Highly Emitting Neutral Dinuclear Rhenium Complexes as Phosphorescent Dopants for Electroluminescent Devices

Highly Emitting Neutral Dinuclear Rhenium Complexes

as Phosphorescent Dopants for Electroluminescent

Devices

By

Matteo Mauro, Elsa Quartapelle Procopio, Yinghui Sun, Chen-Han Chien,

Daniela Donghi, Monica Panigati, Pierluigi Mercandelli,* Patrizia Mussini,

Giuseppe D’Alfonso,* and Luisa De Cola*

1. Introduction

After the seminal works of Baldo,

Thomp-son and Forrest,

_{a large number of papers}

have appeared, dealing with the use of

phosphorescent emitters as triplet

harvest-ing dopants, able to increase the efficiencies

of light emitting devices.

Iridium(III)

complexes have been mainly exploited,

but also other transition metals have been

investigated, including rhenium(I).

The synthesis of a new family of

tricarbonyl rhenium(I) complexes has

recently been reported,

which exhibit

intense photoluminescence in the range

580–620 nm, with the highest quantum

yields ever reported for neutral tricarbonyl

rhenium complexes (up to ca. 10%). These

compounds, of general formula [Re

(m-X)

(CO)

(m-diazine)], are dinuclear and

contain one bridging 1,2-diazine and two

bridging ancillary ligands X (in Scheme 1

the prototypical chloro-complexes with

bridging pyridazine and 4-Me-pyridazine

are shown, which will be indicated here as

compounds 0 and 1). The emission arises

from triplet metal-to-ligand charge-transfer

FULL

P

APER

[*] Dr. P. Mercandelli

Dipartimento di Chimica Strutturale e Stereochimica Inorganica Universita´ degli Studi di Milano

Via Venezian 21, 20133 Milano (Italy) E-mail: [email protected]

Prof. G. D’Alfonso, M. Mauro, E. Quartapelle Procopio, Dr. D. Donghi, Dr. M. Panigati

Dipartimento di Chimica Inorganica Universita´ degli Studi di Milano Metallorganica e Analitica ‘‘L. Malatesta’’ Via Venezian 21, 20133 Milano (Italy) E-mail: [email protected]

DOI: 10.1002/adfm.200900744

Prof. L. De Cola, Dr. Y. Sun

Westfa¨lische Wilhelms Universita¨t Mu¨nster Physikalisches Institut

Mendelstrasse, 7 48149 Mu¨nster, Germany and Center for Nanotechnology (CeNTech) 48149 Mu¨nster (Germany)

E-mail: [email protected] C.-H. Chien

Department of Applied Chemistry National Chiao Tung University Hsinchu, 30056 (Taiwan) Prof. P. R. Mussini

Dipartimento di Chimica Fisica ed Elettrochimica Universita´ degli Studi di Milano

Via Golgi 19, 20133 Milano (Italy)

A series of neutral, dinuclear, luminescent rhenium(I) complexes suitable for

phosphorescent organic light emitting devices (OLEDs) is reported. These

compounds, of general formula [Re2(m-Cl)2(CO)6(m-1,2-diazine)], contain

diazines bearing alkyl groups in one or in both the b positions. Their

electrochemical and photophysical properties are presented, as well as a

combined density functional and time-dependent density functional study of

their geometry, relative stability and electronic structure. The complexes show

intense green/yellow emissions in toluene solution and in the solid state and

some of the complexes possess high emission quantum yields (f = 0.18–0.22

for the derivatives with disubstituted diazines). In butyronitrile glass, at 77 K,

due to the charge transfer character of the lowest (emitting) excited state,

strong blue shift of the emission is observed, accompanied by a strong

increase in the lifetime values. The highest-performing emitting complex,

containing cyclopentapyridazine as ligand, is tested in a polymer-based

light-emitting device, with poly(9-vinylcarbazole) as matrix, as well as in a device

obtained by vacuum sublimation of the complex in the

2,7-bis(diphenylphosphine oxide)-9-(9-phenylcarbazol-3-yl)-9-phenylfluorene

(PCF) matrix. This represents the first example of devices obtained with a

rhenium complex which can be sublimed and is solution processable.

Furthermore, the emission is the bluest ever reported for electrogenerated

luminescence for rhenium complexes.

FULL

P

APER

(

_{MLCT) states. Lifetimes and emission intensities strongly vary}

on changing the ancillary ligands (halide

or hydride

) or the

substituents on the diazines, as a result of a wide variability of the

rate of non-radiative decays. Computational studies

indicate that

the emission intensity can be correlated to an easily computable

parameter, such as the instantaneous interaction energy,

_thus

showing the connection between the accessibility of radiationless

relaxation pathways and the stiffness of the Re

-diazine scaffold, as

expected for transitions involving MLCT states. On this ground it

was possible to predict that highly emitting species might be

obtained on using electron rich 1,2-diazines suitably substituted in

the b positions.

The present work testifies on the exactness of this prediction. In

fact, we have found that the presence of alkyl substituents in both

the 4 and 5 positions of the 1,2-diazines gives [Re

(m-Cl)

(CO)

(m-diazine)] complexes with emission quantum yields about two fold

higher than their mono-substituted analogues. Furthermore, due

to the good solubility and stability upon sublimation of the

complexes and their reversible electrochemical behavior, we have

investigated their use as component in electroluminescent

devices. We report here for the first time a comparison, with

the same Re(I) complex (the best emitting material) of

electro-luminescent devices made by sublimation and by spin coating

using a polymeric matrix.

2. Results and discussion

2.1. Synthesis of Ligands and Complexes

The synthesis of the ligands involved a [4 þ 2] Diels–Alder

cyclo-addition between 1,2,4,5-tetrazine and the proper dienofile

(substituted alkyne or cyclopentene enamine, see Scheme 2).

