A new class of deep-blue emitting Cu( I ) compounds – e ﬀ ects of counter ions on the emission behavior †

Three deep blue emitting Cu( I ) compounds, [Cu(PPh 3 )tpym]PF 6 , [Cu(PPh 3 )tpym]BF 4 , and [Cu(PPh 3 )tpym] BPh 4 (tpym = tris(2-pyridyl)methane, PPh 3 = triphenylphosphine) featuring the tripodally coordinating tpym and the monodentate PPh 3 ligands were studied with regard to their structural and photophysical properties. The compounds only di ﬀ er in their respective counter ions which have a strong impact on the emission properties of the powder samples. For example, the emission quantum yield can be signi ﬁ cantly increased for the neat material from less than 10% to more than 40% by exchanging BPh 4 − with PF 6 − . These e ﬀ ects can be linked to di ﬀ erent molecular packings which depend on the counter ion. In agreement with these results, it was found that the emission properties also strongly depend on the surrounding matrix environment which was elucidated by investigating photophysical properties of the compounds as powders, doped into a polymer matrix, and dissolved in a ﬂ uid solution, respectively. The observed di ﬀ er-ences in the emission behavior can be explained by di ﬀ erent and pronounced distortions that occur in the excited state. These distortions are also displayed by density functional theory (DFT) calculations.


Introduction
][3][4][5][6][7][8][9][10][11][12][13][14] A breakthrough in this field was achieved when the potential of 3 rd row transition metal complexes was recognized for these applications.0][21] As a consequence, the efficiency increases by a factor of four compared to conventional purely organic, fluorescent emitters which can only utilize singlet excitons.21][22][23][24][25][26][27][28][29] Furthermore, developing efficient and long-term stable blue light emitters with this class of compounds remains challenging due to energetically relatively lowlying metal centered dd* states that provide a path for nonradiative decay to the ground state or even molecular decomposition. 302][33][34][35][36][37][38] At first sight, emitters based on Cu(I) seem rather unsuitable for application in electroluminescent devices due to the significantly smaller SOC constant of copper compared to platinum or iridium, 39 for example.[40][41][42][43][44][45][46][47][48][49] This would lead to pronounced saturation effects, if these materials are applied as emitters in electroluminescent devices. 50owever, the long triplet decay time is not effective if the compounds exhibit only a small energy separation ΔE(S 1 -T 1 ) between the first excited singlet S 1 and the triplet state T 1 .Then, an efficient thermal population (up-ISC) of the S 1 state from the energetically lower lying T 1 state can occur at † Electronic supplementary information (ESI) available: Parts of the crystal structures and NMR spectra of 1-3 are given in Fig. S1-S11.CCDC 1415676-1415678.For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c5dt03065j ‡ These authors contributed equally to this work.ambient temperature.5][40][41][42][43][44][45]51,52 This emission mechanism corresponds to a thermally activated delayed fluorescence (TADF). 53In an electroluminescent device, this mechanism allows utilizing all injected excitons for the generation of light.Accordingly, the excitation is harvested and emitted essentially via the singlet S 1 state.54][55][56] Additionally, in Cu(I) complexes the d-shell of the copper ion is fully occupied (d 10 electron configuration) and therefore, low-lying dd* states that could quench the emission and would be a source of molecular decomposition do not occur.This renders Cu(I) complexes promising candidates for the realization of efficient and long-term stable blue light emitters for OLEDs and LEECs.

