Luminescent ion pairs with tunable emission colors for light-emitting devices and electrochromic switches

A class of luminescent ion pairs with tunable emission colors was designed and synthesized for light-emitting devices and electrochromic switches.


Introduction
Luminescent materials have attracted considerable attention because of their wide applications in displays, data recording and storage, chemical sensing, bioimaging, etc. [1][2][3][4][5][6][7][8][9] Emissive materials with tunable luminescence colors are of great significance for their optical applications. Up to now, conventional methods for tuning the photophysical properties of luminescent materials are mainly limited to the modication of the chemical structures, such as the change of p-conjugation skeletons or the introduction of functional groups. [10][11][12][13][14][15][16][17][18][19][20][21] However, this method oen requires complicated synthetic processes and tedious purication procedures. Stimuli-responsive luminescent materials have attracted increasing interest, because they can exhibit tunable emissive properties which are sensitive to external physical stimuli, such as light, temperature, force, electric eld, etc. [22][23][24][25][26][27][28] Among these stimuli, electric eld is an important external stimulus, which can be conveniently combined with the current semiconductor technology in electronic devices. [29][30][31][32][33] Recently, electrochromic luminescent materials that exhibited emission color changing induced by an electric eld have been reported (Fig. 1a). For example, Chidichimo and co-workers have presented p-conjugated ionic liquid crystals in which the direct electrochemical reduction leads to a reversible electrochromic luminescence (ECL) response. 34 Beneduci and co-workers have shown an ECL-active polymer gel based on the conversion of the redox states, in which the emission properties of the polymer depend on the uorophore contents and the voltages applied. 35 We have also reported a series of phosphorescent iridium(III) complexes containing a protonating functional group, which exhibited evident emission color change induced by an electric eld. [36][37][38] However, special functional groups, such as redox active or protonating groups, in these electrochromic luminescent materials may reduce their stability, especially toward redox reagents and pH values. Therefore, it is urgent to explore a new generation of electrochromic luminescent materials with controllable emission colour change. Ion pairs (IPs) consist of an anionic and a cationic component which are bonded together with electrostatic and van der Waals interactions. Employment of emissive anionic and cationic components yields luminescent IPs showing a mixed luminescence color. [39][40][41][42][43] Compared to single luminophores, IPs containing two luminescent chromophores exhibit richer excited states, which results from potential energy transfer between the two luminophores. In addition, IPs exhibit a "single-component" characteristic and can avoid phase separation in the solid state, which usually occurs in a physical mixture of two luminophores.
In this work, we aim to develop luminescent IPs with tunable emissive colors. Owing to the advantageous photophysical properties of phosphorescent iridium(III) complexes, such as high luminescence quantum yields, long emission lifetimes, large Stokes shis, high photostability, and photophysical properties sensitive to the microenvironment, a series of positively-and negatively-charged iridium(III) complexes have been chosen as the cationic and anionic components of luminescent IPs (IP1-IP6, Fig. 1b), respectively. [44][45][46][47][48][49][50][51][52][53] Upon photoexcitation, these IPs exhibited a mixed luminescence from both the cation and the anion. The emission spectra of IP1-IP6 show concentration dependence. In particular, IP6 displayed white emission at a suitable concentration in solution or solid state. Thus, in this contribution, UV-chip (365 nm) excited light-emitting diodes (LEDs) showing orange, light yellow and white emission colors were successfully fabricated using IP6 as the emitter. In addition, interesting and tunable electrochromic luminescence has been observed in solution and quasi-solid devices, and the electrochromic mechanism of these IPs has been investigated.

