Experimental and computational evidence of the intermolecular motifs in the crystal packing of luminescent pentacoordinated gallium(III) complexes

Article abstract of DOI:10.1039/b606895b

This paper reports the synthesis, characterization, photophysical and structural properties of the homologous series of good emitting pentacoordinated GaQ′₂L complexes 1-3, where Q′ is 2-methyl-quinolin-8- olate and L is a phenolate substituted in para position with respect to the oxygen donor atom. A combined approach between the experimental structural analysis (i.e. the molecular fragments involved in intermolecular π-π interactions) and the computational study (i.e. the nature of the molecular orbitals residing therein) is discussed in order to compare the charge transport aptitude of complexes 1-3, in relation to the facing of their LUMO/LUMO, HOMO/HOMO and HOMO/LUMO arising from the molecular packing. The different phenolate ligands significantly change the packing characteristics of 1-3, with indirect effects on the electron hopping kinetics responsible for conduction in amorphous thin films. The observed preference for pyridyl-pyridyl stacking proves a faster electron conduction in comparison to hole conduction in the three complexes studied. The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b606895b

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Experimental and computational evidence of the intermolecular motifs in the
crystal packing of luminescent pentacoordinated gallium(III) complexes†‡
Alessandra Crispini,a Iolinda Aiello,a Massimo La Deda,a Irene De Franco,a Mario Amati,b Francesco Lelj*b
and Mauro Ghedini*a
Received 16th May 2006, Accepted 3rd July 2006
First published as an Advance Article on the web 17th July 2006
DOI: 10.1039/b606895b
This paper reports the synthesis, characterization, photophysical and structural properties of the
homologous series of good emitting pentacoordinated GaQꢀ2L complexes 1–3, where Qꢀ is 2-methyl-
quinolin-8-olate and L is a phenolate substituted in para position with respect to the oxygen donor
atom. A combined approach between the experimental structural analysis (i.e. the molecular fragments
involved in intermolecular p–p interactions) and the computational study (i.e. the nature of the
molecular orbitals residing therein) is discussed in order to compare the charge transport aptitude of
complexes 1–3, in relation to the facing of their LUMO/LUMO, HOMO/HOMO and HOMO/LUMO
arising from the molecular packing. The different phenolate ligands significantly change the packing
characteristics of 1–3, with indirect effects on the electron hopping kinetics responsible for conduction
in amorphous thin films. The observed preference for pyridyl–pyridyl stacking proves a faster electron
conduction in comparison to hole conduction in the three complexes studied.
probably due to the experimental intrinsic difficulties which do
Introduction
not allow routine charge transport measurements, data relating
Organic light emitting diodes (OLEDs) are electro-optic devices
quantitative charge transport processes to molecular structures
are only qualitatively inferred.1b,7
built with two oppositely charged electrodes sandwiching amor-
phous thin films of molecular or polymeric materials. Injection
of holes and electrons through these materials induces a series of
processes, which in the end determine the frequency and intensity
of the emitted light.1 The presently available OLED performances
do not meet all the requirements (e.g. cost, colour, working voltage,
durability) demanded by the continuously emerging different
applications, thus stimulating new investigations on both materials
and architecture devices.2
Emitting species like the coordination compounds MQ3 (M =
Al(III) or Ga(III); Q = quinolin-8-olate anion) display electron
transport properties.8 It is generally accepted that the charge
transport efficiency in organic molecular materials is enhanced by
strong intermolecular interactions between polarizable p-systems
existing in their crystalline state. An accurate analysis of the
structural data from X-ray analysis of MQ3 has shown the presence
˚
of strong p–p stacking interactions (3.5–3.9 A) between quino-
Electroluminescent devices are characterised by an important
figure-of-merit, the external quantum efficiency (gext), defined as
the ratio between the emitted photon and the injected electron.3
The gext depends on a number of variables such as the charge
mobility, the fraction of the recombined charge carriers, the
photoluminescent yield and the refractive indices and absorption
coefficients of the various layers. Therefore, a way to improve gext is,
for example, the appropriate optimisation of the charge transport
ability of the amorphous materials.
line ligands of adjacent molecules, with preferential overlap of
the pyridyl rings.9 As in these complexes the Q pyridyl ring is
the principal location of the LUMO and the Q phenyl ring is the
principal location of the HOMO,4b the observed stackings are con-
sidered the most likely cause for the existence of electron mobility
higher than hole mobility. Along with recent investigations dealing
with different OLED materials,10 tests performed on the emitting
complexes MQꢀ2L (M = Al(III) or Ga(III); Qꢀ = 2-methyl-quinolin-
8-olate anion; L = phenolate or carboxylate monodentate ligand)
proved their ability as electron transport species.
We are currently involved in the synthesis and characterization
of coordination compounds suitable for applications in electro-
luminescent devices,11 and in this context we have now focused
our attention on the analysis of the intermolecular interactions
occurring in the crystalline state of pentacoordinated GaQꢀ2Ln
complexes, 1–3 (Scheme 1), where L is a monodentate phenolate
ligand (1: L1 = OC6H5; 2: L2 = OC6H4-4-CN; 3: L3 = OC6H4-
4-NO2). In this paper we report the synthesis, characterization,
photophysical and structural properties of 1–3, with particular
emphasis on the crystal supramolecular motifs formed by the
planar aromatic units. Moreover, computational TD-DFT studies
are performed and discussed in relation to the experimental
absorption and emission properties of this class of compounds.
Charge transport is one of the most important properties
when molecular materials perform as OLEDs,4 organic field-effect
transistors5 and solar cells.6 However, despite its prominent role,
aCentro di Eccellenza CEMIF.CAL-LASCAMM, CR-INSTM Unita` della
Calabria, Dipartimento di Chimica, Universita` della Calabria, I-87036,
Arcavacata di Rende (CS), Italy. E-mail: m.ghedini@unical.it; Fax: +39
(0)984 492066; Tel: +39 (0)984 492062
bLASCAMM, CR-INSTM Unita` della Basilicata and LaMI, Dipartimento
di Chimica, Universita` della Basilicata, I-85100, Potenza, Italy. E-mail:
lelj@unibas.it; Tel: +39 (0)971 202246
† The HTML version of this article has been enhanced with additional
colour images.
‡ Electronic supplementary information (ESI) available: Optimised geome-
tries of complexes 1–3 and the complete list of the computed electronic
excitations for each complex. See DOI: 10.1039/b606895b
5124 | Dalton Trans., 2006, 5124–5134
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
Microcrystalline complexes 1–3 were thermally characterized
by differential scanning calorimetric (DSC) analysis. In all
cases the trace showed only the melting peak (see Experimen-
tal), which excludes any temperature dependence polymorphism
for 1–3.
The final evidence for the molecular structure of complexes 1–3
has been obtained through single crystal X-ray diffraction analysis
performed on good quality crystals obtained in all cases from slow
diffusion of n-hexane in a chloroform solution.
