
E-mail: antoni.llobet@uab.es
Servicios Xerais de Apoio  Investigación
Universidade da CoruÇa, 15071 A CoruÇa (Spain)
[c]A. Poater, Prof. M. Duran, Prof. M. Solà
Institut de Química Computacional, Universitat de Girona
Campus de Montilivi, 17071, Girona (Spain)
Fax : (+34)972-418-356
[h]Dr. T. Parella
Servei de RMN, Universitat Autònoma de Barcelona
Bellaterra, 08193 Barcelona (Spain)
E-mail: miquel.sola@udg.es
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
[d]Prof. T. D. P. Stack
Department of Chemistry, Stanford University
Stanford, CA 94305–5080 (USA)
5146
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/chem.200500088
Chem. Eur. J. 2005, 11, 5146 – 5156

FULL PAPER
accomplished under mild conditions, which in turn can lead
to the improved efficiency and selectivity of the whole pro-
cess.[4]
ꢀ
In a recent paper we presented a novel aryl C H activa-
tion process involving CuII complexes that leads to the for-
mation of relatively unusual and stable CuIII complexes con-
taining a monoanionic ligand.[5]
Few d8 CuIII complexes have been structurally character-
ized by single-crystal X-ray diffraction and those that have
all exhibit a square-planar type of geometry. The higher oxi-
dation state is stabilized either by the presence of a trianion-
ic ligand derived from the deprotonation of tripeptides,[6]
corroles[7] or confused porphyrins,[8] or by the participation
of several ligands supplying overall at least three negative
charges, such as the trifluoromethanate ligand.[9] The differ-
ent factors that govern CuIII stabilization and its reactivity
are of considerable interest in biology since CuIII is the
active center of certain metalloproteins[10] and also because
CuIII is thought to be responsible for the oxidative degrada-
tion and site-specific cleavage of DNA.[11] Furthermore CuIII
has been postulated as an intermediate in the oxidation of
saturated hydrocarbons using copper complexes as oxygen
and/or hydrogen peroxide activators, a variation of the so-
called Gif chemistry developed by Barton et al.[12]
Scheme 2. Synthetic strategy for the preparation of the triazamacrocyclic
ligand HL3.
HL4, and HL6, which contain Me- and NO2-substituted aryl
groups, are described and characterized for the first time in
this work, whereas the other three have been reported pre-
viously.[5,15] The detosylation process can also be carried out
with HBr/HAcO/phenol to give the same organic com-
pounds but with better yields.[13] The new HLi ligands have
been thoroughly characterized by the usual analytical and
spectroscopic techniques and the data are presented in the
Supporting Information.
Herein we present the synthesis and characterization of a
Under anaerobic conditions at room temperature, the HLi
ligands react instantaneously with CuII salts (triflate or per-
chlorate) in acetonitrile to afford the corresponding unstable
[CuII(HLi)]2+ complexes, which quantitatively disproportion-
ate to generate a colorless CuI and an orange CuIII complex
[Eq. (1)], both of which are diamagnetic.
family of [CuIIILi]2+ macrocyclic complexes, generated via
II
i
2+
ꢀ
C H activation using the corresponding [Cu (HL)] com-
plexes, containing the ligands displayed in Scheme 1. These
ligands have been chosen to enable us to gain some under-
standing of the influence of steric and electronic effects on
the properties of the CuIII complexes.
MeCN, RT
2þ
III
i
2þ
2½CuIIðHLiފ
½Cu ðL ފ2þ þ ½CuIðH2Liފ
ꢀꢀꢀꢀꢀ!
ꢀꢀꢀꢀꢀ
ð1Þ
Diffusion of diethyl ether into the resulting acetonitrile
solution overnight results in the precipitation of red micro-
crystals of [CuIII(Li)]2+ (1–6, i=1–6) in nearly quantitative
yield. These six CuIII complexes, 1–6, are all stable in the
solid state and in both protic (H2O, pH 1–7; MeOH) and
aprotic solvents (MeCN, CH2Cl2) under aerobic conditions
except for 4 which decomposes in solution at room tempera-
ture as a result of the strong electron-withdrawing nature of
the nitro substituent. Under basic conditions the CuIII com-
plexes readily transform into the corresponding dinuclear di-
Scheme 1. Triazamacrocyclic ligands HLi (i=1–6).
phenoxo complexes of general formula [CuI2I(m-OLi)2]2+
,
which will be reported elsewhere.
Results andDiscussion
1H NMR spectroscopic analysis of the reaction mixture,
performed by using CD3CN as the solvent, allowed us to
confirm that the reaction proceeded by disproportionation
by simple integration of the CuI and CuIII resonances. Fur-
thermore, optical titration of the reaction mixture with
excess phenanthroline also confirms a 50% conversion of
CuII to CuI.
Synthesis andX-ray structures : The triazamacrocyclic li-
gands used in this work are depicted in Scheme 1. In gener-
al, these ligands were prepared by the cyclization of the cor-
responding 1,3-bis(bromomethyl)benzene with the appropri-
ate tosylated triamine following a similar methodology (see
Scheme 2 for the particular case of HL3) to that first de-
scribed by Bencini et al.[13] for related compounds. The tosy-
lated triazamacrocycles were then treated with lithium sus-
pended in liquid ammonia[14] to afford the triazamacrocyclic
ligands HLi in moderate-to-good yields. The ligands HL3,
Complex 2 reacts with Clꢀ at room temperature to form a
pentacoordinate copper complex, [CuIII(L2)Cl]+, 7, whose
structure will be discussed below. In a similar manner, 4,
which contains the NO2 electron-withdrawing group, also
reacts readily with Clꢀ stabilizing the resulting nitro com-
Chem. Eur. J. 2005, 11, 5146 – 5156
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5147

X. Ribas, A. Llobet, M. Solà et al.
Table 2. Selected bond lengths [Š]and angles [ o]for complexes 1–3 and 5–7.
plex [CuIII(L4)Cl]+, 8, although
its isolation in the solid state
still remains elusive as a result
of decomposition.
Crystallographic data and se-
lected bond lengths and angles
for complexes 2, 3, 6 and 7 are
reported in Table 1 and Table 2,
respectively. ORTEP views of
these complexes together with
their labeling schemes are
shown in Figure 1 and in the
1[a]
2
3[b]
5[a]
6
7
ꢀ
C1 Cu1
1.848(4)
1.899(4)
1.911(3)
2.000(8)
1.901(2)
1.960(2)
1.952(2)
2.002(2)
1.902(2)
1.961(2)
1.963(2)
1.995(2)
1.905(3)
1.985(3)
1.961(3)
2.031(3)
1.926(8)
1.957(7)
1.963(7)
2.048(7)
1.898(6)
1.958(5)
1.963(5)
1.988(5)
ꢀ
N1 Cu1
ꢀ
N3 Cu1
ꢀ
N2 Cu1
C1-Cu1-N1
C1-Cu1-N3
N1-Cu1-N2
N2-Cu1-N3
C1-Cu1-N2
N1-Cu1-N3
84.600(2)
84.400(2)
105.600(2)
85.200(2)
163.500(2)
169.000(2)
82.800(1)
81.700(1)
95.600(1)
99.700(1)
178.400(1)
157.200(1)
82.000(1)
82.200(1)
97.400(1)
96.400(1)
173.800(1)
156.000(1)
82.900(2)
82.300(2)
96.900(1)
99.700(1)
177.600(1)
160.700(2)
84.100(3)
82.600(4)
94.300(3)
97.600(3)
173.800(3)
161.000(4)
82.200(2)
81.100(2)
94.700(2)
99.300(2)
171.900(2)
154.500(2)
2.556(2) (Cl)[c]
3.301(2) (Cl)[c]
160.7(2) (Cl)[c]
ꢀ
Supporting Information. In Cu O (short)
2.428(3)
2.584(3)
170.9(2)
2.408(2)
2.643(2)
156.6(1)
2.256(2)
2.835(2)
161.0(1)
2.442(4)
2.736(4)
177.3(1)
2.279(5)
3.662(5)
148.01(3)
ꢀ
Cu O (long)
complexes 1, 2, 3, 5, and 6, the
O-Cu-O
CuIII metal center is coordinat-
[a]See reference [5]. [b]There are two different molecules in the unit cell with very similar metric parameters.