X-ray crystal structures
Single crystals suitable for X-ray structure determination could be obtained for all three investigated compounds (1, 2, and 3) from a saturated chloroform solution by layering with n-pentane at ambient temperature.The crystallographic data and structure refinement details are summarized in Table 1; selected bond distances and angles are listed in Table 2.The molecular structures of 1-3 are shown in Fig. 1.The three compounds crystallize in different crystal systems, compound 1 in the monoclinic, compound 2 in the trigonal, and compound 3 in the orthorhombic crystal system.
The copper(I) centers are coordinated by the phosphine ligand PPh 3 and the three N atoms of tpym in a distorted tetrahedral configuration.The bending of the PPh 3 ligand from the C16-Cu1 axis (compare Fig. 1) is different for the three compounds.For compound 2, there is no bending of the PPh 3 ligand (P1-Cu1-C16 = 180.0(0)°).For compound 1, with an angle of P1-Cu1-C16 = 177.6 (0)°, the PPh 3 bending is small, but for compound 3, with an angle of P1-Cu1-C16 = 172.6 (1)°, the bending is clearly displayed.Thus it seems that the bending of the PPh 3 ligand from the C16-Cu1 axis increases with the increasing size of the counter anion (BF 4 < PF 6 < BPh 4 ).The crystal packing diagram of compound 3 with the most pronounced PPh 3 bending reveals that this bending is a result of the interaction of the PPh 3 ligand with the counter anion BPh 4 (see Fig. S3, ESI †).One phenyl group of BPh 4 is directly oriented towards the PPh 3 ligand of the adjacent copper complex.The resulting steric repulsion is decreased by the bending of the PPh 3 ligand.For compounds 1 and 2 with the smaller counter anions PF 6 and BF 4 , the interaction of the PPh 3 group of the Cu(I) complex with its neighboring molecules is rather balanced.This can be supported by the performed DFT calculations (see below) which indicate that in the absence of counter ions and neighboring molecules the angle P1-Cu1-C16 amounts to 180°.
It is remarked that we have performed similar investigations for neutral Cu(I) complexes with tripodal ligands (similar to tpym) previously but with a halide (Cl, Br, or I) instead of the PPh 3 ligand at the fourth coordination site.Also for these compounds a bending of the monodentate (halide) ligand has been observed in the crystal structures. 54

Computational studies
Quantum chemical calculations have been carried out for the cationic complex [Cu(PPh 3 )tpym] + using density functional theory (DFT) and time-dependent density functional theory (TDDFT) with the hybrid functional B3LYP [85][86][87] and the basis  set def2-SVP. 88,89For [Cu(PPh 3 )tpym] + , the ground state (S 0 ), the first excited triplet state (T 1 ), and the first excited singlet state (S 1 ) geometries were calculated.The results are displayed in Fig. 2. Frequency calculations confirm that these three optimized structures are minima on the potential energy surface.
For the S 0 geometry, it was found that the atom P1 of the PPh 3 group lies on the axis that is defined by the atoms C16 and Cu1 (Fig. 2a) (angle P1-Cu1-C16 = 180°).The bond lengths of the three Cu-N bonds are almost equal, amounting to 2.137 Å (Cu1-N1), 2.137 Å (Cu1-N2), and 2.139 Å (Cu1-N3).However, in the T 1 (Fig. 2b) and S 1 state geometries (Fig. 2c), the P1 atom, and thus, the whole PPh 3 group, is bent away from the C16-Cu1 axis.For the T 1 geometry, the PPh 3 group is angled by about 26°(P1-Cu1-C16 = 153.8°)and for the S 1 geometry by about 30°(P1-Cu1-C16 = 149.7°).Also, the three Cu-N bonds are no longer equal in the excited state geometries.In the T 1 geometry, two Cu-N bonds are significantly shorter than in the S 0 geometry, with bond lengths amounting to Cu1-N1 = 2.000 Å and Cu1-N2 = 1.965Å, whereas the length of the third copper-nitrogen bond increases to Cu1-N3 = 2.167 Å.On the other hand, for the S 1 geometry two Cu-N bonds become longer, with two equal bond lengths of Cu1-N1 = 2.149 Å and Cu1-N3 = 2.149 Å, and one Cu-N bond becomes significantly shorter, with a bond length of Cu1-N2 = 1.986Å.These data and other important bond lengths and angles are summarized in Table 3.
Furthermore, it was found that the highest occupied molecular orbital (HOMO) is mainly located at the Cu(I) center whereas the lowest unoccupied molecular orbital (LUMO) is distributed over two of the three pyridine moieties of the tpym ligand (compare Fig. 3).TDDFT calculations reveal that transitions between these two frontier orbitals largely determine the first excited singlet S 1 and triplet T 1 states which leads to the assignment of these states as metal-to-ligand charge-transfer (MLCT) states.This allows us to give an explanation for the occurrence of geometry distortions in the S 1 and T 1 states: On excitation, a significant amount of charge is transferred from the Cu(I) metal center to the ligand.As a consequence, the copper center is formally partially oxidized from Cu(I) to Cu(II).As Cu(II) prefers a planar coordination environment in contrast to Cu(I) (which favors tetrahedral coordination), such an oxidation is connected with pronounced structural reorganizations.1][92] In the case of the cationic Cu(I) complex presented in this study, the distortion is represented by the bending of the PPh 3 group as described above.
The pronounced charge transfer character of the S 1 and T 1 states has another important consequence.Due to the distinct spatial separation of HOMO and LUMO, the spatial overlap between these two frontier orbitals is small.As a consequence, the exchange integral is also small.Accordingly, the energy separation ΔE(S 1 -T 1 ) between the first excited singlet and triplet state is small.From TDDFT calculations it was found that this energy separation amounts only to ΔE(S 1 -T 1 ) = 810 cm −1 .For such a small energy splitting a thermally activated delayed fluorescence (TADF) is expected to occur at ambient temperature, which is also indicated by the studies presented below.