While both complexes 2 and 3 crystallize in the triclinic
space group with almost the same cell parameters, complex 1 is
monoclinic, crystallizing in the P21/c space group. The perspective
drawing of complex 1 is illustrated in Fig. 1 and the selected bond
distances and angles of complexes 1, 2 and 3 are listed in Table 1.
Scheme 1 Molecular structure and proton numbering scheme of the
synthesised complexes 1–3.
The discussion of the experimental structural analysis (i.e. the
molecular fragments involved in intermolecular p–p interactions)
and the computational study (i.e. the nature of the molecular
orbitals residing therein) on the homologous series of compounds
1–3, is reported in this paper as a systematic approach for the study
of the charge transport aptitude based on the LUMO/LUMO,
HOMO/HOMO or HOMO/LUMO facing arising from the
molecular packing.
Results and discussion
Synthesis and structural studies
The synthesis of the GaQꢀ2Ln complexes was performed by adapt-
ing the method suggested by Sapochak et al.10b in order to avoid the
competitive formation of the hexacoordinated tris-chelated GaQꢀ3
derivative. The phenol ligands were activated by sodium hydroxide
in ethanol prior to being added to the reaction mixture containing
gallium(III) nitrate hydrate salt dissolved in water. The addition
of a solution of HQꢀ in ethanol was slowly performed and the
resulting mixture was refluxed overnight. The solid formed upon
cooling, was filtered, washed with water, ethanol and diethyl ether,
before being recrystallized from a chloroform–hexane solution to
yield the title compounds as greenish yellow crystalline powders
in relatively good overall yields. All complexes were characterized
by elemental analysis, IR and 1H NMR spectroscopies (see
Experimental). The infrared spectra show bands characteristic
of the coordinated Qꢀ fragment in the 1610–1400 cm−1 range;
moreover, in the case of 2 the stretching band relative to the –
CN group is observed at 2221 cm−1, while for 3 the intense band at
1303 cm−1 is ascribed to the stretching of the –NO2 group. The 1H
NMR spectra for 1–3 are consistent with the proposed structures,
revealing the magnetic equivalence of the two Qꢀ moieties with only
one set of signals. The chemical shift of the Ha,aꢀ and Hb,bꢀ signals,
corresponding to the hydrogens of the phenol ring (Scheme 1), is
weakly influenced by the presence of the gallium centre. In all cases
a weak shield to lower frequencies can be observed with respect to
the corresponding ligand.
Fig. 1 Perspective of 1 with thermal ellipsoids (50% probability) and
numbering scheme.
In all cases, the GaQꢀ2Ln complexes contain a pentacoordinated
gallium centre. The angles around the Ga(III) ion approximate a
trigonal bipyramid geometry, with the two Qꢀ ligands in an N,N
trans conformation (N–Ga–N angle ranges from 171.4–173.3◦).
˚
The average Ga–O and Ga–N bond distances of 1.87 and 2.09 A
◦
˚
Table 1 Selected bond distances (A) and angles ( ) for compounds 1–3
Complex
1
2
3
Ga–O(1)
Ga–O(2)
Ga–O(3)
Ga–N(1)
1.885(1)
1.881(1)
1.852(1)
2.084(2)
2.091(1)
1.872(1)
1.868(1)
1.851(1)
2.084(2)
2.088(2)
1.868(2)
1.872(3)
1.863(3)
2.081(3)
2.077(3)
The redox potentials were measured versus ferrocene as internal
references (Cp2Fe/Cp2Fe+) using tetrabutylammonium perfluo-
rate (0.1 M) as electrolyte in dichloromethane–acetonitrile (1 : 1)
solutions. The cyclic voltammetric analyses showed no reduction
waves for all complexes inside the solvent window. Oxidation
processes have been noticed as broad irreversible waves above
1300 mV. The non well-defined irreversible oxidation waves and the
precipitation at the Pt working electrode surface prevent further
analysis.
Ga–N(2)
O(1)–Ga–O(2)
O(1)–Ga–O(3)
O(1)–Ga–N(1)
O(1)–Ga–N(2)
O(2)–Ga–O(3)
O(2)–Ga–N(1)
O(2)–Ga–N(2)
N(1)–Ga–N(2)
116.61(7)
117.54(7)
83.78(6)
91.48(6)
125.79(7)
94.25(6)
83.60(6)
173.28(6)
119.18(7)
116.93(7)
83.77(6)
92.43(6)
123.88(7)
91.77(6)
83.66(7)
171.73(6)
118.55(12)
117.85(12)
83.82(11)
91.89(12)
123.59(12)
92.82(12)
83.65(12)
172.35(12)
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 5124–5134 | 5125
©

View Article Online
Table 2 The p–p stacking parameters of compounds 1–3
Compound
[N(1)/C(9)]–[N(1)/C(9)]*
[N(2)/C(19)]–[N(2)/C(19)]*
[C(11)/C(16)]–[C(11)/C(16)]*
C–H-p
a
d
e
˚
h
˚
˚
˚
1
d = 3.47 A
d = 3.46 A
a = 14.9◦
D = 2.90 A
b
˚
S = 1.24 A
S = 0.92 A
ac = 19.7◦
CPh–H–X = 163.1◦
f
˚
˚
d
˚
h
˚
d
˚
2
3
d = 3.43 A
d = 3.55 A
d = 3.65 A
a = 8.90◦
D = 2.66 A
˚
˚
S = 0.98 A
S = 2.40 A
S = 0.57 A
CPh–H–X = 162.6◦
a = 16.4◦
a = 34.2◦
f
˚
˚
d
˚
g
˚
d
˚
d = 3.39 A
d = 3.459 A
d = 3.60 A
D = 2.69 A
˚
˚
S = 1.07 A
S = 2.22 A
S = 0.52 A
CPh–H–X = 158.7◦
a = 17.6◦
a = 31.7◦
a = 8.2◦
a Interplanar distance: distance between the ring planes. b Horizontal displacement: distance between the ring normal and the centroid vector.
c Displacement angle: angle between the ring normal and the centroid vector. *symmetry codes:d 1 − x, 2 − y, 1 − z; e 2 − x, 2 − y, 1 − z; f 2 −
x, 2 − y, 2 − z; g 1 − x, 1 − y, 1 − z; h x, 1 + y, z;
found for compounds 1, 2 and 3, are in good agreement with those
found for other Ga(III) pentacoordinated hydroxyquinolinate
complexes.9b
Intermolecular p–p stacking and hydrogen bonds interactions
are the most remarkable structural features of the GaQꢀ2Ln solids.
An examination of the solid-state packing of complexes 1–3
shows that there are close intermolecular p–p stacking interac-
tions mainly between the pyridyl–pyridyl rings of neighbouring
molecules. The p–p stacking parameters are summarized in
Table 2. Fig. 2 and 3 show the crystal packing and relevant
intermolecular motif interactions in complexes 1 and 2 (complex
3 is isostructural with complex 2).