[c]The values shown here for complex 7 are the shortest Cu Cl distances and the smallest Cl Cu Cl angles
(see Figure 2 for the other values).
ed to the three nitrogen atoms
of the macrocyclic ligand and a
carbon atom of the phenyl ring
and adopts a distorted square-
ꢀ
ꢀ
ꢀ
planar geometry. The shortest coordination distance is
the steric crowding and strain around the tertiary nitrogen
atom of the amine together with solvation effects through
hydrogen bonding are identified as the main factors that
produce this effect.[18]
ꢀ
always the organometallic Cu1 C1 bond, which ranges from
1.848 Š for 1 to 1.926 Š for 6. This is in agreement with the
fact that the macrocyclic ligand, L1, of complex 1 has the
smallest cavity of this series and also with the fact that the
The N-methylation effect of the tertiary amines, besides
ꢀ
ꢀ
longest Cu1 C1 distance is that of complex 6, which con-
the elongation of the Cu Ntert bond, is also structurally man-
tains a ligand, L6, with a tertiary amine. The trans effect pro-
ifested in our case through a geometrical distortion. This
distortion involves the torsion angles between the best-fitted
plane that contains the carbon framework of the aromatic
ring and the best-fitted plane that contains the C1, Cu1, N1,
and N3 atoms (the first coordination sphere of the copper
center with the exception of the nitrogen atom of the terti-
ary amine). In complexes 5 and 6, which contain a tertiary
amine, this torsion angle is 9.4 and 9.68, respectively, where-
as in complexes 2 and 3 this distortion is reduced to 7.4 and
2.18, respectively, thus showing a lower degree of distortion
(the mathematical equations describing the best-fitted
planes as well as the distances of all the atoms to the planes
are presented in the Supplementary Information).
ꢀ
ꢀ
duced by the Cu1 C1 bond generates the longest Cu N dis-
ꢀ
tances (Cu1 N2 bond lengths range from 1.995 Š for 3 to
2.048 Š for 6) and are 0.033–0.095 Š longer than the mean
ꢀ
ꢀ
ꢀ
distances of the other two Cu N bonds (Cu1 N1 and Cu1
N3). This is especially acute for complexes 5 and 6 which
contain the macrocyclic ligands L5 and L6 with the methylat-
ed central amine. This phenomenon has been described in
detail previously mainly by Meyerstein and co-workers[16]
and by Bernhardt;[17] the M N bond strength of tertiary
amines is weaker than that of secondary amines under com-
parable conditions. Herein this phenomenon will from now
on be termed the “N-methylation effect”. The increase in
ꢀ
Table 1. Crystal data and structure refinement of CuIII complexes.
2-(OTf)2
3-(OTf)2
6-(OTf)2
7-(ClO4)
empirical formula
Mr
C16H22Cu1F6N3O6S2
594.03
monoclinic
P21/c
C17H24Cu1F6N3O6S2
608.05
monoclinic
P21/c
C18H26Cu1F6N3O6S2
C28H44Cu2N6O8Cl4
861.70
orthorhombic
Pbca
622.08
monoclinic
P21/c
crystal system
space group
a [Š]9.603(2)
b [Š]8.777(2)
c [Š]26.476(5)
a [8]90
24.727(1)
12.4908(3)
15.8686(4)
90
24.163(5)
14.822(3)
17.658(4)
90
16.959(1)
18.088(2)
22.441(1)
90
b [8]98.08(3)
g [8]90
107.265(1)
90
4680.3(2)
4
127.83(3)
90
4995.0(2)
8
90
90
V [Š3]2209.4(8)
Z
6884.0(8)
8
8
temperature [K]160(2)
l(MoKa) [Š]0.71073
1calcd [gcmꢀ3]1.786
m [mmꢀ1]1.268
R[a]
153(2)
0.71073
1.726
1.199
180(2)
0.71073
1.670
1.129
160(2)
0.71073
1.663
1.604
0.0295
0.0765
0.0409
0.1026
0.0730
0.1569
0.0548
0.0895
[b]
Rw
1
[a] R=ꢀ[F0ꢀFC]/ꢀF0. [b] Rw ={ꢀ[w(F20ꢀFc2)2]/ꢀ(wF04)} =
.
2
5148
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 5146 – 5156

Electronic Properties of Organometallic CuIII Complexes
FULL PAPER
which at the same time bridge two metal centers. Further-
more, this complex exhibits two crystallographically distinct
molecules in its unit cell (their atoms are thus labeled by
adding the letters “a” or “b”, for example, the copper atoms
are labeled Cu1a and Cu1b), although their parameters are
very similar. While the crystal packing in the complexes 1–6
is unremarkable, the chlorine bridge in 7 produces pseudo-
chains along the z axis with alternating “a” and “b” mole-
cules, as shown in Figure 2.
Figure 2. Crystal packing for complex 7 showing the formation of a
ꢀ
pseudochain through the Cu Cl bonds; structural parameters are also
indicated.
Figure 1. ORTEP plot (50% probability) of the cationic moiety of the X-
ray crystal structures of complexes a) 2 and b) 6. Hydrogen atoms have
been omitted for clarity.
Spectroscopic properties: The 1D and 2D NMR spectra of
the CuIII complexes 1–8 and their corresponding ligands
HL1–HL6 were recorded in CD3CN and are presented in the
Supporting Information. The combination of 1D and 2D
NMR spectra has allowed us to unambiguously identify the
resonances of all the protons in the complexes and ligands.
Their assignments are reported in the Experimental Section
and in Table 3 and are consistent with the structures found
in the solid state.
Table 3 also contains the coordination-induced shifts
(CIS) as well as the Methylation-induced shifts (MeIS). The
latter is defined as the difference in chemical shift for a par-
ticular resonance when the central amine is methylated and
when it is not and thus remains a secondary amine. This
shift has been introduced here to analyze the electronic ef-
fects produced by methylation.
A careful examination of the data in Table 3 reveals the
following trends. 1) In both the ligands and complexes, the
Me and NO2 substituents in the aromatic ring significantly
affect the d values of the remaining aromatic protons, very
slightly affect the benzylic resonances, and virtually have no
affect on the aliphatic resonances. 2) In all cases, with the
exception of the aromatic protons, complexation produces a
positive CIS, the highest values being observed for the ben-
zylic and aminic methyl groups (0.6–0.7 ppm). The negative
CIS values observed for the aromatic protons suggest that
the aromatic rings, besides acting as s donors, also act as
weak p acceptors. 3) In the ligands, a negative MeIS is ob-
served for all the methylenic protons (Hb–d) neighboring the
Note also that in all cases the hydrogen atoms bonded to
the secondary amines (2 and 3) are oriented in the same di-
rection and thus the macrocyclic ligand adopts a syn confor-
mation. This syn conformation is also adopted by macrocy-
clic ligands with methylated central amines (5 and 6). The
non-anionic nature of the nitrogen-coordinating atoms is in
line with the longest CuIII N bonds found in this family of
ꢀ
complexes relative to the bond lengths found in other CuIII
complexes that contain deprotonated amides as ligands.[6,19]
ꢀ
ꢀ
In complexes 1–3, 5, and 6 the ClO4 or CF3SO3 counter-
anions or a water molecule (in the case of complex 6) are
weakly associated with the metal center, generating a de-
compressed tetragonally distorted octahedral environment
ꢀ
in which one of the Cu O distances is substantially longer
ꢀ
than the other (see Table 2). In all cases, the longer Cu O
distance is located on the same side of the macrocycle as the
ꢀ
ꢀ
ꢀ
N H or N Me bonds, whereas the shorter Cu O distance is
located on the other side which has a much lower steric en-
cumbrance. This distorted octahedral environment is per-
haps the most salient feature of this family of compounds[5]
in comparison with CuIII complexes described previously[6–9]
and can be attributed to the fact that in our case the higher
oxidation state of the copper center is stabilized by a mono-
anionic ligand.
