Structure-property relationships based on Hammett constants in cyclometalated iridium(iii) complexes: Their application to the design of a fluorine-free FIrPic-like emitter

Article abstract of DOI:10.1039/c3dt52739e

While phosphorescent cyclometalated iridium(iii) complexes have been widely studied, only correlations between oxidation potential E_OX and Hammett constant σ, and between the redox gap (ΔE_REDOX = E_OX - E_RED) and emission or absorption wavelength (λ_abs, λ_em) have been reported. We present now a quantitative model based on Hammett parameters that rationalizes the effect of the substituents on the properties of cyclometalated iridium(iii) complexes. This simple model allows predicting the apparent redox potentials as well as the electrochemical gap of homoleptic complexes based on phenylpyridine ligands with good accuracy. In particular, the model accounts for the unequal effect of the substituents on both the HOMO and the LUMO energy levels. Consequently, the model is used to anticipate the emission maxima of the corresponding complexes with improved reliability. We demonstrate in a series of phenylpyridine emitters that electron-donating groups can effectively replace electron-withdrawing substituents on the orthometallated phenyl to induce a blue shift of the emission. This result is in contrast with the common approach that uses fluorine to blue shift the emission maximum. Finally, as a proof of concept, we used electron-donating substituents to design a new fluorine-free complex, referred to as EB343, matching the various properties, namely oxidation and reduction potentials, electrochemical gap and emission profile, of the standard sky-blue emitter FIrPic.

Full text of DOI:10.1039/c3dt52739e

Volume 43 Number 15 21 April 2014 Pages 5605–5936
Dalton
Transactions
An international journal of inorganic chemistry
www.rsc.org/dalton
ISSN 1477-9226
COVER ARTICLE
Baranoff et al.
Structure–property relationships based on Hammett constants in
cyclometalated iridium(III) complexes: their application to the design of a
fluorine-free FIrPic-like emitter

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Structure–property relationships based on
Hammett constants in cyclometalated iridium(III)
complexes: their application to the design of a
fluorine-free FIrPic-like emitter†
Cite this: Dalton Trans., 2014, 43,
5667
Julien Frey,a Basile F. E. Curchod,b Rosario Scopelliti,a Ivano Tavernelli,b
Ursula Rothlisberger,b Mohammad K. Nazeeruddina and Etienne Baranoff*c
While phosphorescent cyclometalated iridium(III) complexes have been widely studied, only correlations
between oxidation potential EOX and Hammett constant σ, and between the redox gap (ΔEREDOX = EOX
−
ERED) and emission or absorption wavelength (λabs, λem) have been reported. We present now a quantitat-
ive model based on Hammett parameters that rationalizes the effect of the substituents on the properties
of cyclometalated iridium(III) complexes. This simple model allows predicting the apparent redox poten-
tials as well as the electrochemical gap of homoleptic complexes based on phenylpyridine ligands with
good accuracy. In particular, the model accounts for the unequal effect of the substituents on both the
HOMO and the LUMO energy levels. Consequently, the model is used to anticipate the emission maxima
of the corresponding complexes with improved reliability. We demonstrate in a series of phenylpyridine
emitters that electron-donating groups can effectively replace electron-withdrawing substituents on the
orthometallated phenyl to induce a blue shift of the emission. This result is in contrast with the common
approach that uses fluorine to blue shift the emission maximum. Finally, as a proof of concept, we used
electron-donating substituents to design a new fluorine-free complex, referred to as EB343, matching
the various properties, namely oxidation and reduction potentials, electrochemical gap and emission
profile, of the standard sky-blue emitter FIrPic.
Received 1st October 2013,
Accepted 29th November 2013
DOI: 10.1039/c3dt52739e
www.rsc.org/dalton
quantum yields and a relatively short triplet excited state
lifetime.7–13 As a result of these characteristics, they have
Introduction
Comprehensive structure–properties relationships for tran- attracted considerable interest for use as dopants in organic
sition metal complexes are essential owing to their great light-emitting diodes (OLEDs)14–22 and virtually any appli-
importance in catalysis, optoelectronics and sensing cation requiring highly phosphorescent materials.23–31
applications.1–6 Among the many families of complexes of
Interestingly the emission wavelength of iridium(III) com-
interest, cyclometalated iridium(III) complexes possess unique plexes can be tuned over the entire visible spectrum. This
photophysical properties, namely high phosphorescence color tuning is possible because, first, the emission is mainly a
lowest unoccupied molecular orbital (LUMO) to highest occu-
pied molecular orbital (HOMO) transition.32 It is therefore
possible to adjust the color of the emission by changing the
HOMO–LUMO gap, which can be achieved simply by changing
the ligand skeleton.33–38 Second, the HOMO and LUMO are
spatially differentiated.32,39,40 For example, on complexes fea-
turing 2-phenylpyridine (ppy) as a ligand, the HOMO is pre-
dominantly localized across the phenyl ring and the iridium
center, whereas the LUMO resides preferentially on the pyri-
dine.32 In this case, the HOMO–LUMO gap can be tuned by
grafting substituents onto the ligand skeleton and, depending
on the position of the substituents, they will impact the LUMO
or the HOMO energy levels. The substituents can exert both
inductive and mesomeric effects that result in an electron-
aLaboratory of Photonics and Interfaces, Institute of Chemical Sciences and
Engineering, École Polytechnique Fédérale de Lausanne, CH-1015 Lausanne,
Switzerland
bLaboratory of Computational Chemistry and Biochemistry, Institute of Chemical
Sciences and Engineering, École Polytechnique Fédérale de Lausanne,
CH-1015 Lausanne, Switzerland
cSchool of Chemistry, University of Birmingham, Edgbaston, Birmingham B15 2TT,
UK. E-mail: e.baranoff@bham.ac.uk
†Electronic supplementary information (ESI) available: Synthetic procedures,
X-ray crystallographic data, additional figures, theoretical calculation, voltammo-
grams, 1H NMR and 19F NMR spectra, and HRMS of the complexes. CCDC
complex 1: 964310, complex 2: 964311, complex 3: 964312, complex 4: 964313,
complex 6: 964314, complex 7: 964315. For ESI and crystallographic data in CIF
or other electronic format see DOI: 10.1039/c3dt52739e
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 5667–5679 | 5667

View Article Online
Paper
Dalton Transactions
donating or withdrawing character. The electronic character of reasonable, one can legitimately assume that the HOMO
the substituents is quantified in terms of position according energy level and σ also correlate. For Ir(4-CF3-ppy)3, the HOMO
to Hammett σ-values (σmeta or σpara).41 More positive (negative) level is shifted positively compared to Ir(ppy)3 in agreement
values of σ are associated with the more electron-withdrawing with the electron-withdrawing nature of the –CF3 group. Conse-
(donating) substituent.42 Hence acceptor (donor) groups are quently, the absence of an emission wavelength shift between
expected to stabilize (destabilize) the orbital that is involved by those two complexes can be attributed to significant stabiliz-
pulling out (pushing in) electron density from (to) this ation of the LUMO energy level in Ir(4-CF3-ppy)3 compared to
orbital, respectively. Hammett parameters have been para- Ir(ppy)3. This implies that despite being located on the cyclo-
mount for the development of quantitative structure–property metalated phenyl ring where the HOMO is mainly localized,
relationships (QSPR).43 In general, it is possible to build the –CF3 group also impacts the LUMO energy level of the
strong QSPR between the nature of these substituents based complex. A thorough comparison of the –CF3 and –F substitu-
on their Hammett σ-values and the corresponding electro- ents reported by Avilov et al. also stresses the importance of
chemical properties of series of compounds, and one can the position of the substituents.53 Based on a detailed theore-
observe a linear dependence of oxidation (EOX) or reduc- tical analysis, the authors illustrate how the emission pro-
tion (ERED
)
potentials with respect to the Hammett perties of bis-cyclometalated iridium complexes are influenced
by the substitution pattern and conclude that both HOMO and
parameters.1–3,5,44–48
For cyclometalated iridium(III) complexes, although there LUMO energy levels are unequally impacted by those changes.
are reports on the correlation between oxidation potential EOX This result about the position of the substituents is reminis-
and σ,7,50 as well as the correlation between the redox gap cent of the effect of the methoxy group in the report of Watts,
(ΔEREDOX = EOX − ERED) and absorption or emission wave- although this qualitative information is of limited assistance
length (λabs, λem),50,51 there is apparently no general correlation for practical design. Clearly, QSPR that would relate the elec-
between λem and Hammett σ-values. The early work of Watts tronic character of the substituents to the emissive property of
and coworkers nicely illustrates these two paradigms.49 The the complex is still missing for cyclometalated iridium
authors reported the electrochemical and photophysical data complexes.