ꢀ
ꢀ
Fig. 2 Orthogonal views of the QA (a) and the QB (b) chelates stacking in complex 1 (operation symmetries reported in Table 2); crystal packing of
complex 1 showing (c) C–H-p hydrogen bonds between raw of molecules and (d) the generation of secondary repetition motifs.
5126 | Dalton Trans., 2006, 5124–5134
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
ꢀ
ꢀ
Fig. 3 Orthogonal views of the QA (a) and the QB (b) chelates stacking in complex 2 (operation symmetries reported in Table 2); crystal packing of
complex 2 view down the a axis showing (c) the QꢀB stacking contacts and (d) the generation of chains running along the b axis.
In all cases, both the Qꢀ ligands, nominally called the QꢀA (the
N(1) quinolinate) and the QꢀB (the N(2) quinolinate) chelates, are
involved in aromatic interactions with the symmetrical related
rings, while stacks with mixed chelates are not found (Fig. 2a, b
and 3a, b). The primary motif between the planar aromatic units
for both chelates is the offset face-to-face (OFF) type,12 with
the distance between the mean planes passing through the rings
offset (less than 1 A), which causes a nearly complete facing of
˚
the overall aromatic system, involving the pyridyl-fused phenolate
rings in the same interaction (Fig. 2b). Therefore, each molecule of
1 interacts with two symmetrical related molecules through the QꢀA
and the QꢀB chelates, generating extended two-dimensional chains
(Fig. 2c). Among chains, edge-to-face interactions occur between
one hydrogen atom of the L1 ligand and the [C(1)/C(6)] aromatic
ring of the QꢀA chelate. Furthermore, the secondary repetition
motif recognizable is a combination of two OFF and two EF
interactions (Fig. 2d).
The different offset in the OFF interaction between the QꢀB
chelates observed in complexes 2 and 3 (Table 2) is such that
each chelate shows two contacts: a pyridyl–pyridyl [N(2)–C(19)]
interaction, above the mean plane of the QꢀB ligand, and a
phenolate–phenolate [C(11)–C(16)] stacking below it, with the
generation of chains running approximately along the b axis
˚
involved in the stacking within the range 3.3–3.8 A, generally
accepted for the phenomenon of p–p stacking.13 Moreover, the Ln
rings are involved in edge-to-face interactions (EF), often referred
as CH–p hydrogen bonds,14 being the aromatic ring H donor in
the case of complex 1 and H acceptor in both complexes 2 and 3
(Table 2). While the QꢀA–QꢀA stacking is characterized by the same
parameters in the series 1–3, the QꢀB–QꢀB aromatic interaction
differs from complex 1 to complexes 2 and 3. In fact, the QꢀB–
QꢀB stacking in complex 1 is characterized by a small transverse
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 5124–5134 | 5127
©

View Article Online
Table 3 Photophysical data of compounds 1–3
CH2Cl2 solution
KBr pellets
Film
Compound Abs,k/nm (e/M−1cm−1
)
aEm, k/nm, eV (U) Abs, k/nm
Em, k/nm, eV Abs k/nm
Em, k/nm, eV
1
2
3
260(67000)–304(2170,sh)–
507, 2.45 (0.40)
518, 2.39 (0.44)
514, 2.41 (0.25)
265–315(sh)–380
472, 2.63
480, 2.58
480, 2.58
264–320(sh)–362 508, 2.44
261–315(sh)–368 500, 2.48
263–318–340 500, 2.48
315(2570,sh)–365(4500)
260(68370)–304(1690,sh)–
315(2100,sh)–365(2930)
260(68970)–304(11800)–
365(6500,sh)
264–305(sh)–375
263–340
a kex = 365 nm.
(Fig. 3b and c). In both cases the presence of edge-to-face
interactions involving the L2 and L3 rings, respectively, gives
rise to a combination of three OFF and two EF interactions as
a supramolecular secondary motif, connecting three molecules
together (Fig. 3d). These repetition units are linked between p–p
interactions involving the QꢀA chelates of symmetrical molecules.
As a final confirmation of the lack of any polymorphism in the
Ga(III) derivatives 1, 2 and 3, X-ray powder diffraction analysis on
the bulk sample showed it to be identical in all cases to the single
crystal used for structure determination.
and a high quantum yield, with values of about 40% for 1 and
2, and 25% for 3, measured by exciting the solution compounds
at 365 nm. Interestingly, while for 1 and 2 the U-value is almost
independent of the exciting wavelength in the range 290–410 nm,
for 3, with an absorption spectrum which shows a prominent
band at 304 nm, the emission quantum yield measured exciting
at this wavelength is dramatically reduced at 5%. The spectral
characteristics showed in solution by complexes 1–3 are retained
in solid and film samples. Compared to the solution spectra, the
film emission is slightly blue-shifted while a reduced Stokes shift
is noted in the solid samples.
Comparing the electronic spectra of complexes 1–3 with those of
similar pentacoordinated hydroxyquinolinate gallium(III) deriva-
tives containing different monodentate ligands,10b,15 no appreciable
differences can be found. Obviously, this is evidence of the fact that
electronic properties of this class of complexes are mainly due to
the quinolinate ligands. A complete band assignment carried out
by theoretical calculations is reported herein.
DFT and TD-DFT methods have previously been applied to
the study of ground and excited states properties of the AlQ3
complex.16 In particular, the B3LYP/6-31G(d) level of approxima-
tion appears to be more accurate in comparison to analogous ap-
proaches based on pure exchange–correlation functionals such as
BLYP, LB94, BPW91 and LSDA.17 The same level of approxima-
tion has been fruitfully applied to the study of UV-Vis absorption
Photophysical characterization and computational results
Photophysical data of compounds 1–3 in solution, KBr pellets
and a pure thin film are compiled in Table 3.
Absorption spectra of 1 and 2 recorded in dichloromethane
solution are almost indistinguishable, showing an intense band
at 260 nm and a lower-intensity one at 365 nm, while two weak
shoulders are located at 304 and 315 nm. Spectrum of 3 shows the
same intense band at 260 nm observed in the other compounds.
However, the shoulder located at 304 nm in the case of 1 and 2,
is now an evident band centred at the same wavelength, while the
band at 360 nm becomes a shoulder (Fig. 4).
All complexes are luminescent in CH2Cl2 at room temperature,
with a maximum in the range 507–518 nm (Table 3 and Fig. 4)
Fig. 4 Absorption (left) and emission (right) spectra of compounds 1(solid line), 2 (dashed line) and 3 (dotted line) in CH2Cl2 solution.