In complex 7, the CuIII metal center adopts a similar ge-
ometry as in the previous cases except that the axial posi-
tions are occupied by chlorine instead of oxygen atoms,
Chem. Eur. J. 2005, 11, 5146 – 5156
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5149

X. Ribas, A. Llobet, M. Solà et al.
Table 3. Chemical shifts, CIS, and MeIS for the ligands HL2–HL6 and complexes 2–6.[a]
HL2 HL3 HL4 HL5 MeIS[b] HL6 MeIS[b] 2- CIS[c] 3- CIS[c] 4-
(ClO4)2 (OTf)2 (OTf)2
CIS[c] 5-
(ClO4)2
CIS[c] MeIS[b] 6-
(OTf)2
CIS[c] MeIS[b]
Hy 7.25
–
–
7.24
–
6.88
7.27
6.95
6.12
0.02
ꢀ0.16 6.78
–
–
7.27
ꢀ0.05 6.95
6.33
0.03 0.00
ꢀ0.12 0.00
–
Hz 7.11 6.91 7.95 7.07
ꢀ0.13 7.9
6.77
7.11
–
ꢀ0.11 ꢀ0.01
[d][d][d][d]
[d]
[d]
Ha
Hb
6.96
3.77
[d][d][d][d]
[d]
[d]
3.14
–
Ha1 3.87 3.84 3.98 3.9
Ha2
0.03 3.84
0.00 4.61
4.49
0.68 4.65
4.42
0.02 3.27
3.15
0.23 2.10
1.98
0.53 2.90
3.10
0.70 4.75
4.55
0.43 2.64
0.67 4.67
4.52
0.70 0.05
0.36 0.00
0.42 0.15
0.67 0.08
0.72
4.67
4.52
3.37
3.15
2.33
2.02
3.21
2.68
2.91
2.28
0.76
0.76
0.57
0.06
0.05
0.14
Hb1 2.84 2.78 2.77 2.5 ꢀ0.34 2.5 ꢀ0.26 3.09
Hb2 2.63
Hc1 1.68 1.66 1.65 1.63 ꢀ0.05 1.61 ꢀ0.05 1.77
Hc2 2.04
ꢀ0.01 3.05
2.67
0.38 1.96
0.31 2.15
1.95
0.52 3.08
3.24
–
Hd1 2.56 2.52 2.53 2.49 ꢀ0.07 2.46 ꢀ0.06 2.99
0.48 2.97
0.49 ꢀ0.06
Hd2
He
Hf
–
–
–
3.18
–
–
3.12
–
–
2.34
–
–
2.01
–
2
2.30
–
2.31
2.73
–
0.91
ꢀ0.02 ꢀ0.03
ꢀ0.03
–
[a]See Scheme 1 for labels. [b]The methylation induced shift (MeIS) is defined as the difference between the chemical shift of a particular proton in a methylat-
ed compound and the same proton in the corresponding non-methylated compound. For CH2 methylenic groups with magnetically different protons the MeIS is
calculated by taking the mean value. [c]In the CIS of the complexes, the mean value of the chemical shifts of two magnetically different protons of each CH2
methylenic group is subtracted from the chemical shift displayed by the same protons in the free ligand. [d]Not observed.
central amine as a consequence of the electron-donating
effect exerted by the methyl group (see Scheme 3 for the
atomic labeling used in the NMR assignment). The Hd pro-
tributed to a number of factors including the formation of
hydrogen bonds between the secondary amines and solvent
molecules and less metal–ligand orbital overlap as a result
of the steric effects generated by the methylated tertiary
amine which displace the nitrogen bonding electron pair out
of the main plane in which the metal center is located.
Indeed, the latter effect can also be observed in the com-
plexes described in this paper since the methylated com-
plexes 5 and 6 have the highest angles between the planes
defined by the carbon atoms of the aromatic ring and the
C1, Cu1, N1, and N3 atoms (vide supra).
The UV/Vis spectra of complexes 2–7 are presented in
Figure 3 and their lmax values together with the correspond-
Scheme 3. Drawings and 1H NMR labeling scheme for the cationic
moiety of CuIII complexes 2–6.
tons, in the a position, are as expected the ones that are
most affected. 4) In sharp contrast, the opposite effect is ob-
served in the complexes with positive MeIS values for the
Hb–d protons. Furthermore, the most affected proton is now
Hc, which is in a b position with respect to the central
amine, followed by Hd, and finally Hb is only slightly affect-
ed.
Figure 3. UV/Vis spectra for complexes 2–7. The inset shows an enlarge-
ment of the 450–600 nm zone (see text for experimental details).
This inversion of the MeIS for the complexes reveals that
from an electronic point of view the methyl group situated
at the tertiary amine paradoxically acts as an electron-with-
drawing group. This is further supported by the E1/2 redox
potentials of the CuIII/CuII couple bearing the methylated
tertiary amines since they are anodically shifted relative to
those of the non-methylated secondary amine (vide infra).
This effect has previously been observed in aqueous solu-
tions of non-organometallic CuII complexes[16,17] and was at-
ing extinction coefficients are listed in Table 4. The com-
plexes 2–7 display a very intense band below 300 nm and an-
other less intense but broad band in the range of 370–
520 nm (see Figure 3) that can be assigned to two similar
p!dx2ꢀy2 ligand-to-metal charge transfer (LMCT) transi-
tions.[20] As shown in Table 4 and Figure 3 (see also the de-
5150
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 5146 – 5156

Electronic Properties of Organometallic CuIII Complexes
FULL PAPER
Table 4. Calculated and experimental deconvoluted UV/Vis data as-
signed to p1!dx2ꢀy2 and p2!dx2ꢀy2 transitions.
ꢀ
Cu1 C1. This orbital is the HOMO-2 except in complex 4,
in which it is the HOMO-3 owing to the insertion of an orbi-
tal with mainly nitro-group character as the HOMO-2 orbi-
tal, and in complex 7, in which the s1 orbital is the HOMO-
p1!dx2ꢀy2
p2!dx2ꢀy2
Complex lmax(exp.[a]
)
lmax(theor.)
lmax(exp.[a]
) lmax(theor.)
[nm]
[nm]
[nm]
[nm]
ꢀ
6 because of the intercalation of three orbitals with Cu Cl
antibonding character and another orbital of macrocyclic
character related to the Cu N bonding interaction.
The p!dx2ꢀy2 LMCT band is composed of two charge-
transfer excitations, principally from the ligand to the metal.
The main arrival orbital is the LUMO (see Figure 4a), but
the starting orbital changes depending on the system. The
first excitation starts from the p1 orbital (Figure 4c), while
the second one starts from the p2 orbital (Figure 4d). The p1
and p2 orbitals are the HOMO and HOMO-1 orbitals in
2
390(417)
487(303)
277(325)
393(345)
503(198)
328(630)
552
649
496
560
661
469
463(147)
388(559)
380(15)
478(157)
397(285)
386(1260)
608
617
567
623
634
497
3[b]
4
ꢀ
5
6[b]
7
[a]Extinction coefficients are given in parentheses and are measured in
Mꢀ1 cmꢀ1. [b]The energies of the p1 and p2 orbitals are inverted com-
pared with the energies of the other complexes; see text, Figure 5 and
the Supporting Information for details.
convoluted spectra presented in
the Supporting Information),
the p1!dx2ꢀy2 band for com-
plexes 3, 5, and 6 is red-shifted
relative to 2, while for com-
plexes 4 and 7 a blue shift is ob-
served.
Density functional theory
(DFT) calculations (see Com-
putational Details in the Exper-
imental Section) were carried
out for complexes 2–7 to eluci-
date the electronic properties
of these complexes and thus un-
derstand the nature of the tran-
sitions observed in the UV/Vis
spectra. For complex 2, the
DFT-optimized structure nicely
fits the experimental X-ray
structure (see the calculated
structure and the comparison of
the geometrical parameters pre-
sented in the Supporting Infor-
mation) thus validating the ade-
quacy of the theoretical method
employed for the geometry op-
timizations of these particular
Figure 4. 3D representation of the orbitals most involved in the calculated excitations of the UV/Vis spectra of
complex 2: a) the LUMO orbital; b) the starting orbital (s1) involved in the highest energy band; c) the start-
ing orbital (p1) involved in the p1!dx2ꢀy2 excitation; d) the starting orbital (p2) involved in the p2!dx2ꢀy2 exci-
tation. Isosurface values are ꢀ0.05 and 0.05 a.u.
systems.
Time-dependent DFT calcu-
lations (TDDFT) show that the
most intense band below
300 nm involves the LUMO
frontier orbital (Figure 4a) as the main arrival orbital for
complex 2 (only small differences are observed for com-
plexes 2–7) and the s1 orbital (Figure 4b) as the main start-
ing one. The LUMO orbital is mainly a combination of the
dx2ꢀy2 orbital of the copper atom and the p orbitals of the ni-
trogen atoms and the carbon atom that participates in the
complexes 2, 4, and 5. In the other complexes, these two or-
bitals are exchanged, except in complex 7 in which the p1
and p2 orbitals are the HOMO-3 and HOMO-4, respective-
ly. The solvent has a significant effect on the energy of the
p!dx2ꢀy2 LMCT band and a less strong one on the band
below 300 nm. After including this effect in the calculations,
the p!dx2ꢀy2 LMCT bands are blue-shifted, this shift being
more noticeable in solvents with high dielectric constants.
