Unsupported Au(i)...Cu(i) interactions: Influence of nitrile ligands and aurophilicity on the structure and luminescence

Article abstract of DOI:10.1039/b900768g

The synthesis, structural characterization and the study of the photophysical properties of complexes [AuCu(C₆F₅) ₂(NC-CH₃)₂] 1, [AuCu(C₆F ₅)₂(NC-Ph)₂]₂2, and [AuCu(C ₆F₅)₂(NC-CHCH-Ph)₂] 3 have been carried out. The crystal structures of complexes 1-3 consist of dinuclear Au-Cu units built from mediated metallophilic Au(i)...Cu(i) interactions. In the case of complex 2 two dinuclear units interact via an aurophilic interaction leading to a tetranuclear Cu-Au-Au-Cu arrangement. Complex 2 is brightly luminescent in solid state at room temperature and at 77 K with a lifetime in the nanoseconds range, while complexes 1 and 3 do not display luminescence under the same conditions. The presence of the aurophilic interaction in complex 2 seems to be responsible for the blue luminescence observed. DFT and time-dependent DFT calculations agree with the experimental results and support the idea that the origin of the luminescence of these complexes arise from orbitals located in the interacting metals. The Royal Society of Chemistry 2009.

Full text of DOI:10.1039/b900768g

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Unsupported Au(I) ◊ ◊ ◊ Cu(I) interactions: influence of nitrile ligands and
aurophilicity on the structure and luminescence†
Eduardo J. Ferna´ndez,a Antonio Laguna,b Jose´ M. Lo´pez-de-Luzuriaga,*a Miguel Monge,a Manuel Montiel,a
M. Elena Olmosa and Mar´ıa Rodr´ıguez-Castilloa
Received 14th January 2009, Accepted 4th June 2009
First published as an Advance Article on the web 21st July 2009
DOI: 10.1039/b900768g
The synthesis, structural characterization and the study of the photophysical properties of complexes
≡
≡
≡
=
[AuCu(C6F5)2(N C–CH3)2] 1, [AuCu(C6F5)2(N C–Ph)2]2 2, and [AuCu(C6F5)2(N C–CH CH–Ph)2]
3 have been carried out. The crystal structures of complexes 1–3 consist of dinuclear Au–Cu units built
from mediated metallophilic Au(I) ◊ ◊ ◊ Cu(I) interactions. In the case of complex 2 two dinuclear units
interact via an aurophilic interaction leading to a tetranuclear Cu–Au–Au–Cu arrangement. Complex 2
is brightly luminescent in solid state at room temperature and at 77 K with a lifetime in the
nanoseconds range, while complexes 1 and 3 do not display luminescence under the same conditions.
The presence of the aurophilic interaction in complex 2 seems to be responsible for the blue
luminescence observed. DFT and time-dependent DFT calculations agree with the experimental results
and support the idea that the origin of the luminescence of these complexes arise from orbitals located
in the interacting metals.
One of the most successful strategies for the synthesis of
Introduction
these type of materials is an acid–base reaction between basic
gold substrates and acidic heterometal precursors. In our case,
in the last few years we have developed a synthetic procedure
that involves bis(perhalophenyl)aurate(I) anions. The use of these
materials offers the additional advantage of the stability that
perhalophenyl groups provide to the synthesized complexes. Thus,
the use of [Au(C6X5)2]- (X = F, Cl) Lewis basis against silver(I)
or thallium(I) salts5 has led to broad families of compounds
displaying different structural arrangements as dimers, loosely
bound butterfly clusters or 1D-, 2D- and 3D-polymeric species.
These special types of interactions are rather strong (around
200 kJ mol-1) and they display a very strong ionic component
(ca. 80%) due to the ionic nature of the metal counterparts and
an additional metallophilic contribution (ca. 20%), as we have
observed in several ab initio studies. Moreover, in some cases
we have observed additional contributions to the unsupported
interaction between metallic fragments as hydrogen bonding,6g
p–p interactions6a or even Cipso ◊ ◊ ◊ M interactions (Cipso from
C6F5 rings of the aurate units).6b But, although one would
expect that all the coinage metals would be among the most
favorable candidates for these type of materials, a view of the
published literature demonstrates that gold–copper complexes
are almost non-existent. Therefore, we have recently focused our
attention on Cu(I) as the heterometal, and using these [Au(C6F5)2]-
basic units we have reported the first unsupported Au(I) ◊ ◊ ◊ Cu(I)
Polynuclear complexes bearing Au(I) ◊ ◊ ◊ Au(I) (aurophilic) or
Au(I) ◊ ◊ ◊ M (metallophilic) closed-shell interactions represent an
entire interdisciplinary research area whose significance spans
several interests, such as supramolecular structural analysis,1
theoretical calculations,2 photophysical properties3 or potential
applications.4 Among them, unsupported closed-shell interactions
of the type Au(I) ◊ ◊ ◊ M4b,5 (M = Ag(I), Tl(I), Hg(II), Hg(0), etc.) are
an important class of interactions since they govern the formation
of many different types of structural arrangements ranging from
discrete species to supramolecular assemblies. Many of these
complexes bearing unsupported Au ◊ ◊ ◊ M interactions display
fascinating photoluminescent properties that can be tuned from
blue to red emissions or from short-lived (fluorescent) to long-lived
(phosphorescent) emitters, depending on the type and number of
ligands, the heterometal, the environment around the metal centers
or the intermetallic distances.3a Theoretical calculations are being
used to explain the nature of the metallophilic closed-shell
interactions found in the solid state structures and to support and
explain the photophysical properties observed experimentally.6
Moreover, some of these luminescent heterometallic systems can
be applied as sensors for volatile organic compounds4a–c or are
promising candidates for their use as LEDs.4d
aDepartamento de Qu´ımica, Universidad de La Rioja, Grupo de S´ıntesis
Qu´ımica de La Rioja, UA-CSIC. Complejo Cient´ıfico-Tecnolo´gico, 26004-
Logron˜o, Spain. E-mail: josemaria.lopez@unirioja.es; Fax: +34 941 299
621; Tel: +34 941 299 649
+
interaction between aurate fragments and a [Cu(pyrimidine)]n
chain as a first result.7 Using a different strategy we have also
≡
synthesized the polymeric complex [Au2Cu2(C6F5)4(N C–CH3)2]n
bDepartamento de Qu´ımica Inorga´nica, Instituto de Ciencia de Materiales
de Arago´n, Universidad de Zaragoza-CSIC, 50009-Zaragoza, Spain
≡
in a transmetalation reaction between [Au2Ag2(C6F5)4(N C–
† Electronic supplementary information (ESI) available: X-ray crystallo-
graphic data in CIF format for 1–3; phase and modulation curves and
normalized residuals of complex 2 for lifetime measurements; theoretical
excitations for model 2a. CCDC reference numbers 708198–708200. For
ESI and crystallographic data in CIF or other electronic format see DOI:
10.1039/b900768g
CH3)2]n and CuCl.8 The isolation of these derivatives displaying
unsupported Au(I) ◊ ◊ ◊ Cu(I) interactions illustrates an easy access
to a new class of Au(I)–Cu(I) compounds presenting unsup-
ported metallophilic interactions and interesting photophysical
properties.