5128 | Dalton Trans., 2006, 5124–5134 This journal is The Royal Society of Chemistry 2006
©

View Article Online
Fig. 5 Qualitative description of complex 1 Kohn–Sham orbitals close in energy to the HOMO and the LUMO.
and emission features of 8-hydroxyquinoline ligands with different
phenylazo substituents on carbon 5 and their zinc and aluminium
complexes.11d Pentacoordinated complexes of aluminium with
ligands and structure similar to the GaQꢀ2Ln complexes 1–3
(Scheme 1) have been investigated with the same computational
approach.11e On the basis of these considerations, the B3LYP/6-
31G(d) computations have been applied to these pentacoordinated
Ga(III) complexes after ground state energy minimization. Fig. 5
shows the Kohn–Sham (KS) orbital energies and shapes of some
of the highest occupied and lowest unoccupied orbitals of 1. Their
qualitative description can be extended to 2 and 3.
A general feature of all the reported orbitals is the substantial
localization on the ligands, with a very small contribution from the
metal centre. This is a typical feature of this class of compounds
and it has already been pointed out in the cases of AlQ3, ZnQ2
and AlQꢀ2L.11e,16,18 The orbitals reported in Fig. 5 can be grouped
in orbitals localized on the Qꢀ chelants and orbitals localized on
the Ln ligand. The Qꢀ orbitals have been labeled as “q” and consist
of pairs of almost-degenerate orbitals, which are mostly produced
by the slightly interacting Qꢀ fragment orbitals, delocalized on
both chelants. The Ln orbitals have been labeled as “p” orbitals,
and they are very similar to the fragment orbitals of the pyridyl-
fused phenolate moiety. Fig. 6 shows the fragment orbitals that
can be assigned to each KS orbital in 1–3. It is possible to note
that, in the case of 1, the HOMO is localized on L1 (p2), whereas
the LUMO is localized on the Qꢀ ligands (q2). This characteristic
is not retained in the case of 2 and 3. Moreover, Fig. 6 shows the
energy trend of these orbitals in 1–3.
Fig. 6 Energy level scheme of relevant Kohn–Sham orbitals of complexes
1, 2 and 3.
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 5124–5134 | 5129
©

View Article Online
Table 4 Energies of the Kohn–Sham orbitals reported in Fig. 5 and 6.
Interestingly, in the case of compound 1, the excited states S1 and
S2 belong to the p → q charge transfer class, with low oscillator
strengths. These transitions can be assigned to monoelectronic
excitations from the p2 to the q2 orbital. On the other hand, the
excited state S3 is a q-localized transition (q1 → q2 monoelectronic
excitations) with larger oscillator strength. Thus, according to the
computations, the lowest energy experimental absorption band
should not be associated with the HOMO–LUMO transition. This
is the only difference between 1 and compounds 2 and 3, for which
the first excited state is a q1 → q2 transition with larger oscillator
strength. Furthermore, the first intense computed excitation at
399–400 nm for 1–3 is always a q1 → q2 transition, as proved by
the constant energy gap between q1 and q2 orbitals, of 3.68 eV in
all compounds. We can conclude that the substituents on the Ln
ligands do not significantly influence the electronic structure of the
Qꢀ chelants, leaving the q-localized electronic transitions within an
extremely narrow interval of wavelength.
A similar observation can be extended to the fluorescence
process. The experimental similarity in the fluorescence peak wave-
lengths of the three complexes can be explained as a consequence
of the localisation of the emissive electronic transitions on the
Qꢀ chelants. As reported in Table 5, the lowest energy excited
state (S1) in complex 1 should produce a significantly different
emission wavelength peak and quantum yield respect to the other
complexes. In fact, the S1 energy (relative to the ground state)
changes from 2.86 eV (434 nm) in complex 1 to 3.09 eV (400 nm)
in complexes 2 and 3. This gap of 0.23 eV is not confirmed by
our experiments, where the fluorescence emission peak spreads in
a range of 0.06 eV in CH2Cl2 solution, 0.05 eV in KBr pellets
and 0.04 eV in films (Table 3). Furthermore, the shape of the
emission peaks (Fig. 4) is comparable for all complexes, proving a
similarity in the emission processes. The extremely low oscillator
strength assigned to the S1 state of complex 1 should imply a barely
emissive complex, which does not correspond to the experimental
data. Therefore we can conclude that the emission process as well
as the lowest energy absorption processes are dominated by the Qꢀ
chelants. Strong evidence for the predominance of the Qꢀ chelants
in the fluorescence process has already been observed in the case
of similar pentacoordinated aluminium complexes,11e for which a
low energy monodentate phenolate → quinolinate excited state is
present. As a consequence of this finding, after light absorption
and population of the q1 → q2 excited states (from S3 to S6 in
complex 1), the non-radiative internal conversion towards the S1
All the values are in eV
1
2
3
p1
q1
q1
−6.20 p1
−5.59 q1
−5.45 p2
−6.92 p1
−5.84 p2
−5.73 q1
−6.99
−5.98
−5.88
p2 (HOMO) −5.20 q1 (HOMO) −5.70 q1 (HOMO) −5.73
q2 (LUMO)
−1.77 q2 (LUMO)
−1.69 q2
−2.02 q2 (LUMO)
−1.94 q2
−0.77 p3
−0.74 q3
q3
−2.05
−1.97
−1.65
−0.80
−0.77
q2
q3
q3
−0.54 q3
−0.50 q3
The presence of the electron attracting groups –CN and –NO2
in the L2 and L3 ligands, leads to a stabilization of the p orbitals.
As a consequence, in 2 and 3 the HOMO becomes localized on the
Qꢀ chelants (one of the q1 orbitals), whereas the LUMO continues
to be a q2 localised orbital. In the case of compound 3, the virtual
orbital p3 (not reported in the case of compounds 1 and 2) is
computed. It consists in a p orbital with antibonding character
between the oxygen and nitrogen atoms of the –NO2 group. Its
presence is important to understand the absorption spectrum of
this compound, somehow different from those of 1 and 2 (vide
infra). Table 4 collects the energies of the reported orbitals.
According to TD-DFT computations, the electronic spectra
of the three compounds show two principal absorptions (see
the ESI for the complete list of all the computed electronic
transition energies and oscillator strengths‡). The first absorption
is computed at 399 nm (3.10 eV) for 1 and 400 nm (3.10 eV)
for 2 and 3, while the second one is more intense and in all
cases is found between 242 and 247 nm (5.11–5.01 eV). These
electronic transitions can be assigned to changes in the electronic
density within the Qꢀ chelants and, in this sense, can be considered
localized on these chelants. In the case of 3, the presence of the
–NO2 group leads to an intense additional absorption computed
at 304 nm (4.08 eV) associated to an electronic transition from the
p2 quinolinate-localized orbital to the p3 orbital localized on the –
NO2 group. The computed spectrum reproduces the experimental
spectrum well, allowing an accurate peak attribution.
Table 5 reports information about the six lowest energy excited
states. These states can be classified in two classes: q-localized and
p → q charge transfer.