The corresponding complexes [Re

(m-Cl)

(CO)

(m-diazine)] (1–6,

Schemes 1 and 3) were obtained in good isolated yields by refluxing

[ReCl(CO)

] with 0.5 equiv. of the 1,2-diazine, in toluene. The

complexes are stable in solution at room temperature, also in a

coordinating solvent as acetonitrile, and their purity was attested by

spectroscopic and elemental analysis. All the complexes display

four intense bands in the n(CO) IR region, as typical of this class of

compounds.

_The

_{H resonances of the ortho-hydrogens of}

coordinated diazines exhibit a significant downfield shift (0.3–

0.6 ppm) with respect to the corresponding free diazines.

2.2. Electrochemical Characterization

The complexes 1–6 were investigated by cyclic voltammetry (see

Figure 1 and Table 1) and the results have confirmed the main

Scheme 1. Structures of the diazine complexes 0 and 1.

Figure 1. CV characteristics obtained for complexes 1–6 and for the analogous complex with pyridazine (compound 0), at 0.2 V s
1_{scan rate,}

in MeCN, with 0.1MTBAPF6as the supporting electrolyte, at 298 K, with

ohmic drop compensation. Thick lines: CV including first oxidation and reduction peaks; thin lines: first oxidation peak at 2 V s
1_.

Scheme 3. Structures of complexes 2–6.

FULL

P

APER

findings of the previous study

on the related [Re

(m-Cl)

(CO)

(m-diazine)] complexes. For comparison, the data concerning the

complex bearing unsubstituted pyridazine (compound 0) are also

reported in Table 1 and Figure 1. A more complete list of the

electrochemical data is provided in Table S1 in the Supporting

Information.

The first reduction peak, which is centered on the diazine

ligand,

is monoelectronic and reversible, both from the chemical

and the electrochemical point of view.

The peak potentials

regularly shift in the negative direction with an increase in the

number of alkyl substituents, consistently with an increasingly

electron-rich aromatic site.

The metal-centered oxidation corresponds to a simultaneous

two-electron transfer, as in the former study,

according to the

neatly doubled i

/c parameter with respect to the reduction peak.

The process appears irreversible, both chemically and

electro-chemically,

but in the upper range of the explored scan rates

return peaks emerge for some of these species (namely

compounds 1, 2, 3, and 5), as shown in Figure 1. This clearly

points to an (electrochemically quasi-reversible) electron transfer

step and a subsequent chemical step (EC mechanism, or better

E

C, taking into account the bi-electronic nature of the

process),

fast enough to be competitive with the reduction

of the oxidation product in the backward scan. In agreement with

this hypothesis, the ratio between forward and backward peak

currents for complexes 1, 2, 3, and 5 regularly decreases with

increasing scan rate (due to the decrease of the reaction time), so

that it has been possible to evaluate the pseudo-first order kinetic

constants k

I

, that are reported in Table S1 of the Supporting

Information.

Table 1 also reports the electrochemical HOMO–LUMO gaps,

which regularly increase with increasing alkyl substitution, as a

result of the decrease of the reduction potentials. The comparison

with the spectroscopic (absorption) gaps (also reported in Table 1)

is discussed in the following.

2.3. Photophysical Characterization

Electronic absorption spectra for three representative complexes 1,

4, and 6 are depicted in Figure 2 (the others are shown in Figure S3

of the Supporting Information). Table 2 summarizes the

mean-ingful photophysical data for all the new complexes. At room

temperature, in dichloromethane solution, all the complexes show

spectra dominated by two main absorption features in the UV–vis

region. The intense band (e  0.9–1.02  10

_M

_cm

_{) at higher}

energy is independent of number and nature of the alkyl

substituents on the pyridazine (pydz) ring and little sensitive to

the solvent (Figure S4 in the Supporting Information). On the base

of time-dependent density functional theory (TD DFT)

computa-tions it can be attributed to a superposition of d–d excitacomputa-tions from

the ‘‘t

’’ set of the two Re atoms (the six HOMOs, showing a Re–Cl

p

character

) to the ‘‘e

’’ set of the two Re atoms (the LUMO þ 2 to

LUMO þ 5 orbitals, showing in addition a large C

_{O p}

character).

These transitions, being centered on the [Re

(m-Cl)

(CO)

] moiety

of the complexes, do not show any significant dependence on the

nature of the diazine ligand. In addition, the charge redistribution

associated to these excitations does not lead to a large variation of

the dipole moment of the molecule, in accord with the observed

little sensitivity to the solvent of this absorption.

Figure 2. Absorption spectra of 1 (---), 4 (–~–) and 6 (—) in dichloro-methane solution at room temperature.

Table 1. First cathodic and anodic peak potentials (Ep,c andEp,a), and

electrochemical (Eg) and UV (Eg,UV) energy gaps. Potentials are referred to

the Fcþ_{jFc couple. The data for complexes 0 and 1 are from the previous}

work [6].