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 7509–7518 | 7509
©

View Article Online
Scheme 1 Synthesis of complexes 1–3.
In addition, regarding properties and applications, the use
of Cu(I) as the heterometal interacting with Au(I) would also
open new fields of research. These fields include, for example,
biological applications, as DNA spectroscopic probes through
the use of N-donor ligands that are constituents of biological
substrates;9 emitting materials for molecular light-emitting devices
(MOLEDs),4d or catalytic applications, since the use of weakly
coordinating anions such as [AuR2]- would enhance the reactivity
of Cu(I) catalytic sites.10
In order to study systematically the unsupported Au(I) ◊ ◊ ◊ Cu(I)
interactions we have chosen nitrile ligands for several reasons:
these ligands can be easily bonded to or displaced from Cu(I)
centers; changes in the alkyl or aryl substituents from the
nitrile ligands keeping the rest of the molecular components
unchanged may produce structural and electronic changes leading
to different structural arrangements and photophysical prop-
erties. Herein we report the synthesis and structural charac-
through a transmetalation reaction, in a 1 : 2 molar ratio (1 : 1
with respect to silver) in acetonitrile. After filtration of the AgCl
formed, complex 1 was isolated as a white solid by precipitation
from the acetonitrile solution by addition of Et2O and subsequent
cooling. In this way, in the presence of an excess of acetonitrile,
≡
complex [AuCu(C6F5)2(N C–CH3)2] (1), in which a Cu(I) center
is joined to two nitrile ligands, is obtained.
The reaction of complex 1 with excess or two equivalents of
nitrile ligands as benzonitrile or cinnamonitrile in toluene, leads
to the substitution of both acetonitrile molecules leading to com-
≡
≡
plexes [AuCu(C6F5)2(N C–Ph)2]2 (2) or [AuCu(C6F5)2(N C–
=
CH CH–Ph)2] (3), respectively (Scheme 1).
The IR spectra of 2 and 3 in Nujol mulls show absorptions
arising from C6F5 groups in the range 1495–1500, 949–962 and
-
779–784 cm-1. In these complexes, the absorptions due to the
organic ligands coordinated to the Cu(I) centers can be assigned.
These bands appear at significantly different energies from the
absorptions of the uncoordinated organic ligands. For example,
≡
terization of the heterometallic complexes [AuCu(C6F5)2(N C–
CH3)2] 1, [AuCu(C6F5)2(N C–Ph)2]2 2, and [AuCu(C6F5)2(N C–
complex 2 displays a band at 2260 cm-1 due to the n(C N)
≡
≡
≡
=
CH CH–Ph)2] 3 (see Scheme 1). We have analyzed the influence
stretching vibration while, in the free ligand, this band appears
-1
≡
of acetonitrile, benzonitrile and cinnamonitrile on the structural
arrangements found in the solid state and, consequently, on the
photophysical properties of these complexes. We have finally
interpreted the results by the use of DFT and time-dependent
DFT calculations that support the experimental findings.
at 2226 cm . Complex 3 displays the corresponding n(C N)
stretching vibration at 2238 cm-1 and this absorption appears at
2218 cm-1 in the case of the free ligand. The IR spectra of complex
1 cannot be obtained due to the loss of acetonitrile molecules and
subsequent decomposition.
The 1H NMR spectra of complexes 1–3 in CD3CN display
the signals corresponding to the organic nitrile species (see
Experimental section). The chemical shifts are similar to those
of the free organic ligands as is expected due to dissociation in
solution.
These complexes also display similar 19F NMR spectra in which
the signals corresponding to the C6F5 groups bonded to Au(I) in
Results and discussion
Synthesis
≡
The starting material [AuCu(C6F5)2(N C–CH3)2] (1) was ob-
tained by reacting [Au2Ag2(C6F5)4(N C–CH3)2]n with CuCl
≡
7510 | Dalton Trans., 2009, 7509–7518
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
the [Au(C6F5)2]- units are observed at -114.8 (Fo), -161.6 (Fp) and
-162.8 (Fm) ppm in all cases.
Complexes 1–3 behave as 1 : 1 electrolytes in 5 ¥ 10-4 M acetoni-
trile solutions (KM = 112 (1); 234 (2) and 112 (3) X-1 cm2 mol-1),
which suggests that metallophilic interactions are not present in
solution and a separation in ionic counterparts is achieved. It is
also worth mentioning that the large value obtained for complex
2 (234 X-1 cm2 mol-1) could be, perhaps, related to an equilib-
rium in solution between the 1 : 1 electrolyte formed between
cationic [Cu(NCPh)2]+ and anionic [Au(C6F5)2]-, and a possible
2[Cu(NCPh)2]+ : {[Au(C6F5)2]2}2- 2 : 1 electrolyte, through associa-
tion between aurate(I) units via aurophilic interactions in solution,
similar to those observed in the solid state structure (see below).
Crystal structures†
Crystal structures of complexes 1–3 were determined by X-ray
diffraction from single crystals obtained by slow diffusion of n-
hexane into a solution of the complex in toluene. Although the
three complexes are in principle very similar and their stoichiome-
≡
tries correspond to the general formula [AuCu(C6F5)2(N C–R)2],
their crystal structures show significant differences. Thus, while
complex 1 crystallizes in the P1 space group of the triclinic system
as discrete dinuclear molecules and without solvent, complex 2
forms tetranuclear Au2Cu2 units through aurophilic interactions
and crystallizes also without solvent, but in the centrosymmetric
Fig. 2 Molecular structure of complex 3 with the labelling scheme for
the atom positions. Hydrogen atoms have been omitted for clarity and
ellipsoids are drawn at the 30% level.
˚
2.5876(5) A) where one of the pentafluorophenyl ligands of the
¯
adjacent [Au(C6F5)2]- units acts as a bridge between Au and Cu,
P1 space group, and compound 3 crystallizes in the same
space group as 2, but as discrete dinuclear molecules and as a
and also longer than in some clusters11,12 (2.589 and 2.584 A).
˚
≡
=
solvate ([AuCu(C6F5)2(N C–CH CH–Ph)2]·0.5PhMe). The half
a molecule of toluene in 2 is disordered over two positions with
the same occupancy for both of them.
The higher steric demand of the alkyl substituent in complex
3 forces the molecules to pack away (shorter intermolecular
˚
Au–Cu distance of 5.641 A), while in 1 the dinuclear units are
The crystal structures of 1 and 3 (Fig. 1 and 2) consists of discrete
located at a shorter, but also non interacting Au–Cu distance of
-
≡
˚
molecules [AuCu(C6F5)2(N C–R)2] in which the [Au(C6F5)2]
3.757 A along the x crystallographic axis forming a “intermittent
+
≡
anion and the [Cu(N C–R)2] cation are held together through
an unsupported Au ◊ ◊ ◊ Cu contact of 2.9335(11) A in 1 and
2.6727(4) A in 3, both of them being considerably longer than
those present in [Au2Cu2(C6F5)4(N C–CH3)2]n, (2.5741(6) and
polymer” (see Fig. 3). In contrast, the intermolecular Au–Cu
˚
˚
8
≡
Fig. 1 Molecular structure of complex 1 with the labelling scheme for the
Fig. 3 Packing in complex 1 viewed along the crystallographic y axis.
atom positions. Ellipsoids are drawn at the 30% level.