of a series of homologous homoleptic tris-cyclometalated
We recently reported our initial progress towards such
iridium complexes [Ir(R-ppy)3], ppy = 2-phenylpyridine and QSPR in an homologous series of halogen substituted bis-
substituents on the phenyl R = H, 4-methyl, 4-propyl, 4-tert- cyclometalated complexes [Ir(2-(2,4-diX-phenyl)pyridine) (pic)],
butyl, 4-fluoro, 4-trifluoromethyl, 4-methoxy, 5-methoxy. The with X = H, F (FIrPic), Cl, Br and pic = picolinate. We were able
correlations of the Hammett parameters σ with the oxidation to link the Hammett parameters of the substituents to the
potentials EOX and with the emission maxima λem are shown wavelength of emission maxima.54 Despite the higher values
in Fig. S1† (experimental data are given in Table 1). Whereas of σm for chlorine and bromine compared to fluorine, the
the first correlation appears very convincing, the latter corre- emissions of the chloro and bromo substituted complexes are
lation with λem presents large discrepancies for the complexes red-shifted with respect to FIrPic. Nevertheless, we obtained a
[Ir(4-CF3-ppy)3] and [Ir(5-OMe-ppy)3]. The trifluoromethyl sub- very good correlation of σm versus EOX along the series, as
stituent, a stronger acceptor than fluorine, results in green expected. To obtain a convincing correlation with λem, it was
emission similar to the parent complex [Ir(ppy)3], while the necessary to introduce the Hammett parameter that accounts
methoxy substituent, usually considered as an electron-donat- for the substituents effect in ortho/para position, σp (≈ σo). The
ing group, results in red or blue shifted emission compared to new relation (σm − σp) versus λem provided a quasi-quantitative
[Ir(ppy)3], depending on its position.
rationalization of the unexpected shift of emission along the
The oxidation potentials of cyclometalated iridium com- series: the chloride and bromide substituents (σp = 0.22) are
plexes can be related to the corresponding HOMO energy much stronger electron-withdrawing elements towards the
levels.50,52 Since correlations between EOX and σ-values seem para position than the fluoride substituent (σp = 0.06) and
therefore –Cl and –Br stabilize the LUMO, which is mainly
localized on the pyridine ring that is in the ortho/para position
of the substituents, much more than –F. This energetic stabil-
ization of the LUMO has been reproduced with theoretical
calculations.
Table 1 Experimental data from ref. 49 (λem at 77 K), values of the
Hammett parameters41 and the condition on the Hammett parameters
according to eqn (12) and (13) (see below)
We report now a quantitative model based on Hammett
parameters that rationalizes the effect of the substituents on
the properties of cyclometalated iridium(III) complexes. This
model allows for predicting the apparent redox potentials as
well as the electrochemical gap of complexes based on phenyl-
pyridine ligands with good accuracy. In particular, the model
accounts for the unequal effect of the substituents on both
the HOMO and the LUMO energy levels. Consequently,
the model can also be used to anticipate the emission maxima
λem
λem
Complex
(nm) (cm−1
)
EOX
σm
σp
Condition
Ir(ppy)3
494
493
496
497
468
494
20 243 0.77
0
0
0.00
0.05
0.03
0.04
0.30
0.05
0.31
Ir(4-Me-ppy)3
Ir(4-Pr-ppy)3
Ir(4-tBu-ppy)3
Ir(4-F-ppy)3
Ir(4-CF3-ppy)3
20 284 0.70 −0.07 −0.17
20 161 0.67 −0.07 −0.15
20 121 0.66 −0.10 −0.20
21 368 0.97
20 243 1.08
20 790 0.75
18 553 0.55
0.34
0.43
0.06
0.54
Ir(4-OMe-ppy)3 481
Ir(5-OMe-ppy)3 539
0.12 −0.27
0.12 −0.27 −0.35
5668 | Dalton Trans., 2014, 43, 5667–5679
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
of the corresponding complexes with improved reliability. General method for the synthesis of 2,3′-bipyridine chloro-
Importantly, this model provides an improved theoretical bridged dimer complexes
frame that can help in further developing iridium emitters. In
[Ir(COD)(μ-Cl)]2 (500 mg, 0.74 mmol) was suspended in 10 mL
particular, we demonstrate in a series of phenylpyridine emit-
of 2-ethoxyethanol. The suspension was filled with nitrogen by
ters that electron-donating groups can effectively replace elec-
3 cycles vacuum/nitrogen. The ligand L (2.98 mmol, 4.0 equiv.)
was added, and 1 mL of 2-ethoxyethanol was used for rinsing.
tron-withdrawing substituents on the orthometallated phenyl
to induce a blue shift of the emission. This result is in contrast
The mixture was filled again with nitrogen by 3 cycles vacuum/
with the common approach that uses fluorine to blue shift the
nitrogen. The flask was sealed and heated at 130 °C. After 3 h,
the mixture was cooled down. The volume of solvent was
emission maximum.
Finally, as a proof of concept, we used electron-donating
reduced to half under vacuum and 20 mL of methanol was
substituents to design a new fluorine-free complex, referred to
added. The resulting precipitate was filtered, washed with
as EB343, matching the various properties, namely oxidation
methanol, and dried to afford the chloro-bridged iridium
dimer [Ir(L)2(μ-Cl)]2.
[Ir(L5)2(μ-Cl)]2. L was 2′,6′-difluoro-2,3′-bipyridine (L5). The
and reduction potentials, electrochemical gap and emission
profile, of the standard sky-blue emitter FIrPic.
chloro-bridged iridium dimer [Ir(L5)2(μ-Cl)]2 was obtained as a
pale yellow solid (824 mg, 95%). 1H NMR (CDCl3, 400 MHz):
δ 9.05 (broad d, 4H, J = 5.8 Hz); 8.36 (broad d, 4H, J = 8.2 Hz);
7.98 (broad dd, 4H, J = 7.9 Hz); 6.96 (broad dd, 4H, J = 6.4 Hz);
5.22 (s, 4H).
Experimental
Materials and general considerations
All materials and solvents were of reagent quality and used as
[Ir(L6)2(μ-Cl)]2. L was 2′,6′-dimethoxy-2,3′-bipyridine (L6).
received. [Ir(COD)Cl]2 was purchased from Strem, IrCl3·xH2O The chloro-bridged iridium dimer [Ir(L6)2(μ-Cl)]2 was obtained
from Heraeus, 2-phenylpyridine from Aldrich, and acetyl- as a yellow solid (877 mg, 89%). 1H NMR (CDCl3, 400 MHz):
acetone and picolinic acid from Fluka. 1, 2, and FIrPic were syn- δ 9.04 (broad d, 4H, J = 5.4 Hz); 8.47 (broad d, 4H, J = 8.2 Hz);
thesized as previously reported.54,55 1H and 13C NMR spectra 7.66 (broad ddd, 4H, J = 8.5, 1.2 Hz); 6.62 (broad dd, 4H, J =
were recorded using a Bruker AV 400 MHz spectrometer. 6.1 Hz); 4.85 (s, 4H); 3.96 (s, 12H); 3.59 (s, 12H).
Chemical shifts δ (in ppm) are referenced to residual solvent
[Ir(L7)2(μ-Cl)]2. L was 6′-methoxy-2′-methyl-2,3′-bipyridine
peaks. 19F NMR spectra were recorded using a Bruker AV (L7). The chloro-bridged iridium dimer [Ir(L7)2(μ-Cl)]2 was
200 MHz spectrometer. Coupling constants are expressed in obtained as a yellow solid (795 mg, 87%). 1H NMR (CDCl3,
hertz (Hz). High-resolution mass spectra (HRMS) and elemen- 400 MHz): δ 9.17 (broad d, 4H, J = 5.3 Hz); 8.01 (broad d, 4H,
tal analyses have been performed by the Service d’Analyse of J = 8.3 Hz); 7.76 (broad ddd, 4H, J = 8.3 Hz); 6.73 (broad ddd,
EPFL. UV-visible spectra were recorded in a 1 cm path length 4H, J = 5.3 Hz); 5.00 (s, 4H); 3.61 (s, 12H); 2.74 (s, 12H).
quartz cell on a Cary 100 spectrophotometer. Emission spectra
were recorded on a Fluorolog 3-22 using a 90° optical geome-
try. Voltammetric measurements employed a PC controlled
AutoLab PSTAT10 electrochemical workstation and were
Synthesis of [Ir(L7)2(Pic)] EB343
2-Picolinic acid (98%, 98 mg, 2.00 × 10−4 mol, 4.0 eq.) and
TBAOH (170 mg, 4.2 eq.) were dissolved in CH2Cl2–MeOH,
carried out in an argon-filled glove box, oxygen and water
7/3 volume to volume, and this solution was added to a sus-
<5 ppm. All experiments were realized using 0.1 M tetrabutyl-
pension of [Ir(L7)2(μ-Cl)]2 (250 mg, 2.00 × 10−4 mol) CH2Cl2
ammonium hexafluorophosphate in anhydrous N,N-dimethyl-
(10 mL). The solution was gently refluxed at 40 °C under
formamide as an electrolyte using a set of glassy carbon and
nitrogen overnight. After cooling down to room temperature,
two Pt wires as the working, counter, and reference electrodes,
the solution was evaporated to dryness. The crude material
respectively. Ferrocene was used as an internal standard. A
was dissolved in pure CH2Cl2 and filtered over a pad of neutral
scan rate of 1000 mV s−1 has been applied. Before each
silica. The volume of the solution was reduced under vacuum
measurement, samples were stirred for 15 s and left to equili-
and the main fraction was slowly precipitated with hexane.
brate for 5 s.