Table 5 The computed lowest energy excited states of compounds 1–3. The reported composition has to be considered a qualitative description in terms
of monoelectronic excitation
1
2
3
State Wavelength (f )a/nm Composition State Wavelength (f )/nm Composition
State Wavelength (f )/nm Composition
S1
S2
S3
S4
S5
S6
S7
434 (0.0007)
417 (0.0005)
399 (0.0819)
389 (0.0052)
378 (0.0044)
370 (0.0040)
317 (0.0001)
p2–q2
p2–q2
q1–q2
q1–q2
q1–q2
q1–q2
p1–q1
S1
S2
S3
S4
S5
S6
S7
400 (0.0805)
392 (0.0049)
388 (0.0031)
378 (0.0019)
377 (0.0039)
368 (0.0042)
296 (0.0016)
q1–q2
p2–q2
S1
S2
400 (0.0809)
390 (0.0091)
377 (0.0038)
369 (0.0029)
368 (0.0007)
356 (0.0045)
334 (0.0003)
q1–q2
q1–q2
q1–q2
q1–q2
p2–q2
p2–q2
q1–p3
mixed q1–q2 and p2–q2 S3
p2–q2
q1–q2
q1–q2
q1–q3
plus other ones
S4
S5
S6
S7
a The f is oscillator strength.
5130 | Dalton Trans., 2006, 5124–5134
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
state is slower than the emission from the populated (pumped)
states, which is in agreement with the apparent low electronic
interaction between the Ln ligands and Qꢀ chelants discussed
previously.
In the case of 1, we can predict that the electron conduction
is better than the hole one. In fact, the observed conductive
stacking of the two chelants QꢀA and QꢀB does not allow a good
overlap between the Qꢀ phenolate rings and, therefore, between
the localized Qꢀ-HOMO. Since the Qꢀ-LUMO (localized on the
Qꢀ pyridyl rings) shows a better overlap, an enhanced electron
transport should be expected. Another important point to be
considered is the localization of the HOMO of 1 on its Ln phenolate
ligand (Fig. 5). If this vacuum-calculated molecular localization is
retained in the crystalline phase, then the hole could be trapped on
the Ln phenolate ligand, with consequent reduced mobility. In the
case of 2 and 3, the QꢀA–QꢀA and QꢀB–QꢀB stackings are favourable
to electron transport; in fact, the QꢀB chelant forms an infinite p–p
stack and two kinds of overlap can be observed along the same
stack. The QꢀB phenyl ring forms p-interactions in one direction,
while the pyridyl ring overlaps in the opposite one. The phenyl–
phenyl [C(11)–C(16)] stacking is a point of difference with AlQ3,
and it seems to allow a much quicker hole hopping between pairs
of molecules. Whether this hopping can promote an improved
hole transport in these materials with respect to AlQ3 is debatable.
In these materials quick hole hopping is limited to a couple of
adjacent molecules, while in AlQ3 the hole is strongly confined on
a single molecule. On the contrary, electron transport could be
much more extended: in fact, in the case of the QꢀA chelant, only
pyridyl rings are involved in p-interactions. Thus, it is reasonable
to believe that the hole transport is much slower than the electron
movement. However, a better overlap of the QꢀB phenyl rings could
imply a hole delocalization larger than that of AlQ3, increasing the
hole stability.
With reference to the absorption spectra, the presence of an
intense electronic excitation around 304 nm, assigned to the p2 →
p3 transition involving the –NO2 chromophore, is the only relevant
difference between complexes 1–2 and complex 3. As already
reported in the discussion of the experimental spectra, this band
seems to be overlapped by the lowest energy band at 365 nm
(Fig. 4), which is attributed to q1 → q2 excitations. This evidence
could explain the lowest quantum yield observed in the case of
complex 3. The lowest population of the emissive Qꢀ-localised
excited states, combined with a small interaction between the Ln
and Qꢀ ligands (with consequent slow intersystem crossing towards
the emissive state) could lead to a reduced quantum yield. This is
supported by the different U-value for 3 measured when exciting
the complex on the band corresponding to the p2 → p3 (304 nm)
or to the q1 → q2 (365 nm) transition. This limitation should
not be encountered in electroluminescence experiments, when
the population of the excited states is achieved by hole–electron
recombination only mediated by the Qꢀ chelants.
Intermolecular interactions and charge-transport ability
Recently, the conductive properties of quinolin-8-olate aluminium
and zinc complexes have been related to the nature of the
p–p contacts between adjacent molecules in their crystalline
states.9c,18,4b Although OLEDs are based on amorphous films, the
use of solid state characterization can provide a helpful guide
to the knowledge of the molecular packing preferences of the
corresponding molecular materials.
Conclusion
In the reported crystalline b-phase of AlQ3, each molecule
forms p–p stacking with different Q chelants, giving rise to infinite
interacting pairs of adjacent molecules; these different conductive
interactions do not seem to be favourable to the hole transport.8d In
fact, while the Q-HOMO is mostly localized on the Q phenyl ring,
barely overlapped in all of the stacking patterns, the Q-LUMO is
mostly localized on the pyridyl rings, which are more involved in
p–p interactions due to their better overlaps. In the AlQ3 a-phase
the stacking interactions involve one chelant and are limited to
a pair of adjacent AlQ3 molecules. Moreover, in this case, the
distance between the Q phenyl rings of the stacking chelants is
longer than that observed for the pyridyl rings. Therefore, it has
been proposed that the involvement of the Q pyridyl rings in the
p–p interactions found in the crystalline phases of AlQ3 is the
principal cause of the electron transport character of the AlQ3
crystals and, by extension, of its amorphous phases.4b
In this paper, a detailed analysis of the p–p stacking interactions
existing in the crystalline state of the GaQꢀ2Ln compounds, 1–3, is
widely discussed in relation to the possible influence of the presence
of different Ln ligands in the molecule. Although the presence of
substituents on the Ln ligands generates a difference in the stacking
motifs of the QꢀA and QꢀB chelants, as already discussed in the
structural studies section, none of them are directly involved in
p–p interactions. Therefore, since conduction is associated with
strong stacking between aromatic rings, in the case of complexes
1–3, only the Qꢀ ligands involved in p-interactions could be mainly
responsible for the conduction.
The role of the substituted phenolate Ln ligands in pentacoordi-
nated 2-methyl-quinolin-8-olate gallium(III) complexes is studied
with respect to their absorption and emission properties, as well
as to the molecular packing of their crystalline solids. Ab initio
computations aimed at the interpretation of the spectroscopic data
and the prediction of the conduction properties in amorphous thin
films were also carried out.
The studied complexes show similar absorption and emission
spectra. The low energy excited states responsible for the emission
process are associated with changes in electronic density, mostly
localised on the 2-methyl-quinolin-8-olate chelants, with negligible
contributions from the Ln ligands. This is in agreement with
the acquired knowledge about analogous aluminium complexes.