Complex Ep,c[V] Ep,aI[V] Eg[eV] Eg,UV[eV]

0
1.345 1.315 2.66 3.31 1
1.457 1.271 2.73 3.41 2
1.460 1.275 2.74 3.42 3
1.441 1.290 2.73 3.42 4
1.583 1.308 2.89 3.54 5
1.571 1.267 2.84 3.50 6
1.571 1.324 2.90 3.54

Table 2. Absorption and emission spectral data of 1 – 6 at room temperature.

labs[a] [nm] (e10–4M
1cm
1) lem[b] [nm] lem[c] [nm] F[d] t[b] [ms] t[d] [ms] t[c] [ms]

1 264 (0.90), 364 (0.81) 574 513 0.08 0.39 2.0 34.72 2 263 (1.02), 363 (0.94) 575 510 0.09 0.43 2.2 33.78 3 263 (0.92), 363 (0.87) 573 505 0.10 0.43 2.4 30.90 4 263 (0.92), 351 (0.86) 550 500 0.18 0.44 4.9 37.17 5 263 (1.02), 355 (0.95) 550 496 0.19 0.43 5.1 36.56 6 261 (0.97), 351 (0.89) 547 500 0.22 0.39 5.3 41.66

[a] In aerated CH2Cl2. [b] In aerated toluene (lex¼ 366 nm).) [c] In butyronitrile glass at 77 K (lex¼ 366 nm).) [d] In deaerated toluene; F versus fac-Ir(ppy)3

in deaerated CH2Cl2(Fref¼ 0.40). [e] The data for 1 are slightly different from those previously published [6], particularly for the lifetime, probably due to

FULL

P

APER

The lower energy (351–364 nm) absorption bands have slightly

weaker intensity (e  0.81–0.94  10

M

cm

) and can be

assigned to singlet–singlet spin-allowed metal-to-ligand charge

transfer transitions (MLCT), by analogy with the mononuclear

tricarbonyl Re(I) complexes containing chelating diimine ligands.

This assignment is also supported by the typical strong solvent

dependence of the charge-transfer Re(d)–L(p

) absorption band.

Indeed, as already reported for compound 0,

_{a blue shift is}

observed upon increasing solvent polarity (from 371 nm in toluene

to 336 nm in MeCN for 1) (Fig. S4, Supporting Information). The

position of the MLCT transition is dependent on the mono- or

di-alkyl substitution on the aromatic ring. Going from 1–3 to 4–6, the

presence of two weak electron-donating groups makes charge

transfer onto the substituted pyridazine more difficult, giving a

blue shift in the absorption maxima (Fig. 2).

This trend nicely agrees with the variation of the

electro-chemical energy gap, E

. Taking into account that the oxidation

potential values are almost independent on the nature of the

ligand, the observed decrease of the l

maximum could be

attributed to the more negative reduction potential of the

complexes containing dialkylated pyridazine. As evidenced by

TD DFT computations

and illustrated in Figure 3, the electronic

transitions responsible for the MLCT absorption band involve as

starting orbitals the HOMO–1 and the HOMO–3 (i.e., not the

HOMO) and as final orbitals both the two low lying p

orbitals of

the diazine, LUMO and LUMO þ 1 (i.e., not only the LUMO).

However, according to the data reported in Table S4 of the

Supporting Information for the three species 0, 1, and 4, the shift in

energy associated to the mono- and the dialkyl substitution at the b

position of the pyridazine ring is similar for LUMO and LUMO þ 1

on one hand, and for HOMO, HOMO–1, and HOMO–3 on the

other. As a consequence, a correlation between l

and E

can be

evidenced (see Fig. S5, Supporting Information).

Upon excitation in the range 340–400 nm, all the complexes

show bright, broad, and featureless emission in the green–yellow

region of the visible spectrum (range 547–575 nm), at room

temperature in diluted air-equilibrated and deaerated toluene

solutions. The photoluminescence spectra for complexes 1, 4, and

6 are depicted in Figure 4 (the others are shown in Fig. S6 of the

Supporting Information) and the data are summarized in Table 2.

It has been checked that the emission is independent on the

excitation wavelength and that all the complexes are photostable.

The position of the emission is not affected by the nature of the

alkyl substituents, but only by their number: the

mono-alkylpyridazine complexes 1–3 show a l

maximum at ca.

575 nm, while the disubstituted analogues emit at ca. 550 nm.

The excited state responsible for this intense emission can be

confidently described as a triplet metal-to-ligand charge transfer

(

_{MLCT) level. The photoluminescence quantum yields (PLQY)}

are in the range 8–10% for the mono-alkylated derivatives, and 18–

22% for the dialkylated ones. Taking into account that the pydz

derivative emits at 600 nm with PLQYof ca. 5%,

it results that the

introduction in the b positions of each weak electron-donating

alkyl group roughly doubles the PLQYand causes a blue shift of ca.

25 nm. These findings suggest a sort of additive effect of the

substitution, both on the energy and on the intensity of the

emission.

To the best of our knowledge, these PLQYs represent the highest

values for neutral tricarbonyl Re(I) complexes.

Figure 3. Partial molecular orbital diagrams for the complexes [Re2

(m-Cl)2(CO)6(m-L)], L ¼ pydz (0), 4-Mepydz (1), and 4,5-Me2pydz (4). The

electronic transitions responsible for the d–d and the MLCT absorption bands are highlighted, along with the HOMO–LUMO energy gap. The values reported were computed by TD DFT in the gas phase. Computations done in the presence of solvents of different polarity result in a qualitatively similar picture (values for dichloromethane and acetonitrile can be found in Table S3 of the Supporting Information).

Figure 4. Emission spectra of complexes 1 (---), 4 (–~–) and 6 (—) in deaerated toluene at room-temperature (lex¼ 366 nm). Inset: emission

spectra of complexes 1 (---), 4 (–~–) and 6 (—) in butyronitrile glass at 77 K (lex¼ 366 nm).

FULL

P

APER

The complexes showing the highest PLQYs (4–6) have also the

longest emission lifetimes (Table 2). The values of F

/t

(¼ k

 h

, where the latter term represents the efficiency of the

intersystem crossing process) are in the range 6.3–8.1  10

s

.

The

MLCT nature of the transition involved in the emission is

supported by the dioxygen quenching of the emission (see Table 2):

on going from deaerated to aerated solutions, the lifetimes drop

from 2–5 ms to ca. 0.4 ms, independent of the mono- or dialkyl

substitution.

