Peripheral halo-functionalization in [Cu(N^N)(P^P)]+ emitters: influence on the performances of light-emitting electrochemical cells

A series of heteroleptic [Cu(N^N)(P^P)][PF6] complexes is described in which P^P = bis(2-(diphenylphosphino)phenyl)ether (POP) or 4,5-bis(diphenylphosphino)-9,9-dimethylxanthene (xantphos) and N^N = 4,4’-diphenyl-6,6’-dimethyl-2,2’-bipyridine substituted in the 4-position of the phenyl groups with atom X (N^N = 1 has X = F, 2 has X = Cl, 3 has X = Br, 4 has X = I; the benchmark N^N ligand with X = H is 5). These complexes have been characterized by multinuclear NMR spectroscopy, mass spectrometry, elemental analyses and cyclic voltammetry; representative single crystal structures are also reported. The solution absorption spectra are characterized by high energy bands (arising from ligand-centred transitions) which are red-shifted on going from X = H to X = I, and a broad metal-to-ligand charge transfer band with λmax in the range 387–395 nm. The ten complexes are yellow emitters in solution and yellow or yellow-orange emitters in the solid-state. For a given N^N ligand, the solution photoluminescence (PL) spectra show no significant change on going from [Cu(N^N)(POP)] to [Cu(N^N)(xantphos)]; introducing the iodo-functionality into the N^N domain leads to a red-shift in λmax em compared to the complexes with the benchmark N^N ligand 5. In the solid state, [Cu(1)(POP)][PF6] and [Cu(1)(xantphos)][PF6] (fluorosubstituent) exhibit the highest PL quantum yields (74 and 25%, respectively) with values of τ1/2 = 11.1 and 5.8 μs, respectively. Light-emitting electrochemical cells (LECs) with [Cu(N^N)(P^P)][PF6] complexes in the emissive layer have been tested. Using a block-wave pulsed current driving mode, the best performing device employed [Cu(1)(xantphos)] and this showed a maximum luminance (Lummax) of 129 cd m −2 and a device lifetime (t1/2) of 54 h; however, the turn-on time (time to reach Lummax) was 4.1 h. Trends in performance data reveal that the introduction of fluoro-groups is beneficial, but that the incorporation of heavier halo-substituents leads to poor devices, probably due to a detrimental effect on charge transport; LECs with the iodo-functionalized N^N ligand 4 failed to show any electroluminescence after 50 h.


Experimental
General 1 H, 13 C and 31 P NMR spectra were recorded using a Bruker Avance III-500 NMR spectrometer. 1 H and 13 C NMR chemical shifts were referenced to the residual solvent peaks with respect to δ(TMS) = 0 ppm and 31 P NMR chemical shifts with respect to δ(85% aqueous H 3 PO 4 ) = 0 ppm.Solution absorption and emission spectra were measured using an Agilent 8453 spectrophotometer and a Shimadzu RF-5301PC spectrofluorometer, respectively; a Bruker esquire 3000plus instrument was used to record electrospray ionization (ESI) mass spectra.Quantum yields (CH 2 Cl 2 solution and powder) were measured using a Hamamatsu absolute photoluminescence (PL) quantum yield spectrometer C11347 Quantaurus-QY.Emission lifetimes and powder emission spectra were measured with a Hamamatsu Compact Fluorescence lifetime Spectrometer C11367 Quantaurus-Tau, using an LED light source with λ exc = 365 nm.Quantum yields and PL emission spectra in thin films were recorded using a Hamamatsu absolute quantum yield C9920.The preparation of the thin film samples consisted of deposition on a quartz plate (1 cm 2 ) of the complex with addition of the ionic liquid 1-ethyl-3-methylimidazolium hexafluoridophosphate [Emim][PF 6 ].
Compounds 1-5 were prepared using reported methods [45][46][47] and the NMR spectroscopic data matched with those reported.POP was purchased from Acros and xantphos from Fluorochem.[Cu(MeCN) 4 ][PF 6 ] was prepared by the published method. 48Cu( 1   (5 ml).A suspension of xantphos (59 mg, 0.1 mmol) and 1 (37.2 mg, 0.10 mmol) in CH 2 Cl 2 (5 ml) was added and the mixture turned red and then orange while it was stirred for 2 h at RT.The solution was filtered, the solvent was removed and the crude material was ground to a powder, washed with hexane (2 × 15 ml) and dried under vacuum to give  (30 ml) and the mixture was stirred for 1.5 h at room temperature.Compound 5 (84.1 mg, 0.25 mmol) was added and the mixture turned orange as it was stirred for another 2 h.Additional POP (26.9 mg, 0.05 mmol) was added; stirring was continued for another 1 h during which the solution turned yellow.After filtration, the solvent was removed from the filtrate; the solid residue was washed with hexane (2 × 30 ml) and Et 2 O (7 × 30 ml) and dried under vacuum to give [Cu(3) (POP)][PF 6 ] (130 mg, 0.12 mmol, 48%) as a yellow solid. 1