According to our computations, the lowest fluorescence quantum
yield recorded in complex 3 is not directly related to electronic
effects induced by the –NO2 group on the Ln ligand, but is
associated with the low population of the emissive states in
fluorescence experiments, due to light absorption of the –NO2
chromophore at the excitation wavelength used.
The 2-methyl-quinolin-8-olate chelants are mainly responsible
for the electric conduction in the solid phase of these materials.
The crystallographic analysis has pointed out that conductive p–p
contacts between adjacent molecules are only accomplished by the
2-methyl-quinolin-8-olate chelants. However, since the different Ln
ligands are somehow involved in the different stacking motifs
of the Qꢀ chelants observed in the crystal packing of 1–3, they
exert indirect effects on the electron hopping kinetics responsible
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 5124–5134 | 5131
©

View Article Online
◦
1
for conduction in amorphous thin films. In particular, in 1 the
pyridyl–pyridyl stack involves both Qꢀ chelants, suggesting a better
electron conduction with respect to hole conduction, where the
holes are trapped on the L1 ligand. In 2 and 3, while one Qꢀ
chelant shows the same aromatic interactions as 1, the other shows
a pyridyl–pyridyl stack (LUMO–LUMO facing each other) in one
direction, favourable to the electron transport, and a phenolate–
phenolate stack (HOMO–LUMO facing each other) in the other
direction, favourable to the hole transport, but limited to a couple
of adjacent molecules. While the electron transport is much more
extended, it seems reasonable to conclude that in 2 and 3 the
electron conduction is a faster process when compared to the hole
conduction.
GaQꢀ2(OC6H5), 1. Yellow solid, yield: 50%. Mp 267 C. H
NMR (CDCl3, 300 MHz), ppm: 8.19 (d, J = 8.27 Hz, 2H, H4),
7.46 (t, J = 7.93 Hz, 2H, H6), 7.33 (d, J = 8.28 Hz, 2H, H3), 7.15
(m, 5H, H5, H7 and Hc), 6.80 (d, J = 2.07 Hz, 2H, Hb,bꢀ), 6.52 (d,
J = 2.42 Hz, 2H, Ha,aꢀ), 3.15 (s, 6H, –CH3). IR (KBr, cm−1): 3052,
2992, 2928, 1610, 1591, 1576, 1506, 1466, 1431, 1282, 1114, 883,
834, 753, 648. Anal. Calcd for C26H21O3N2Ga: C, 65.17; H, 4.42;
N, 5.85. Found: C 65.03; H, 4.42; N, 5.57.
GaQꢀ2(OC6H4-4-CN), 2. Yellow solid, yield: 55%. Mp 264 ◦C.
1H NMR (CDCl3, 300 MHz), ppm: 8.25 (d, J = 8.41 Hz, 2H, H4),
7.49 (t, J = 8.07 Hz, 2H, H6), 7.38 (d, J = 8.42 Hz, 2H, H3), 7.21
(d, J = 7.37 Hz, 2H, H5 or H7), 7.17 (d, J = 7.71 Hz, 2H, H7 or
H5), 7.10 (d, J = 8.77 Hz, 2H, Hb,bꢀ), 6.48 (d, J = 8.77 Hz, 2H,
H
a,aꢀ), 2.98 (s, 6H, –CH3). IR (KBr, cm−1): 3046, 2221 (m −CN),
Experimental
1598, 1576, 1504, 1465, 1432, 1303, 1269, 1115, 860, 845, 835, 755,
651, 530. Anal. Calcd for C27H20N3O3Ga: C, 64.32; H, 4.00; N,
8.33. Found: C, 64.15; H, 3.90; N, 8.31.
Materials
All the commercially available chemicals were used without further
purification. All solvents utilised for photophysical characterisa-
tion were Fluka spectroscopic grade solvents.
GaQꢀ2(OC6H4-4-NO2), 3. Yellow solid, yield 60%. Mp 273 ◦C.
1H NMR (CDCl3, 300 MHz), ppm: 8.26 (d, J = 8.53 Hz, 2H, H4),
7.75 (d, J = 8.87 Hz, 2H, Hb,bꢀ), 7.50 (t, J = 7.84 Hz, 2H, H6),
7.39 (d, J = 8.53 Hz, 2H, H3), 7.22 (d, J = 8.18 Hz, 2H, H5 or
H7), 7.18 (d, J = 7.51 Hz, 2H, H7 or H5), 6.45 (d, J = 9.2 Hz, 2H,
Equipment
H
a,aꢀ), 2.99 (s, 6H, –CH3). IR (KBr, cm−1): 3063, 1586, 1496, 1451,
IR spectra (KBr pellets) were recorded on a Perkin–Elmer
Spectrum One FT-IR spectrometer equipped for reflectance
measurements. 1H NMR spectra were recorded on a Bruker
WH-300 spectrometer in CDCl3 solutions, with TMS as internal
standard. Elemental analyses were performed with a Perkin–Elmer
2400 analyzer CHNS/O.
1430, 1342, 1303 (m −NO2), 1270, 1115, 869, 852, 834, 755, 674,
531. Anal. Calcd for C26H20N3O5Ga: C, 59.58; H, 3.85; N, 8.02.
Found: C, 59.34; H, 3.83; N, 8.32.
Crystallography
Thermal analysis was monitored with a Zeiss Axioscope polar-
ising microscope equipped with a Linkam CO 600 heating stage
and a Perkin–Elmer DSC-6 differential scanning calorimeter with
a heating and cooling rate of 10.0 ◦C min−1 following calibration
with indium.
Absorption spectra were recorded with a Perkin–Elmer Lambda
900 spectrophotometer. Corrected luminescence spectra were ob-
tained with a Perkin–Elmer LS-50B spectrofluorimeter, equipped
with a Hamamatsu R-928 photomultiplier tube. Photolumines-
cent quantum yields were measured with the method described by
Demas and Crosby19 using [Ru(bipy)3]Cl2 (bipy = 2,2ꢀ-bipyridine;
U = 0.028 in aerated water20) as standard. The experimental
uncertainty in the band maximum for absorption and lumines-
cence spectra is 2 nm; that for luminescence intensity is 15%. All
emissions are confirmed by excitation spectra.
Details of the crystal data collection are listed in Table 6. X-Ray
data for 1 and 2 were collected on a Bruker-Nonius X8 Apex CCD
area detector equipped with graphite monochromator and Mo Ka
radiation (k = 0.71073), and data reduction was performed using
the SAINT programs; absorption corrections based on multiscan
were obtained by SADABS.21 X-Ray data for 3 were collected on
a Siemens R3m/V automated four-circle diffractometer equipped
with graphite-monochromated Mo Ka radiation (k = 0.71073).
The data were corrected for Lorentz and polarization effects. An
empirical absorption correction was applied using a method based
upon azimuthal (W) scan data.