Low-temperature emission spectra and lifetimes were

mea-sured in butyronitrile rigid matrix at 77 K for all the complexes

(Table 2). The emission bands (shown in the inset of Figure 4 for 1,

4, and 6) maintain the structureless shape observed at room

temperature, but their maxima display pronounced hypsochromic

shifts. This rigido-chromism,

_{usually observed for emission}

from polar excited states, arises from the increase of the energy of

the

_{MLCTstates, less stabilized by the lack of solvent mobility. The}

emission lifetimes in frozen conditions (t

) are much longer than

in the fluid state (30.9–41.7 ms), as expected by lowering the

temperature. This behavior can be explained taking into account

that a strong mixing of the

MLCT and the triplet ligand centered,

_{LC, states is possible at low temperature due to the rising in energy}

of the

MLCT levels.

The best performing complex, 6, was selected for the

construction of electroluminescent devices.

2.4. Electroluminescent Devices

Neutral rhenium complexes have been rarely employed in

electroluminescent devices, either due to their very low emission

quantum yields or their scarce solubility.

Furthermore, there are

only few reports on OLEDs based on sublimation of such

emitters.

To the best of our knowledge, thus far no comparison

between devices made by sublimation and spin-coating has

appeared in the literature for any type of rhenium compounds. We

decided to compare such an electroluminescent behavior using the

best photoluminescent complex, 6, which is very soluble in organic

solvents and is stable towards sublimation. It is also interesting to

note that there is only one previous report concerning the use of a

dinuclear rhenium complex in an electroluminescent device

(which was solution-processed).

We built up polymer-based light-emitting diodes with 6 under

inert conditions, using the studied compound as dopant in a

poly(9-vinylcarbazole) (PVK) matrix. Devices were made with a

layer of PEDOT:PSS acting as a hole-transporting material. The

device structure can be represented as follows: glass/ITO/

PEDOT:PSS (80 nm)/PVK þ 6 (10 wt%) (60 nm)/TPBI (20 nm)/

Ba (5 nm)/Al (80 nm) (full details about these acronyms are given

in the Experimental Section).

The electroluminescence (EL) spectrum of the device with 6 is

depicted in Figure 5. The EL maximum is centered at 550 nm and

was only slightly shifted with respect to the photoluminescence

spectrum (Table 2), indicating that the same optical transition is

responsible for the light emission. The device showed weak voltage

dependence. With increasing voltage, the EL did not increase

dramatically while a weak emission resulting from the host

polymer matrix was observed,

suggesting that incomplete

energy transfer from the host polymeric matrix to the guest

rhenium complex takes place. The Commission International De

L’Eclairage (CIE) coordinates of the emitted light are x ¼ 0.40

and y ¼ 0.54, corresponding to a yellow emission color as observed

by eye.

Figure 6 presents the current–voltage and luminance–voltage

characteristics of the device. The turn-on voltage (V

), that is, the

voltage needed to reach 1 cd m

, is in our device configuration

about 12 V. The maximum luminance is 960 cd m

at a driving

voltage of 27 V. The value of maximum brightness observed for the

EL devices studied in this paper is higher than the 121 cd m

previously reported for an EL device based on the Re(CO)

Cl-(mopvb) (mopvb ¼ (trans-4-methyl-4

-(2-4-octadecyloxylphenyl)vin-yl)-2,2

_{-bipyridine) complex incorporated into PVK host with a}

device configuration of ITO/PVK:Re(CO)

Cl(mopvb)/BCP/LiF/

Al (BCP ¼ 2,9-dimethy-4,7-diphenyl-1,10-phenanthroline),

_and

higher than 730 cd m

for a Re complexes with 2,2

_{-bipyridine-5,5}

-diyl with triphenylamine and 1,3,4-oxadiazole moietes doped in host

material ofpolucarbonate (PC) with device struture of ITO/PVK:

Re: PC/Al.

The device showed a maximum current efficiency of 2.03 cd A

(0.34 lm W

) with a luminance of 40 cd m

at a driving voltage of

18.5 V. The values of maximum current efficiency observed for the

EL devices studied in this paper is close to the best previously

reported value (2.1 cd A

_).

_{The external quantum efficiency}

(EQE) maximum of the device is around 0.6%, a value comparable

Figure 5. EL spectra with increased voltage from 12 V to 20 V of the solution-processed devices made with complex 6.

Figure 6. TheI–V–L curves of the solution-processed device made with 6. The structure of the device is shown in the inset.

FULL

P

APER

to or even better than those reported for previous polymer-based

neutral rhenium electroluminescent devices,

_{which encourages}

further studies of similar materials.

The same complex was employed to construct an OLED by

vacuum sublimation. A light-emitting device with the

configura-tion glass/ITO/NPB (30 nm)/PCF:7 wt% 6 (30 nm)/BCP (10 nm)/

Alq

(30 nm)/LiF (10 A˚)/Al (100 nm) was fabricated for this

purpose. NPB and Alq

were utilized as hole- and

electron-transport layers, respectively; BCP was employed as a

simulta-neous electron-transport and hole-block layer. Bipolar host

material PCF

_{, which features diphenylphosphine oxide groups}

appended to a carbazole/fluorene hybrid, was used as the host

material for the rhenium complex. Figure 7 shows the

corresponding EL spectrum at a bias of 7 V, with the CIE color

coordinates of 0.26, 0.54 representing a green–yellow emission. To

the best of our knowledge the emission of this device

(l

¼ 514 nm) is the bluest ever reported for rhenium

complexes. When we increased the driving voltage to 13 V, the

CIE coordinates of the emission color remained almost constant

and the EL spectrum did not change. The current density–voltage–

luminance (I–V–L) characteristics of the PCF-based device are

depicted in Figure 8. The device has a turn-on voltage of 4.1 V and,

at a maximum luminance of 1 983 cd m

, its driving voltage was

merely 12.5 V. These values are much lower than the driving

voltages of the PVK-based device. We ascribe the reduced driving

voltage of the PCF-based device to facile hole/electron injections

within this bipolar host. The maximum current efficiency of

11.0 cd A

(6.3 lm W

) was achieved at a luminance of 27 cd m

.