Crystallography
Data were collected on a Bruker Kappa Apex2 diffractometer with data reduction, solution and refinement using the programs APEX 49 and CRYSTALS. 50Structural analysis was carried out using Mercury v. 3.7. 51,52

Device preparation
LECs were prepared on pre-patterned indium tin oxide (ITO) covered glass substrates.The substrates were previously cleaned using subsequent sonication with soap, deionized water and 2-propanol.After drying with a N 2 flow, the substrates were placed in a UV Ozone cleaner (Jelight 42-220) for 20 minutes.An Ambios XP-1 profilometer was used to determine the layer thickness.Following, 80 nm of poly(3,4ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) (CLEVIOS™ P VP AI 4083, aqueous dispersion, 1.3-1.7%solid content, Heraeus) was coated in order to avoid the formation of pinholes and to improve the reproducibility of the cells.Subsequently, the emitting layer was deposited by spin-coating from a 2-butanone solution of the emitting compound with the addition of the ionic liquid 1-ethyl-3-methylimidazolium hexafluoridophosphate [Emim][PF 6 ] (>98.5%,Sigma-Aldrich) in a 4 to 1 molar ratio.The active layer spin conditions were optimized to form 100 nm thick films.70 nm of aluminium, acting as the top electrode, were thermally evaporated onto the device using an Edwards Auto500 evaporator integrated into an inert atmosphere glovebox (<0.1 ppm O 2 and H 2 O, MBraun).The active areas of the devices are 6.5 mm 2 .The devices were then characterized under inert atmosphere conditions at room temperature.

Device characterization
The devices were measured by applying a pulsed current ( J = 50 A m −2 , 1 kHz, 50% duty cycle) and monitoring the voltage and luminance versus time by using a True Colour Sensor MAZeT (MTCSiCT sensor) with a Botest OLT OLED Lifetime-Test system.The electroluminescence (EL) spectra were measured using an Avantes AvaSpec-2048 fiber optic spectrometer during device measurements.The turn-on time (t max ) is defined as the time to reach the maximum luminance (Lum max ) and the lifetime (t 1/2 ) is the time to reach one-half of Lum max after this value is attained.

Electrochemistry
The electrochemical behaviour of the [Cu(N^N)(P^P)][PF 6 ] complexes was investigated using cyclic voltammetry and data are given in Table 1 and a typical CV is shown in Fig. 8.The single oxidation wave for each complex is attributed to the copper-centred oxidation process.For the POP-containing complexes, this process is irreversible whilst with xantphos, it is quasi-reversible provided that the potentials do not rise above ∼+1.1 V (Fig. 8).The presence of the halo-substituents has a negligible effect on E ox 1=2 .The values of E ox 1=2 ∼ +0.9 V are compared with +0.82 and +0.81 V for [Cu(6,6′-Me 2 bpy) (POP)][BF 4 ] and [Cu(6,6′-Me 2 bpy)(xantphos)][BF 4 ] (MeCN solution, vs. Fc/Fc + ). 33Reduction processes were poorly defined within the solvent accessible window.Complex cation  maximum shown in Fig. 9 from 266 nm to 278 nm, the lowest energy being the complex containing the iodo-functionality.The same trend is observed for the [Cu(N^N)(POP)][PF 6 ] series (λ max shifts from 266 to 278 nm, Fig. S4 †) and reflects that observed for the free N^N ligands. 45The broad absorption in the range 387-395 nm is attributed to metal-to-ligand charge transfer (MLCT).2).Complexes containing N^N ligands 1, 2, 3 or 5 undergo a blue shift on comparing the solution with the solid state (except for complexes with ligand 5 in thin film).This is consistent with other [Cu(N^N)(POP)] + emitters, 22,24,30 although the opposite trend has been observed when N^N = 2,2′:6′,2″-terpyridine. 58The PLQY of degassed solutions of [Cu(N^N)(POP)][PF 6 ] are generally higher than that of the xantphos-containing analogues (Table 2).However, in the solid state, due to packing interactions, luminescence properties can be significantly influenced by the type of substituent; this is more prevalent in powders than in thin films.An explanation for this behaviour has previously been proposed on the basis of the flattening that the pseudo-tetragonal geometry of the complexes experience while passing from the electronic ground state (S 0 ) to the emitting excited state. 44This flattening, which is more favoured in a fluid medium, is hindered in the crystalline state (powder) and is partially hindered in thin films.The highest PLQY values are exhibited by the fluoro-functionalized complexes; powdered [Cu(1)(POP)][PF 6 ] has a PLQY = 74% with a lifetime τ 1/2 = 11.1 μs (Tables 2 and  S1 †), while the thin film has a PLQY = 13%.The range of values of τ 1/2 (Table 2) is similar to that observed for [Cu(N^N) (POP)][PF 6 ] and [Cu(N^N)(xantphos)][PF 6 ] complexes in which N^N = 6-methyl-2,2′-bipyridine, 6-ethyl-2,2′-bipyridine, 6-phenyl-2,2′-bipyridine or 6,6′-dimethyl-2,2′-bipyridine, 22,24 and for [Cu(6,6′-Me 2 bpy)(POP)][BF 4 ] in PMMA thin-films. 33he solid-state emission data indicate that the introduction of the remote fluoro-substituent enhances PL (compare complexes with N^N = 1 versus 5), but that replacement of the fluorine atom by a heavier congener in ligands 2, 3 or 4 is detrimental.