All structures were solved by the Patterson method
(SHELXS/L program in the SHELXTL-NT software package)22
and refined by full-matrix least squares based on F2. All non-
hydrogen atoms were refined anisotropically and hydrogen atoms
were included as idealized atoms riding on the respective carbon
atoms with C–H bond lengths appropriate to the carbon atom
hybridization.
Films were obtained by spin coating of 1–3 dichloromethane
solutions on a quartz support.
Syntheses
CCDC reference numbers 607556–607558. For crystallographic
data in CIF or other electronic format see DOI: 10.1039/b606895b
Sodium hydroxide pellets (1.95 mmol) was added to a solution of
the suitable phenol (1.95 mmol) in ethanol (5 mL). The resulting
mixture was slowly transferred to a solution of gallium(III)
nitrate (1.95 mmol) dissolved in water (20 mL). The mixture was
stirred further for several minutes and a solution of 2-methyl-8-
hydroxyquinoline (620 mg, 3.9 mmol) in ethanol (5 mL) was slowly
added. The reaction mixture was refluxed overnight. After cooling,
a yellow solid was filtered, washed with water, ethanol and diethyl
ether. The crude product was recrystallized from CHCl3–Et2O (1 :
3) solution.
Computational details
Molecular geometries were optimized using the Kohn–Sham
DFT23 with the 6-31G(d) basis set and the Becke three-parameters
hybrid exchange–correlation functional known as B3LYP.24 An-
alytical evaluation of the energy second derivative matrix w.r.t.
Cartesian coordinates (Hessian matrix) at the same level of
approximation confirmed the nature of minima of the energy
surface points associated to the optimized structures.
5132 | Dalton Trans., 2006, 5124–5134
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
Table 6 Crystallographic data for compounds 1–3
1
2
3
Empirical formula
FW
C26H21N2O3Ga
479.17
Monoclinic
P21/c
15.736(1)
10.9021(8)
13.209(1)
90
103.836(3)
90
C27H20N3O3Ga
504.18
Triclinic
P-1
9.199(2)
9.670(2)
15.109(3)
106.774(4)
91.538(4)
114.381(3)
1155.5(4)
2
C26H20N3O5Ga
524.17
Triclinic
P-1
9.077(2)
9.410(2)
15.134(3)
105.23(3)
92.54(3)
113.09(3)
1131.4(4)
2
Crystal system
Space group
˚
a/A
˚
b/A
˚
c/A
a/◦
b/◦
c /◦
3
˚
V/A
2200.2(3)
4
1.447
Z
D(calcd)/g cm−3
l/mm−1
1.449
1.225
1.539
1.261
1.281
Temp./K
298
984
2.29–30.51
55913
298
516
1.43–24.71
10323
298
536
2.44–25.05
4286
F(000)
h/◦
No. measured reflns
No. unique reflns
Reflns with I > 2r(I)
Refined parameters
Goodness-of-fit
R indices
6727 [R(int) = 0.0345]
3933 [R(int) = 0.0164]
4009 [R(int) = 0.0314]
4990
3628
3043
291
1.026
308
1.109
316
1.035
R1a
0.0361
0.0932
0.0269
0.0762
0.0450
0.1083
wR2 b
R indices (all data)
R1
wR2
0.0544
0.1052
0.0291
0.0798
0.0603
0.1124
a R1 = R(|Fo| − |Fc|)/R|Fo|. b wR2 = [Rw(Fo − Fc2)2/Rw(Fo2)2]1/2
.
2
TD-DFT25 allowed the computation of vertical excitation
energies, oscillator strengths and excited state compositions in
terms of monoelectronic excitations between occupied and virtual
orbitals. These computations were performed at the B3LYP/6-
31G(d) level of approximation.
All the calculations were performed using the Gaussian98
program.26 Fig. 5 has been produced by the program Molekel
4.3 for Windows platforms.27
3 J. Slinker, D. Bernards, P. L. Houston, H. D. Abrun˜a, S. Bernhard and
G. G. Malliaras, Chem. Commun., 2003, 2392.
4 (a) M. A. Baldo, D. F. O’Brien, Y. You, A. Shoustikov, S. Sibley, M. E.
Thompson and S. R. Forrest, Nature, 1998, 395, 151; (b) A. Curioni,
M. Boero and W. Andreoni, Chem. Phys. Lett., 1998, 294, 263; (c) B. C.
Lin, C. P. Cheng, Z.-Q. You and C.-P. Hsu, J. Am. Chem. Soc., 2005,
127, 66.
5 H. Sirringhaus, P. J. Brown, R. H. Friend, M. N. Nielsen, K. Bechgaard,
B. M. W. Langeveld-Voss, A. J. H. Spiering, R. A. J. Janssen, E. W.
Meijer, P. Herwig and D. M. de Leeuw, Nature, 1999, 401, 685.
6 J. J. M. Halls, C. A. Walsh, N. C. Greenham, E. A. Marseglia, R. H.
Friend, S. C. Moratti and A. B. Holmes, Nature, 1995, 376, 498.
7 (a) P. E. Burrows, Z. Shen, V. Bulovic, D. M. McCarty, S. R. Forrest,
J. A. Cronin and M. E. Thompson, Appl. Phys. Lett., 1996, 79, 7991;
(b) W. Brutting, S. Berleb and M. Muckl, Org. Electron., 2001, 2, 1;
(c) A. P. Kulkarni, C. J. Tonzola, A. Babel and S. A. Jenekhe, Chem.
Mater., 2004, 16, 4556.
8 (a) C. W. Tang, S. A. VanSlyke and C. H. Chen, J. Appl. Phys., 1989, 65,
361; (b) C. Hosokawa, H. Tokailin, H. Higashi and T. Kusumoto, Appl.
Phys. Lett., 1992, 60, 1220; (c) B. J. Chen, W. Y. Lai, Z. Q. Gao, C. S.
Lee, S. T. Lee and W. A. Gambling, Appl. Phys. Lett., 1999, 75, 4011;
(d) M. Brinkmann, G. Gadret, M. Muccini, C. Taliani, N. Maschiocchi
and A. Sironi, J. Am. Chem. Soc., 2000, 122, 5147; (e) L. S. Hung and
C. H. Chen, Mater. Sci. Eng., R, 2002, 39, 143; (f) B. J. Chen, X. W.
Sun and Y. K. Lin, Appl. Phys. Lett., 2003, 82, 3017.
Acknowledgements
This research was supported by the Italian Ministero
dellꢀIstruzione, dell’Universita` e della Ricerca (MIUR) through
FIRB and the Centro di Eccellenza CEMIF.CAL grants. F. L.
and M. A. thank PRIN 2002, CIPE CLUSTER 14 SP 5 and CIPE
CLUSTER 26 SP 8 for financial support. M. A. thanks Consorzio
INSTM for the contribution to his fellowship.