Photophysical properties
The photoluminescence properties of the complexes were investigated. The PL spectra of the cationic (C1Cl-C3Cl) and anionic (n-Bu 4 NA1-n-Bu 4 NA4) complexes in acetonitrile solution are presented in Fig. 1d and Fig. S3, † peaking at 524, 591, 638, 550, 595, 475 and 523 nm, respectively. For the IPs, their emissions are contributed to by both cations and anions. Interestingly, the PL spectra of the IPs exhibit concentration dependence (Fig. S1 †). Taking IP1 as an example (Fig. 1c), at a relatively low concentration of 10 À6 M, the emission peak of IP1 is around 524 nm, which originated from C1, because the quantum yield of C1 (21%) is much higher than that of A1 (17%). As the concentration of solution increases, the green emission from C1 was quenched by the anionic complex A1, due to energy transfer from C1 to A1. When the concentration of IP1 was increased to 10 À3 M, the PL spectrum is dominated by the emission from the anionic component. The emission spectra of IP1-IP6 were also measured in various solvents (THF, ethyl acetate, CHCl 3 , CH 2 Cl 2 , CH 3 OH), and there was no evident change in different solvents at the same concentration of IPs as shown in Fig. S2. † The energy transfer process between the two components of IP1 was investigated by Stern-Volmer quenching analysis ( Fig. 1d and e). The quenching study was based on a bimolecular quenching model, s 0 /s ¼ 1 + K q s 0 [Q], where s and s 0 are the emission lifetimes of n-Bu 4 NA1 with and without the quencher C1Cl, K q is the experimental quenching rate constant and [Q] is the molar concentration of the quencher. The lifetimes of n-Bu 4 NA1 in acetonitrile solution with various amounts of the quencher, C1Cl, were measured. The concentration of n-Bu 4 NA1 was kept at 1.0 Â 10 À5 M, while that of C1Cl varied from 0 to 4.0 Â 10 À4 M. The Stern-Volmer plot of the mixture of the two complexes in solution revealed a good linear relationship between s 0 /s and [Q]. The calculation yields a K q value of 1.63 Â 10 À4 M À1 s À1 . This quenching effect can be attributed to the intermolecular triplet-triplet energy transfer. 43,46 The results indicate that the energy transfer quenching process is very efficient between the two ionic complexes.

White LED devices
For IP6, the two luminescent components peak at 475 nm and 580 nm. At a concentration of 1.0 Â 10 À3 M, the ion pair shows orange emission. With the decrease of concentration in CH 3 CN, the greenish-blue emission increases gradually ( Fig. 2a and c). When the concentration decreases to 6.5 Â 10 À6 M, the solution exhibits white emission with the Commission Internationale de L'Eclairage (CIE) of (0.28, 0.30) and a quantum efficiency of 0.22. As the concentration continues to decrease, the solution shows weak blue emission.
IP6 was doped into a polyethylene-polypropylene glycol polymer to investigate the photophysical properties of the solid lm. At different doping concentrations, the lm showed different emission colors (Fig. 2b, and S4 †). At a concentration of about 1.9 mg IP6 in 800 mg polymer, the solid lm exhibited orange emission. With the decrease of the doping concentration, the solid lm showed light yellow emission gradually. Finally, when the concentration decreased to about 0.7 mg IP6 in 800 mg polymer, the solid lm showed white emission with a quantum efficiency of 0.16. IP6, therefore, is a potential candidate material for white LED (WLED).
Compared to the widely used inorganic phosphors, organic emitters have a greater advantage due to their large absorption cross section, which can reduce the phosphor consumption and cut down the costs. 59 Here, to fabricate the near-UV excited LED, powdered organic emitter IP6 was dispersed in a polyethylenepolypropylene glycol polymer and then coated onto the surface of commercially available 365 nm UV LED chips. Bright orange, light yellow and white emission were obtained when polyethylene-polypropylene glycol polymers doped with different concentrations (1.9 mg/800 mg, 0.9 mg/800 mg and 0.7 mg/800 mg) of IP6 were coated onto the UV chip as emitters. We can clearly see the bright orange, light yellow and white emissions from the prepared devices with CIE coordinate values of (0.51, 0.47), (0.40, 0.41) and (0.31, 0.36), respectively (Fig. 2d).