Figure 5 shows the energies of the orbitals and the wave-
lengths of the two calculated excitations that produce the
ꢀ
Cu C bond. In 7 there is an extra contribution from the pz
orbital of the chlorine atom.
The s1 orbital depicted in Figure 4b is composed essential-
ꢀ
ly of the py orbitals of the atoms contained in the axis N2
Chem. Eur. J. 2005, 11, 5146 – 5156
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5151

X. Ribas, A. Llobet, M. Solà et al.
the LUMO, that is, the LUMO is slightly destabilized, which
is not surprising if one considers that this methyl group be-
haves as an electron donor. In line with this, the p1 and p2
orbitals in complex 3 are significantly destabilized owing to
their proximity to the aromatic group. In fact, the p1 orbital
is destabilized more than the p2 orbital because the methyl
group bonded to the aromatic carbon contributes to the p1
orbital with an antibonding interaction whereas it does not
participate in the p2 orbital. This effect is responsible for the
relative inversion of the energies of the p1 and p2 orbitals in
complexes 3 and 6 in which the methyl substituent is
bonded to the phenyl group.
Figure 5. Calculated qualitative energy diagram of the orbitals involved
in the p–dx2ꢀy2 LMCT bands of complexes 2, 3, and 5. The orbital energies
are given in atomic units (hartrees) and the wavelengths of the excita-
tions in nm.
The significant destabilization of the p1 and p2 orbitals in
complex 3 account for the larger red shift observed for com-
plex 3 than for complex 5.
The presence of a nitro group (complex 4) in the aromatic
ring causes a blue shift of the p1!dx2ꢀy2 transition relative to
complex 2 (exptl: 113 nm, calcd: 56 nm). This blue shift re-
sults from a greater stabilization of the p1 and p2 orbitals
than of the LUMO giving a larger energy difference as a
result of the electron-accepting nature of the nitro group.
In complex 7, two bands were found experimentally at
375 and 522 nm. The band at 375 nm splits into two bands
centered at 328 nm and 386 nm (Gaussian deconvolution).
In this case, theoretical calculations were useful to under-
stand and assign the origin of these transitions. It has been
found that the band at 375 nm is a result of a p!dx2ꢀy2
LMCT transition similar to that occurring in complexes 2–6
but the band at 522 nm arises from three transitions that
p!dx2ꢀy2 LMCT bands for complexes 2, 3, and 5, whereas
Table 4 and Figure S12 (Supporting Information) reveal the
main features of the experimental and calculated LMCT
transitions of complexes 2–7. As can be observed from the
data in Table 4, the experimental p1!dx2ꢀy2 transition corre-
lates well with the TDDFT calculated one with a shift of
141–167 nm for complexes 2, 3, 6, and 7 and of 219 nm for
4. Although the absorption wavelengths calculated by
TDDFT are clearly overestimated, the errors are compara-
ble to those found in similar studies.[21] These errors may be
attributed to the size of the basis set,[22] the inadequate de-
scription of solvent effects, which are very relevant in charg-
ed species, and the different localized or delocalized charac-
ter of the initial and final electronic states.[21a] However, we
want to emphasize that our main goals are to tentatively
assign the observed transitions and to understand the origin
of the shifts experienced by the different bands when going
from complex 2 to complexes 3–7. In our opinion, these
shifts are well reproduced by our calculations and an under-
standing of the origin of the shifts is provided by the analy-
sis of the molecular orbitals that are most involved in the
electronic transitions. Finally, the p2!dx2ꢀy2 transition has a
similar degree of correlation to that of the p1!dx2ꢀy2 transi-
tion, with the exception of 3 and 6, which are shifted in op-
posite directions.
ꢀ
start from antibonding s* and p* orbitals of the Cu Cl
bond (HOMO, HOMO-1, and HOMO-2) (see Figure 6) and
finish at the LUMO orbital.
Our calculations indicate that the p1!dx2ꢀy2 LMCT transi-
tion in complex 7 is blue-shifted relative to the same transi-
tion in 2 (exptl: 62 nm, calcd: 83 nm). This blue shift results
from the substantial destabilization of the LUMO orbital
and at the same time an important stabilization of the p1
and p2 orbitals in this complex. The destabilization of the
LUMO orbital is a result of the large charge transfer
(0.62 e) from the chlorine anion to complex 2 that occurs in
the formation of complex 7.
Methylation of the aromatic ring of species 2 to yield
complex 3 has a greater effect on the p1!dx2ꢀy2 LMCT band
than the methylation of the central amine in complex 5. In
both cases, the band is red-shifted, but the magnitude of the
shift in the former is much larger (exptl: 97 nm, calcd:
97 nm for 3; exptl: 3 nm, calcd 8 nm for 5). Methylation in
both places (complex 6) causes a red shift that is nearly the
sum of the two shifts described above (exptl: 113 nm, calcd:
109 nm for 6).
By using complex 2 as a reference, it is found that the en-
ergies of the p1 and p2 orbitals remain almost constant after
the methylation of the amine (complex 5), while the LUMO
is slightly stabilized, as expected from the fact that the
methyl group behaves as an electron-acceptor group (vide
supra) in this particular species (see Figure 5). Methylation
of the aromatic ring in complex 3 has the opposite effect on
Finally, our calculations also correctly predict much
ꢀ
higher intensities for the p!dx2ꢀy2 bands of the Cu Cl com-
plex 7 than for complexes 2–6 (see oscillator strength values
in Table S5 and Figure S12 in the Supporting Information).
Redox properties: The redox properties of complexes 1–8
were investigated by means of cyclic voltammetric and cou-
lometric techniques and their electrochemical parameters
are given in Table 5. Figure 7 presents the cyclic voltammo-
grams of complexes 2–6 and reveals a chemically and elec-
trochemically quasireversible CuIII/CuII redox couple. Coulo-
metric experiments show that upon reduction to oxidation
state ii, the complexes described in this work are not stable,
undergoing irreversible decomposition reactions. The CuIII/
CuII redox couple as expected depends on the electronic
nature of the phenyl substituents as well as on the nature of
5152
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 5146 – 5156

Electronic Properties of Organometallic CuIII Complexes
FULL PAPER
Figure 7. Cyclic voltammograms of complexes 2–6 (see text for experi-
mental details).
lize CuIII by shifting the E1/2 anodically by 38 mV (E1/2
=
ꢀ0.201 V; Ep,a =ꢀ0.152 V, Ep,c =ꢀ0.250 V, DE=98 mV). Ter-
tiary amines have a lower s-donating capacity than secon-
dary amines owing to a lower overlap capacity and also sol-
vation effects (N-methylation effect) and this is usually man-
ifested in a cathodic shift of the E1/2 redox potential. This is
observed in complex 5 in which the central secondary amine
of the L2 ligand has been replaced by a methylated tertiary
amine and has a significant cathodic shift of 60 mV (E1/2
=
ꢀ0.103 V; Ep,a =ꢀ0.056 V, Ep,c =ꢀ0.150 V, DE=94 mV). In
complex 6 there are two competing electronic effects, the
enhanced electron-donating capacity of the methylated
phenyl group versus the reduced s-donating ability of the
tertiary amine; overall a 34 mV cathodic shift is observed
Figure 6. 3D representations of the starting orbitals most involved in the
excitation at 522 nm in the experimental UV/Vis spectra of complex 7: a)
HOMO; b) HOMO-1; and c) HOMO-2 orbitals. Isosurface values are
ꢀ0.05 and 0.05 a.u.
for complex
6
(E1/2 =ꢀ0.129 V;
E
p,a =ꢀ0.084 V, Ep,c
=
ꢀ0.174 V, DE=90 mV).
Furthermore, the electronic effects correlate with the ki-
netic stability of the corresponding CuII complexes; thus
complex 4 has the largest I (Ip,a/Ip,c), whereas complex 3 has
the lowest, showing that electron-withdrawing groups stabi-
lize the CuII oxidation state whereas electron-donating
groups destabilize CuII and thus the coupled chemical reac-
tion becomes more significant after electron transfer.
The behavior of complex 7 is radically different displaying
an irreversible cathodic wave at ꢀ0.400 V. The cyclic vol-
tammogram of complex 8 reveals a reversible process in the
cyclic voltammetry timescale but it has the largest peak
Table 5. Electrochemical parameters (versus SSCE) of complexes 1–8.