The suspension was filtered off, washed with hexane, and
dried under vacuum to afford EB343 as a yellow solid (199 mg,
X-ray crystal structure determination
1
70%). H NMR (CDCl3, 400 MHz): δ 8.78 (broad ddd, 1H, J =
The data collection for the crystal structures was measured at 5.7, 1.5 Hz); 8.30 (broad d, 1H, J = 7.7 Hz); 8.01 (broad d, 1H,
low temperature [100(2) K] using Mo Kα radiation on a Bruker J = 8.5 Hz); 7.97 (broad d, 1H, J = 8.5 Hz); 7.89 (ddd, 1H, J =
APEX II CCD equipped with a kappa geometry goniometer. 7.7, 1.5 Hz); 8.72 (m, 3H); 7.45 (broad ddd, 1H, J = 5.7, 1.0 Hz);
The data were reduced using EvalCCD56 and then corrected for 7.35 (ddd, 1H, J = 5.4, 2.2, 1.5 Hz); 7.09 (ddd, 1H, J = 5.7,
absorption.57 The solutions and refinements were performed 1.3 Hz); 6.86 (ddd, 1H, J = 6.0, 1.3 Hz); 5.59 (s, 1H); 5.31
using SHELX.58 The structures were refined using full-matrix (s, 1H); 2.79 (s, 3H); 3.72 (s, 3H); 2.80 (s, 3H); 2.74 (s, 3H).
least-squares based on F2 with all non-hydrogen atoms aniso- ESI-TOF HRMS: MH+ m/z: calc. 714.1694 found 714.1661. Anal.
tropically defined. Hydrogen atoms were placed in calculated calcd for C30H26IrN5O4: C, 50.55; H, 3.68; N, 9.83. Found C,
positions by means of the “riding” model.
50.45; H, 3.64; N, 9.86.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 5667–5679 | 5669

View Article Online
Paper
Dalton Transactions
Results and discussion
Structure–property relationships based on Hammett parameters
The electrochemical gap ΔEREDOX is usually well correlated
with the emission maximum (optical gap) λem. However they
are not strictly equivalent as the optical gap is smaller than the
electrochemical gap.59 The relation between ΔEPHOTO and
ΔEREDOX is of the form:
Scheme 1 Definition of the substitution pattern of the cyclometalated
phenylpyridine ligand, as used in this work.
X
ΔEPHOTO ¼ ΔEREDOX
þ
αi
ð1Þ
cyclometalated iridium complexes, we can reasonably assume
that such a relation exists using a similar approach to the oxi-
dation potentials when the LUMO of the complex is mainly
localized on the neutral ring of the orthometallated ligand.
The nitrogen atom N (Scheme 1) is now considered as a
reduced representation of the LUMO of the complex. The
reduction potential of the corresponding complex is then
expressed as a linear function of the Hammett parameters as
follows:
i
where the additional term collects factors not directly deriva-
ble from ΔEPHOTO and ΔEREDOX, such as Coulomb terms (elec-
tron repulsion, singlet/triplet states), solvation energy
differences, and energy differences of relevant orbitals, for the
four species involved: the complex in the ground state, the
complex in the excited state, the oxidized complex, and the
reduced complex. Nevertheless, within a same family of mole-
cules, it is usually assumed that the difference between electro-
chemical and optical gaps is fairly constant, and eqn (2) can
be used:
ERED ¼ CσN þ D
ð6Þ
where σN is the sum of Hammett parameters with respect to N
for all substituents on the ligand. Developing the expression
according to the nomenclature introduced in Scheme 1, we
obtain the relation:
ΔEPHOTO / ΔEREDOX ¼ EOX ꢀ ERED
ð2Þ
When correlations between EOX and σ are reported in the
literature, the value of σm or σp is chosen based on the position
of the substituent with respect to the carbon C coordinated to
the iridium. This carbon atom is therefore a minimal represen-
tation of the HOMO of the complex. The oxidation potential of
the corresponding complex is then expressed as a linear func-
tion of the Hammett parameters as follows:
ERED ¼ C½ðσm þ σp þ σm þ σp Þ þ βðσp þ σm þ σp
R3′
R4′
R5′
R6′
R2
R3
R4
þ σm Þꢁ þ D
ð7Þ
R5
In order to develop a reliable predictive model, we focused
on fac-tris-homoleptic cyclometalated iridium(III) complexes to
avoid the interplay of ancillary ligands. It should be noted that
no correlation is reported in the literature with substituents on
the pyridine ring, as well as no correlation with reduction
potentials. We used data from the literature11,49,60 to obtain
the values for A, B, C, D, α, and β from a multiple linear
regression (Fig. 1). We did not take into account any solvent
effect or differences in measurement conditions and we re-
calculated all potentials vs. ferrocene.61,62 We also used the
potentials of [Ir(ppy)3] as reported in ref. 11 as reference data
EOX ¼ AσC þ B
ð3Þ
where σC is the sum of Hammett parameters with respect to C
for all substituents on the ligand. As discussed in the Intro-
duction, the substituents on the cyclometalated phenyl can
impact the energy of the LUMO localized on the pyridine ring.
Similarly, substituents on the pyridine are expected to impact
the energy of the HOMO. However the impact will be attenu-
ated due to the longer distance between the substituent on the
pyridine ring and the carbon C, so that σC is rewritten as:
X
X
σC
¼
σPhen þ α
σPyr
ð4Þ
C
C
where σPhen are the Hammett parameter values with respect
C
to C of the substituents on the phenyl, σPyr are the Hammett
C
parameter values with respect to C of the substituents on the
pyridine, and α is the attenuation factor. Considering that σp =
σo and replacing according to the nomenclature introduced in
Scheme 1, we can combine the relation with eqn (3) to obtain
the general expression of EOX as a function of the Hammett
parameters of the substituents:
EOX ¼ A½ðσm þ σp þ σm þ σp Þ þ αðσp þ σm þ σp
R2
R3
R4
R5
R3′
R4′
R5′
þ σm Þꢁ þ B
ð5Þ
R6′
Although no relation between reduction potentials ERED
and Hammett parameters has been reported before for lines are of the form a + bx.
●
○
); the
Fig. 1 Correlation between σC and EOX
(
) and σN and ERED
(
5670 | Dalton Trans., 2014, 43, 5667–5679
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
for other data sets. We obtained eqn (8) for EOX in V vs. ferro-
cene, and eqn (9) for ERED in V vs. ferrocene:63
Similarly, a substituent in position 3 needs to fulfill the
condition in eqn (13) to result in a larger redox gap, and hence
an expected blue shifted emission compared to [Ir(ppy)3]:
ꢀ
ꢁ
X
X
X
EOX ꢂ 0:7
σPhen þ 0:8
σPyr þ 0:3
ð8Þ
C
C
σp ꢀ 0:7σm > 0
ð13Þ
ꢀ
ꢁ
X
In the same way as for the substituents on the phenyl,
simple relations can be derived for substituents on the pyri-
dine. For a substituent on position 5′:
ERED
ꢂ
σPyr þ 0:5
σPhen ꢀ 2:7
ð9Þ
N
N
Using eqn (2), we obtain the general relation for the redox
gap ΔEREDOX and the Hammett parameters σ in eqn (10):
σm ꢀ 0:5σp < 0
and for a substituent on position 4′:
σp ꢀ 0:5σm < 0
ð14Þ
ΔEREDOX ꢂ f0:7½ðσm þ σp þ σm þ σp
Þ
R2
R3
R4
R5
þ 0:8ðσp þ σm þ σp þ σm Þꢁ þ 0:3g
R3′
R4′
R5′
R6′
ð15Þ
ꢀ f½ðσm þ σp þ σm þ σp
Þ
ð10Þ
R3′
R4′
R5′
R6′
þ 0:5ðσp þ σm þ σp þ σm Þꢁ ꢀ 2:7g
Compared to the case of substituents on the phenyl, the
first term here represents the impact of the substituent on the
LUMO energy level, while the second term is the impact on the
HOMO energy level. It is interesting to note that for substitu-
ents on the pyridine, the effect on the remote orbital, the
HOMO, is smaller than the effect on the LUMO of a substitu-
ent on the phenyl (coefficient 0.5 compared to 0.7).