Electrochromic switches
Considering that IPs consist of luminescent ions with opposite charges, the electric eld is anticipated to regulate the luminescent behavior. In order to demonstrate this behavior, IP1 was used to perform the electrochromic luminescence experiment as illustrated in Fig. 3a. A concentration of 10 À5 M of the ion pair was chosen, because the emission intensity of the two peaks of IP1 was almost equal at this concentration. Two platinum electrodes were immersed in a CH 3 CN solution of IP1 with a distance of 20 mm between each other. Before applying a voltage, the solution of IP1 exhibited yellow emission. Subsequently, upon applying a voltage of 3 V onto the electrodes, the emission color of the solution near the anode changed from yellow to red within 30 s, which is in accordance with its cationic component (C1). Meanwhile, the yellow emission near the cathode changed to green, which is same as that of the anionic component (A1). Such a change of emission color gradually extends to the middle of the two electrodes, appearing as a clear boundary.
Then once the voltage was removed, the original yellow emission was recovered aer stirring the solution and the electrochromic luminescence reappeared when the stirring was stopped. This observation indicates that IP1 shows an excellent reversibility for electrochromic luminescence. To explore the mechanism of this phenomenon, the emission spectra of IP1 near the anode and cathode aer electrical stimuli are measured and shown in Fig. 3b. It was found that the emission spectra of the solution near the anode and cathode are almost the same as those of the anionic (n-Bu 4 NA1) and cationic iridium(III) complex (C1Cl), respectively. Therefore, we believe that the anionic component directionally moves to the anode and the cationic one shis to the cathode because the electrostatic interaction was broken under an electric eld, resulting in the emission color changes near the two electrodes.
To have a better understanding of these observations, a schematic model was proposed in Fig. 4e. Before applying a voltage, the IPs were disordered and uniformly distributed in the solution, and the interaction between cationic and anionic components is mainly through electrostatic force, which can be broken upon applying a voltage, leading to migration of cationic and anionic components to the cathode and anode, respectively. Aer removing the voltage and stirring the solution, the two components held together again. This mechanism was conrmed by the NMR experiments. Because of the asymmetry of the ancillary ligand in C1Cl, the two C^N ligands are non-chemically equivalent. There are two peaks attributed to the proton in the b-position with each integral of 1 (Fig. S5 †). For n-Bu 4 NA1, the two C^N ligands are chemically equivalent owing to the symmetry of the ancillary ligands CN À in the structure. Eight peaks from the C^N ligand could be observed in the NMR spectrum. The integral of the proton in the a-position is 2. In the original NMR spectrum of IP1, the integral ratio of protons at the a-position to the b-position is 2 : 1, indicating that the anion and cation are equal. Aer applying a voltage, the integral ratio in the NMR spectrum of the solution near the cathode changed to 0.56 : 1.00, while that near the anode changed to 2.00 : 0.82. These results demonstrated the migration of anions and cations to anode and cathode, respectively, under the electric eld.
To prove that the formation of IPs is necessary for realizing this electrochromism phenomenon, a control experiment using a mixed solution of n-Bu 4 NA1 and C1Cl (molar ratio 1 : 1) was carried out. Upon applying a voltage of 3 V to the electrodes for about 2 minutes, no emission color change was observed ( Fig. S6 †), which may be because the small counterions Cl À and n-Bu 4 N + moved to the anode and cathode preferentially compared to the bulky ionic iridium(III) complexes. Hence, our design strategy based on luminescent IPs is effective to realize the electrochromic luminescence.
Furthermore, color-tunable electrochromic luminescence has been realized by using other IPs (IP2-IP6, Fig. 3). For example, the emission color of IP2 can be changed from yellow to blue at the anode and orange at the cathode. For IP3, the emission color at the anode changed from red to blue, while that near the cathode remained red because of efficient energy transfer. For IP4, the emission color changed from green to blue at the anode, and remained green near the cathode. For IP5, the emission color at the cathode changed from yellow to orange, and that near the anode remained yellow. For IP6, the emission color at the cathode changed from orange to blue-green, and that near the anode remained orange.
Based on the above results, we further realized the electrochromic luminescence in a quasi-solid state. A device with a simple sandwiched structure, in which a quasi-solid lm was coated between two Pt electrodes, has been fabricated. The yellow-emitting quasi-solid lm was prepared by mixing ion pair IP1 with DMSO and SiO 2 particles (300-400 mesh). The IPs are able to freely migrate within the inorganic network channels of the SiO 2 skeleton and the phosphorescence properties of IPs would not be affected by the presence of SiO 2 particles. Under a voltage of 3 V, the luminescent color near the anode changed from yellow to red and that near the cathode was altered to green, which is consistent with the electrochromic luminescence observed in solution. When reversing the direction of the electric eld, the emission colors near the two electrodes exchanged within 20 minutes (Fig. S7 †).
Next, we carried out another experiment to further conrm that the mechanism of electrochromic luminescence is due to the migrations of the ions under an electric eld, as shown in Fig. 4a and b. At rst, a quasi-solid gel was prepared by adding several drops of DMSO into SiO 2 particles and then coated between two Pt electrodes. IP1 was doped into the middle of the gel, which showed a yellow emission. Upon applying a voltage of 3 V for 10 minutes, green emission color appeared and gradually spread to the cathode. Meanwhile, the red emission color spread to the anode gradually. Next, the direction of the electric eld was exchanged, and the emission colors near the two edges were switched. This phenomenon can be repeated. The above results demonstrated that the electrochromic luminescence is induced by the migration of the ions.
Finally, a simple sandwiched structure device, in which a quasi-solid lm (about 2 mm in thickness) doped with IP1 was coated between two ITO electrodes, has been fabricated as shown in Fig. 4c and d. At rst, a letter pattern of "IAI" was coated onto the ITO electrodes. The color of the pattern was orange. Then a voltage of 3 V was applied to the ITO electrodes for about 3-4 seconds. The emission color of the lm near the anode changed to red immediately. Meanwhile, a green color appeared near the cathode. This phenomenon can be repeated very well.