Photophysical studies
In Fig. 4, electronic absorption spectra of the compounds [Cu(PPh 3 )tpym]PF 6 (1), [Cu(PPh 3 )tpym]BF 4 (2), and [Cu(PPh 3 ) tpym]BPh 4 (3) are displayed.In addition, the absorption spectra of the tpym and the PPh 3 ligands are also shown.All spectra were recorded under ambient conditions for the compounds dissolved in dichloromethane, except for PPh 3 for which the absorption was measured in acetonitrile. 93he absorption spectra of compounds 1-3 show similar spectral shapes.All spectra exhibit intense high energy absorption bands in the wavelength region below 280 nm with peaks at 229 nm, 248 nm, and 260 nm, respectively.The corresponding electronic transitions are identified to originate from ligand centered (LC) π-π* transitions of the tpym and the PPh 3 ligands, which show intense absorptions in this spectral range.At longer wavelengths, distinctly weaker absorption bands are observed which are not present in the spectrum of the tpym or PPh 3 ligands.Therefore, these bands can be attributed to transitions that are of metal-to-ligand charge transfer (MLCT) character.43][44][45]51,52 It is not surprising that the absorption spectra of all three compounds exhibit similar spectral shapes as the investigated substances differ only in their respective counter anions.In a dilute solution, the counter anions do not interact with the Cu(I) complexes and therefore, do not influence the absorption behavior.Furthermore, the BF 4 − and PF 6 − ions are not expected to show absorption in the investigated wavelength range.In contrast, the phenyl groups of BPh 4 − show distinct absorption in the wavelength range between 230 nm and 260 nm which explains the differences in the spectra of substance 3 when compared to those of 1 and 2.
In fluid dichloromethane (DCM) solution, the Cu(I) complexes are not emissive at ambient temperature, even if oxygen is carefully removed from the solution by repeatedly applying a freeze-pump-thaw process.In contrast, emission is observed for complexes doped into a polymethylmethacrylate (PMMA) matrix.The corresponding emission spectra are displayed in Fig. 5 and are found to be essentially identical for all the compounds.They are broad and featureless, being in agreement with the MLCT character of the emissive state, [31][32][33][34][35]40,[63][64][65] with a peak at λ max = 470 nm. Also, th emission quantum yield of Φ PL = 7% is equal for all the compounds (compare Table 4).These results show that the counter anions do not have an influence on the emission behavior of the Cu(I) complexes Fig. 4 Absorption spectra of the three compounds 1-3 and the tpym ligand recorded in dichloromethane solution under ambient conditions.PPh 3 absorption has been measured under the same conditions but in acetonitrile solution.93 doped into the PMMA matrix which presumably is due to the spatial separation of the counter anions and the Cu(I) complex in the PMMA matrix at low doping concentrations.The relatively small value of Φ PL is related to the non-rigid environment in the PMMA polymer.This allows significant quenching of the excited states as explained below.Interestingly, for the powder samples with different counter anions one finds distinct differences in the emission properties.The spectra of the three compounds are clearly shifted relative to each other with emission maxima lying at λ max (1) = 466 nm, λ max (2) = 449 nm, and λ max (3) = 452 nm (Fig. 5).Also, the emission quantum yields vary distinctly amounting to Φ PL (1) = 43%, Φ PL (2) = 19%, and Φ PL (3) = 7%, at ambient temperature.
Presumably, these differences are a consequence of the interaction of the counter anions with the Cu(I) complex due to their proximity in the solid phase.In the crystals of the compounds, these interactions are displayed in different packings and different geometries of the [Cu(PPh 3 )tpym] + complex, depending on the counter anion (compare crystal structures discussed above and ESI Fig. S1-S3 †).For example, the orientation of the PPh 3 group compared to the rest of the molecules is strongly dependent on the counter anion and thus, influences the emission properties.5][96][97] It is reasonable to assume that in the powder phase, the counter anions have similar effects on the compounds' geometries.
A comparison of the emission properties in different matrices reveals an interesting trend.In the powder phase, compound 1 exhibits an emission quantum yield of 43%, but when doped into a PMMA matrix the quantum yield amounts only to 7% and in solution, the quantum yield is ≪1%.This is related to increasing distortions that Cu(I) complexes undergo on excitation with decreasing matrix rigidity. 9,31,40,55,60As already discussed in the previous section, the compounds show structural distortion on excitation, especially, a bending of the PPh 3 group away from the C16-Cu1 axis.As a consequence, the potential energy curves for the excited and the ground states are shifted with respect to each other.Accordingly, the non-radiative rate increases due to an increase of the Franck-Condon factors that govern these deactivation processes to the ground state. 9,31,98,99In matrices with low rigidity, such as fluid solutions, this effect is particularly pronounced.In more rigid environments, geometry distortions are much less distinct resulting in higher quantum yields.In powder, geometry distortions upon excitation are partly suppressed resulting in higher quantum yields of the Cu(I) compounds.Interestingly, here it is shown that also the counter anion (and the resulting molecular packing) has an influence on the emission quantum yield of the powder samples.Thus, the counter anions prevent large geometry distortions if they limit the available space for distortions in the molecular packing.
It is remarked that the emission spectra of the compounds as powders show a clear blue shift of the order of 10 nm on heating from T = 77 K to ambient temperature (1: 540 cm −1 , 2: 620 cm −1 , 3: 480 cm −1 ).This is an indication that a thermally activated delayed fluorescence (TADF) occurs.According to this emission mechanism, the singlet state S 1 is populated at ambient temperature and contributes to the emission (Fig. 6).As this state lies energetically higher, a blue shift of the emission with increasing temperature is expected.The energy separation of ΔE(S 1 -T 1 ) = 810 cm −1 , as found from TDDFT calculations, is in fairly good agreement with the experimental values resulting from the spectral shifts and therefore supports this assignment.Furthermore, the thermal population of the singlet state should result in an increase of the radiative rate on heating as the spin-allowed S 1 → S 0 transition carries significantly more allowedness than the spin-forbidden T 1 → S 0 transition.Indeed, on heating, a slight increase of the radiative rate from k r (77 K) = 2.7 × 10 4 s −1 to k r (300 K) = 3.1 × 10 4 s −1 is observed for compound 1.However, this increase is smaller than that observed for other TADF materials.Presumably, this is mainly due to the fact that the triplet state itself exhibits comparably fast deactivation to the ground state with a decay time of only 26 µs for compound 1.]41