Hydrogen atoms have been omitted for clarity.
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 7509–7518 | 7511
©

View Article Online
˚
of leading forces. The Au ◊ ◊ ◊ Cu contacts, with Au–Cu separations
distance in 3 is about 0.26 A shorter than in 1, and the Au–Cu
˚
of 2.6163(12) and 2.6092(12) A, are stronger than in the poly-
separation in the latter is also longer than that found in the related
7
7
˚
≡
meric [Cu{Au(C6F5)2}(N C–Me)(m2-C4H4N2)]n (2.8216(6) A)
≡
polymeric compound [Cu{Au(C6F5)2}(N C–Me)(m2-C4H4N2)]n
˚
˚
and also stronger than in complexes 1 (2.9335(11) A) and 3
(2.6727(4) A). The presence of phenyl substituents in the nitrile
(2.8216(6) A), which also displays unsupported Au ◊ ◊ ◊ Cu interac-
˚
tions. The lengthening of the Au–Cu distance in the acetonitrile
derivative 1 if compared to the cinnamonitrile compound 3 can
be probably attributed to the higher donor capability of the
N-donor ligand in 1 due to the presence of a –Me instead of
ligands, with a higher withdrawing capability than the substituents
in 1 or 3, makes the copper even more acidic than in complex
3 and, hence, leads to shorter Au ◊ ◊ ◊ Cu lengths. The slightly
shorter Au–Cu distances present in the acetonitrile derivative
=
a –CH CH–Ph substituent, what makes the copper center less
8
˚
≡
[Au2Cu2(C6F5)4(N C–CH3)2]n, (2.5741(6) and 2.5876(5) A) could
acidic in 1. Consequently, the attraction for the bis(aryl)aurate
anion decreases and leads to a longer Au–Cu distance. In both
structures the Cipso atom of one of the C6H5 groups (C1) maintains
an interaction with the copper centre of different strength with
seem to be contradictory; nevertheless, an additional factor has
to be taken into account in this last case: the presence of
pentafluorophenyl bridging ligands between Au and Cu, which
get the metal centers even closer. As in the structures of 1 and
3 one of the carbon atoms bonded to gold in each [Au(C6F5)2]-
fragment maintains an interaction with a copper centre, displaying
˚
Cu–C distances of 2.919 (1) or 2.638 A (3). Such contacts are
responsible for the narrow C–Au–Cu angle in which the interacting
C atom is implied (68.96◦ (1) or 66.36◦ (3)). The other one displays
a value of 110.60◦ (1) and 114.78◦ (3). If the Au ◊ ◊ ◊ Cu contact is
considered, both 1 and 3 exhibit a distorted planar T-frame Cu(I)
center formed by its additional coordination to the nitrogen atoms
˚
Cu–C distances of 2.723 and 2.706 A, intermediate between those
observed in 1 and 3. The corresponding C–Au–Cu angles are again
narrow (70.0 and 69.4◦) if compared to the other two C–Au–Cu
angles (107.1 and 108.3◦). Again, the copper centers show a nearly
planar environment (see Table 3) by coordination to the nitrogen
atoms of two benzonitrile ligands and considering the Au ◊ ◊ ◊ Cu
contact. The Cu–N bond lengths, ranging form 1.869(8) to and
˚
of two nitrile molecules (Cu–N: 1.875(15) and 1.941(14) A in 1,
and 1.866(3) and 1.871(3) A in 3). Finally, the gold(I) atoms are
˚
linearly coordinated to two pentafluorophenyl groups with typical
Au–C bond lengths (see Tables 2 and 4) showing an additional (and
rather uncommon) Au ◊ ◊ ◊ Cu contact within the dinuclear unit.
The crystal structure of 2 (Fig. 4) consists of a tetranuclear
˚
1.877(8) A, are all equal and similar to those commented above
for 1 and 3. Finally, if the metal–metal interactions are taken into
account, the environment of the gold(I) atoms could be described
as distorted square-planar, although with significant deviations
from linearity in the Cu–Au–Au angles (165.68(3) and 153.02(3)◦),
which is more pronounced in Au2, and with both planes forming
an angle of about 48◦.
compound formed by two [Au(C6F5)2]- and two [Cu(N C–
≡
Ph)2]+ units joined together through unsupported Au ◊ ◊ ◊ Cu
and Au ◊ ◊ ◊ Au contacts, leading to an unusual Cu–Au–Au–
Cu arrangement (opposite to Coulomb’s rule) where repulsive
instead of attractive forces could be expected. An analogous
situation was previously described by us for the Au–Tl complex
{[Tl(bipy)]2[Au(C6F5)2]2}n6e, where a Tl–Au–Au–Tl disposition of
metals is found, although it may be influenced by the presence of
bridging bipy ligands.
Photophysical properties
The absorption spectra of complexes 1–3 in acetonitrile solutions
(4.5 ¥ 10-6 M) are shown in Fig. 5. Complex 1 displays absorption
bands between 200 (e = 5.7 ¥ 104) and 260 nm (e = 8.2 ¥
103 mol-1 dm3 cm-1). Complex 2 displays a similar profile with
bands between 200 (e = 13.4 ¥ 104), 260 (e = 1.9 ¥ 104), and a less
intense band at 270 nm (e = 7.1 ¥ 103 mol-1 dm3 cm-1) showing a
band edge at lower energy than that of complex 1. Finally, complex
3 shows an absorption spectrum with bands also placed between
200 (e = 15.3 ¥ 104) and 250 (e = 4.9 ¥ 104), and a low energy band
Fig. 4 Molecular structure of complex 2 with the labelling scheme for
the atom positions. Hydrogen atoms have been omitted for clarity and
ellipsoids are drawn at the 30% level.
˚
Fig. 5 Absorption spectra of 1 (blue), 2 (green) and 3 (red) in 4.5 ¥ 10-6
CH3CN at room temperature. Inset: absorption spectra of free benzonitrile
and cinnamonitrile ligands in 4.5 ¥ 10-6 M CH3CN solutions).
M
Surprisingly, a Au ◊ ◊ ◊ Au interaction (of 3.0011(5) A) also
appears in 2 and it is considerably shorter than in the Au–Tl
polymeric compound (3.4092(3) A) despite the apparent absence
˚
7512 | Dalton Trans., 2009, 7509–7518
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
at 270 nm (e = 9.8 ¥ 104 mol-1 dm3 cm-1) showing also a lower
energy band edge if compared to those of complexes 1 or 2.
These profiles are very similar to the one observed for the
gold(I) precursor complex NBu4[Au(C6F5)2] (which displays bands
between 200 and 264 nm with similar intensities) and that were
assigned to p–p* transitions in the pentafluorophenyl rings.5a,6d,7
Nevertheless, the bands at 270 nm observed for compounds 2 and
3, respectively, are not found in the precursor NBu4[Au(C6F5)2],
(see Fig. 5). In this sense, the d10 → d9s1 transitions of the
free Cu+ ion are at much lower energies: 21 390 cm-1 (467 nm)
and 26 260 cm-1 (381 nm) for the spin forbidden and allowed
transitions, respectively.13 Previous work of Che and co-workers14
showed that the transition [3ds* → 4ps] in a dinuclear cop-
per complex occurs at a similar energy, 276 nm (e = 10.8 ¥
103 mol-1 dm3 cm-1). Nevertheless, the formation of dinuclear cop-
per species in solution is highly unlikely to be responsible for this
behaviour. Similarly, absorptions arising from remaining Au–Cu
interactions in solution are also unlikely since the three complexes
behave as 1 : 1 electrolytes when conductivity measurements are
carried out (see Experimental section).
