Desymmetrizing Heteroleptic [Cu(P^P)(N^N)][PF6] Compounds: Effects on Structural and Photophysical Properties, and Solution Dynamic Behavior

The preparation, characterization and electrochemical and photophysical properties of a series of desymmetrized heteroleptic [Cu(P^P)(N^N)][PF6] compounds are reported. The complexes incorporate the chelating P^P ligands bis(2-(diphenylphosphanyl)phenyl)ether (POP) and (9,9-dimethyl-9H-xanthene-4,5-diyl)bis(diphenylphosphane) (xantphos), and 6-substituted 2,2′-bipyridine (bpy) derivatives with functional groups attached by –(CH2)n– spacers: 6-(2,2′-bipyridin-6-yl)hexanoic acid (1), 6-(5-phenylpentyl)-2,2′-bipyridine (2) and 6-[2-(4-phenyl-1H-1,2,3,triazol-1-yl)ethyl]-2,2′-bipyridine (3). [Cu(POP)(1)][PF6], [Cu(xantphos)(1)][PF6], [Cu(POP)(2)][PF6], [Cu(xantphos)(2)][PF6], and [Cu(xantphos)(3)][PF6] have been characterized in solution using multinuclear NMR spectroscopy, and the single crystal structure of [Cu(xantphos)(3)][PF6].0.5Et2O was determined. The conformation of the 6-[2-(4-phenyl-1H-1,2,3,triazol-1-yl)ethyl]-substituent in the [Cu(xantphos)(3)]+ cation is such that the α- and β-CH2 units reside in the xanthene ‘bowl’ of the xantphos ligand. The 6-substituent desymmetrizes the structure of the [Cu(P^P)(N^N)]+ cation and this has consequences for the interpretation of the solution NMR spectra of the five complexes. The NOESY spectra and EXSY cross-peaks provide insight into the dynamic processes operating in the different compounds. For powdered samples, emission maxima are in the range 542–555 nm and photoluminescence quantum yields (PLQYs) lie in the range 13–28%, and a comparison of PLQYs and decay lifetimes with those of [Cu(xantphos)(6-Mebpy)][PF6] indicate that the introduction of the 6-substituent is not detrimental in terms of the photophysical properties.

Improving the photophysical properties of [Cu(PˆP)(NˆN)] + complexes can be approached by structural modification of either the PˆP or NˆN domains. Typically, the PˆP ligand is a wide bite-angle bisphosphane such as bis (2-(diphenylphosphanyl)phenyl)ether (POP) or (9,9-dimethyl-9H-xanthene-4,5-diyl)bis(diphenylphosphane) (xantphos) (Scheme 1) and these commercially available ligands remain the most popular choices. While structure-property relationships may be developed [12][13][14], enhanced PLQY is most often the result of trial-anderror structural variation of the NˆN ligand. Synthetically, it is easier to vary the functionalities in the latter than in the PˆP domain. It has also been shown that intramolecular π-stacking interactions in [Cu(PˆP)(phen)] + and [Cu(PˆP)(4,7-Ph 2 phen)] + (4,7-Ph 2 phen = 4,7-diphenyl-1,10phenanthroline) lead to increased PLQY values because of inhibition of the flattening of the coordination sphere in the excited state [15]. In addition to enhancing photophysical behaviour, one of the challenges in the design of [Cu(PˆP)(NˆN)] + emitters is to minimize the tendency for ligand-redistribution reactions. One approach has been to use macrocyclic ligands to produce pseudorotaxanes [16]. The covalent linkage of the PˆP and NˆN domains is an attractive way forward but appears to have been little explored. This approach will involve the use of longer chains as linkers, and strategies that might be developed involve condensation reactions between appropriate functionalities on the PˆP and NˆN domains, or the use of click chemistry. In this paper, we consider the consequences on the structural, dynamic and photophysical properties of copper(I) iTMCs of the desymmetrization caused by the introduction of model linker chains at the 6-position of a bpy (ligands 1-3 in Scheme 1). We have previously demonstrated that the introduction of 6-methyl, 6-ethyl, 6-phenyl or 6-phenylthio substituents is beneficial to the photophysical properties of [Cu(POP)(bpy)] + and [Cu(xantphos)(bpy)] + derivatives [17][18][19]. However, to the best of our knowledge, little is known about the effects of introducing longer, and potentially sterically demanding, 6-substituents. Scheme 1. Structures of the PˆP and NˆN ligands with labelling for the NMR assignments.