References
1 (a) J. R. Sheats, H. Antoniadis, M. Heuschen, W. Leonard, J. Miller,
R. Moon, D. Roitman and A. Stocking, Science, 1996, 273, 884; (b) U.
Mitschke and P. Bauerle, J. Mater. Chem., 2000, 10, 1471; (c) E. Holder,
B. M. W. Langeveld and U. S. Schubert, Adv. Mater., 2005, 17, 1109.
2 (a) P. E. Burrows, L. S. Sapochak, D. M. McCarty, S. R. Forrest and
M. E. Thompson, Appl. Phys. Lett., 1994, 64, 2718; (b) A. Elschner,
H. W. Heuer, F. Jonas, S. Kirchmeyer, R. Wehrmann and K. Wussow,
Adv. Mater., 2001, 13, 1811; (c) D. Ma, J. M. Lupton, R. Beavington,
P. L. Burn and I. D. W. Samuel, Adv. Funct. Mater., 2002, 12, 507; (d) J.
Liu, Q. G. Zhou, Y. X. Cheng, Y. H. Geng, L. X. Wang, D. G. Ma,
X. B. Jing and F. S. Wang, Adv. Mater., 2005, 17, 2974; (e) H. Kanno,
R. J. Holmes, Y. Sun, S. Kena-Cohen and S. R. Forrest, Adv. Mater.,
2006, 18, 339.
9 (a) H. Schmidbaur, J. Lattenbauer, L. Dallas, W. G. Muller and O.
Kumberger, Z. Naturforsh., B, 1991, 46, 901; (b) Y. Wang, W. Zhang, Y.
Li, L. Ye and G. Yang, Chem. Mater., 1999, 11, 530; (c) L. S. Sapochak,
A. Padmaperuma, N. Washton, F. Endrino, G. T. Schmett, J. Marshall,
D. Fogarty, P. E. Burrows and S. R. Forrest, J. Am. Chem. Soc., 2001,
123, 6300; (d) M. Brinkmann, B. Fite, S. Pratontep and C. Chaumont,
Chem. Mater., 2004, 16, 4627.
10 (a) P. Moore, S. A. VanSlyke and H. J. Gysling, US Patent, 5,484, 922,
1996; (b) L. S. Sapochak, P. E. Burrows, D. Garbuzov, D. M. Ho, S. R.
Forrest and M. E. Thompson, J. Phys. Chem., 1996, 100, 17766; (c) T.
Enochido, M. Tamano and S. Okutsu, EU Patent, EP 0 757 088 A2,
1997; (d) G. Wang, F. Liang, Z. Xie, G. Su, L. Wang, X. Jing and
F. Wang, Synth. Met., 2002, 131, 1; (e) Y. Qiu, Y. Shao, D. Zhang and
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 5124–5134 | 5133
©

View Article Online
X. Hong, Jpn. J. Appl. Phys., 2000, 39, 1151; (f) X. T. Tao, H. Suzuki,
T. Wada, S. Miyata and H. Sasabe, J. Am. Chem. Soc., 1999, 121, 9447.
11 (a) M. Ghedini, M. La Deda, I. Aiello and A. Grisolia, Synth. Met.,
2003, 138, 189; (b) M. Ghedini, M. La Deda, I. Aiello and A. Grisolia,
J. Chem. Soc., Dalton Trans., 2002, 3406; (c) M. Ghedini, M. La Deda,
I. Aiello and A. Grisolia, Inorg. Chim. Acta, 2004, 357, 33; (d) M. La
Deda, A. Grisolia, I. Aiello, A. Crispini, M. Ghedini, S. Belviso, M.
Amati and F. Lelj, Dalton Trans., 2004, 2424; (e) M. La Deda, I. Aiello,
A. Grisolia, M. Ghedini, M. Amati and F. Lelj, Dalton Trans., 2006,
330.
21 SMART, SAINT and SADABS, Bruker AXS Inc., Madison, WI, USA,
1997.
22 SHELXTL-NT Crystal Structure Analysis Package, Version 5.1, Bruker
AXS Inc., Madison, WI, USA, 1999.
23 R. G. Parr and W. Yang, Density Functional Theory of Atoms and
Molecules, Oxford University Press, Oxford, 1989.
24 A. Becke, J. Chem. Phys., 1993, 98, 5648.
25 (a) M. Casida, Time dependent density functional response theory
for molecules in Recent Advances in Density Functional Methods, ed.
D. P. Chong, World Scientific, Singapore, 1995, vol. 1, p. 155; (b) M. E.
Casida, C. Jamorski, K. C. Casida and D. R. Salahub, J. Chem. Phys.,
1998, 108, 4439.
12 V. Russel, M. Scudder and I. Dance, J. Chem. Soc., Dalton Trans., 2001,
789.
13 (a) C. A. Hunter and J. K. M. Sanders, J. Am. Chem. Soc., 1990, 112,
5525; (b) C. J. Janiak, J. Chem. Soc., Dalton Trans., 2000, 3885.
14 M. Nishio, CrystEngComm, 2004, 6, 130.
15 S. Tan, B. Zhao, Y. Zou, Z. Xiao, X. Wang, G. Yu, Y. Liu and D. Zhu,
J. Mater. Sci., 2004, 39, 1405.
16 (a) G. P. Kushto, Y. Iizumi, J. Kido and Z. H. Kafafi, J. Phys. Chem.
A, 2000, 104, 3670; (b) M. D. Halls and H. B. Schlegel, Chem. Mater.,
2001, 13, 2632; (c) A. Degli Esposti, M. Brinkmann and G. Ruani,
J. Chem. Phys., 2002, 116, 798; (d) M. Amati and F. Lelj, Chem. Phys.
Lett., 2002, 107, 2560.
17 M. Amati and F. Lelj, J. Phys. Chem. A, 2003, 358, 144.
18 L. S. Sapochack, F. E. Benincasa, R. S. Schofield, J. L. Baker, K. K. C.
Riccio, D. Fogarty, H. Kohlmann, K. F. Ferris and P. E. Burrows,
J. Am. Chem. Soc., 2002, 124, 6119.
26 M. J. Frisch, G. W. Trucks, H. B. Schlegel, 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 and J. A. Pople, GAUSSIAN 98 (Revision A.11), Gaussian,
Inc., Pittsburgh, PA, 2001.
19 J. N. Demas and G. A. Crosby, J. Phys. Chem., 1971, 75, 991.
20 K. Nakamaru, Bull. Chem. Soc. Jpn., 1982, 5, 2697.
27 P. Flu¨kiger, H. P. Lu¨thi, S. Portmann and J. Weber, MOLEKEL 4.0,
Swiss Center for Scientific Computing, Manno, Switzerland, 2000.
5134 | Dalton Trans., 2006, 5124–5134
This journal is
The Royal Society of Chemistry 2006
©