≡
Fig. 7 Theoretical model systems [AuCu(C6F5)2(N C–CH3)2] (1a),
≡
≡
=
[AuCu(C6F5)2(N C–Ph)2]2 (2a) and [AuCu(C6F5)2(N C–CH CH–Ph)2]
(3a) (hydrogen atoms are omitted for clarity).
of the metal–metal distances when the temperature is lowered,
which would lead to a reduction of the HOMO–LUMO band gap
and, consequently, a red shift of the emission energy. In this case,
this rigidochromism could be related to the short distances in the
Cu(I)–Au(I)–Au(I)–Cu(I) metal arrangement where further con-
traction provoked by decreasing temperature is unlikely.
Therefore, these absorptions seem not to have their origins in
the copper center. Instead, they could be assigned to intraligand
(nitrile) transitions influenced by the metal centers. Taking into
As in the other heterometallic complexes reported, the complex
behaves as non-luminescent in fluid solutions, recovering the
property when the solvent is evaporated. This result is interpreted
in terms of a rupture of the metal–metal interactions provoked
by the solvent and their recovery when the solvent is evaporated,
suggesting that the absence of these interactions give rise to the loss
of the luminescence. Taking into account that the three complexes
display gold–copper interactions; that all three have copper in a
similar coordination environment and bonded to the same type
of ligands; that the number of ligands around copper is relatively
low and an attack by basic solvents is possible; and that only in
the case of complex 2 luminescence is observed, it is likely that the
main difference, i.e. the gold–gold interactions that appear in 2,
are of capital importance for its luminescent behavior.
≡
account that the absorptions due to the nitrile group (C N)
appear below 190 nm we propose that these absorptions in solution
are due to p–p* transitions between orbitals in the phenyl groups
of the nitrile substituents. In fact, the absorption spectra of these
ligands (benzonitrile and cinnamonitrile) in acetonitrile display
bands at 270 and 272 nm, respectively (see inset in Fig. 5).
From the three complexes, only in the case of complex
≡
[AuCu(C6F5)2(N C–Ph)2]2 (2) is luminescence in the solid state
observed (Fig. 7). It emits at 442 nm (exc. 392 nm) at room temper-
atureandalso emits at asimilar energy(exc. 371nm) atliquidnitro-
gen temperature (77 K) (Fig. 6). This curious non-dependence on
the temperature, which is termed as luminescence rigidochromism
in other luminescent gold–heteropolynuclear systems,6e,15 has its
origin in an effect that relates the emission features with the envi-
ronmental rigidity in charge transfer states formed in transitions
from a fluid to a rigid solvent.16 Different to that in solution,
in the solid state this phenomenon is not fully understood, since
in transitions mostly metal-based one would expect a shortening
The lifetime measurements, determined by the phase-
modulation technique in the solid state at room temperature
displays two components within the nanosecond time scale [t1 =
2
65 ns, t2 = 17 ns; c = 8.6], indicating that the emission
probably originates from an excited state of singlet parentage and,
consequently, is tentatively assigned as fluorescence. Similarly,
the Stokes-shift between excitation and emission peaks is only
2886 cm-1, which is indicative of a small distortion of the excited
state if compared to the ground state and, therefore, supporting
the assignation of fluorescence for this transition. Interestingly,
≡
the lifetime of the previously reported [Cu{Au(C6F5)2}(N C–
Me)(m2-C4H4N2)]n,7 which shows a polymeric copper–pyrimidine
structure with [Au(C6F5)2]- units interacting with each copper
center, is much longer (10.3 ms) and assigned as phosphorescence.
In that complex, the luminescent behaviour was assigned to arise
from a MMLCT (metal(gold)–metal(copper)-to-ligand-charge-
transfer) state. The observed quenching of the luminescence in
basic solvents by formation of a five-coordinate copper complex
that stabilizes the excited state and destabilizes the ground state
supports this assignation.17 This quench was observed even in
glassy (EtOH : MeOH : CH2Cl2 8 : 2 : 1) solutions at 77 K. In
this case, a glassy solution of complex 2 shows an emission
at 420 nm, suggesting that the formation of the five-coordinate
copper complex is not produced and, consequently, the quench is
not observed, or, if it is formed, the luminescence is not arising only
Fig. 6 Excitation and emission spectra of complex 2 in solid state at room
temperature.
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 7509–7518 | 7513
©

View Article Online
from the gold–copper interactions. In fact, theoretical TD-DFT
calculations agree with this assignation (see theoretical studies
part).
Therefore, taking into account the previous comments, what
we propose is that the luminescence in this complex arises from
a transition between orbitals mostly metal-based and for which
the higher contribution is from orbitals of the interacting gold(I)
centers, although a p–p* transition in the perhalophenyl groups
cannot be excluded. In this sense, previous luminescence studies
on a heteronuclear gold–silver complex with pentafluorophenyl
groups showed that in acetone solutions, where the metal–
metal interactions are not present, luminescence appeared at a
slightly higher energy (405 nm) by excitation at 332 nm that
arose from pentafluorophenyl localized p–p* excited states.5a In
contrast to our case, the precursor complex NBu4[Au(C6F5)2] is
not luminescent in solid state, which reinforces the interacting
gold(I) centers as the main contributors to the orbitals responsible
for this behaviour in the solid state.
Theoretical results
Fig. 8 Frontier molecular orbitals and HOMO–LUMO gaps for model
systems 1a, 3a and 2a (left to right).
In view of the results reported in the photophysical stud-
ies section, single point DFT calculations on model sys-
≡
≡
tems [AuCu(C6F5)2(N C–CH3)2] (1a), [AuCu(C6F5)2(N C–
Ph)2]2 (2a) and [AuCu(C6F5)2(N C–CH CH–Ph)2] (3a) and
orbital placed at the [Au(C6F5)2]- fragment and the LUMO orbital
≡
=
≡
≡
time-dependent DFT calculations on model [AuCu(C6F5)2(N C–
placed at the cinnamonitrile ligand. Model [AuCu(C6F5)2(N C–
Ph)2]2 (2a) have been carried out. These theoretical models
represent the solid state structures of compounds 1–3 since all the
metallophilic interactions responsible for the observed dinuclear
(complexes 1 and 3) or tetranuclear (complex 2) arrangements
are considered in the model systems. All models have been built
up from the X-ray diffraction results (Fig. 7). In order to check
that the use of these model systems is qualitatively correct we
Ph)2]2 (2a) displays the shortest HOMO–LUMO gap (3.03 eV) and
interesting differences regarding the shape of the frontier orbitals
are observed. Thus, the main difference arises from the HOMO
shape that in the case of model 2a is centred on the interacting
Cu ◊ ◊ ◊ Au ◊ ◊ ◊ Au ◊ ◊ ◊ Cu metals with a higher contribution from the
gold centers and showing a ndz2s* character.