Complex
E1/2 [mV][a]
Ep,a [V]
Ep,c [V]
DE [mV]
Ip,a/Ip,c
1
2
3
4
5
6
7
8
–
ꢀ163
ꢀ201
6
ꢀ103
ꢀ129
–
–
ꢀ0.444
ꢀ0.210
ꢀ0.250
ꢀ0.063
ꢀ0.150
ꢀ0.174
ꢀ0.400
ꢀ0.394
–
94
98
139
94
90
–
ꢀ0.116
ꢀ0.152
0.076
ꢀ0.056
ꢀ0.084
–
0.9
0.6
0.9
0.9
0.8
–
–
246
ꢀ271
ꢀ0.148
1.0
splitting by far (E1/2 =ꢀ0.271 V;
E
p,a =ꢀ0.148 V, Ep,c
=
[a] E1/2 values referenced to the ferrocene/ferrocenium couple can be
found in Table S6 of the Supporting Information.
ꢀ0.394 V, DE=246 mV).
the central macrocyclic amine. To evaluate these electronic
Conclusions
effects, the CuIII/CuII redox couple of complex 2 (E1/2
=
ꢀ0.163 V; Ep,a =ꢀ0.116 V, Ep,c =ꢀ0.210 V, DE=94 mV) was
used as a reference. Electron-withdrawing substituents in
the phenyl ring destabilize the oxidation state iii and thus
the E1/2 of complex 4 is strongly shifted cathodically by
169 mV (E1/2 =0.006 V; Ep,a =0.076 V, Ep,c =ꢀ0.063 V, DE=
139 mV). On the other hand electron donors in the phenyl
ring, as in complex 3, produce the opposite effect and stabi-
A family of Cu complexes in the unusual oxidation state of
+3 and containing triazamacrocyclic ligands has been de-
scribed and their structural, spectroscopic, and redox prop-
erties thoroughly investigated. This has allowed us to unrav-
el the detailed electronic structure of these CuIII complexes
which in turn permitted us to rationalize their redox and
spectroscopic properties. As a result of this detailed work it
Chem. Eur. J. 2005, 11, 5146 – 5156
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5153

X. Ribas, A. Llobet, M. Solà et al.
elemental analysis calcd (%) for C15H25N3 (247.38): C 72.83, H 10.19, N
16.99; found: C 73.01, H 10.23, N 16.85.
has been possible to measure the electronic effects of N-
methylation in a quantitative manner.
A similar procedure was used to synthesize the rest of the ligands.
LigandHL 4: Yield: 35%. 1H NMR (200 MHz, CDCl3, 258C): d=8.20 (s,
1H; Hx), 7.91 (s, 2H; Hz), 3.98 (s, 4H; Ha), 2.79 (t, J(H,H)=5 Hz, 4H;
Hb), 2.56 (t, J(H,H)=6 Hz, 4H; Hd), 1.70 ppm (quint, J(H,H)=6 Hz,
4H; Hc); 13C NMR (50.3 MHz, CDCl3, 258C): d=143.17, 131.41, 121.46
(Caromatic), 52.54 (Ca), 46.43 (Cb), 44.49 (Cd), 29.12 ppm (Cc); ESI-MS
(CH3CN): m/z(%): 279 [HL4 +H]+; elemental analysis calcd (%) for
C14H22N4O2 (278.35): C 60.41, H 7.97, N 20.13; found: C 60.85, H 8.24, N
20.50.
LigandHL 6: Yield: 40%. 1H NMR (200 MHz, CDCl3, 258C): d=7.39 (s,
1H; Hx), 6.84 (s, 2H; Hz), 3.86 (s, 4H; Ha), 2.57 (t, J(H,H)=6 Hz, 4H;
Hb), 2.42 (t, J(H,H)=4 Hz, 4H; Hd), 2.30 (s, 3H; Hf), 2.01 (s, 3H; He),
1.63 ppm (quint, J(H,H)=5 Hz 4H; Hc); 13C NMR (50.3 MHz, CDCl3,
258C): d=140.49, 137.56, 127.01, 121.49 (Caromatic), 55.84, 52.64 (CH2-CH2-
CH2), 44.92 (Ca), 40.80 (Ce), 27.36 (Cc), 21.16 ppm (Cf); IR (KBr): n˜ =
3292 (s), 2916 (s), 2789 (s), 1604 (m), 1459 (s), 1127 (s), 730 cmꢀ1 (s);
ESI-MS (CH3CN): m/z(%): 262 [HL6 +H]+; elemental analysis calcd
(%) for C16H27N3 (261.41): C 73.51, H 10.41, N 16.07; found: C 73.73, H
10.62, N 15.87.
Experimental Section
Materials: All reagents used in this work were obtained from Aldrich
Chemical Co. and were used without further purification. Reagent-grade
organic solvents were obtained from SDS and high purity deionized
water was obtained by passing distilled water through a nanopure Mili-Q
water purification system.
Instrumentation andmeasurements : IR spectra were recorded with a
Mattson Satellite FT-IR spectrometer using KBr pellets or by using a
MKII Golden Gate Single Reflection ATR System. UV/Vis spectroscopy
was performed by using a Cary 50 Scan (Varian) UV/Vis spectrophotom-
eter with 1 cm quartz cells. NMR spectra were recorded with a Bruker
500 MHz or 200 Model Avance spectrometer. Samples were analyzed in
either CDCl3 or in [D3]acetonitrile, with residual solvent protons and/or
tetramethylsilane as internal references. Elemental analyses were per-
formed with a CHNS-O Elemental Analyser EA-1108 from Fisons. The
ESI-MS experiments were performed on a Navigator LC/MS chromato-
graph from Thermo Quest Finnigan using methanol or acetonitrile as a
mobile phase.
Synthesis of complexes: The synthesis of complexes [CuIII(L1)](ClO4)2 (1-
(ClO4)2), [CuIII(L2)](ClO4)2 (2-(ClO4)2), and [CuIII(L5)](ClO4)2 (5-(ClO4)2)
have been reported previously.[5] However, the synthesis of all complexes
have been reproduced in this work and in the case of 2-(ClO4)2, good
crystals for structural resolution were obtained.
Cyclic voltammetric (CV) experiments were performed in a PAR 263A
EG&G or IJ-Cambria IH-660 potentiostat with a three-electrode cell.
Glassy carbon disk electrodes (3 mm diameter) from BAS were used as
the working electrode, platinum wire as the auxiliary and SSCE as the
reference electrode (unless explicitly mentioned in the text, the potentials
given are always with respect to this reference electrode). The ferrocene/
ferrocenium couple was used as the internal reference with E1/2 (versus
SSCE)=371 mV. The complexes were dissolved in previously degassed
0.1m (nBu4N)(PF6) in acetonitrile as the supporting electrolyte. All E1/2
values reported in this work were estimated from cyclic voltammetry as
the average of the oxidative and reductive peak potentials (Ep,a +Ep,c)/2
(scan rate=100 mVsꢀ1, room temperature). The concentration of the
complexes was 2 mm. Bulk electrolyses were carried out in a three-com-
partment cell using carbon felt from SOFACEL as the working electrode.
[CuIII(L3)](OTf)2, 3-(OTf)2: Equimolar amounts of ligand HL3 (15.4 mg,
0.062 mmol) and CuII(ClO4)2·6H2O (23.0 mg, 0.062 mmol) were dissolved
in CH3CN (2 mL) under magnetic stirring and under either N2 or Ar.
After 30 min, the resulting orange solution was filtered through celite.
Slow diffusion of diethyl ether into the filtered solution over 24 h led to
the formation of orange crystals of 3-(OTf)2 in 50% isolated yield
(19.0 mg, 0.031 mmol). 1H NMR (500 MHz, CD3CN, 258C): d=6.96 (m,
2
2
2H; Ha), 6.78 (s, 2H; Hz), 4.65 (d, J(H,H)=16 Hz, 2H; Ha1), 4.42 (d, J-
(H,H)=16 Hz, 2H; Ha2), 3.27 (m, 2H; Hb1), 3.15 (m, 2H; Hb2), 3.10 (m,
2H; Hd2), 2.90 (m, 2H; Hd1), 2.31 (s, 3H; Hf), 2.10 (m, 2H; Hc1),
1.98 ppm (m, 2H; Hc2); IR (KBr): n˜ =3140 (m), 2925 (m), 2857 (m), 1456
(w), 1433 (w), 1279 (s), 1255 (s), 1170 (s), 1029 (s), 636 (m), 517 cmꢀ1
(w); UV/Vis (CH3CN): lmax (e)=482 nm (sh, 322 mꢀ1 cmꢀ1); ESI-MS
(CH3CN): m/z (%): 458 [3-(OTf)]+; elemental analysis calcd (%) for
C17H24N3CuF6S2O6 (608.1): C 33.58, H 3.98, N 6.91, S 10.55; found: C
33.75, H 4.06, N 6.67, S 10.44.