Conclusion and outlook
In this contribution, we have presented a new class of emitter materials based on a Cu(I) complex with a tripodal ligand.As powders, the three compounds [Cu(PPh 3 )tpym]X (X = PF 6 (1), BF 4 (2), and BPh 4 (3)) display bright emission in the deep-blue range of the spectrum.For example, at ambient temperature, the emission of [Cu(PPh 3 )tpym]PF 6 (1) peaks at 466 nm and the emission quantum yield amounts to 43%.However, if this compound is doped into a polymer PMMA matrix, the emission is slightly red-shifted to 470 nm, but the quantum yield is drastically reduced to 7%.In fluid dichloromethane solution, the compound is not emissive.This indicates that molecular reorganizations on excitation which can easily occur in nonrigid environments are a major source of non-radiative relaxation to the ground state.For other Cu(I) compounds, it has been shown that limiting such distortions can lead to drastic increases of the emission quantum yields even in solution.For example, for the cationic complexes [Cu(POP)(dmbpy)] + and [Cu(POP)(tmbpy)] + (POP = bis[2-(diphenylphosphino)-phenyl]ether, dmbpy = 4,4′-dimethyl-2,2′-bipyridyl, tmbyp = 4,4′,6,6′-tetramethyl-2,2′-bipyridyl) 40 it has been demonstrated that the introduction of two sterically demanding methyl groups can significantly reduce geometry distortions on excitation and therefore, cause a drastic increase of the emission quantum yield by almost a factor of ten.This strategy could also be applied in future investigations to increase the quantum yield of the compounds presented in this study, for example, by introducing sterically demanding groups that prohibit the bending distortion of the triphenylphosphine group.Furthermore, we have demonstrated that different counter ions and molecular packings can have a strong impact on the emission behavior of solid samples.For example, the powder of the compound [Cu(PPh 3 )tpym]PF 6 exhibits an emission quantum yield of 43% and an emission maximum of 466 nm at ambient temperature, whereas for [Cu(PPh 3 )tpym]BPh 4 the emission is blue shifted to 452 nm and the emission quantum yield is decreased to 7%.These results should be taken into consideration, if solid state samples of other Cu(I) complexes are being investigated.