As we have mentioned in the experimental photophysical results,
≡
≡
have optimized a simplified model system [AuCu(C6F5)2(N CH)2]
complex [AuCu(C6F5)2(N C–Ph)2]2 2 is the only one that displays
at DFT B3LYP level of theory and using a Cs symmetry in
order to keep the computational costs feasible. The optimized
geometry reproduces fairly well the structures found in the solid
photoluminescent properties upon excitation with UV light. This
trend would also be supported with the theoretical results. Thus,
when no aurophilic interactions are present (complexes 1 and 3) a
large HOMO–LUMO gap is observed in which charge transfer
absorptions arising from [Au(C6F5)2]- fragments to the metals
(complex 1) or to the nitrile ligands (complex 3) may occur, but at
rather high energy. Nevertheless, when an aurophilic interaction is
present (complex 2) the HOMO–LUMO gap is reduced due to the
destabilization of the HOMO orbital and the stabilization of the
LUMO orbital. These trends arise from the combination of the
ndz2 or the ns/np atomic orbitals of the metals that leads to the
corresponding more stable sigma-bonding orbital and less-stable
sigma-antibonding orbital (see Fig. 9). Therefore, it is likely that
this shortening in the HOMO–LUMO gap would lead to a more
energetically available excited state responsible for luminescent
emissions in the visible range.
˚
state showing a Au–Cu interaction of 2.60 A and one Cu–Cipso
interaction of 2.24 A. The Cs symmetry prevents the nitrile ligands
˚
adopting a similar relative disposition as found in the X-ray
analyses of the complexes, but other structural parameters such as
distances and angles are comparable (see ESI for details†).
We first studied the electronic structures of models 1a–3a
obtained through single-point DFT calculations. With these
results we can build up qualitative molecular orbital diagrams
for each model system and the shape of the frontier orbitals can
be checked. Thus, we can check the contribution of each part of
the molecule to the frontier orbitals. These DFT calculations can
be used to explain how the structural differences among complexes
1–3 affect the electronic structures by the qualitative analysis of the
HOMO–LUMO gap. Fig. 8 displays qualitative molecular orbital
diagrams for each model system.
In a second step we have also carried out time-dependent DFT
calculations on model system 2a. These calculations permit us
to reproduce the absorption spectrum of this complex by the
computation of the first few spin-allowed singlet excitations. A
careful analysis of the oscillator strengths and the shape of the
orbitals involved in each excitation relays important information
about the photophysical properties of this complex. Thus, the
obtained theoretical excitation profile can be compared with
the experimental excitation spectra in solid state where the
≡
Model system [AuCu(C6F5)2(N C–CH3)2] 1a shows a large
HOMO–LUMO gap of 4.464 eV. The HOMO orbital for model
1a is located at the [Au(C6F5)2]- fragment, meanwhile the LUMO
orbital is mainly placed at the Cu(I) and Au(I) centers with some
contribution from the nitrile and pentafluorophenyl groups. In
the case of model [AuCu(C6F5)2(N C–CH CH–Ph)2] 3a the
HOMO–LUMO gap is intermediate (3.285 eV) with the HOMO
≡
=
7514 | Dalton Trans., 2009, 7509–7518
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
2
Fig. 9 Qualitative formation of ndz s antibonding and ns/np bonding
orbitals through Cu ◊ ◊ ◊ Au ◊ ◊ ◊ Au ◊ ◊ ◊ Cu interactions (the z-axis is placed
along the metal–metal interaction direction).
Fig. 11 Most intense theoretical singlet excitations in the high energy
region.
metallophilic Cu–Au–Au–Cu arrangement is kept. Fig. 10 displays
the first few singlet excitations (red bars) that are compared with
the excitation profile (blue line) (see Table 5).
This is in agreement with the absence of visible luminescence in
complexes 1 and 3 or even in 2 in solution, where the gold–gold
interactions are loss.
Experimental
Instrumentation
Infrared spectra were recorded in the 4000–200 cm-1 range
on a Perkin-Elmer FT-IR Spectrum 1000 spectrophotometer,
using Nujol mulls between polyethylene sheets. Conductivities
were measured in ca. 5 ¥ 10-4 M acetonitrile solutions with a
Jenway 4510 conductimeter. C, H, N analyses were carried out
with a C.E. Instrument EA-1110 CHNS-O microanalyser. Mass
spectra were recorded on a HP-5989B Mass Spectrometer API-
Electrospray with interface 59987A. 1H and 19F NMR spectra were
recorded on a Bruker ARX 300 in CD3CN. Chemical shifts are
quoted relative to SiMe4 (1H external) and CFCl3 (19F, external).
Absorption spectra in solution were registered on a Hewlett
Packard 8453 Diode Array UV-visible spectrophotometer. Ex-
citation and emission spectra as well as lifetime measurements
were recorded with a Jobin-Yvon Horiba Fluorolog 3-22 Tau-3
spectrofluorimeter.
Fig. 10 Comparison between theoretical first few singlet excitations (red
bars A–I) with the experimental excitation profile in solid state (black line).
Although excitations A–C appear at lower energy and could not
be related to the experimental excitation spectrum, the theoretical
singlet excitations D–I agree well with the experimental excitation
energy. Therefore, the orbitals involved in these singlet transitions
may be related to the fluorescent behaviour observed for complex
2, especially those excitations with larger oscillator strengths,
G and H.
The most intense high energy transitions G and H arise from the
HOMO orbital located at the Cu ◊ ◊ ◊ Au ◊ ◊ ◊ Au ◊ ◊ ◊ Cu interacting
metal centers (mainly at the gold atoms) and arrive to LUMO + 4
and LUMO + 6 orbitals that are placed at the nitrile ligands and
the metals, respectively. These two singlet excitations would be
responsible for the observed high energy fluorescent emission for
complex 2, that would arise from a mixed metal-centered (MC) and
metal-to-ligand-charge-transfer (MLCT) transition in agreement
with the results described in the photophysical studies section
(Fig. 11). Obviously, the rupture of the gold–gold interactions
would increase the HOMO–LUMO gap and would change the
shape of the occupied frontier orbitals, preventing that transition.
Crystallography
The crystals were mounted in an inert oil on glass fibers
and transferred to the cold gas stream of a Nonius Kappa
CCD diffractometer equipped with an Oxford Instruments low-
temperature attachment. Data were collected by monochromated
˚
Mo Ka radiation (l = 0.71073 A). Scan type w and f.
Absorption corrections: numerical (based on multiple scans).