These attenuation values can be compared to the extent of
delocalization of the HOMO and LUMO over the entire C^N
ligand obtained from DFT(M06)/IEF-PCM(ACN) calculations.
The extent of HOMO on the pyridine is much smaller than the
one of LUMO on the orthometallated phenyl (Fig. 2). While
33% of the HOMO is localized on the orthometallated phenyl
of [Ir(ppy)3], only 7% is found on the pyridine (the remaining
contribution is found on the iridium). This proportion is
reversed for the LUMO of [Ir(ppy)3], where now 60% is loca-
lized on the pyridine, while as much as 38% lies on the phenyl
ring. Therefore it is not surprising that the impact on the
HOMO energy level of a substituent on the pyridine will be
smaller than the impact on the LUMO energy level of a substi-
tuent on the phenyl. It is very satisfying that our simple and
naïve model can qualitatively reproduce such results from
advanced theoretical calculations.
R2
R3
R4
R5
¼ σTOTAL þ 3
For the parent complex [Ir(ppy)3], all σ values are equal to 0
and the result is a ΔEREDOX ≈ 3 eV.11 Eqn (10) is the general
quantitative relationship linking the electronic character of the
substituents, as defined from their Hammett parameters, and
the electrochemical gap of fac-tris-homoleptic iridium com-
plexes based on 2-phenylpyridine cyclometalated ligand. The
correlation between ΔEPHOTO, that is, λem, and σTOTAL is shown
in Fig. S2 (see ESI†). Although the values of A, B, C, D, α, and β
are expected to vary with the ligand skeleton, and should be
redefined in the case of cycles with a different number of
atoms than six, the relation is qualitatively useful for other
cyclometalated complexes (see below and ESI†).
With eqn (10) in hand, we now look at the effect of a single
substituent, and its impact on the redox gap depending on its
position, in particular to obtain blue-shifted emission com-
pared to the parent complex [Ir(ppy)3]. Within the same family
of molecules, we assume that the difference between electro-
chemical and optical gaps is fairly constant. Using this simpli-
fication, the condition to obtain a blue shifted emission
compared to [Ir(ppy)3] is a larger redox gap:
Using eqn (12) and (13) to calculate the conditions on the
Hammett parameters in the series of complexes reported by
Watts and coworkers,49 the observed emissions are now as
ΔEREDOX Blue > ΔEREDOX ppy
ð11Þ
For a substituent R4, only σmR4 and σpR4 are non-zero. Using
σmR4 = σm and σpR4 = σp in eqn (10), we obtain the simple
relation:
σm ꢀ 0:7σp > 0
ð12Þ
A substituent fulfilling this condition will result in a larger
redox gap than [Ir(ppy)3] and thus in an expected blue shift of
the emission. The first term, here σm, represents the influence
of the substituent on the HOMO energy level and corresponds
to the commonly used correlation between Hammett para-
meters and EOX. The second term, here σp, represents the influ-
ence of the substituent on the LUMO energy level. This
influence is less than unity, coefficient 0.7, which is well in
line with the standard rule of thumb that substituents on the
orthometallated phenyl influence more the HOMO than the
LUMO.
Fig. 2 Kohn–Sham HOMO (left) and LUMO (right) of the fac-tris homo-
leptic compound [Ir(ppy)3], obtained at the DFT(M06)/IEF-PCM(ACN)
level of theory.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 5667–5679 | 5671

View Article Online
Paper
Dalton Transactions
expected (Table 1). In particular the effect of the –CF3 group complexes and quasi-quantitative to qualitative in other cases.
on position 4 is now understood in a very simple manner: due These relations clearly explain the unusual position effect of
to its high σp value, this substituent influences the LUMO as substituents such as –CF3 and –OMe. This result quantitatively
much as the HOMO and the resulting HOMO–LUMO gap is demonstrates that a strong acceptor will not necessarily result
the same as for [Ir(ppy)3] (condition = 0.05, λem[Ir(4-CF3-ppy)3] in blue-shifted emission, and that electron-donating groups
≈ λem[Ir(ppy)3]). The methoxy case is also explained for the can be used on the orthometallated phenyl to obtain a blue
case when on position 4 the condition is positive (blue shift) shift of the emission.
and on position 5 the condition is negative (red shift). It
should be stressed that the model calculates the redox gap and
Donor groups on orthometallated phenyl for a blue shift of
not the emission gap ΔEPHOTO, which explains the similar
emission
value of the condition for the methoxy group on position 4
The most common strategy to blue shift the emission of cyclo-
metalated iridium complexes is to use acceptor substituents
on the orthometallated phenyl ring, often using fluorine sub-
stituents on positions 2 and 4. However, those complexes are
particularly unstable in electroluminescent devices. In
addition to the high energy content of the blue exciton result-
ing in a thermally accessible antibonding metal-centered
state,74 the strong electron-withdrawing character of fluorine
leads to chemical degradation in operational devices.75–77 The
possibility to replace the acceptor fluorine with a donor group
in blue phosphorescent emitters is therefore particularly
attractive. To further explore the effect of electron-donating
groups on the emission maximum of iridium complexes and
to exploit the possibility of replacing fluorine acceptor substi-
tuents by methoxy donor groups for blue emission, we pre-
pared a series of novel complexes (Scheme 2) and compared
them with known emitters, namely [Ir(ppy)2(acac)] (1), [Ir(diFp-
py)2(acac)] (2), and FIrPic.
compared to fluoride, despite the fact that fluoride substituted
compounds show a more pronounced blue-shift of the emis-
sion (see below).
To evaluate the applicability of the relations derived for fac-
tris-homoleptic iridium(III) complexes to other systems based
on cyclometalated 2-phenylpyridine ligands, we applied them
to a series of [Ir(R-ppy)2(acac)],9,55,64–67 [Ir(R-ppy)2(pic)],54,68–70
and [Pt(R-ppy)(acac)]71,72 complexes, with acac
= acetyl-
acetonate, pic = picolinate, and R-ppy = substituted 2-phenyl-
pyridine. The results are shown in ESI (Fig. S3–S5†). The
numeric values obtained from [Ir(R-ppy)3] have been directly
used in eqn (4), (6), and (10) without new calculations. It is
therefore remarkable that the oxidation potentials are still well
correlated with σC in the cases of [Ir(R-ppy)2(acac)] and [Ir(R-
ppy)2(pic)] complexes. However, correlations between ΔEREDOX
and λem are much less satisfactory. Besides the various
measurement conditions, in particular the solvent,73 and the
difference between ΔEREDOX and ΔEPHOTO (see eqn (1)), substi-
tuents possibly resulting in extended aromatic ligand systems,
such as carboxylic acid, ester, acetyl, and dimethylamino
groups, are more red shifted than anticipated by the model
based on redox potentials. In the case of [Pt(R-ppy)(acac)], the
correlation between ΔEREDOX and λem is now remarkable, while
the correlation between the reduction potentials and σN is less
satisfactory. λem for both data sets used the same solvent at
77 K, in contrast to various solvents at room temperature for
the [Ir(R-ppy)2(acac)] and [Ir(R-ppy)2(pic)] series, which may
explain the apparent improved correlation.
The model will fail when the HOMO and LUMO are not
localized on, respectively, the phenyl and pyridine of the main
ligand. For example, when a chromophoric ancillary ligand
such as 2,2′-bipyridine is used, correlation would be found
only between EOX and σC because the LUMO is localized on the
ancillary ligand. A similar failure of the model was observed
when the frontier orbitals are localized on a substituent
instead of the main ligand.68
We presented here a very simple relationship between
Hammett parameters and electrochemical properties of cyclo-
metalated iridium complexes. In particular, eqn (12)–(15) are
straightforward to apply and give a reliable condition on the
substituents to tune the redox potentials and the emission
wavelength. Due to the small data set used to derive the
numerical values and the approximations made concerning
the experimental measurement conditions, the relations are
reasonably quantitative for tris-cyclometalated ppy-based Scheme 2 Chemical structure of the complexes studied in this work.