Ligand Syntheses
Compound 1 has previously been reported and was prepared according to the published method [20]. The route shown in Scheme 2 was used to synthesize compound 2. The first step was lithiation of the methyl group in 6-methyl-2,2 -bipyridine (6-Mebpy) using lithium diisopropylamide (LDA) prepared in situ. The intermediate was then treated with 1-bromo-4-phenylbutane to give 2. The base peak at m/z 303.10 in the electrospray (ESI) mass spectrum arose from the [M + H] + ion. The 1 H-and 13 C{ 1 H}-NMR spectra were assigned using 2D methods, and HMQC and HMBC spectra are shown in Figures S1 and S2 in the Supplementary Materials. Compound 3 was prepared by adapting a literature procedure [21] and the synthetic route is summarized in Scheme 3. The base peak (m/z 328.08) in the ESI mass spectrum corresponded to the [M + H] + ion, and the 1 H and 13 C{ 1 H} NMR spectra (assigned using COSY, NOESY and HMQC and HMBC methods) were fully in accord with the structure shown in Scheme 3. Figure 1 illustrates the 1 H NMR spectrum of 3, and additional spectra are presented in Figures S3 and S4

Synthesis and Characterization of Heteroleptic Copper(I) Complexes
[  6 ] . 0.5Et 2 O were grown from a CH 2 Cl 2 of the compound layered with diethyl ether. The compound crystallizes in the triclinic space group P-1. Disorder of the phenyl ring of the coordinated ligand 3 (see the experimental section) meant that this part of the structure was refined isotropically, as seen in the ORTEP representation in Figure S5 in the Supplementary Materials. Table 1 presents important bond lengths and angles, and P-C bond lengths are typical, lying in the range 1.827(3) to 1.834(3) Å. Atom Cu1 in the [Cu(xantphos)(3)] + cation is in a distorted tetrahedral environment, with the largest angle in the coordination sphere being P2-Cu1-N2 = 118.78(7) o . Figure 2a,b show two views of the [Cu(xantphos)(3)] + cation, and is noteworthy that the 6-substituent lies over and to one side of the 'bowl' shaped cavity of the xanthene; this is relevant to the subsequent NMR spectroscopic discussion. The space-filling diagram in Figure 2c illustrates that the conformation of the chain is such that is desymmetrizes the structure. It is tempting to suggest that this is associated with a stacking of the triazole ring over one arene ring of the xanthene. However, the distance between the ring-centroids is 4.07 Å indicating that this is, at best, a weak π-interaction. Figure 2b reveals that two of the PPh 2 phenyl rings are aligned to give a π-stacking interaction. The angle between the ring planes is 10.0 • , and the centroid...centroid distance is 3.85 Å, making this an effective interaction and one that is typical in many [Cu(xantphos)(NˆN)] + complexes [19]. Figure  S6 illustrates this interaction and also highlights the proximity of bpy proton H A6 (see Scheme 1) to the stacked rings which is pertinent to the NMR spectroscopic discussion below.  Attempts to grow X-ray quality crystals of the remaining four complexes were unsuccessful, and we therefore modelled one of the POP-containing complexes in order to gain information about the structural relationship between the substituent attached to the bpy domain, and the POP ligand. The structure of [Cu(POP)(1)] + was minimized, first at a molecular mechanics (MM2) level, and this geometry was used as an input to a DFT level energy minimization (B3LYP 6-31G*) [22]. Figure

NMR Spectroscopic Properties and Dynamic Behaviour
The 1 H-and 13 C{ 1 H}-NMR spectra of the complexes were recorded in acetone-d 6 6 ], the signal for bpy H A6 is broadened, and presumably this is associated with the proximity of this proton to two phenyl rings of the PPh 2 groups of the xantphos ligand (see Figure S6). The third notable distinction between the 1 H spectra is the splitting of the signals for the phenyl rings into two sets (labelled D and D'), a phenomenon more pronounced in [Cu(xantphos)(1)][PF 6 ] than [Cu(POP)(1)][PF 6 ]. The difference is also apparent in the HMQC spectra, expansions of which are displayed in Figure 5.  Non-dissociative dynamic processes to be considered include: (i) rotation of the phenyl rings around P-C bonds which has a very low energy barrier; (ii) conformational change of the 6-substituent from the left-to the right-side of the complex as defined in Scheme 4a; (iii) the conformational change of the POP backbone (Scheme 4b), or conformational change of the xanthene unit in xantphos (Scheme 4c). We may assume that process (i) occurs in all the complexes at 298 K. Processes (ii) and (iii) may be coupled or independent, and the series of complex cations described here reveals different scenarios. First, we compare  6 ], in addition to a NOESY crosspeak to H B5 (bpy), the protons in the CH 2 group attached to the bpy unit (H a in Scheme 1) show a NOESY crosspeak with H D2 , but not with H D2 . The data are consistent with the 6-substituent undergoing the conformational change shown in Scheme 4a. This in not coupled with inversion of the xanthene unit (i.e., the process in Scheme 4c is not fast on the NMR timescale at 298 K). In contrast, in the NOESY spectrum of [Cu(POP)(1)][PF 6 ] (Figure 6b), EXSY peaks are observed between H D2 and H D2 , indicating that phenyl rings D and D' exchange, but the process in slow enough on the NMR timescale at 298 K that signals for phenyl groups in two chemically different environments remain distinct. Significantly, the signal for the CH 2 group attached to bpy in [Cu(POP)(1)][PF 6 ] is an overlapping doublet of doublets (Figure 7b), indicating that these protons are diastereotopic. Figure 7b shows part of the NOESY spectrum of [Cu(POP) ( 1) (Figure 8a). Critically, there are EXSY peaks between the two methyl groups of the CMe 2 unit of xantphos (Figure 8b) confirming that the xanthene unit undergoes a conformation change (Scheme 4c) which is slow on the NMR timescale at 298 K. The NMR spectroscopic data are consistent with this being combined with the 'flip' of the 6-substituent (Scheme 4a). It is possible that this is associated with the presence of the aromatic triazole unit in the middle of the substituent chain (see structural discussion and Figure 2).