The EL efficiencies of this dinuclear rhenium-based device are

amongst the highest reported for rhenium-based OLEDs. Notably,

the current efficiency of the vacuum-deposited device was 5.5

times higher than that of its spin-coated counterpart. This large

enhancement in device performance can be attributed to more

balanced charge flux and better exciton confinement within the

emission layer in vacuum-deposited device.

3. Conclusions

We have prepared a family of [Re

(m-Cl)

(CO)

(m-diazine)]

complexes, designed for optimizing the interaction energy and

then reducing radiationless deactivation pathways. Our strategy

results in air stable and highly emitting compounds. The emission

quantum yields are much higher than usually observed for neutral

Re(I) complexes. The good processability and solubility in organic

solvents of these compounds, as well as their electrochemical

properties encouraged their possible testing in electroluminescent

devices and we made two types of OLED fabricated by solution

processing and sublimation of the metal complex. The results

obtained suggest that this class of triplet emitters can be processed

in different ways and give similar emission energies for both types

of devices. The electroluminescence obtained with the sublimed

devices is one of the most efficient ever reported and such findings

suggest that rhenium complexes could play an important role for

the development of new triplet emitters for OLEDs.

4. Experimental

Synthesis: All the reactions were carried out under N2using the Schlenk

technique. All the solvents were deoxygenated and dried by standard methods before use, while commercial deuterated solvents were used as received. [ReCl(CO)5] was prepared according to a literature method [20].

4-methyl-pyridazine (Lancaster), hydrazine monohydrate (Fluka), 2-butyne (Fluka), formamidine acetate, 1-octyne, 2-hexyne, 3,3-dimethyl-1-butyne and 1-pyrrolidino-1-cyclopentene (all Aldrich) were used as received.

Synthesis of the 1,2-diazines: All the alkylated pyridazines, except the commercially available 4-methylpyridazine, were prepared according to a literature procedure, involving as first step the synthesis of 1,2,4,5-tetrazine (from hydrazine hydrate and formamidine acetate) and then its reaction with RCCR0_{alkynes (see Figure 2) [10]. For the synthesis of}

6,7-dihydro-5H-cyclopentapyridazine, tetrazine was reacted with 1-pyrrolidino-1-cyclopen-tene instead of the reactants used in the original work [10]. The purity of the ligands was checked by1H-NMR spectroscopy.

Synthesis of the Complexes 1–6: The complexes (see Schemes 1 and 3 for their structures and abbreviations) were prepared from [ReCl(CO)5],

using the method previously reported [6]. The crude products were purified by column chromatography (silica gel, CH2Cl2/n-hexane, 4:1), affording

pure yellow powders in high isolated yields (50–60%).

Characterization: 1_{H-NMR spectra were recorded on a Bruker DRX400}

spectrometer, in CD2Cl2solution, at 300K. IR spectra were acquired on a

Bruker Vector 22 FT instrument, in toluene solution. Elemental analyses were performed on a Perkin Elmer CHN2400 instrument.

[Re2(m-Cl)2(CO)6(m-4-methylpyridazine)] (1) [21]: 1H-NMR: d 2.68 (s,

3H, CH3), 7.86 (dd,1H), 9.64 (d, 1H,J ¼ 5.8 Hz), 9.67 (d, J ¼ 1.3 Hz, 1H). FT-IR n(CO) ¼ 2050 (m), 2034 (s), 1947 (s), 1915 cm
1_{(s). Anal. calcd for}

C11H6Cl2N2O6Re2: C 18.73, H 0.86, N, 3.97; found: C 19.08, H 0.79, N 4.05.

[Re2(m-Cl)2(CO)6(m-4-n-hexylpyridazine)] (2):1H-NMR: d 0.94 (t, 3H,

CH3), 1.58–1.28 (br m, 6H, CH2CH2CH2), 1.80 (quint, 2H, CH2), 2.92 (t,

2H,CH3), 7.84 (dd, 1H), 9.63 (d,J ¼ 1.2 Hz, 1 H), 9.65 (d, J ¼ 5.8 Hz, 1 H).

FT-IR n(CO) ¼ 2050 (m), 2034 (s), 1947 (s), 1915 cm
1_{(s). Anal. calcd for}

Figure 7. Electroluminescence spectra, at different bias, of a sublimation-processed device containing 6 in a PCF matrix.

Figure 8. I–V–L curves of the device made by sublimation of complex 6 in a PCF matrix. The structure of the device is shown in the inset.

FULL

P

APER

C16H16Cl2N2O6Re2: C 24.78, H 2.08, N 3.61; found: C 25.08, H 1.93,

N 3.61.

[Re2(m-Cl)2(CO)6(m-4-ter-butylpyridazine)] (3): 1H-NMR: d 1.53 (s, 9H,

C(CH3)3), 7.95 (dd, 1H), 9.67 (dd,J ¼ 6.2 Hz, 1H), 9.73 (dd, J ¼ 2.1 Hz,

1H). FT-IR n(CO) ¼ 2051 (m), 2034 (s), 1948 (s), 1916 cm
1 (s). Anal. calcd for C14H12Cl2N2O6Re2: C 22.49, H 1.62, N 3.75%; found: C 22.63, H

1.65, N 3.71.

[Re2(m-Cl)2(CO)6(m-4,5-dimethylpyridazine)] (4): 1H-NMR: d 2.56 (s,

6H, CH3), 9.50 (s, 2H). FT-IR n(CO) ¼ 2050 (m), 2033 (s), 1946 (s),

1914 cm
1(s). Anal. calcd for C12H8Cl2N2O6Re2: C 20.03, H 1.12, N 3.89;

found: C 20.49, H 1.25, N 3.84.