Sample preparations: Caution! Some complexes described in this section
contain the perchlorate anion; even though we have not suffered any ex-
plosion, perchlorate salts are potentially explosive.
Ligands: All ligands have been prepared following two different methods
previously used for the synthesis of HL1 and HL2[5, 13,14] although HL2 and
HL5 have previously been prepared following different synthetic
routes.[15] The remaining ligands have been synthesized in this work for
the first time.
The labels used in the NMR assignment for this and all the other com-
plexes are given in Scheme 3.
The same procedure was followed using CuII(ClO4)2 as the CuII salt and
the complex was obtained in quantitative yield. Complexes 4 and 6 were
synthesized following the same procedure, except for the addition of the
appropriate ligand, and yields identical to that of 3 were obtained. Thus
only their analytical and spectroscopic data will be given.
LigandHL 3: N-(3-Aminopropyl)propane-1,3-diamine (3.45 g, 25.7 mmol)
was tosylated with TsCl (16.45 g, 85.0 mmol) in THF/H2O (200 mL) in
the presence of NaOH (4.21 g, 103.1 mmol) at 608C overnight to give a
52% yield of tritosylated amine (7.95 g, 13.40 mmol) after purification by
column chromatography (silica gel, CH2Cl2/ethyl acetate 92:8). 3,5-Bis-
(bromomethyl)toluene (1.11 g, 4.01 mmol) and the tritosylated amine
(2.38 g, 4.01 mmol) in CH3CN (250 mL) was refluxed for 24 h to give,
prior to purification by column chromatography on silica gel (CH2Cl2/
ethyl acetate 98:2), the desired cyclized tosylated product HL3 (2.28 g,
3.22 mmol) in 80% yield. Detosylation of tosylated HL3 was achieved by
reduction with lithium in liquid ammonia (approx. 250 mL) at ꢀ708C.
Lithium metal was added until the solution turned deep blue. After
1 hour, NH4Cl (15 g, 0.28 mol) was added and NH3 carefully evaporated.
Extraction with CHCl3/1m HCl, evaporation of the aqueous phase, basifi-
cation with 30% NaOH, extraction with CHCl3, and final purification on
silica gel (CH2Cl2/MeOH/NH4OH 80:20:5) yielded ligand HL3 (0.42 g,
1.71 mmol, 48% yield). 1H NMR (200 MHz, CDCl3, 258C): d=7.47 (s,
1H; Hx), 6.86 (m, 2H; Hz), 3.84 (s, 4H; Ha), 2.76 (t, J(H,H)=5.4 Hz, 4H;
Hb), 2.53 (t, J(H,H)=5.8 Hz, 4H; Hd), 2.31 (s, 3H; Hf), 1.69 ppm (quint,
J(H,H)=5.6 Hz, 4H; Hc); 13C NMR (50.3 MHz, CDCl3, 298 K): d=
141.07, 137.63, 127.65, 121.56 (Caromatic), 52.78 (Ca), 46.44 (Cb), 44.01 (Cd),
28.84 (Cc), 21.17 ppm (Cf); ESI-MS (CH3CN): m/z(%): 248 [HL3 +H]+;
[CuIII(L4)](OTf)2, 4-(OTf)2: Yield: 50%. 1H NMR (200 MHz, CD3CN,
2
2
258C): d=7.9 (s, 2H; Hz), 4.75 (d, J(H,H)=16 Hz, 2H; Ha1), 4.55 (d, J-
(H,H)=16 Hz, 2H; Ha2), 3.12 (m, 2H; Hd1), 2.97 (m, 2H; Hd2), 2.64 (m,
4H; Hb1, Hb2), 1.96 ppm (m, 4H; Hc1, Hc2); UV/Vis (CH3CN): lmax (e)=
400 nm (sh, 318 mꢀ1 cmꢀ1); ESI-MS (CH3CN): m/z (%): 490 [4-(OTf)]+.
[CuIII(L6)](OTf)2, 6-(OTf)2: Yield: 50% (18.3 mg, 0.029 mmol) 1H NMR
(500 MHz, CD3CN, 258C): d=7.71 (m, 2H; Ha), 6.77 (s, 2H; Hz), 4.67
(d, 2J(H,H)=16 Hz, 2H; Ha1), 4.52 (d, 2J(H,H)=16 Hz, 2H; Ha2), 3.37
(m, 2H; Hb1), 3.21 (m, 2H; Hd1), 3.15 (dt ,3J(H,H)=3.3 Hz, 2J(H,H)=
13.3 Hz, 2H; Hb2), 2.91 (s, 3H; He), 2.68 (m, 2H; Hd2), 2.33 (m, 2H; Hc1),
2.28 (s, 3H; Hf), 2.02 ppm (m, 2H; Hc2); IR (KBr): n˜ =3139 (m), 2925
(m), 2857 (m), 1452 (w), 1435 (w), 1279 (s), 1259 (s), 1170 (s), 1031 (s),
636 (m), 518 cmꢀ1 (w); UV/Vis (CH3CN): lmax (e)=496 nm (sh,
218 mꢀ1 cmꢀ1); ESI-MS (CH3CN): m/z(%): 472 [6(OTf)]+; elemental
analysis calcd (%) for C17H24N3CuF6S2O6 (608.1): C 34.75, H 4.21, N 6.75,
S 10.31; found: C 34.53, H 4.32, N 6.67, S 10.32.
[CuIII(L2)(Cl)](OTf), 7-(OTf): Complex 2-(OTf)2 (0.022 g, 0.037 mmol)
was dissolved in CH3CN (2 mL) and NaCl (1 equiv, 0.0022 g) in H2O
5154
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 5146 – 5156

Electronic Properties of Organometallic CuIII Complexes
FULL PAPER
(0.5 mL) was added. After the addition of NaCl the original orange solu-
tion turned red immediately. The solution was then magnetically stirred
for 30 min and filtered. Diffusion of diethyl ether into the filtrate led to
the formation of red crystals of 7(OTf) in 73% isolated yield (0.012 g,
0.026 mmol). 1H NMR (400 MHz, CD3CN, 258C): d=7.12 (t, 3J(H,H)=
7.5 Hz, 1H), 6.81 (d, 3J(H,H)=3.6 Hz, 2H)), 6.07 (m, 2H; NH), 4.38
(m,2H), 4.25 (m, 2H), 3.14 (m, 3H), 2.80 (m, 6H), 1.89 (m, 1H),
1.70 ppm (m, 2H); IR (KBr): n˜ =3126, 3090, 3033, 2926, 2854, 1459,
1423, 1144, 1116, 1078, 627 cmꢀ1; UV/Vis (CH3CN): lmax (e)=375 (1900),
522 nm (540 mꢀ1 cmꢀ1). Crystals of [CuIII(L2)(Cl)](ClO4), 7-(ClO4), suita-
ble for X-ray diffraction were synthesized by the same procedure.
tion.[37,38] Electrons in lower shells were treated within the frozen-core
approximation.[27] A set of auxiliary s, p, d, f, and g functions, centered in
all nuclei, was introduced in order to fit the molecular density and Cou-
lomb potential accurately in each SCF cycle.[39] For the particular case of
2, the experimental and theoretical bond lengths and angles differ by less
than 0.07 Š and 3.08, respectively. The standard deviation for the bond
distances is 0.03 Š, and for the angles 1.818. The standard deviation is
calculated using Equation (2), where CV means calculated value, EV ex-
perimental value (X-ray data), and N is the number of bond distances or
angles taken into account. See the Supporting Information for the bond
distances and angles used to calculate snꢀ1
.
[CuIII(L4)(Cl)](OTf), 8-(OTf): This complex was prepared in solution
only. UV/Vis (CH3CN): lmax (e)=365 nm (2900mꢀ1 cmꢀ1), 503 nm
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u
(308mꢀ1 cmꢀ1
)
N
u P
2
u
t
ðCVꢀEVÞ
Nꢀ1
ð2Þ
X-ray structure determination: The structures of 1-(ClO4)2 and 5-(ClO4)2
i¼1
snꢀ1
¼
have been published previously.[5]
An orange crystal of 2-(OTf)2, a red crystal of 6-(OTf)2, and a parallepip-
ed dark-red crystal of 7-(ClO4) were placed on a STOE imaging plate dif-
fraction system (I.P.D.S) equipped with an Oxford Cryosystems Cryo-
stream Cooler Device. These crystals were studied by X-ray diffraction
using graphite-monochromated MoKa radiation (l=0.71073 Š) operating
in the f scan method. The cell measurement temperature was 180(2) K
for 1-(ClO4)2 and 6-(OTf)2, and 160(2) K for 2-(OTf)2 and 7-(ClO4).