General remarks
Syntheses and handling of air-and moisture-sensitive substances were carried out using standard Schlenk-and glovebox-techniques.Solvents were dried using standard procedures 100 and stored over Al 2 O 3 /molecular sieves 3 Å/R3-11G catalyst (BASF).The starting materials were obtained from commercial sources and used as received.The following materials were prepared according to literature procedures: Copper(I) chloride, 101 tetrakis(acetonitrile)copper(I) hexafluorophosphate, 102 tetrakis(acetonitrile)copper(I) tetrafluoroborate, 103 and tris(2-pyridyl)methane (tpym). 84MR spectra were recorded at 300 K on a Bruker DPX 250, Bruker ARX 300, Bruker DRX 400, Bruker ARX 500, or Bruker DRX 500 using CDCl 3 or CD 3 CN as the solvent.Chemical shifts are given with respect to tetramethylsilane ( 1 H, 13 C) and 85% phosphoric acid ( 31 P).Calibration of 1 H and 13 C NMR spectra was accomplished with the deuterated solvent residual signals. 31P NMR spectra were calibrated externally ( phosphoric acid).The numbering of the hydrogen and carbon atoms of the three compounds is shown in Fig. 7.
Electrospray ionization (ESI) mass spectra were recorded on a Thermo Fisher Scientific LTQ FT Ultra using acetonitrile as the solvent.IR spectra were recorded on a Bruker Alpha FT-IR spectrometer using powder samples at ambient temperature.Intensities of the bands were characterized as follows: vs = very strong (0-50% transmission), s = strong (50-70% transmission), m = medium (70-90% transmission), w = weak (90-100% transmission).Elemental analysis was performed on an Elementar vario MICRO cube.
UV-Vis absorption measurements were carried out using a Varian Cary 300 double beam spectrometer.Emission spectra were recorded with a Fluorolog 3-22 (Horiba Jobin Yvon) spectrophotometer which was equipped with a cooled photo-multi- plier (RCA C7164R).For the decay time measurements, the same photomultiplier was used in combination with a FAST ComTec multichannel scaler PCI card with a time resolution of 250 ps.As the excitation source for the decay time measurements, a pulsed diode laser (Picobrite PB-375L) with an excitation wavelength of λ exc = 378 nm and a pulse width <100 ps was used.For absolute measurements of photoluminescence quantum yields at ambient temperature, a Hamamatsu Photonics (C9920-02) system was applied.Doping of polymethylmethacrylate (PMMA) films was performed by dissolving the respective complex (<1 wt%) and the polymer in dichloromethane.After this, the solution was spin-coated onto a quartz-glass plate.All calculations were carried out with Gaussian09. 104As the functional B3LYP [85][86][87] and as the basis set def2-SVP 88,89 were used.As the starting geometry for optimization, a structure obtained from X-ray measurements was used.The optimization of the S 0 and T 1 structures was made by DFT methods and the optimization of the S 1 structure was made by TDDFT methods.No symmetry constraints were applied.Vibrational frequency calculations confirm that all three optimized structures are minima on the potential energy surface.
The data collection for the single crystal structure determinations was performed on a Bruker D8 QUEST diffractometer by the X-ray service department of the Faculty of Chemistry, University of Marburg.The D8-QUEST is equipped with a Mo-K α X-ray microsource (Incotec), a fixed chi goniometer and a PHOTON 100 CMOS detector.Bruker software (Bruker Instrument Service, APEX2, SAINT) was used for data collection, cell refinement and data reduction. 105The structures were solved with SIR-97, 106 refined with SHELXL-2014 107 and finally validated using PLATON 108 software, all within the WinGX 109 software bundle.Absorption corrections were applied within the APEX2 software (multi-scan). 105Graphic representations were created using Diamond 3. 110 C-bound H-atoms were constrained to the parent site.In all graphics the displacement ellipsoids are shown for the 50% probability level, hydrogen atoms are shown with an arbitrary radius.CCDC 1415676-1415678 contain the supplementary crystallographic data for the structures reported in this paper.

