P. Pander et al.
Dyes and Pigments 184 (2021) 108857
measurements recorded in solution indicate a monoexponential decay
(see Figs. S3–S7 for decay in solution and Figs. S8–S12 for decay in thin
film). This is likely due to inhomogeneity of the films and is a common
behaviour. To facilitate comparison between the photoluminescence
lifetime in solution and in the solid state a weighted average lifetime is
used in the latter case [26]. There is significantly less difference between
a solid state complex’s properties at room temperature and 77 K than
there is in solution. This is because there is less vibrational/rotational
freedom for the molecule in solid films even at room temperature and
this does not change significantly at lower temperatures. Furthermore,
as expected in a polymer matrix, the host is not completely frozen at 77
K and this allows for much more freedom to the molecules than a
2-MeTHF glass at same temperature. This is manifested as only a small
improvement of the vibrational resolution in photoluminescence spectra
vs. room temperature. Importantly, the emission spectra of 3-doped
films remain nearly the same at both room temperature and 77 K.
fast intersystem crossing the excited Sn states at the S1 state geometry are
not discussed as they are of negligible importance for this study.
In all cases the lowest singlet excitation is related predominantly to
the HOMO→LUMO transition with the same dominant contribution of
LUMO→HOMO in the T1→S0 transition (Table S2). However, HOMO-
1→LUMO and HOMO-2→LUMO transitions also contribute to the singlet
and triplet excitations. Interestingly, the contribution of the HOMO-
→LUMO transition to the lowest singlet excitation in 5 is the largest
among all at 95% while the value for complexes 1–3 remains at ≈85%.
Excitation energies and the energies of triplet excited states clearly
follow the trends observed in experimental data. S0→S1 energy at S0
geometry shows good agreement with the experimental absorption
spectra, however the S0→T1 energy at T1 geometry shows a systematic
0.3–0.4 eV offset from the experiment. These calculations clearly
explain the emissive properties of the complexes. In all compounds the
HOMO and the LUMO both span over the π-conjugated system of the
ligands with both having pronounced metal orbital admixtures. Apart
from contributions of the central atom to both frontier orbitals, the
HOMO spans over the dpm ligand and the phenyl arm of the extended
phenylpyridine ligand, while the LUMO spans almost exclusively on the
extended phenylpyridine ligand. This behaviour reflects the relative
electron-deficient and electron-rich regions of the molecule. The HOMO
and LUMO surfaces and spin density distribution in T1 state for com-
plexes 1, 2, and 4 presented in this work show perfect resemblance to
their analogues studied by Bossi et al. [16].
3.3. Electrochemistry
The electrochemical behaviour of all complexes shows good corre-
lation with their photophysical properties in the ground state as corre-
sponding electrochemical and optical energy gaps follow the same trend
(Fig. 2). Electrochemical processes reveal how a ground state molecule
reacts in a chemical reaction involving the receipt or donation of elec-
trons. The electrical potential at which these processes occur can be
transposed to an energy scale representing the HOMO and LUMO levels
of the molecule (as their approximation), or more precisely their ion-
isation potential (IP) and electron affinity (EA), respectively [27,28].
The ground state behaviour seen in electrochemistry will reflect the
degree of conjugation present in the complex. In the case of molecules 1,
2 and 3 their respective oxidation and reduction potentials are nearly
identical. This clearly suggests that their conjugation is limited to the
phenylpyridine moiety and it does not extend to the peripheral phenyl
ring in 3. The outer phenyl group is thus orthogonal to the rest of the
phenylpyridine ligand. This view is also supported by computational
analysis (Fig. 3, S16).