5672 | Dalton Trans., 2014, 43, 5667–5679
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
To effectively replace electron-withdrawing fluorine, in elemental analysis and high-resolution mass spectrometry) are
addition to maintaining identical emission maximum it is available in ESI.†
also preferable to have identical molecular orbital energy
levels. In this case, the architecture of the whole electrolumi-
X-ray single crystal structures
nescent device could be kept identical, simply replacing fluor- Single crystals of 1–4, 6, and 7 were obtained by slow diffusion
ine-based emitters with donor-based emitters. To tune the of hexane or methanol in a dichloromethane solution of the
oxidation and reduction of donor-substituted complexes complex. Structures are shown in Fig. 3 and selected bond
towards higher potentials, we introduced a nitrogen atom in lengths and angles are given in Table 2. Crystallographic para-
place of the R3-substituent of the cyclometalating ligand. This meters are given in Table S2.† As expected, all complexes
modification produces
a negative shift of both HOMO adopt a distorted octahedral geometry around the iridium
and LUMO.78–80 Combining the donor-substituents with 2,3′- center with N-binding pyridines in trans positions relative to
bipyridine we developed complexes with similar electro- each other. Overall the geometry around the metal is not sig-
chemical and emission properties as complexes based on nificantly influenced by the ligand used. Whether the latter is
2,4-difluorophenylpyridine.
phenylpyridine or 2,3′-bipyridine, similar bond lengths and
angles are observed throughout the series. Similarly, there is
no net variation in the Ir–O bond length between the iridium
Synthesis
The ligands used in this study were prepared by conventional center and the acetylacetonate ancillary ligand. However, the
Suzuki coupling. It is important to stress that very clean main ligand is markedly impacted by the R2 and R4 substitu-
ligands, which are indispensable for the isolation of pure ents at the cyclometalating ring. Firstly, modification of the
iridium dimers, could only be obtained after several chromato- substitution pattern translates into an elongation of the C1–C′
2
graphy purifications (see the Experimental part for details). bond that links both rings of the cyclometalating ligand as evi-
Dimer precursors to complexes 1–4 were prepared according to denced by comparing complexes 2–4 and 6–7 with 1. Secondly,
the report of Watts and coworkers.81 Stoichiometric amounts torsions of up to 6.0° are observed for the C′ –C′ –C1–C2 angle
3
2
of the cyclometalating ligand and IrCl3 were refluxed overnight when methyl and methoxy substituents are used, in contrast to
at 130 °C in a mixture of 2-ethoxyethanol–water. Unfortunately, complexes 1 and 2. This is particularly evident for complexes 4
these standard conditions proved inefficient when 2,3′-bipyri- and 7 where a methyl group is introduced at the R2-position.
dine ligands were used, and the corresponding dimers could This deformation is also accompanied by a severe torsion
be isolated in low yields only (for example with the dimer-pre- between the R2-group and the plane of the cyclometalating
cursor of 3, isolated yield is 6%). Although we do lack clear evi- ligand (C′ –C1–C2–R2 angle) as observed for complexes 3–4 and
2
dence regarding the side-products that are formed during the 6–7 in contrast to 1 and 2. Finally, R2 and R4 substituents tend
reaction, we suspect that protons released in solution, as a to deviate from the plane of the cyclometalating ring, as exem-
result of cyclometalation, induce cleavage of the methoxy plified by the R2–C2–X–C4 and C2–X–C4–R4 torsion angles, with
groups of the ligands. This prompted us to explore an alterna- a more pronounced effect when methyl groups are used.
tive source of iridium in order to avoid such degradation.
Electrochemical properties
Encouraged by the recent success we experienced using [Ir
(COD)(μ-Cl)]2 for the synthesis of tris-heteroleptic com- The electrochemical properties of the complexes were recorded
plexes,55,82 we tested similar conditions to prepare the 2,3′- in dimethylformamide. Data are summarized in Table 3. As
bipyridine-based dimers. Contrary to IrCl3 that cyclometalates expected, each complex shows quasi-reversible to irreversible
via electrophilic aromatic substitution, COD-precursors of one-electron processes in both positive and negative bias (see
iridium usually proceed initially by oxidative addition owing to Fig. S6†). The anodic peak potential is ascribed to oxidation of
the low oxidation degree of the metal.83 This particular reactivity the iridium(III) metal center to iridium(IV) while the two catho-
is well exploited in catalytic C–H bond activations where very dic peaks are assigned to reduction of the phenylpyridine or
active and selective complexes are needed.84 Gratifyingly, when 2,3′-bipyridine ligands depending on the complex.32 For EB343
stoichiometric amounts of 2′,6′-dimethoxy-2,3′-bipyridine or 6′- and FIrPic, an additional third reduction is observed within
methoxy-2′-methyl-2,3′-bipyridine are reacted with [Ir(COD) the range of potential scan; they are attributed to the reduction
(μ-Cl)]2 in 2-ethoxyethanol at 130 °C, very clean dimers are of the picolinate ligand (Fig. 4).
obtained quickly (usually in about 3 hours) and in good yields
As evidenced by comparing the electrochemical data of 3
(see ESI†). With the various dimers in hand, both acac and pic and 4 with reference complex 1, decorating the phenyl ring
complexes become readily accessible following a standard pro- with methoxy and methyl groups increases the electrochemical
cedure. By gently refluxing the chloro-bridged complex with gap. A similar effect is seen for 2 when electron-withdrawing
acetylacetone or picolinic acid in the presence of a base, the fluorine groups are used. Nevertheless, in contrast to fluorines
final compounds could be isolated in good yields by simple fil- that induce a greater stabilization of the HOMO than the
tration over neutral silica to avoid cleavage of the ancillary LUMO due to the large positive value of σm, the combination
ligand55 followed by recrystallization from a dichloromethane– of methoxy and methyl substituents operates by destabilizing
methanol or dichloromethane–hexane mixture. Full experi- the LUMO more than the HOMO due to the large negative
mental details and characterizations (1H and 19F NMR, value of σp, as expected from the Hammett parameter-based
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 5667–5679 | 5673

View Article Online
Paper
Dalton Transactions
Fig. 3 ORTEP drawings from left to right of (top) 1, 2, 3, 4 and (bottom) 6 and 7. The labels for selected atoms as used in Table 2 are shown on the
general structure of the complexes.
Table 2 Average of selected bond lengths (Å) and angles (°) for
complexes 1, 2, 3, 4, 6 and 7
Table 3 Oxidation and reduction potentials, electrochemical gap of
complexes 1–7, FIrPic, and EB343 in dimethylformamide (0.1
TBAPF6). Potentials are listed in V vs. the Fc+/Fc couple. The peak-to-
peak separations between the anodic and the cathodic peak potentials
(ΔE = Epa − Epc) are indicated in parentheses in mV
M
Bond lengths
1
2
3
4
6
7
Ir–N
Ir–C
Ir–O
2.037
1.992
2.159
1.458
2.042
1.995
2.147
1.466
2.039
1.985
2.153
1.465
2.036
1.990
2.160
1.473
2.037
1.983
2.147
1.460
2.042
1.981
2.156
1.473
EOX
ERED
ΔEREDOX
σTOTAL
C1–C′
2
1
2
3
4
5
6
7
0.41 (95)
0.76 (115)
0.42 (145)
0.35 (225)
1.08 (−)
0.74 (155)
0.74 (145)
0.93 (165)
−2.60 (125)
3.01
3.20
3.19
3.07
3.37
3.43
3.33
3.26
0
Bond angles
N–Ir–N
C–Ir–C
C–Ir–O
O–Ir–O
−2.44 (115)
0.60
0.62
0.36
173.86 176.37 176.85 175.93 177.26 173.11
−2.77 (−), −3.29 (−)
−2.72 (145)
−2.29 (145), −2.57 (225)
−2.69 (−)
−2.59 (145), −2.88 (−)
−2.32 (185), −2.76 (180),
−3.08 (200)
−2.28 (150), −2.62 (170),
−2.98 (200)
90.82
91.25
87.27
88.67
91.42
88.95
90.60
90.57
88.47
88.00
92.31
87.66
90.86
90.55
88.33
89.19
91.36
88.45
Torsion angles
EB343
C′ –C′ –C1–C2
2.03
—
0.90
1.30
4.39
5.40
4.78
2.84
1.79
4.06
6.02
12.76
3
2
C′ –C1–C2–R2
2
FIrPic
0.93 (150)
3.21
R2–C2–X–C4
C2–X–C4–R4
—
—
179.67 179.21 177.45 177.72 175.83
178.52 179.27 179.40 178.74 176.48
structure–property relationships discussed above. Electron- bipyridine was used to obtain deep-blue emissions.78–80 As
donating groups can therefore produce the desired blue-shift expected, modification of the substitution pattern impacts the
in emission (see photophysical properties).