[Re2(m-Cl)2(CO)6(m-4 methyl-5-propylpyridazine)] (5): 1H-NMR: d 1.47

(t, 3H, CH3), 1.79 (pt, 2H, CH2), 2.59 (s, 3H, CH3), 2.86 (t, 2H,CH2), 9.45

(s, 1H), 9.49 (s, 1H). FT-IR n(CO) ¼ 2050 (m), 2033 (s), 1947 (s), 1914 cm
1(s). Anal. calcd for C14H12Cl2N2O6Re2: C 22.49, H 1.62, N 3.75;

found: C 22.75, H 1.60, N 3.65.

[Re2(m-Cl)2(CO)6(m-6,7-dihydro-5H-cyclopentapyridazine)] (6): 1

H-NMR: d 2.42 (quint, 2H,CH2), 3.27 (t, 4 H, -CH2-CH2-CH2-), 9.67 (s,

2H). FT-IR n(CO) ¼ 2050 (m), 2033 (s), 1946 (s), 1914 cm
1 (s). Anal. calcd for C13H8Cl2N2O6Re2: C 21.34, H 1.10, N 3.83; found: C 21.56, H

1.24, N 3.68.

Electrochemical Measurements: The cyclovoltammetric study of the complexes was performed at scan rates typically ranging 0.02 to 10 V s
1, in HPLC-grade CH3CN solutions at 0.00025–0.001Mconcentration in each

substrate, deareated by N2 bubbling, with tetrabutylammonium

hexa-fluorophosphate TBAPF6 (Fluka) 0.1Mas the supporting electrolyte, at

298 K. The ohmic drop was compensated by the positive feedback technique [22]. The experiments were carried out using an AUTOLAB PGSTAT potentiostat (EcoChemie, The Netherlands) run by a PC with GPES software. The working electrode was a glassy carbon (AMEL, ø ¼ 1.5 mm) cleaned by diamond powder (Aldrich, ø ¼ 1 mm) on a wet cloth (STRUERS DP-NAP); the counter electrode was a Pt wire; the reference electrode was a saturated calomel electrode (SCE), having in our working medium a difference of –0.385 V vs. the FcþjFc couple (the intersolvental redox potential reference currently recommended by IUPAC)[23] and þ0.032 V vs. the Me10FcþjMe10Fc couple (an improved

intersolvental reference under investigation) [24].

Computational Details: Geometries were optimized by means of density functional calculations. The parameter-free hybrid functional PBE0 [25] was employed along with the standard valence double-z polarized basis set 6-31G(d,p) for C, H, Cl, N, and O. For Re the Stuttgart–Dresden effective core potentials were employed along with the corresponding valence triple-zbasis set. Dissociation and interaction energy values reported in the Supporting Information (Table S2) for compound 0, 1, and 4 are counterpoise corrected [26].

In order to simulate the absorption electronic spectrum down to 230 nm the lowest 30 singlet excitation energies were computed by means of time-dependent density functional calculations. Calculations were done also in the presence of solvent (dichloromethane and acetonitrile, Table S3) described by the conductor-like polarizable continuum model (CPCM) [27]. All the calculations were done with Gaussian 03 [28].

Spectroscopy: Absorption spectra were measured with a Varian Cary 5000 double-beam UV–Vis–NIR spectrometer and baseline corrected. Steady-state emission spectra were recorded on a HORIBA Jobin-Yvon IBH FL-322 Fluorolog 3 spectrometer equiped with a 450 W Xenon arc lamp, double grating excitation and emission monochromators (2.1 nm mm
1; 1 200 grooves mm
1_{) and a Hamamatsu R928 photomultiplier tube.}

Emission and excitation spectra were corrected for source intensity (lamp and grating) and emission spectral response (detector and grating) by standard correction curves. Time-resolved measurements were performed using the time-correlated single-photon-counting (TCSPC) option on the Fluorolog 3. NanoLEDs (402 nm; FWHM < 750 ps) with repetition rates between 10 kHz and 1 MHz used to excite the sample. The excitation source were mounted on the sample chamber at 908 to a double grating emission monochromator (2.1 nm mm
1 dispersion; 1 200 grooves mm
1) and collected by a TBX-4-X single-photon-counting detector. The photons collected at the detector were correlated by a time-to-amplitude converter (TAC) to the excitation pulse. Signals were collected using an IBH

DataStation Hub photon counting module and data analysis was performed using the commercially available DAS6 software (HORIBA Jobin Yvon IBH). The goodness of fit was assessed by minimizing the reduced chi-square function (x2) and visual inspection of the weighted residuals. Luminescence quantum yields (Fem) were measured in optically dilute solution (optical

density < 0.1 at the excitation wavelength) and compared to reference emitters by the method of Demas and Crosby [29]. The fac-Ir(ppy)3complex

was used as reference in deaerated dichloromethane solution at room temperature (Fem¼ 0.40) [30]. All the solvents were spectrophotometric

grade and freshly distilled. Deaerated samples were prepared by the freeze-pump-thaw technique.

OLED Preparation and Characterization; Spin-Coated Devices: The solution processed devices were made using poly(9-vinylcarbazole) (PVK, Mw¼ 1 100 000), which was obtained from Aldrich.

Poly(3,4-ethylenedioxythiophene): poly(styrenesulfonate) (PEDOT:PSS, 1: 6 disper-sion in water, electronic grade AI4083) and 1,3,5-tris[N-(phenyl)benzimid-azole]benzene (TPBI) were purchased from HC Starck and from Sensient Imaging Technologies, respectively.