Analysis of the spin density of the T1 state calculated at the lowest
triplet excited state geometry is crucial for understanding the emissive
properties of 4 vs. 1 and 2, but also gives an important insight into the
properties of 5 (Fig. 4). Generally speaking, the T1 spin density spans
over the extended phenylpyridine ligand unit with a sound contribution
of the metal centre. However, the distribution of the spin density in 4 is
nearly identical to that of 1 and 2. This again indicates that in fact the
outer phenyl ring in 4 does not take part in emission, limiting the
conjugation to the phenylpyridine itself. Interestingly, substitution of
the cyclopentene ring by the benzene ring in 5 not only extends spin
density over that unit but also turns on the contribution of the other
outer benzene ring that is not contributing to the spin density in 4. The
turn-on of the second benzene unit fully explains the strong redshift of
Complex 4 shows very different electrochemical properties from 1
and 2 despite all three having nearly identical emission spectra. This is
because in the ground state of 4 the outer benzene ring is conjugated, at
least partly, with the phenylpyridine moiety. This results in a narrower
electrochemical energy gap by reducing the oxidation potential and
increasing the reduction potential. In the excited state however, the
peripheral phenyl ring does not play a significant role (Fig. 4). Finally, a
further extension of the ground state conjugation is observed in 5,
showing the smallest electrochemical energy gap. 5 however, clearly
shows a red-shifted emission from other emitters, especially 4.
photoluminescence emission in 5. Extension of the π-conjugated system
achieved in 5 effectively destabilises the quinoid conjugation of the T1
state of molecules 1–4 towards benzenoid conjugation in complex 5.
3.5. Devices
The T1 energy of all complexes but 5 is relatively large, close to 2.7
eV, with the emission onset entering the blue spectral region [29]. For
this reason a host comprised of poly (N-vinylcarbazole) (PVK) and 1,
3-bis [2-(4-tert-butylphenyl)-1,3,4-oxadiazo-5-yl]benzene (OXD-7)
(50:50 w/w) is used as this mixture has previously shown good perfor-
mance in blue PhOLED devices based on FIrPic-like metalorganic
emitters [30]. The use of OXD-7 in relation to the popular 2-(bipheny-
l-4-yl)-5-(4-tert-butylphenyl)-1,3,4-oxadiazole (PBD) is necessary due to
the low triplet energy of the latter which quenches high energy triplet
states of emitters. The use of the PVK:OXD-7 blend for all molecules is
justified by the need for consistency, as by using the same host it is easier
to make comparisons between emitters. 1,3,5-tri (m-pyridin-3-ylphenyl)
benzene (TmPyPb) is used as an electron transport layer. A simple de-
vice structure has been employed, without the use of hole trans-
port/electron blocking layers: ITO | Heraeus Clevios HIL 1.3 N (45 nm) |
PVK:OXD-7 (50:50 w/w) co. 5% (wt.) emitter (70 nm) | TmPyPb (50
nm) | LiF (0.8 nm) | Al (100 nm). In general, using a high molecular
weight PVK, M = 1.1 × 106 g molꢀ 1, (PVKH) layer (10 nm thick) in
between hole injection layer (HIL) and emissive layer (EML) gives a
good charge balance [31–33], however, this approach can only be used
with EML solution-processed from non-chlorinated solvents, such as
toluene. Due to the limitations in solubility of OXD-7 as well as the
3.4. Calculations
In order to understand the different emissive behaviour of the
investigated organometallic compounds, basic quantum mechanical
calculations have been performed. In the ground state all complexes but
5 are planar around the Pt (II) centre and within the π-conjugated system
of ligands, while only the aliphatic groups remain off-plane (Fig. 3 and
S13-S17). 5 however shows its extended phenylpyridine ligand to be
twisted (see ESI, Fig. S18). This helicene-like twist is caused by repulsion
of hydrogen atoms in adjacent benzene rings of the ligand π-conjugated
structure. All geometries calculated at the lowest triplet excited state,
except for 3 and 5, are in general almost identical to the ground state
structures. In case of 3 in the T1 state the C–C bond axis connecting the
single phenyl substituent is distorted out of plane along with the rest of
the phenylpyridine ligand (Fig. S16). This behaviour explains the
different vibronic pattern of the emission spectrum in 3 compared to the
analogues 1 and 2.
To characterise T1→S0 transitions the T1 was calculated at the T1
geometry to reflect changes to the structure in the excited state. Due to
7