The structures were solved by direct methods and refined on
F2 using the program SHELXL-97.18 If the crystal structure
¯
of complex 1 is refined in the P1 space group (even with
disorder) a second copper center is symmetry generated resulting
in a 1-D polymer built via Au ◊ ◊ ◊ Cu interactions, but with
a sequence ◊ ◊ ◊ Au ◊ ◊ ◊ Cu ◊ ◊ ◊ Cu ◊ ◊ ◊ Au ◊ ◊ ◊ Cu ◊ ◊ ◊ Cu ◊ ◊ ◊ Au ◊ ◊ ◊ and
˚
with distances shorter than 1 A between the Cu atoms. Thus, it had
to be refined in the P1 space group. All non-hydrogen atoms were
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 7509–7518 | 7515
©

View Article Online
Table 1 Data collection and structure refinement details for 1–3
Compound
1
2
3·0.5toluene
Chemical formula
Crystal habit
C16H6AuCuF10N2
Orange prism
0.25 ¥ 0.2 ¥ 0.18
Triclinic
P1
6.6903(1)
7.6776(2)
8.7851(2)
84.411(2)
88.645(2)
79.218(1)
441.170(17)
1
2.547
6761.73
314
-100
72
9.616
10 897
4070
0.0419
C52H20Au2Cu2F20N4
Colorless prism
0.18 ¥ 0.16 ¥ 0.12
Triclinic
C30H14AuCuF10N2·0.5C7H8
Colorless prism
0.4 ¥ 0.3 ¥ 0.1
Crystal size/mm
Crystal system
Space group
Triclinic
¯
¯
P1
P1
˚
a/A
13.423(1)
13.459(1)
16.492(2)
101.794(6)
98.641(7)
119.560(6)
2424.13(4)
2
2.194
1601.73
1512
-100
54
8.1519(1)
14.5421(2)
14.9953(2)
62.108(1)
78.587(1)
77.368(1)
1523.65(3)
2
1.960
899.01
862
-100
60
˚
b/A
˚
c/A
a/◦
b/◦
g /◦
3
˚
U/A
Z
Dc/g cm-3
M
F(000)
T/◦C ◦
2qmax
/
m(Mo Ka)/mm-1
No. of reflections measured
No. of unique reflections
Rint
7.019
5.596
30 949
7817
39 910
10 180
0.0584
0.0464
0.1292
721
288
1.150
3.312
0.069
Ra (I > 2s(I))
wRb (F2, all refl.)
No. of parameters
No. of restraints
Sc
0.0389
0.0920
268
96
1.050
2.252
0.0280
0.0638
525
164
1.024
-3
˚
Max. Dr/e A
0.975
ꢀ
ꢀ
ꢀ
ꢀ
2
2
2
a R: (F) = ꢀFo| - |Fcꢀ/ |Fo |. b wR: (F2) = [ {w(Fo - Fc2)2}/ {w(Fo2)2}]0.5; w-1 = s (Fo2) + (aP)2 + bP, where P = [Fo + 2Fc2]/3 and a and b
ꢀ
are constants adjusted by the program. c S = [ {w(Fo - Fc2)2}/(n - p)]0.5, where n is the number of data and p the number of parameters
2
◦
◦
˚
˚
Table 4 Selected bond lengths [A] and angles [ ] for complex 3
Table 2 Selected bond lengths [A] and angles [ ] for complex 1
Au–C(11)
Au–C(1)
Au–Cu
C(11)–Au–C(1)
C(11)–Au–Cu
C(1)–Au–Cu
2.028(14)
2.064(12)
2.9335(11)
179.4(5)
110.7(3)
69.0(3)
Cu–N(2)
Cu–N(1)
1.875(15)
1.941(14)
Au–C(11)
Au–C(1)
Au–Cu
C(11)–Au–C(1)
C(11)–Au–Cu
C(1)–Au–Cu
2.031(3)
2.053(3)
2.6727(4)
176.12(13)
114.78(8)
66.36(9)
Cu–N(1)
Cu–N(2)
1.866(3)
1.871(3)
N(2)–Cu–N(1)
N(2)–Cu–Au
N(1)–Cu–Au
149.8(8)
95.9(5)
114.1(4)
N(1)–Cu–N(2)
N(1)–Cu–Au
N(2)–Cu–Au
155.88(12)
108.73(9)
94.10(9)
◦
˚
Table 3 Selected bond lengths [A] and angles [ ] for complex 2
Au(1)–C(1)
Au(1)–C(11)
Au(2)–C(31)
Au(2)–C(21)
Au(1)–Cu(1)
Au(2)–Cu(2)
2.052(9)
2.066(9)
2.062(10)
2.079(9)
2.6163(12) Au(1)–Au(2)
2.6092(12)
Cu(1)–N(1)
Cu(1)–N(2)
Cu(2)–N(3)
Cu(2)–N(4)
1.869(8)
1.870(8)
1.870(8)
1.877(8)
3.0011(5)
Table 5 TD-DFT singlet excitation calculations for model
≡
[AuCu(C6F5)2(N C–Ph)2]2 (2a)
Excitation
l
calc/nm
Oscil. str. (s)a
Contributionsb
C(1)–Au–C(11)
C(31)–Au(2)–C(21)
Cu(1)–Au(1)–Au(2)
Cu(2)–Au(2)–Au(1)
N(1)–Cu–N(2)
175.5(3)
N(3)–Cu(2)–N(4)
N(1)–Cu(1)–Au(1) 107.2(2)
N(2)–Cu(1)–Au(1)
N(3)–Cu(2)–Au(2)
N(4)–Cu(2)–Au(2) 104.8(2)
155.3(3)
A
B
C
D
E
482.3
470
432.3
377.4
354.3
0.0681
0.11
0.0218
0.0107
0.0208
318a → 319a (96)
318a → 320a (95)
318a → 322a (97)
315a → 320a (90)
313a → 320a (50)
312a → 319a (17)
312a → 319a (53)
312a → 320a (16)
318a → 323a (88)
318a → 325a (78)
312a → 320a (42)
312a → 321a (25)
176.5(3)
165.68(3)
153.02(3)
152.9(4)
96.5(2)
96.5(2)
F
353.4
0.0263
anisotropically refined with the only exception of C14 (otherwise
non-positive definite), and hydrogen atoms were included using a
riding model. Further details on the data collection and refinement
methods can be found in Table 1. Selected bond lengths and angles
are shown in Tables 2–4 and crystal structures of 1–3 can be seen
in Fig. 1–4. CCDC 708198–708200 contains the supplementary
crystallographic data for this paper.†
G
H
I
337.6
335.4
333.6
0.0507
0.297
0.0152
a Value is |coeff| ¥ 100. b Oscillator strength shows the mixed representa-
2
tion of both velocity and length representations.
7516 | Dalton Trans., 2009, 7509–7518
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
DFT and TD-DFT calculations
7.21–7.30 (m, 1H, CH–Ph), 7.26 (m, 5H, Ph) ppm; MS: m/z (%)
530.5 [Au(C6F5)2]-, 1124.5 [Au2Cu(C6F5)4]- (ES-).
The molecular structures used in the theoretical studies on
≡
≡
[AuCu(C6F5)2(N C–CH3)2] (1a), [AuCu(C6F5)2(N C–Ph)2]2 (2a)
Acknowledgements
≡
=
and [AuCu(C6F5)2(N C–CH CH–Ph)2] (3a) were taken from the
X-ray diffraction data for complexes 1–3, respectively. Keeping all
distances, angles and dihedral angles frozen, single-point DFT
calculations were performed on the models. In them, the single-
point ground-state calculations and the subsequent calculations
of the electronic excitation spectra, the B3-LYP functional19
as implemented in TURBOMOLE20 was used. The excitation
energies were obtained at the density functional level by using
the time-dependent perturbation theory approach (TD-DFT),21–25
which is a DFT generalization of the Hartree–Fock linear response
(HF-LR) or random-phase approximation (RPA) method.26
In all calculations, the Karlsruhe split-valence quality basis
sets27 augmented with polarization functions28 were used (SVP).