Their structures were solved by direct methods using the SIR92 pro-
gram,[23] and subsequent difference Fourier maps. The models were re-
fined by least-squares procedures on F2 with the aid of SHELX-97.[24]
To simulate the UV/Vis spectra an uncontracted triple-& basis set was
used for all the atoms. For the copper atom this basis set was augmented
by an extra polarization function.[37,38] We have checked in several com-
plexes that the inclusion of diffuse functions in the basis set and relativis-
tic effects in the hamiltonian only produce slight changes in the positions
of the bands, and therefore to reduce the computational cost these effects
have not been taken into account.[40]
The effect of solvent on the UV/Vis spectra was considered by using the
conductor-like screening model (COSMO) of Klamt and Schüürmann,[41]
as implemented by Pie and Ziegler in the ADF program.[42] The radius
chosen for the solvent [2.75 Š for acetonitrile (e=37.50)][43] was obtained
by calculating the molecular volume with the Gaussian 98 package.[44]
The radii used for the C, N, Cl, H, O, and Cu atoms were 2.00, 1.40, 1.20,
1.18, 1.30, and 1.50 Š, respectively.[42]
Statistical disorder was found on the atoms labeled N2 and C5 in com-
plex 1 and is located on the two sites with a ratio of occupancy close to
50%. Disorder was also found on the SO3 group of the triflate anion in
complex 6-(OTf)2 and was found to be statistically distributed between
the two sites in a ratio of occupancy equal to 50%; some restraints were
introduced on interatomic distances and angles in order to regularize the
geometry of the SO3 group. Further, disordered perchlorate oxygen
atoms were also found in complex 7-(ClO4), which contains two almost
identical molecules of the chloro cationic complex.
All molecular orbitals were constructed by using the MOLEKEL pro-
gram.[45]
Data reduction, Lorentz polarization, and empirical absorption correc-
tions were performed using the Sadabs package. Structures were solved
by direct methods using SHELXS-86,[25] and refined by full-matrix least-
squares analyses with SHELXL-93.[26]
Acknowledgements
CCDC-261432 (2), CCDC-261433 (3), CCDC-261434 (6), and CCDC-
261435 (7) contain the supplementary crystallographic data for this
paper. These data can be obtained free of charge from the Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
This research has been financed by the MCYT of Spain through projects
BQU2003-02884, BQU2003-01677 and BQU2002-04112-C02-02. A.L.
and M.S. are grateful to the CIRIT Generalitat de Catalunya (Spain) for
the Distinction award and A.L. is grateful for aid through SGR2001-UG-
291. A.L. also thanks Johnson and Matthey for the loan of platinum
metal. X.R. is grateful for a Juan de la Cierva contract from the MEC.
R.X. and A.P. are grateful for the award of doctoral grants from the
UdG and the MEC, respectively.
Computational details: The reported calculations were carried out by
using the Amsterdam density functional (ADF) package developed by
Baerends and co-workers[27–29] and vectorized by Ravenek.[30] The numer-
ical integration scheme employed was that of te Velde and Baerends.[31]
Both geometry optimizations and energy evaluations were performed
using a generalized gradient approximation (GGA) that includes Beckeꢁs
GGA exchange correction[32] and Perdewꢁs GGA correlation correc-
tion.[33] Experimentally it was found that these mononuclear species are
diamagnetic. For this reason, optimizations were carried out for neutral
closed-shell singlet ground-state structures. However, we have checked
the relative stabilities of the singlet and triplet states of all the calculated
complexes. In all cases, the triplet state has a higher energy. This is in
agreement with the sharp peaks in the diamagnetic regions of the NMR
spectra. The difference between the two states is around 25 kcalmolꢀ1 in
all the cases studied.
[1]a) J. P. Collman, L. S. Hegedus, J. R. Norton, R. G. Finke in Princi-
ples and Applications of Organotransition Metal Chemistry, 2nd ed.,
University Science, Mill Valley, 1987, Chapter 7; b) A. E. Martell,
D. T. Sawyer, Oxygen Complexes and Oxygen Activation by Transi-
tion Metals, Texas A & M University, 1987; c) M. D. Fryzuk, S. A.
Johnson, Coord. Chem. Rev. 2000, 200–202, 379–409; d) M. Al-
brecht, G. van Koten, Angew. Chem. 2001, 113, 3866–3898; Angew.
Chem. Int. Ed. 2001, 40, 3750–3781; e) W. D. Jones, Acc. Chem. Res.
2003, 36, 140–146; f) M. E. van der Boom, D. Milstein, Chem. Rev.
2003, 103, 1759–1792.
[2]a) J. A. Labinger, J. E. Bercaw, Nature 2002, 417, 507–514; b) J. M.
Thomas, R. Raja, G. Sankar, R. G. Bell, Acc. Chem. Res. 2001, 34,
191–200; c) H. Chen, S. Schlecht, T. C. Semple, J. F. Hartwig, Sci-
ence 2000, 287, 1995–1997; d) R. H. Crabtree, Chem. Rev. 1995, 95,
987–1007; e) A. E. Shilov, G. B. Shul’pin, Chem. Rev. 1997, 97,
2879–2932.
The UV spectra were simulated by the time-Dependent DFT method
(TDDFT),[34] using the LB94 potential of van Leeuwen and Baerends.[35]
The 2000.02 release of the ADF package was used for all geometry opti-
mizations and the calculated UV/Vis spectra were obtained by using the
2002.03 release of the ADF program.[36]
In the geometry optimizations we used an uncontracted triple-& basis set
to describe the 3s, 3p, 3d, 4s, and 4p orbitals of copper. For carbon
(2s,2p), nitrogen (2s,2p), oxygen (2s,2p), and hydrogen (1s), double-&
basis sets were employed and augmented by an extra polarization func-
[3]M. J. Gradassi, N. W. Green, FuelProcess. Techno.l 1995, 42, 65–83.
[4]M. Gandelman, L. J. W. Shimon, D. Milstein, Chem. Eur. J. 2003, 9,
4295–4300.
Chem. Eur. J. 2005, 11, 5146 – 5156
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5155

X. Ribas, A. Llobet, M. Solà et al.
[5]X. Ribas, D. A. Jackson, B. Donnadieu, J. Mahía, T. Parella, R.
Xifra, B. Hedman, K. O. Hodgson, A. Llobet, T. D. P. Stack, Angew.
Chem. 2001, 113, 3117–3120; Angew. Chem. Int. Ed. 2002, 41, 2991–
2994.
[6]L. L. Diaddario, W. R. Robinson, D. L. Margerum, Inorg. Chem.
1983, 22, 1021–1025.
[26]SHELX-93, G. M. Sheldrick, University of Gçttingen (Germany),
1998.
[27]E. J. Baerends, D. E. Ellis, P. Ros, Chem. Phys. 1973, 2, 41–51.
[28]C. Fonseca Guerra, O. Visser, J. G. Snijders, G. te Velde, E. J. Baer-
ends, Methods and Techniques for ComputationalChemistry , STEF,
Cagliari, 1995, p. 305.
[7]I. H. Wasbotten, T. Wondimageng, A. Ghosh, J. Am. Chem. Soc.
2002, 124, 8104–8116.
[8]a) H. Furuta, H. Maeda, A. Osuka, J. Am. Chem. Soc. 2000, 122,
803–807; b) M. Stepien, L. Latos-Grazynski, Chem. Eur. J. 2001, 7,
5113–5117.
[29]G. te Velde, F. M. Bickelhaupt, E. J. Baerends, C. Fonseca Guerra,
S. J. A. Van Gisbergen, G. J. Snijders, T. Ziegler, J. Comput. Chem.
2001, 22, 931–967.
[30]W. Ravenek, Algorithms and Applications on Vector and Parallel
Computers, Elsevier, Amsterdam, 1987.
[9]a) M. A. Willert-Porada, D. J. Burton, N. C. Baenziger, J. Chem.
Soc., Chem. Commun. 1989, 1633–1634; b) D. Naumann, T. Roy,
K. F. Tebbe, W. Crump, Angew. Chem. 1993, 105, 1555–1556;
Angew. Chem. Int. Ed. Engl. 1993, 32, 1482–1483; c) R. Eujen, B.