LUMO levels of the complexes as well, because of the elec-
Introducing a nitrogen atom in place of the R3-substituent tronic communication between both cyclometalating and
of the cyclometalating ligand produces a negative shift of both N-coordinating pyridines. This is manifest on the first reduction
HOMO and LUMO as shown by the apparent electrochemical potentials of 5, 6 and 7, which are lowered by ca. 100–150 mV
potentials measured for complexes 5–7. More specifically, compared to 2, 3 and 4, respectively.
changing the HOMO-localized ring from phenyl to pyridine
By comparing the electrochemical properties of 2 with 7,
shifts the oxidation potential positively by ca. 300 mV (the one can realize that the 6′-methoxy-2′-methyl-2,3′-bipyridine
difference measured between complexes 4 and 7 is attributed ligand is quasi-isoelectronic to a 2,4-difluorophenylpyridine.
to the poor reversibility of the oxidation of 7). This observation As a result, the main ligand used in FIrPic can now be replaced
is in line with previous reports where 2′,6′-difluoro-2,3′- while retaining the properties of the complex. This is strikingly
5674 | Dalton Trans., 2014, 43, 5667–5679
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Fig. 4 Cyclic voltammogram comparison between EB343
( ) and
FIrPic (—) recorded in dimethylformamide at a scan rate of 1000 mV s−1
.
evident from the electrochemistry of EB343. The new fluorine-
free complex displays oxidation and reduction potentials that
are virtually identical to that of the standard sky-blue emitting
FIrPic complex (Fig. 4).
Fig. 5 UV-visible absorption spectra of complexes in acetonitrile:
(a) 1 (—), 2 ( ), 3 ( ), 4 ( ); (b) 1 (—), 5 ( ), 6 ( ), 7 ( ); (c) FIrPic (—),
EB343 ( ).
Table 4 Data for complexes 1–7, EB343, and FIrPic
UV-visible absorption spectra
λem a (nm)
RT
λem b (nm)
77 K
The UV-visible absorption spectra of all complexes were
measured in solution in acetonitrile (Fig. 5 and Table 4). All
spectra display a similar overall profile and, by analogy with
previously reported assignations, the intense absorption bands
in the UV region (λabs ≈ 250–300 nm and ε ≈ 40 × 103 M−1
cm−1) are attributed to ligand centered (LC) 1(π–π*) tran-
sitions,9 while the weaker and lower-energy absorption bands
correspond to metal-to-ligand charge transfer (1MLCT) tran-
sitions.85 Finally, the lowest energy transitions with ε < 103
λabs (nm) (ε (103 L mol−1 cm−1))a
1
258 (43.2), 342 (8.6), 398 (4.0),
456 (2.3), 487 (0.9)
252 (40.3), 328 (10.2), 383 (4.5),
428 (2.0), 458 (0.7)
260 (34.2), 274 (37.3), 326 (22.2),
398 (4.2), 473 (0.6)
277 (46.3), 320 (20.7), 358 (8.24),
401 (4.1), 484 (0.7)
246 (41.8), 265 (47.3),
370 (4.4), 439 (0.4)
268 (38.1), 333 (22.9),
384 (4.8), 457 (0.2)
270 (37.1), 286 (34.8), 322 (20.2),
388 (3.7), 456 (0.3)
268 (45.6), 281 (45.1), 320 (21.4),
380 (4.8), 449 (0.2)
253 (40.9), 282 (24.5), 316 (11.9),
375 (4.8), 455 (0.4)
528
506
472
486
495
447
461
464
456
458
2
491
3
500
4
519
5
(465) 480
470
3
M−1 cm−1 are MLCT transitions caused by direct population
of the emitting triplet state.54 Compared to the model complex 1,
the presence of substituents has practically no influence on
the intensity of the LC and 3MLCT bands. In contrast, the
methoxy groups have a significant impact on the intensity of
6
7
(484) 497
(466) 492
469
EB343
FIrPic
1
the MLCT bands, which more than doubles upon addition of
two methoxy groups (at λabs = 335 nm: ε ≈ 9 × 103 M−1 cm−1
for 1 and 2; ε ≈ 15 × 103 M−1 cm−1 for 4 with one –OMe group;
ε ≈ 22 × 103 M−1 cm−1 for 3 with two –OMe groups).
a In acetonitrile. b In 2-methyltetrahydrofuran.
As expected from the Hammett parameter-based model dis-
cussed above, the onset of absorption (defined as ε = 100 M−1
cm−1) is blue-shifted as methoxy groups are added to the main
Luminescence spectra
ligand. However, in contrast to the electrochemical gaps which With the exception of 6, when excited at 298 K within the
are identical for 2 and 3, as anticipated by the values of σTOTAL
,
ligand (π–π*) and MLCT absorption bands, the complexes
the onset of absorption of 2 is 13 nm (∼600 cm−1) blue-shifted show broad and poorly structured emissions that are character-
compared to 3. This means that the additional term ∑ai relat- istic of complexes containing a combination of cyclometalated
ing ΔEPHOTO and ΔEREDOX in eqn (1) is not constant on going ligands with acetylacetonate as the ancillary ligand (Fig. 6). At
from 2 to 3, although in both cases the skeleton of the main 77
K
in 2-methyltetrahydrofuran (2-MeTHF), the spectra
ligand is phenylpyridine.
become highly structured and clear vibronic progressions are
The replacement of the 2-phenylpyridine core with 2,3′- seen for all complexes (Fig. 7). Similar to the absorption
bipyridine simply leads to 20–25 nm (∼1000 cm−1) blue shift spectra, the emission is blue-shifted as methoxy groups are
of the onset of absorption without significant other changes in added to the main ligand, 519 nm for 4 and 500 nm for 3 com-
the absorption spectra.
pared to 528 nm for reference 1 at room temperature, but not
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 5667–5679 | 5675

View Article Online
Paper
Dalton Transactions
the standard sky-blue emitting FIrPic complex. The adiabatic
E0–0 transition energies computed at the (U)DFT/M05-2X/
IEF-PCM(ACN) level for EB343 and FIrPic are close, 3.03 eV vs.
3.14 eV, respectively. In both cases, the emission exhibits a
LC-MLCT character (see ESI†). The photoluminescence
quantum yield and lifetime of the excited state of EB343 and
FIrPic in solution in degassed acetonitrile at room temperature
are Φ = 0.54 and 0.61 and τ = 1.29 and 1.69 μs, respectively.
Finally, to explore further the colours of emission offered by
this new fluorine-free ligand, we used the chloro-bridged
iridium dimer [Ir(L7)2(μ-Cl)]2 to screen ancillary ligands
(Fig. S12†).86
Conclusions
We have described a simple quantitative model based on
Hammett parameters that rationalizes the effect of the substi-
tuents on the properties of cyclometalated iridium(III) com-
plexes. This model allows predicting the apparent redox
potentials as well as the electrochemical gap of homoleptic
complexes based on phenylpyridine ligands with good accu-
racy. Importantly, the model accounts for the unequal effect of
the substituents on both the HOMO and the LUMO energy
levels by taking into account both the σm and σp values of the
Hammett constant for a given substituent. Consequently, the
model can also be used to anticipate the emission maxima of
the corresponding complexes with improved reliability. In par-
ticular, it was possible to rationalize in a very simple manner
the influence on emission maxima of trifluoromethyl and
methoxy groups and to demonstrate that a strong acceptor will
not necessarily result in a blue-shifted emission, while elec-
tron-donating groups can be used on the orthometallated
phenyl to obtain a blue-shift of the emission.
Fig. 6 Emission spectra of complexes in degassed acetonitrile at room
temperature: (a) 1 (—), 2 ( ), 3 ( ), 4 ( ); (b) 1 (—), 5 ( ), 6 ( ), 7 ( );
(c) FIrPic (—), EB343 ( ).
The main strength of this model is the simplicity of the
approach and of the relations developed from it, which makes
its application to a range of complexes based on cyclometa-
lated 2-phenylpyridine straightforward. Further numeric
improvements could be obtained by increasing the size of the
database and having measurements performed under identical
conditions. Conceptual improvements are expected by sepa-
rating inductive and mesomeric effects and by distinguishing
between para and ortho positions, especially due to the poss-
ible impact of steric hindrance of substituents on the pro-
Fig. 7 Emission spectra of complexes in 2-methyltetrahydrofuran at
77 K: (a) 1 (—), 2 ( ), 3 ( ), 4 ( ); (b) 1 (—), 5 ( ), 6 ( ), 7 ( ); (c) FIrPic (—),
EB343 ( ).
as much as with fluorine substituents, 491 nm, in contrast to perties of the complexes.87 These improvements could result
the electrochemical gaps which are identical for 2 and 3. This in a reliable relationship between Hammett constants and
means that, as for the absorption, the additional term ∑ai emission wavelength maxima. Finally, generalization beyond
relating ΔEPHOTO and ΔEREDOX in eqn (1) is not constant on ppy could be obtained by combining the model in this work
going from 2 to 3.