Electroluminescence
The electroluminescence behaviour of the complexes was tested by incorporating them into LEC devices.For LEC characterization, the turn-on time (t max ) is defined as the time to reach the maximum luminance (Lum max ).The time to reach one-half of the maximum luminance is referred to as t 1/2 (the device lifetime).The devices were operated using a block-wave    pulsed current driving mode (as described in the Experimental section), which was selected in order to enhance the device response.Under these conditions, the voltage required to maintain the current density decreases versus time due to the formation of p-and n-doped regions, which reduces the resistance of the active layer.The electroluminescence (EL) spectra recorded for the LECs showed maxima in the 565-585 nm range (yellow emission) for all complexes (Fig. S9 †).
On the one hand, LECs containing [Cu(4)(POP)] + , [Cu(4) (xantphos)] + and [Cu(3)(POP)] + , contain the complexes with iodo-and bromo-substituted N^N ligands, did not show any EL after 50 hours (Fig. S10-S12 †).However, the LEC with [Cu(3)(xantphos)] + (which also has a bromo-functionalized N^N) showed EL, although this is rather low.The device characteristics for this LEC are depicted in Fig. S13.† This LEC showed a fast t max (10 s), although a rather low Lum max of 10 cd m −2 .Moreover, the device exhibited a luminance decay, and hence a poor device lifetime (t 1/2 = 4.3 min).These results seem to indicate that the attached bromo or iodo atoms have a detrimental effect on the device performances.In view of the weak EL for one bromo-containing complex, it would appear that introducing an iodo-group leads to poorer performances than a bromo-group.On the other hand, LECs containing [Cu(1)(POP)] + , [Cu(1)(xantphos)] + , [Cu(2)(POP)] + , [Cu(2)(xantphos)] + , [Cu(5)(POP)] + and [Cu(5)(xantphos)] + exhibited a typical LEC behaviour under bias.This consists of an increase of luminance accompanied by a fast decrease of the voltage.The luminance and voltage behaviours are graphically depicted in Fig. 12 and S14, † respectively, and the performance parameters are summarized in Table 3.
The t max was reached in 6.5, 4.1, 3.2, 1.5, 0.1 and 5.  3).These results, therefore, indicate that the best LECs are obtained from complexes with the xantphos ligand instead of  ).b λ exc = 365 nm.c Biexponential fit using the equation τ 1/2 (av) = ∑A i τ i /∑A i where A i is the pre-exponential factor for the lifetime (see Table S1).d Thin films consisted of the [Cu(N^N)(P^P)][PF 6 ] complex mixed with the ionic liquid (IL) 1-ethyl-3-methylimidazolium hexafluoridophosphate in a molar ratio 4 : 1 (complex : IL).View Article Online the devices that employ complexes with POP ligands.This is in agreement with the results presented in previous reports. 22,24rom a comparison of the devices, a trend relating to peripheral halo-substituents can be determined.LECs containing complexes with a fluoro-substituted N^N ligand (LECs with [Cu(1)(POP)] + and [Cu(1)(xantphos)] + ) performed better than those with analogous chloro-functionalized ligands, which is consistent with the tendency for devices containing the iodo-ligand 4 or the bromo-ligand 3 to perform poorly.Hence, the results show that the LEC performance is strongly influenced by the attached halogen atoms, and improves on going from iodo-to fluoro-functionalization which is in line with the trend for PLQY in thin-films (see Table 2) for each series (1 > 2 > 3 > 4).Hence, the observed trend suggests that PLQY is the limiting factor for the electroluminescence behavior of the devices when the ligands are functionalized with I or Br.However, we note that although [Cu(4)(xantphos)] + exhibits a PLQY of 6% in thin films, the LEC containing [Cu(4)(xantphos)] + did not show any electroluminescence whereas this is not the case for [Cu(2)(POP)] + with a PLQY of 4% (thin film).Moreover, [Cu(3)(xantphos)] + with a PLQY of 10% shows a rather poor performance in LECs.The origin of this behaviour remains unclear, but our results are consistent with the detrimental effect observed in LECs employing [Ir(C^N) 2 (N^N)] + emitters with a bromophenyl unit in the 4-position of the bpy ligand, which has been reported before. 59his effect has not been studied in depth.