Hoge, D. J. Brauer, J. Organomet. Chem. 1996, 519, 7–20.
[10]a) T. D. P. Stack, Dalton Trans. 2003, 1881–1889; b) N. W. Aboelella,
E. A. Lewis, A. M. Reynolds, W. W. Brennessel, C. J. Cramer, W. B.
Tolman, J. Am. Chem. Soc. 2002, 124, 10660–10661.
[11]S. K. Burke, Y. Xu, D. W. Margerum, Inorg. Chem. 2003, 42, 5807–
5817.
[12]D. H. R. Barton, N. C. Delanghe, H. Patin, Tetrahedron 1997, 53,
16017–16028.
[31]G. te Velde, E. J. Baerends, J. Comp. Phys. 1992, 99, 84–98.
[32]A. D. Becke, Phys. Rev. A 1988, 38, 3098–3100.
[33]J. P. Perdew, Phys. Rev. B 1986, 33, 8822–8824.
[34]a) E. K. U. Gross, J. F. Dobson, M. Petersilka, Density Functional
Theory (Ed.: R. F. Nalewajski), Springer, Heidelberg, 1996;
b) S. J. A. van Gisbergen, J. G. Snijders, E. J. Baerends, Comput.
Phys. Commun. 1999, 118, 119–138.
[35]R. van Leeuwen, E. J. Baerends, Phys. Rev. A 1994, 49, 2421–2431.
[36]ADF2002.03, E. J. Baerends, J. A. Autschbach, A. Bꢂrces, C. Bo,
P. M. Boerrigter, L. Cavallo, D. P. Chong, L. Deng, R. M. Dickson,
D. E. Ellis, L. Fan, T. H. Fischer, C. Fonseca Guerra, S. J. A. van Gis-
bergen, J. A. Groeneveld, O. V. Gritsenko, M. Grüning, F. E. Harris,
P. van den Hoek, H. Jacobsen, G. van Kessel, F. Kootstra, E. van -
Lenthe, V. P. Osinga, S. Patchkovskii, P. H. T. Philipsen, D. Post,
C. C. Pye, W. Ravenek, P. Ros, P. R. T. Schipper, G. Schreckenbach,
J. G. Snijders, M. Solà, M. Swart, D. Swerhone, G. te Velde, P. Ver-
nooijs, L. Versluis, O. Visser, E. van Wezenbeek, G. Wiesenekker,
S. K. Wolff, T. K. Woo, T. Ziegler, Vrije Universiteit Amsterdam,
Amsterdam, 2002.
[13]A. Bencini, M. I. Burguete, E. Garcia-EspaÇa, S. V. Luis, J. F. Mira-
vet, C. Soriano, J. Org. Chem. 1993, 58, 4749–4753.
[14]C. Bazzicalupi, A. Bencini, A. Bianchi, V. Fusi, E. Garcia-EspaÇa, P.
Paoletti, P. Paoli, B. Valtancoli, Inorg. Chem. 1993, 32, 4900–4908.
[15]a) K. Hiraki, J. Tsutsumida, Y. Fuchita, Chem. Lett. 1986, 337–340;
b) K. Hiraki, Y. Fuchita, Y. Ohta, J. Tsutsumida, K. I. Hardcastle, J.
Chem. Soc., Dalton Trans. 1992, 833–836; c) A. Chellini, G. B. Gio-
venzana, R. Pagliarin, G. Palmisano, M. Sisti, Helv. Chim. Acta
2000, 83, 793–800.
[37]G. J. Snijders, E. J. Baerends, P. Vernooijs, At. Nucl. Data Tables
1982, 26, 483–509.
[16]a) G. Golub, H. Cohen, P. Paoletti, A. Bencini, L. Messori, I. Berti-
ni, D. Meyerstein, J. Am. Chem. Soc. 1995, 117, 8353–8361; b) D.
Meyerstein, Coord. Chem. Rev. 1999, 185–186, 141–147.
[17]a) P. V. Bernhardt, J. Am. Chem. Soc. 1997, 119, 771–774; b) P. V.
Bernhardt, Inorg. Chem. 2001, 40, 1086–1092; c) P. V. Bernhardt, P.
Comba, T. W. Hambley, Inorg. Chem. 1993, 32, 2804–2809; d) M. M.
Bernardo, M. J. Heeg, R. R. Schroeder, L. A. Ochrymowycz, D. B.
Rorabacher, Inorg. Chem. 1992, 31, 191–198.
[38]P. Vernooijs, E. J. Baerends, Slater Type Basis Functions for the
Whole Periodic System. Internal Report, Vrije Universiteit of Am-
sterdam, Amsterdam, 1981.
[39]J. Krijn, E. J. Baerends, Fit Functions in the HFS Method. Internal
Report (in Dutch), Vrije Universiteit of Amsterdam, Amsterdam,
1984.
[40]E. van Lenthe, E. J. Baerends, J. Comput. Chem. 2003, 24, 1142–
1156.
[18]It is important to bear in mind the existence of a certain controversy
on the purity of the compounds reported in the work published by
Meyerstein (see ref. [16a]) and Bernhardt (see ref. [17a]).
[19]a) J. Hanss, H.-J. Krüger, Angew. Chem. 1996, 108, 2989–2991;
Angew. Chem. Int. Ed. Engl. 1996, 35, 2827–2830; b) J. Hanss, A.
Beckmann, H.-J. Krüger, Eur. J. Inorg. Chem. 1999, 163–172.
[20]E. I. Solomon M. E. Hanson in Inorganic Electronic Structure and
Spectroscopy, Vol. 2 (Eds.: E. I Solomon, A. B. P. Lever), Wiley,
1999, Chapter 1.
[21]a) L. Bernasconi, J. Blumberger, M. Sprik, R. Vuilleumier, J. Chem.
Phys. 2004, 121, 11885–11899; b) X. Ribas, J. C. Dias, J. Morgado,
K. Wurst, E. Molins, E. Ruiz, M. Almeida, J. Veciana, C. Rovira,
Chem. Eur. J. 2004, 10, 1691–1704; c) Y. Yamaguchi, S. Yokoyama,
S. Mashiko, J. Chem. Phys. 2002, 116, 6541–6548; d) Z.-L. Cai, K.
Sendt, J. R. Reimers, J. Chem. Phys. 2002, 117, 5543–5549.
[22]S. J. A. van Gisbergen, A. Rosa, G. Ricciardi, E. J. Baerends, J.
Chem. Phys. 1999, 111, 2499–2506.
[41]A. Klamt, G. Schüürmann, J. Chem. Soc., Perkin Trans. 2 1993, 799–
805.
[42]C. C. Pye, T. Ziegler, Theor. Chem. Acc., 1999, 101, 396–408.
[43] Handbook of Chemistry and Physics, 83rd ed. (Ed.: D. R. Lide),
CRC Press, Boca Raton FL, 2002–2003.
[44]Gaussian 98 (Revision A.11), M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski,
J. A. Montgomery, Jr., R. E. Stratmann, J. C. Burant, S. Dapprich,
J. M. Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J.
Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli,
C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson, P. Y. Ayala, Q.
Cui, K. Morokuma, P. Salvador, J. J. Dannenberg, D. K. Malick,
A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. Cioslowski, J. V.
Ortiz, A. G. Baboul, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz,
I. Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A.
Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W.
Gill, B. Johnson, W. Chen, M. W. Wong, J. L. Andres, C. Gonzalez,
M. Head-Gordon, E. S. Replogle, J. A. Pople, Gaussian, Inc., Pitts-
burgh PA, 2001.
[23]A. Altomare, G. Cascarano, G. Giacovazzo, A. Guagliardi, M. C.
Burla, G. Polidori, M. Camalli, J. Appl. Crystallogr. 1994, 27, 435–
436.
[24]SHELX-97 (includes SHELXS97, SHELXL97, CIFTAB), Programs
for Crystal Structure Analysis (Release 97–2), G. M. Sheldrick, Uni-
versity of Gçttingen (Germany), 1998.
[45]MOLEKEL 4.0, P. Flükiger, H. P. Lüthi, S. Portmann, J. Weber,
Swiss Center for Scientific Computing, Manno (Switzerland), 2000.
[25]SHELXS-86, Programs for the Solution of Crystal Structures; G. M.
Sheldrick, University of Gçttingen (Germany), 1998.
Received: January 25, 2005
Published online: June 30, 2005
5156
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 5146 – 5156