with concepts such as the ligand electrochemical parameter.88,89
In this work, we used this model to develop a series of emit-
As for their electrochemical properties, 2 and 7 have also a
very similar emission spectrum. As a result, the main ligand ters where electron-donating groups, methyl and methoxy, can
used in FIrPic can be replaced by the 6′-methoxy-2′-methyl-2,3′- effectively replace electron-withdrawing fluorine substituents
bipyridine ligand in EB343. The emission spectra of EB343 on the orthometallated phenyl to induce a blue shift of the
compared to FIrPic are shown in Fig. 6c at room temperature emission. This result is in contrast with the common approach
and Fig. 7c at 77 K. The new fluorine-free complex displays that relies almost exclusively on fluorine to blue shift the emis-
only a slightly blue shifted and broader emission compared to sion maximum of a given complex. The blue shift induced by
5676 | Dalton Trans., 2014, 43, 5667–5679
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
fluorine is mainly due to the strong acceptor character towards 15 S. Bettington, M. Tavasli, M. R. Bryce, A. S. Batsanov,
the meta position, resulting in a stronger stabilization of the
A. L. Thompson, H. A. Al Attar, F. B. Dias and A. P. Monkman,
HOMO than the LUMO, while the blue shift induced by
J. Mater. Chem., 2006, 16, 1046.
methoxy is mainly due to the strong donor character towards 16 C. Adachi, M. A. Baldo, M. E. Thompson and S. R. Forrest,
the para position, resulting in a stronger destabilization of the J. Appl. Phys., 2001, 90, 5048.
LUMO than the HOMO. The work also shows that the assump- 17 R. N. Bera, N. Cumpstey, P. L. Burn and I. D. W. Samuel,
tion that the difference between electrochemical and optical Adv. Funct. Mater., 2007, 17, 1149.
gaps is fairly constant within the same family of molecules 18 P. T. Chou and Y. Chi, Chem.–Eur. J., 2007, 13, 380.
should be used with caution.
Finally, as a proof of concept, we used a 2,3′-bipyridine as
the skeleton of the main ligand with donor substituents and
19 S. Lamansky, P. Djurovich, D. Murphy, F. Abdel-Razzaq,
H. E. Lee, C. Adachi, P. E. Burrows, S. R. Forrest and
M. E. Thompson, J. Am. Chem. Soc., 2001, 123, 4304.
were able to match the various properties, namely oxidation 20 E. Orselli, G. S. Kottas, A. E. Konradsson, P. Coppo,
and reduction potentials, electrochemical gap and emission
profile, of the standard sky-blue emitter FIrPic with a new fluo-
rine-free complex referred to as EB343.
R. Fröhlich, L. De Cola, A. van Dijken, M. Buchel and
H. Börner, Inorg. Chem., 2007, 46, 11082.
21 J. A. G. Williams, A. J. Wilkinson and V. L. Whittle, Dalton
Trans., 2008, 2081.
22 J.-H. jou, M.-F. Hsu, W.-B. Wang, C.-L. Chin, Y.-C. Chung,
C.-T. Chen, J.-J. Shyue, S.-M. Shen, M.-H. Wu, W.-C. Chang,
C.-P. Liu, S.-Z. Chen and H.-Y. Chen, Chem. Mater., 2009,
21, 2565.
Acknowledgements
We acknowledge financial support for this work by Solvay S.A. 23 K. K. Lo, A. W. Choi and W. H. Law, Dalton Trans., 2012, 41,
and the European Union (HetIridium, CIG 322280).
6021.
24 H. J. Bolink, H. Brine, E. Coronado and M. Sessolo, Adv.
Mater., 2010, 22, 2198.
25 R. D. Costa, E. Orti, H. J. Bolink, S. Graber, S. Schaffner,
M. Neuburger, C. E. Housecroft and E. C. Constable,
Adv. Funct. Mater., 2009, 19, 3456.
Notes and references
1 C. Alexiou and A. B. P. Lever, Coord. Chem. Rev., 2001, 26 J. D. Nguyen, J. W. Tucker, M. D. Konieczynska and
216–217, 45. C. R. J. Stephenson, J. Am. Chem. Soc., 2011, 1.
2 J. K. Hino, L. Della Ciana, W. J. Dressick and B. P. Sullivan, 27 A. Valore, E. Cariati, C. Dragonetti, S. Righetto, D. Roberto,
Inorg. Chem., 1992, 31, 1072.
R. Ugo, F. De Angelis, S. Fantacci, A. Sgamellotti, A. Macchioni
3 M. Lanznaster, A. Neves, A. J. Bortoluzzi, A. M. Assumpcao,
and D. Zuccaccia, Chem.–Eur. J., 2010, 16, 4814.
I. Vencato, S. P. Machado and S. M. Drechsel, Inorg. Chem., 28 J. Sun, W. Wu and J. Zhao, Chem.–Eur. J., 2012, 18,
2006, 45, 1005. 8100.
4 C. Makedonas and C. A. Mitsopoulou, Eur. J. Inorg. Chem., 29 B. Happ, A. Winter, M. D. Hager and U. S. Schubert, Chem.
2007, 4176. Soc. Rev., 2012, 41, 2222.
5 B. H. Solis and S. Hammes-Schiffer, J. Am. Chem. Soc., 30 C. Ulbricht, B. Beyer, C. Friebe, A. Winter and U. S. Schubert,
2011, 133, 19036. Adv. Mater., 2009, 21, 4418.
6 A. R. Dick, J. W. Kampf and M. S. Sanford, J. Am. Chem. 31 L. Sun, A. Galan, S. Ladouceur, J. D. Slinker and E. Zysman-
Soc., 2005, 127, 12790. Colman, J. Mater. Chem., 2011, 21, 18083.
7 L. Flamigni, A. Barbieri, C. Sabatini, B. Ventura and 32 P. J. Hay, J. Phys. Chem. A, 2002, 106, 1634.
F. Barigelletti, Top. Curr. Chem., 2007, 281, 143.
8 K. A. King, P. J. Spellane and R. J. Watts, J. Am. Chem. Soc.,
1985, 107, 1431.
33 T. Sajoto, P. I. Djurovich, A. Tamayo, M. Yousufuddin,
R. Bau, M. E. Thompson, R. J. Holmes and S. R. Forrest,
Inorg. Chem., 2005, 44, 7992.
9 S. Lamansky, P. Djurovich, D. Murphy, F. Abdel-Razzaq, 34 A. Tsuboyama, H. Iwawaki, M. Furugori, T. Mukaide,
R. Kwong, I. Tsyba, M. Bortz, B. Mui, R. Bau and
M. E. Thompson, Inorg. Chem., 2001, 40, 1704.
10 M. S. Lowry and S. Bernhard, Chem.–Eur. J., 2006, 12, 7970.
J. Kamatani, S. Igawa, T. Moriyama, S. Miura, T. Takiguchi,
S. Okada, M. Hoshino and K. Ueno, J. Am. Chem. Soc.,
2003, 125, 12971.
11 A. B. Tamayo, B. D. Alleyne, P. I. Djurovich, S. Lamansky, 35 Y. S. Yeh, Y. M. Cheng, P. T. Chou, G. H. Lee, C. H. Yang,
I. Tsyba, N. N. Ho, R. Bau and M. E. Thompson, J. Am.
Chem. Soc., 2003, 125, 7377.
Y. Chi, C. F. Shu and C. H. Wang, ChemPhysChem, 2006, 7,
2294.
12 Y. You and S. Y. Park, Dalton Trans., 2009, 1267.
13 E. Baranoff, J. H. Yum, M. Graetzel and M. K. Nazeeruddin,
J. Organomet. Chem., 2009, 694, 2661.
14 M. A. Baldo, M. E. Thompson and S. R. Forrest, Nature,
2000, 403, 750.
36 S. Bettington, M. Tavasli, M. R. Bryce, A. Beeby, H. Al-Attar
and A. P. Monkman, Chem.–Eur. J., 2007, 13, 1423.
37 Z. Q. Gao, B. X. Mi, H. L. Tam, K. W. Cheah, C. H. Chen,
M. S. Wong, S. T. Lee and C. S. Lee, Adv. Mater., 2008, 20,
774.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 5667–5679 | 5677

View Article Online
Paper
Dalton Transactions
38 R. Wang, D. Liu, H. Ren, T. Zhang, X. Wang and J. Li, 62 V. V. Pavlishchuk and A. W. Addison, Inorg. Chim. Acta,
J. Mater. Chem., 2011, 21, 15494. 2000, 298, 97.
39 E. Baranoff, S. Fantacci, F. De Angelis, X. Zhang, 63 Although equations are given with coefficients rounded to
R. Scopelliti, M. Grätzel and M. K. Nazeeruddin, Inorg.
Chem., 2011, 50, 451.