The device structure consisted of a 120 nm transparent indium tin oxide (ITO) layer as the bottom electrode, supported on a glass substrate. The ITO was treated for 10 minutes with UV/O3(UVO Cleaner 144AX, Jelight

Company) prior to any further processing. A 80 nm PEDOT:PSS layer was deposited, from a water solution of the polymer, on top of the ITO, using a spincoater P6700 from Specialty Coating Systems. The device was then annealed at 180 8C for 2 minutes. Then, the emissive layer was spin-coated from a CH2Cl2solution containing 5 mg ml
1of PVK, 10 wt% of Re complex

with respect to PVK mass. The polymer–CH2Cl2solution was stirred at room

temperature overnight before the Re complex addition. The PVK:Re complex in CH2Cl2solution was stirred for a further hour at room temperature, and

filtered through a 5 mm PTFE filter (Millex, Millipore) prior to spinning. To get a polymer layer with a thickness of 55–80 nm the solution was spin-coated (Delta6 RC spincoater from Suss Microtec) at 3 800 rpm (10 s), followed by 950 rpm (25 s). An electron-transport layer was prepared via thermal evaporation of TPBI (20 nm) at a pressure of 2.0–5.0  10
6mbar at a deposition rate of 2 A˚ s
1using a MBraun evaporation chamber. The barium electrode (5 nm) was evaporated on TPBI film in a vacuum chamber at a pressure of 2.0–5.0  10
6_{mbar at a rate of 2 A˚ s}
1_{. The aluminum top}

electrode (80 nm) was then immediately evaporated on top of the barium (5 nm) without reducing the vacuum.

All measurements were run in an inert atmosphere directly after device fabrication. Voltage scans were then performed from zero to a preset positive voltage and then back to zero. The OLEDs were characterized by attaching a computer-controlled low-noise single-channel direct-current (DC) power source that can act as both voltage source and current source, and a voltage meter or current meter (Keithley 2600, Keithley Instruments). Light from the diode was coupled to a photodiode and read out by an electrometer/high-resistance meter (Keithley 6517, Keithley Instruments). The output of the data was handled by a Labview (National Instruments)-based program. Calibration of the photodiode was done at a fixed current with a luminance meter (LS-100 Minolta). For every diode with a different spectral distribution of light, the photocurrent as measured by the photodiode was correlated to the light output in candles per square meter by this calibration. When recording an EL spectrum, a fiber-optic coupled spectrometer (USB2000, Ocean Optics) was used. The emission was corrected for the wavelength dependence of the spectrometer.

Vacuum-deposited devices: The host molecule 2,7-bis(diphenylphos-phine oxide)-9-(9-phenylcarbazol-3-yl)-9-phenylfluorene (PCF) was pre-pared using previously reported procedures [31]. The hole-transport material 4,40_{-bis[N-(1-naphthyl)-N-phenylamino]biphenyl (NPB), the}

hole-blocker 2,9-dimethy-4,7-diphenyl-1,10-phenanthroline (BCP), and the electron-transport material tris(8-hydroxyquinolinate) aluminum (Alq3)

were all purchased from LumTec Corp. and used without further purification.

The EL devices were fabricated through vacuum deposition (10
6torr) of the materials onto ITO glass (sheet resistance: 25 V square
1). All of the organic layers were deposited at a rate of 1.0 A˚ s
1. The cathode was completed through thermal deposition of LiF (10 A˚; deposition rate: 0.1 A˚ s
1_{) and then capping with Al metal (100 nm) through thermal}

FULL

P

APER

evaporation (deposition rate: 4.0 A˚ s
1_{). The current density–voltage–}

luminance relationships of the devices were measured using a Keithley 2400 source meter and a Newport 1835C optical meter equipped with an 818ST silicon photodiode. The EL spectrum was obtained using a Hitachi F4500 spectrofluorimeter.

Acknowledgements

G.D., M.P., D.D. and M.M. thank Italy’s MIUR for financial support (FIRB 2003, RBNE033KMA, Molecular compounds and hybrid nanostructured material with resonant and nonresonant optical properties for photonic devices). L.D.C and Y.S thank Ciba for some financial support. Supporting Information is available online from Wiley InterScience or from the author.

Received: April 29, 2009 Published online: July 2, 2009

[1] M. A. Baldo, M. E. Thompson, S. R. Forrest,Pure Appl. Chem. 1999, 71, 2095.

[2] P.-T. Chou, Y. Chi,Chem. Eur. J. 2007, 13, 380.

[3] C. Adachi, M. A. Baldo, M. E. Thompson, S. R. Forrest,J. Appl. Phys. 2001, 90, 5048.

[4] For OLED devices, see for instance: a) Y. Li, Y. Liu, J. Guo, F. Wu, W. Tian, B. Li, Y. Wang,Synth. Met. 2001, 118, 175. b) X. Li, D. Zhang, W. Li, B. Chu, L. Han, J. Zhu, Z. Su, D. Bi, D. Wang, D. Yang, Y. Chen,Appl. Phys. Lett. 2008,92, 83302. c) J. Li, Z. Si, C. Liu, C. Li, F. Zhao, Y. Duan, P. Chen, S. Liu, B. Li, Semicond. Sci. Technol. 2007, 22, 553. d) S. Ranjan, S.-Y. Lin, K.-C. Hwang, Y. Chi, W.-L. Ching, C.-S. Liu, y.-T. Tao, C.-H. Chien, S.-M. Peng, G.-H. Lee,Inorg. Chem. 2003, 42, 1248.