The Stuttgart effective core potentials in TURBOMOLE were
used for Au and Cu.29 Calculations were performed without any
assumption of symmetry for 1a–3a.
The D.G.I.(MEC)/FEDER (CTQ2007-67273-C02-02) project
is acknowledged for financial support. M. Rodr´ıguez-Castillo
thanks the C.A.R. for a grant. Professor Vivian W. W. Yam is
thanked for fruitful discussions.
References
1 Modern Supramolecular Gold Chemistry, ed. A. Laguna, Wiley-VCH,
Weinheim, 2008.
2 (a) P. Pyykko¨, Chem. Rev., 1997, 97, 597; (b) P. Pyykko¨, Angew. Chem.,
Int. Ed., 2004, 43, 4412; (c) P. Pyykko¨, Inorg. Chim. Acta, 2005, 358,
4113; (d) P. Pyykko¨, Chem. Soc. Rev., 2008, 37, 1967.
3 (a) E. J. Ferna´ndez, A. Laguna and J. M. Lo´pez-de-Luzuriaga, Dalton
Trans., 2007, 1969; (b) J. M. Forward, J. P. Fackler, Jr., Z. Assefa, in
Optoelectronic Properties of Inorganic Compounds, ed. D. M. Roundhill
and J. P. Fackler, Jr. , Plenum, New York, 1999, pp. 195–226.
4 (a) E. J. Ferna´ndez, A. Laguna, J. M. Lo´pez-de-Luzuriaga and M.
Monge, Spanish Patent P200001391, 2003; (b) E. J. Ferna´ndez, J. M.
Lo´pez-de-Luzuriaga, M. Monge, M. E. Olmos, J. Pe´rez, A. Laguna,
A. A. Mohamed and J. P. Fackler, Jr., J. Am. Chem. Soc., 2003, 125,
2022; (c) E. J. Ferna´ndez, J. M. Lo´pez-de-Luzuriaga, M. Monge, M. E.
Olmos, R. C. Puelles, A. Laguna, A. A. Mohamed and J. P. Fackler, Jr.,
Inorg. Chem., 2008, 47, 8069; (d) H. V. R. Dias, H. V. K. Diyabalanage,
M. G. Eldabaja, O. Elbjeirami, M. A. Rawashdeh-Omary and M. A.
Omary, J. Am. Chem. Soc., 2005, 127, 7489.
Synthesis
≡
Preparation of [AuCu(C6F5)2(N C–Me)2] (1). To an acetoni-
trile solution (20 ml) of [Au2Ag2(C6F5)4(N C–Me)2]n (96 mg,
8
≡
0.070 mmol) was added CuCl (14 mg, 0.141 mmol) and a
precipitate is observed (AgCl). The mixture was stirred for 2 h and
the solid was eliminated by filtration. The solvent of the solution
was evaporated to ca. 5 ml and addition of Et2O (20 ml) at low
temperature led to precipitation of complex 1 as a white solid.
Yield: 70%. KM = 112 X-1 cm2 mol-1. Elemental analysis (%) calcd.
for 1 (C16H6AuCuF10N2): C 28.40, H 0.89, N 4.14; found: C 28.28,
H 0.78, N 4.33; 19F (298 K, CD3CN) d = -162.84 (m, 2F, Fm),
5 (a) E. J. Ferna´ndez, M. C. Gimeno, A. Laguna, J. M. Lo´pez-de-
Luzuriaga, M. Monge, P. Pyykko¨ and D. Sundholm, J. Am. Chem.
Soc., 2000, 122, 7287; (b) E. J. Ferna´ndez, A. Laguna, J. M. Lo´pez-de-
Luzuriaga, M. Monge, M. Montiel, M. E. Olmos, R. C. Puelles and
J. C. Sa´enz, Dalton Trans., 2005, 1162; (c) E. J. Ferna´ndez, P. G. Jones,
A. Laguna, J. M. Lopez-de-Luzuriaga, M. Monge, M. E. Olmos and
R. C. Puelles, Organometallics, 2007, 26, 5931; (d) E. J. Ferna´ndez, A.
Laguna, J. M. Lo´pez-de-Luzuriaga, M. Monge, M. E. Olmos and J.
Pe´rez, J. Am. Chem. Soc., 2002, 124, 5942; (e) V. J. Catalano, B. L.
Bennett and H. M. Kar, J. Am. Chem. Soc., 1999, 121, 10235; (f) A.
Burini, J. P. Fackler, Jr., R. Galassi, B. R. Pietroni and R. J. Staples,
Chem. Commun., 1998, 95; (g) A. Burini, R. Bravi, J. P. Fackler, Jr., R.
Galassi, T. A. Grant, M. A. Omary, B. R. Pietroni and R. J. Staples,
Inorg. Chem., 2000, 39, 3158; (h) A. Burini, A. J. P. Fackler, Jr., R.
Galassi, T. A. Grant, M. A. Omary, M. A. Rawashdeh-Omary, B. R.
Pietroni and R. J. Staples, J. Am. Chem. Soc., 2000, 122, 11264; (i) V. J.
Catalano, M. A. Malwitz and B. C. Noll, Chem. Commun., 2001, 581.
6 (a) E. J. Ferna´ndez, A. Laguna, J. M. Lo´pez-de-Luzuriaga, M. Monge,
P. Pyykko¨ and N. Runeberg, Eur. J. Inorg. Chem., 2002, 750; (b) E. J.
Ferna´ndez, A. Laguna, J. M. Lo´pez de Luzuriaga, M. Monge, M. E.
Olmos and R. C. Puelles, J. Phys. Chem. B, 2005, 109, 20652; (c) E. J.
Ferna´ndez, J. M. Lo´pez-de-Luzuriaga, M. Monge, M. Montiel, M. E.
Olmos, J. Pe´rez, A. Laguna, F. Mendizabal, A. A. Mohamed and
J. P. Fackler, Jr., Inorg. Chem., 2004, 43, 3573; (d) E. J. Ferna´ndez, A.
Laguna, J. M. Lo´pez de Luzuriaga, F. Mendizabal, M. Monge, M. E.
Olmos and J. Pe´rez, Chem.–Eur. J., 2003, 9, 456; (e) E. J. Ferna´ndez,
P. G. Jones, A. Laguna, J. M. Lo´pez-de-Luzuriaga, M. Monge, M. E.
Olmos and J. Pe´rez, Inorg. Chem., 2002, 41, 1056; (f) E. J. Ferna´ndez, A.
Laguna, J. M. Lo´pez-de-Luzuriaga, M. Monge, M. E. Olmos and M.
Montiel, Inorg. Chem., 2007, 46, 2953; (g) E. J. Ferna´ndez, A. Laguna,
J. M. Lo´pez-de-Luzuriaga, M. Monge, M. Nema, M. E. Olmos, J.
Pe´rez and C. Silvestru, Chem. Commun., 2007, 571; (h) M. A. Carvajal,
S. Alvarez and J. J. Novoa, Chem.–Eur. J., 2004, 10, 2117; (i) M. A.
Carvajal, J. J. Novoa and S. Alvarez, Theor. Chem. Acc., 2006, 116,
472; (j) S.-L. Zheng, C. L. Nygren, M. Messerschmidt and P. Coppens,
Chem. Commun., 2006, 3711; (k) S.-L. Zheng, A. Volkov, C. L. Nygren
and P. Coppens, Chem.–Eur. J., 2007, 13, 8583.