40 G. Treboux, J. Mizukami, M. Yabe and S. Nakamura,
J. Photopolym. Sci. Technol., 2008, 21, 347.
41 C. Hansch, A. Leo and R. W. Taft, Chem. Rev., 1991, 91,
165.
42 L. P. Hammett, J. Am. Chem. Soc., 1937, 59, 96.
43 A. R. Katritzky, M. Kuanar, S. Slavov, C. D. Hall, M. Karelson,
I. Kahn and D. A. Dobchev, Chem. Rev., 2010, 110, 5714.
44 T. Kubota and H. Miyazaki, Bull. Chem. Soc. Jpn., 1966, 39,
2057.
the closest 0.1, the coefficients have been all calculated
from the results of the correlation, and not from the
rounded numbers.
64 K. H. Lee, H. J. Kang, S. J. Lee, J. H. Seo, Y. K. Kim and
S. S. Yoon, Synth. Met., 2011, 161, 1113.
65 R. Ragni, E. A. Plummer, K. Brunner, J. W. Hofstraat,
F. Babudri, G. M. Farinola, F. Naso and L. De Cola, J. Mater.
Chem., 2006, 16, 1161.
66 E. Baranoff, S. Suarez, P. Bugnon, H. J. Bolink, C. Klein,
R. Scopelliti, L. Zuppiroli, M. Grätzel and M. K. Nazeeruddin,
ChemSusChem, 2009, 2, 305.
45 M. N. Ackermann, K. B. Moore, A. S. Colligan, 67 E. Baranoff, J. H. Yum, I. Jung, R. Vulcano, M. Gratzel and
J. A. Thomas-Wohlever and K. J. Warren, J. Organomet.
Chem., 2003, 667, 81.
46 J. E. Heffner, C. T. Wigal and O. A. Moe, Electroanalysis,
1997, 9, 629.
47 N. Leventis, A. M. M. Rawaswdeh, G. H. Zhang, I. A. Elder
and C. Sotiriou-Leventis, J. Org. Chem., 2002, 67, 7501.
M. K. Nazeeruddin, Chem.–Asian J., 2010, 5, 496.
68 M. Marin-Suarez, B. F. E. Curchod, I. Tavernelli,
U. Rothlisberger, R. Scopelliti, I. Jung, D. Di Censo,
M. Gratzel, J. F. Fernandez-Sanchez, A. Fernandez-
Gutierrez, M. K. Nazeeruddin and E. Baranoff, Chem.
Mater., 2012, 24, 2330.
48 M. L. Dell’arciprete, C. J. Cobos, J. P. Furlong, D. O. Martire 69 M. L. Xu, R. Zhou, G. Y. Wang and J. Y. Yu, Inorg. Chim.
and M. C. Gonzalez, ChemPhysChem, 2007, 8, 2498. Acta, 2009, 362, 2183.
49 K. Dedeian, P. I. Djurovich, F. O. Garces, G. Carlson and 70 I. S. Shin, S. Yoon, J. I. Kim, J. K. Lee, T. H. Kim and
R. J. Watts, Inorg. Chem., 1991, 30, 1685. H. Kim, Electrochim. Acta, 2011, 56, 6219.
50 S. Okada, K. Okinaka, H. Iwawaki, M. Furugori, 71 J. Brooks, Y. Babayan, S. Lamansky, P. I. Djurovich,
M. Hashimoto, T. Mukaide, J. Kamatani, S. Igawa,
A. Tsuboyama, T. Takiguchi and K. Ueno, Dalton Trans.,
2005, 1583.
I. Tsyba, R. Bau and M. E. Thompson, Inorg. Chem., 2002,
41, 3055.
72 G. J. Zhou, Q. Wang, X. Z. Wang, C. L. Ho, W. Y. Wong,
D. G. Ma, L. X. Wang and Z. Y. Lin, J. Mater. Chem., 2010,
20, 7472.
73 I. R. Laskar, S. F. Hsu and T. M. Chen, Polyhedron, 2006,
25, 1167.
51 J. Li, P. I. Djurovich, B. D. Alleyne, M. Yousufuddin,
N. N. Ho, J. C. Thomas, J. C. Peters, R. Bau and
M. E. Thompson, Inorg. Chem., 2005, 44, 1713.
52 B. D’Andrade, S. Datta, S. R. Forrest, P. Djurovich,
E. Polikarpov and M. E. Thompson, Org. Electron., 2005, 6, 74 T. Sajoto, P. I. Djurovich, A. B. Tamayo, J. Oxgaard,
11.
W. A. Goddard and M. E. Thompson, J. Am. Chem. Soc.,
2009, 131, 9813.
75 V. Sivasubramaniam, F. Brodkorb, S. Hanning, H. P. Loebl,
V. van Elsbergen, H. Boerner, U. Scherf and
M. Kreyenschmidt, Cent. Eur. J. Chem., 2009, 7, 836.
53 I. Avilov, P. Minoofar, J. Cornil and L. De Cola, J. Am. Chem.
Soc., 2007, 129, 8247.
54 E. Baranoff, B. F. E. Curchod, F. Monti, F. Steimer,
G. Accorsi, I. Tavernelli, U. Rothlisberger, R. Scopelliti,
M. Grätzel and M. K. Nazeeruddin, Inorg. Chem., 2012, 51, 76 V. Sivasubramaniam, F. Brodkorb, S. Hanning, H. P. Loebl,
799.
V. van Elsbergen, H. Boerner, U. Scherf and M. Kreyenschmidt,
55 E. Baranoff, B. F. E. Curchod, J. Frey, R. Scopelliti,
J. Fluorine Chem., 2009, 130, 640.
F. Kessler, I. Tavernelli, U. Rothlisberger, M. Grätzel and 77 I. R. de Moraes, S. Scholz, B. Lussem and K. Leo, Org.
M. K. Nazeeruddin, Inorg. Chem., 2012, 51, 215. Electron., 2011, 12, 341.
56 A. J. M. Duisenberg, L. M. J. Kroon-Batenburg and 78 S. J. Lee, K. M. Park, K. Yang and Y. Kang, Inorg. Chem.,
A. M. M. Schreurs, J. Appl. Crystallogr., 2003, 36, 220. 2009, 48, 1030.
57 R. H. Blessing, Acta Crystallogr., Sect. A: Fundam. Crystal- 79 F. Kessler, R. D. Costa, D. Di Censo, R. Scopelliti,
logr., 1995, 51, 33.
58 G. M. Sheldrick, Acta Crystallogr., Sect. A: Fundam. Crystal-
logr., 2008, 64, 112.
E. Ortí, H. J. Bolink, S. Meier, W. Sarfert, M. Grätzel,
M. K. Nazeeruddin and E. Baranoff, Dalton Trans., 2012,
41, 180.
59 A. A. Vlcek, E. S. Dodsworth, W. J. Pietro and A. B. P. Lever, 80 S. B. Meier, W. Sarfert, J. M. Junquera-Hernández,
Inorg. Chem., 1995, 34, 1906.
60 V. V. Grushin, N. Herron, D. D. LeCloux, W. J. Marshall,
V. A. Petrov and Y. Wang, Chem. Commun., 2001, 1494.
M. Delgado, D. Tordera, E. Ortí, H. J. Bolink, F. Kessler,
R. Scopelliti, M. Grätzel, M. K. Nazeeruddin and
E. Baranoff, J. Mater. Chem. C, 2013, 1, 58.
61 C. M. Cardona, W. Li, A. E. Kaifer, D. Stockdale and 81 F. O. Garces, K. A. King and R. J. Watts, Inorg. Chem., 1988,
G. C. Bazan, Adv. Mater., 2011, 23, 2367.
27, 3464.
5678 | Dalton Trans., 2014, 43, 5667–5679
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
82 D. Tordera, M. Delgado, E. Ortí, H. J. Bolink, J. Frey, 86 E. Baranoff, I. Jung, R. Scopelliti, E. Solari, M. Grätzel
M. K. Nazeeruddin and E. Baranoff, Chem. Mater., 2012, 24,
and M. K. Nazeeruddin, Dalton Trans., 2011, 40,
1896.
6860.
83 S. M. W. Rahaman, S. Dinda, T. Ghatak and J. K. Bera, 87 L. Rincon and R. Almeida, J. Phys. Chem. A, 2012, 116,
Organometallics, 2012, 31, 5533. 7523.
84 K. I. Goldberg and A. S. Goldman, in ACS Symposium Series 88 S. S. Fielder, M. C. Osborne, A. B. P. Lever and W. J. Pietro,
885, American Chemical Society, Washington, DC, 2004. J. Am. Chem. Soc., 1995, 117, 6990.
85 B. Schmid, F. O. Garces and R. J. Watts, Inorg. Chem., 1994, 89 H. Masui and A. B. P. Lever, Inorg. Chem., 1993, 32,
33, 9.
2199.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 5667–5679 | 5679