Conclusions
In summary, we have developed a class of tunable emissive materials based on luminescent ion-paired iridium(III) complexes. The photophysical properties of the IPs have been studied in detail. Orange, light yellow and white light-emitting devices were successfully fabricated by coating the polymer lms doped with IP6 on commercially available ultraviolet LEDs. Furthermore, a new strategy for the design of electricresponsive materials which display tunable and reversible electrochromic luminescence was presented. Changing the moiety of cationic or anionic iridium(III) complexes moiety can be used to tune the electrochromic luminescence colors of these ion pairs. To demonstrate the potential practical applications, a solid-lm electrochromic switch device using IPs, which showed a fast and reversible emission color change, has been fabricated successfully. These results showed that electrochromic luminescent materials based on ion pairs will be promising candidates for applications in optoelectronic elds.

General
Unless otherwise stated, all starting materials and reagents were purchased from commercial suppliers and were used without further purication. Silica particles (Aerosil 200, 12 nm) were purchased from Degussa Company (Evonik Industries AG, Rellinghauser Strabe 1-11, 45128 Essen, Germany). CH 2 Cl 2 , CH 3 CN and CH 3 OH were dried under reux over CaH 2 or sodium for several hours at 70-75 C, distilled at these conditions and used fresh. NMR spectra were recorded on a Bruker Ultra Shield Plus 400 MHz NMR instrument ( 1 H: 400 MHz, 13 C { 1 H}: 100 MHz, 19 F{ 1 H}: 377 MHz). Chemical shis (ppm) were reported relative to tetramethylsilane (TMS). Mass spectra were obtained on a Bruker autoex MALDI-TOF/TOF mass spectrometer and a ESI-MS (LCQ Fleet, Thermo Fisher Scientic). UV-Vis absorption spectra were recorded on a UV-1700 Shimadzu UV-Vis spectrophotometer. The photoluminescence spectra and emission lifetimes were recorded on an Edinburgh FL920 spectrouorometer system.

Synthesis of 3
A mixture of 2-aminothiophenol (4.0 g) and 4-hydroxybenzaldehyde (4.12 mL) in DMF (10 mL) was heated to reux under nitrogen atmosphere at 110 C for 72 h. Next, the reaction mixture was cooled down to room temperature, and then the mixture was extracted with H 2 O and ethyl acetate. The solvent was removed under reduced pressure and the product was recrystallized with ethanol.

Synthesis of 4
A mixture of 3, imidazole (200 mg), tert-butylchlorodiphenylsilane (2 mL) and DMF (4.6 mL) in a two-necked ask was heated to reux under nitrogen atmosphere at 40 C for 24 h. Next, the reaction mixture was cooled down to room temperature, and then the mixture was extracted with H 2 O and ethyl acetate. The solvent was removed under reduced pressure and the product was recrystallized with ethanol. 1

Synthesis of 5
Water (56.0 mL), anhydrous ethanol (112.0 mL) and acetic acid (112.0 mL) were added into a bottle. Then, iron powder (13.15 g, 0.235 mol) was added and the temperature was adjusted to 75 C. Lastly, benzaldehyde (5.0 g, 0.033 mol) was poured into the bottle slowly and stirred for 15 min. Then the reaction mixture was cooled down to room temperature, the crude product was ltered and the ltrate was collected, which was then extracted with CH 2 Cl 2 and water. Next, the organic phase was washed with saturated sodium bicarbonate solution (300 mL Â 3 times) and deionized water (300 mL Â 3 times), and then dried over anhydrous magnesium sulfate.

Synthesis of [(ppy) 2 Ir(m-Cl) 2 Ir(ppy) 2 ]
A mixture of IrCl 3 $3H 2 O (200.0 mg, 0.56 mmol) and 2-phenylpyridine (ppy) (173.8 mg, 0.16 mmol) in 2-ethoxyethanol/water (8.0 mL, 3 : 1 v/v) was heated to reux under nitrogen atmosphere for 24 h. Next, the reaction mixture was cooled down to room temperature, a large amount of water was added into the mixture, which was ltered. Then the residue was dried to give the desired yellow powder. Yield: 353.8 mg (59.0%). Because of the asymmetry of the ancillary ligand in C1, the two C^N ligands are non-chemically equivalent. In addition, a coupling process exists between the uorine nucleus and carbon nucleus, and we did not obtain the 13