´
-161.64 (t, 1F, Fp, JF
(298 K, CD3CN) d = 1.95 (s, 3H, CH3) ppm; MS: m/z (%) 531.3
[Au(C6F5)2]-, 1125.4 [Au2Cu(C6F5)4]- (ES-).
= 19.4 Hz), -114.80 (m, 2F, Fo) ppm; 1H
–F
p
o
≡
[AuCu(C6F5)2(N C–Ph)2]2 (2). To a toluene solution (15 ml)
≡
of [AuCu(C6F5)2(N C–Me)2] (1) (95 mg, 0.140 mmol) was added
≡
N C–Ph (60 ml, exc.) and after 30 min of stirring the solvent was
evaporated to ca. 5 ml. Addition of n-hexane gave rise to a white
solid. Yield: 50%. KM = 234 X-1 cm2 mol-1. Elemental analysis (%)
calcd. for 2 (C52H20Au2Cu2F20N4): C 38.99, H 1.26, N 3.50; found:
C 38.99, H 1.40, N 3.41; 19F (298 K, CD3CN) d = -162.86 (m,
2F, Fm), -161.67 (t, 1F, Fp, JF
= 19.4 Hz), -114.81 (m, 2F, Fo)
–F
p
o
ppm; 1H (298 K, CD3CN) d = 7.30 (m, 2H, H3, H5), 7.44 (m, 1H,
H4), 7.48 (m, 2H, H2, H6) ppm; MS: m/z (%) 530.6 [Au(C6F5)2]-,
1124.6 [Au2Cu(C6F5)4]- (ES-).
≡
=
[AuCu(C6F5)2(N C–CH CH–Ph)2] (3). To a solution of
≡
[AuCu(C6F5)2(N C–Me)2] (1) (90 mg, 0.133 mmol) in 15 ml of
≡
=
toluene was added N C–CH CH–Ph (80 ml, exc.). The mixture
was stirred for 30 min and the solvent was evaporated to ca. 5 ml.
Addition of n-hexane (15 ml) led to precipitation of complex 3
as a white solid. Yield: 60%. KM = 112 X-1 cm2 mol-1. Elemental
analysis (%) calcd. for 3 (C30H14AuCuF10N2): C 42.25, H 1.65, N
3.28; found: C 42.57, H 1.81, N 3.80; 19F (298 K, CD3CN) d =
7 E. J. Ferna´ndez, A. Laguna, J. M. Lo´pez-de-Luzuriaga, M. Monge, M.
Montiel and M. E. Olmos, Inorg. Chem., 2005, 44, 1163.
8 E. J. Ferna´ndez, A. Laguna, J. M. Lo´pez-de-Luzuriaga, M. Monge, M.
Montiel, M. E. Olmos and M. Rodr´ıguez-Castillo, Organometallics,
2006, 26, 3639.
-162.86 (m, 2F, Fm), -161.67 (t, 1F, Fp, JF
= 19.4 Hz), -114.80
–F
p
o
(m, 2F, Fo) ppm; 1H (298 K, CD3CN) d = 5.86 (m, 1H, CH–CN),
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 7509–7518 | 7517
©

View Article Online
9 (a) P. C. Ford, E. Cariati and J. Bourassa, Chem. Rev., 1999, 99, 3625;
(b) D. R. McMillin and K. M. McNett, Chem. Rev., 1998, 98, 1201.
10 Y. Zhang, W. Sun, A. M. Santos and F. E. Ku¨hn, Catal. Lett., 2005,
101, 35.
11 V. G. Albano, C. Castellari, C. Femoni, M. C. Iapalucci, G. Longoni,
M. Monari and S. Zacchini, J. Cluster Sci., 2001, 12, 75.
12 H.-J. Haupt, O. Seewald, U. Florke, V. Buss and T. Weyhermuller,
J. Chem. Soc., Dalton Trans., 2001, 3329.
13 C. E. Moore, Natl. Stand. Ref. Data Ser. (US, Natl. Bur. Stand).
NSRDS-NBS, 1971, 35, 51.
14 Z. Mao, M. Y. Chao, Z. Hui, C. M. Che, W. F. Fu, K. K. Cheung and
N. Zhu, Chem.–Eur. J., 2003, 9, 2885.
18 G. M. Sheldrick, SHELXL-97, Program for refinement of crystal
structures, University of Go¨ttingen, Germany, 1997.
19 (a) A. D. Becke, J. Chem. Phys., 1992, 96, 215; (b) A. D. Becke, J. Chem.
Phys., 1993, 98, 5648; (c) C. Lee, W. Yang and R. G. Parr, Phys. Rev.
Lett., 1998, B37, 785.
20 R. Ahlrichs, M. Ba¨r, M. Ha¨ser, H. Horn and C. Ko¨lmel, Chem. Phys.
Lett., 1989, 162, 165.
21 R. Bauernschmitt and R. Ahlrichs, Chem. Phys. Lett., 1996, 256,
454.
22 R. Bauernschmitt and R. Ahlrichs, J. Chem. Phys., 1996, 104, 9047.
23 R. Bauernschmitt, M. Ha¨ser, O. Treutler and R. Ahlrichs, Chem. Phys.
Lett., 1997, 264, 573 and references therein.
15 (a) S. Wang, G. Garzo´n, C. King, J. C. Wang and J. P. Fackler, Jr., Inorg.
Chem., 1989, 28, 4623; (b) M. Wrighton and D. L. Morse, J. Am. Chem.
Soc., 1974, 96, 998.
16 M. K. Itokazu, A. S. Polo and N. Y. Murakami Iha, J. Photochem.
Photobiol., A, 2003, 160, 27.
17 (a) S. Sakaki, H. Mizutani and Y. Kase, Inorg. Chem., 1992, 31, 4575;
(b) M. K. Eggleston, D. R. McMillin, K. S. Koenig and J. A. Pallenberg,
Inorg. Chem., 1997, 36, 172; (c) R. M. Everly, R. Ziesel, J. Suffert and
D. R. McMillin, Inorg. Chem., 1991, 30, 559; (d) K. Shinozaki and Y.
Kaizu, Bull. Chem. Soc. Jpn., 1994, 67, 2435.
24 E. K. U. Gross and W. Kohn, Adv. Quantum Chem., 1990, 21, 255.
25 M. E. Casida, in Recent advances in density functional methods, ed.
D. P. Chong, World Scientific, Singapore, 1995, vol. 1.
26 J. Olsen and P. Jørgensen, in Modern electronic structure theory, ed.
D. R. Yarkony, World Scientific, London, 1995, vol. 2.
27 A. Scha¨fer, H. Horn and R. Ahlrichs, J. Chem. Phys., 1992, 97,
2571.
28 T. H. Dunning, Jr., J. Chem. Phys., 1994, 100, 5829.
29 D. Andrae, U. Haeussermann, M. Dolg, H. Stoll and H. Preuss, Theor.
Chim. Acta, 1990, 77, 123.
7518 | Dalton Trans., 2009, 7509–7518
This journal is
The Royal Society of Chemistry 2009
©