Conclusions
We report a series of [Cu(N^N)(POP)][PF 6 ] and [Cu(N^N) (xantphos)][PF 6 ] complexes in which N^N is either the benchmark ligand 5 (Scheme 1) or is functionalized on the periphery with a halo-substituent (ligands 1-4).

Fig. 7
Fig. 7 Arrangement of [Cu(5)(xantphos)] + cations along the 6-fold screw axis; the accommodation of the xantphos domain within the cavity created by the twisted 4,4'-diphenylbpy unit of a neighbouring cation is shown in space-filling representation.

Fig. 12
Fig. 12 Luminance for glass/ITO/PEDOT:PSS/active layer/Al devices measured by applying a block-wave pulsed current of 50 A m −2 at a frequency of 1 kHz and a duty cycle of 50%.The active layer consisted of different [Cu(N^N)(P^P)][PF 6 ] complexes mixed with the ionic liquid 1-ethyl-3-methylimidazolium hexafluoridophosphate.
The complexes have been fully characterized by mass spectrometry, solution NMR spectroscopy, and cyclic voltammetry.The single crystal structures of several of the complexes confirm the expected distorted tetrahedral environment of the Cu(I) centre, and the chelating nature of the N^N and P^P ligands.The solution absorption spectra are characterized by high energy bands arising from ligand-centred transitions; these bands are red-shifted on going from [Cu(5)(P^P)][PF 6 ] to [Cu(1)(P^P)][PF 6 ] (1 contains the fluoro-substituent).A characteristic MLCT band appears around 390 nm for each heteroleptic complex.[Cu(N^N)(POP)][PF 6 ] and [Cu(N^N)(xantphos)][PF 6 ] complexes are yellow emitters in solution whilst their powdered samples emit in the yellow or yellow-orange region.Changing the P^P ligand while retaining the same N^N domain has little effect on the solution PL spectrum.Going from [Cu(5)(P^P)][PF 6 ] to [Cu(1)(P^P)][PF 6 ] leads to a red-shift in λ max em .In the solid state, [Cu(1)(POP)][PF 6 ] and [Cu(1)(xantphos)][PF 6 ] (fluoro-substituent) exhibit the highest PL quantum yields (74 and 25%, respectively) with values of τ 1/2 = 11.1 and 5.8 μs, respectively.The ten complexes have been tested in the LEC configuration.LECs with the iodo-functionalized ligand 4 did not show any electroluminescence after being under bias for 50 h.An overview of the performance data demonstrates that the introduction of the fluoro-groups is beneficial, and the best performing device employed [Cu(1)(xantphos)] + (Lum max = 129 cd m −2 and device t 1/2 = 54 h); however, a long turn-on time of 4.1 h was observed.We propose that the poor performance of LECs with chloro-or bromo-substituents 59 relates to their lower PL quantum yield in thin films on going from fluoro-to iodofunctionalized ligands.

Table 1
Cyclic voltammetric data for [Cu(N^N)(P^P)][PF 6 ] complexes referenced to internal Fc/Fc + = 0. V; CH 2 Cl 2 (freshly distilled or degassed HPLC grade) solutions with [ n Bu 4 N][PF 6 ] as the supporting electrolyte and a scan rate of 0.1 V s −1 .Processes are quasi-reversible unless otherwise stated (ir = irreversible)

Table 3
Performance parameters obtained for glass/ITO/PEDOT:PSS/active layer/Al devices by applying a block-wave pulsed current of 50 A m −2 at a frequency of 1 kHz and duty cycles of 50%.All copper complexes in the emissive layers are [PF 6 ] − salts