[5] For polymer light emitting devices see for instance: a) L. Qian, D. Bera, P. H. Holloway,Appl. Phys. Lett. 2007, 90, 103511. b) Y.-Y. Lu¨, C.-C. Ju¨, D. Guo, Z.-B. Deng, K.-Z. Wang, J. Phys. Chem. C 2007, 111, 5211. c) G. David, P. J. Walsh, K. C. Gordon,Chem. Phys. Lett. 2004, 383, 292. d) W. K. Chan, P. K. Ng, X. Gong,Appl. Phys. Lett. 1999, 75, 3920. e) Y. P. Wang, W. F. Xie, B. Li, W. L. Li,Chin. Chem. Lett. 2007, 18, 1501. f) B. Li, M. Li, Z. Hong, W. Li, T. Yu, H. Wei,Appl. Phys. Lett. 2004, 85, 4786.

[6] D. Donghi, G. D’Alfonso, M. Mauro, M. Panigati, P. Mercandelli, A. Sironi, P. Mussini, L. D’Alfonso,Inorg. Chem. 2008, 28, 4243.

[7] M. Panigati, D. Donghi, G. D’Alfonso, P. Mercandelli, A. Sironi, L. D’Alfonso,Inorg. Chem. 2006, 26, 10909.

[8] Instantaneous interaction energy (DEint) is the difference between the

energy of the complex (ML) and that of the constituting metal and ligand fragments (M and L), computed employing the frozen geometry of the complex. A stronger metal–ligand interaction corresponds to a more negative interaction energy value.

[9] In particular, we found a stronger metal–ligand interaction on going from pyridazine to 4-methylpyridazine to 4,5-dimethylpyridazine (DEint¼ –283,

–288 and –293 kJ mol(1, respectively). More details can be found in the Supporting Information (Table S2).

[10] J. Sauer, D. K. Heldmann, J. Hetzenegger, J. Krauthan, H. Sichert, J. Schuster,Eur. J. Org. Chem. 1998, 2885.

[11] These statements are based on the presence of symmetrical return peaks, the 57 mV half-peak widths, the nearly zero Epvs log v slopes, as well as

on the typical reversible peak morphological parameters.

[12] In the previous study it was shown that the position of the reduction peaks is well rationalized in terms of Hammett constants of the diazine sub-stituents, giving a neat linear relationship with a rather high slope, on which the novel data here acquired nicely fit (Figure S1, Supporting Information).

[13] In particular, the dEp/dlog v slopes (Table S1, Supporting Information) are

small but significant and the half-peak widths are higher than the expected 30 mV.

[14] A. J. Bard, L. R. Faulkner, inElectrochemical Methods: Fundamentals and Applications, Wiley, New York 2001.

[15] J. M. Save´ant, in Elements of Molecular and Biomolecular Electrochemistry, Wiley, New York 2006.

[16] The symmetry labels ‘‘t2g’’ and ‘‘eg’’ are here used as an approximate

description of the overall shape of the (mainly) metal-centered molecular orbitals. The actual symmetry of the species under study is lower than Oh

(in particular, it is C2vfor 0 and 4, and Csfor 1) and all the transitions

described are indeed symmetry allowed. A list of all the excitations contributing to the bands discussed in the paper for compound 0, 1, and 4 are reported in the Supporting Information (Table S3), along with their description in terms of singly excited configurations. Both gas-phase and solution values are reported.

[17] P. J. Giordano, M. S. Wrighton,J. Am. Chem. Soc. 1979, 101, 2888. [18] M. Wrighton, D. L. Morse,J. Am. Chem. Soc. 1974, 96, 998.

[19] P. J. Giordano, S. M. Fredericks, M. Wrighton, D. L. Morse,J. Am. Chem. Soc. 1978, 100, 2257.

[20] S. P. Schmidt, W. C. Trogler, F. Basolo,Inorg. Synth. 1985, 23, 41. [21] The data for this compound have already been published [6] and are

reported here for comparison.

[22] A. J. Bard, L. R. Faulkner, inElectrochemical Methods: Fundamentals and Applications, Wiley, New York 2002, pp. 648.

[23] a) G. Gritzner, J. Kuta,Pure Appl. Chem. 1984, 56, 461. b) G. Gritzner, Pure Appl. Chem. 1990, 62, 1839.

[24] a) I. Noviandri, K. N. Brown, D. S. Fleming, P. T. Gulyas, P. A. Lay, A. F. Masters, L. Phillips,J. Phys. Chem. B 1999, 103, 6713. b) J. Ruiz, D. Astruc, Comptes Rendus Acad. Sci, Se´rie IIc: Chemie 1998, I, 21. c) L. Falciola, A. Gennaro, A. A. Isse, P. R. Mussini, M. Rossi,J. Electroanal. Chem. 2006, 593, 47.

[25] C Called PBE1PBE in Gaussian: a) C. Adamo, V. Barone,J. Chem. Phys. 1999,111, 6158. b) J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 1996,77, 3865. c) J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 1997, 78, 1396.

[26] N. R. Kestner, J. E. Combariza, inReviews in Computational Chemistry, Vol. 13(Eds: K. B. Lipkowitz, D. B. Boyd ), Wiley-VCH, New York 1999, pp. 99– 132.

[27] a) V. Barone, M. Cossi,J. Phys. Chem. A 1998, 102, 1995. b) M. Cossi, V. Barone,J. Chem. Phys. 2001, 115, 4708. c) M. Cossi, N. Rega, G. Scalmani, V. Barone,J. Comput. Chem. 2003, 24, 669.

[28] Gaussian 03, Revision C.02, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, 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, C. Gonzalez, J. A. Pople, Gaussian, Inc, Wallingford CT 2004.

[29] G. A. Crosby, J. N. C. Demas,J. Am. Chem. Soc. 1970, 92, 7262. [30] A. B. Tamayo, B. D. Alleyne, P. I. Djurovich, S. Lamansky, I. Tsyba, N. N. Ho,

R. Bau, M. E. Thompson,J. Am. Chem. Soc. 2003, 125, 7377.