Syntheses
General procedure for compounds 1 and 2. The ligand tpym was dissolved in a minimum amount of acetonitrile and [Cu(MeCN) 4 ]X (X = PF 6 or BF 4 ) was added.After stirring for 10 min, PPh 3 was added to the solution.The reaction solution was stirred at room temperature for 1 d.The solvent was removed in vacuo and the crude product was triturated with diethyl ether (2× with 10 mL).Single crystals were obtained in chloroform by layering with n-pentane.

Fig. 2
Fig. 2 Optimized geometries of the ground state S 0 (a), the first excited triplet state T 1 (b), and the first excited singlet state S 1 (c) of [Cu(PPh 3 ) tpym] + .Calculations were performed on the B3LYP/def2-SVP level of theory.All hydrogen atoms are omitted for clarity.

Fig. 3
Fig. 3 Highest occupied (HOMO) and lowest unoccupied molecular orbitals (LUMO) for [Cu(PPh 3 )tpym] + calculated for the ground state geometry.Calculations were performed on the B3LYP/def2-SVP level of theory.All hydrogen atoms are omitted for clarity.

Fig. 5 Table 4
Fig. 5 Normalized emission spectra of the investigated compounds 1-3 doped into a PMMA matrix and as powders.All spectra were recorded under ambient conditions.The samples were excited at λ exc = 350 nm.

Fig. 6
Fig. 6 Emission decay paths for compound 1 at different temperatures.At low temperatures (T = 77 K), only the lowest excited triplet state T 1 emits, while at ambient temperature (T = 300 K), an additional radiative decay path via thermal population of the S 1 state is opened.