Electrochemical and Photophysical Properties
The electrochemical behaviour of the copper(I) compounds was investigated using cyclic voltammetry and Table 2 [23]. Although the copper oxidation for each of the compounds reported in Table 2 6 ], and are consistent with the steric demands of the 6-substituents hindering the flattening of the copper coordination sphere which accompanies oxidation from Cu + to Cu 2+ . Each compound also undergoes a reversible bpy-centred reduction, and Figure S24   The solution absorption spectra of the heteroleptic compounds are displayed in Figure 9 and absorption maxima are given in Table 3. Absorptions below ca. 330 nm arise from ligand-centred transitions. These regions of the spectra of the two POP-containing compounds are similar, and those of the xantphos-containing complexes are also comparable, with the more intense high-energy bands for [Cu(xantphos)(3)][PF 6 ] being consistent with the presence of the triazole unit. Each of the five complexes exhibits a low-intensity, broad absorption with λ max in the range 381-384 nm which is assigned to metal-to-ligand charge-transfer (MLCT).  When de-aerated solutions of the compounds are excited into the MLCT band (λ exc = 365 nm), they emit very weakly in the orange region. All solution PLQYs were <1%. As has been described for related heteroleptic [Cu(PˆP)(NˆN)] + complexes [18], the emissions are assigned to dπ(Cu)→π*(bpy) ( 3 MLCT) transitions. The emissions gain in intensity upon going from solution to powdered samples and we focus only on the solid-state data. The emission bands are unstructured ( Figure 10) and values of λ max em are given in Table 4. PLQY values range from 13% for [Cu(POP)(2)][PF 6 ] to 28% for [Cu(xantphos)(1)][PF 6 ]. As Table 4 shows, the xantphos-containing compounds have higher PLQY values than those in which PˆP is POP. A biexponential fit was used for the lifetime decays (see Table 4), and values of τ were all of the same order of magnitude (5.  [18], indicating that the switch from a 6-methyl to longer chain substituent is not detrimental to the photophysical properties.  a A biexponential fit to the lifetime decay was used because a single exponential gave a poor fit; τ is calculated from τ = τ = √ ∑ A i τ i / ∑ A i where A i is the pre-exponential factor for the lifetime; values of τ(1), τ(2), A 1 and A 2 are given in the right-hand columns of the table.

Compound 2
The reaction was carried out in flame-dried glassware on a Schlenk line under an N 2 atmosphere. Diisopropylamine (152 mg, 0.212 mL, 1.50 mmol, 1.0 eq.) was dissolved in dry THF (5 mL) and the mixture was cooled to −78 • C. n-Butyllithium in hexanes (408 mg, 0.600 mL, 2.5 M, 1.50 mmol, 1.0 eq.) was added and the reaction mixture was stirred for 1 h. A solution of 6-methyl-2,2 -bipyridine (93.2 mg, 0.25 mmol, 1.0 eq.) in dry THF (10 mL) was added and the mixture turned dark blue. After the reaction mixture was stirred for another 3 h, a solution of 1-bromo-4-phenylbutane in dry THF (10 mL) was added and the mixture was allowed to warm up to room temperature overnight keeping the vessel submerged in the cooling bath. Then, saturated aqueous NH 4 Cl (15 mL) was added and the mixture was extracted with CH 2 Cl 2 (3 × 35 mL). The combined organic fractions were dried over MgSO 4 and filtered. After the solvent was removed under reduced pressure, the residue was purified by flash column chromatography using basic alumina to give 6-(5-phenylpentyl)-2,2 -bipyridine (2) (138 mg, 0.46 mmol, 30%) as a colourless oil. For the chromatography, three columns were run: first column: cyclohexane/ethyl acetate, gradient from 3 to 20% ethyl acetate in cyclohexane over 16

Crystallography
Single crystal data were collected on a Bruker APEX-II diffractometer with data reduction, solution and refinement using the programs APEX [27] and CRYSTALS [28]. The program CSD Mercury 2020.1 [29] was used for the structure analysis and structural figures. SQUEEZE [30] was used to treat the solvent region, and the electron density removed equated to half a molecule of Et 2 O per complex cation. The anion in [Cu(xantphos)(3)][PF 6 ] . 0.5Et 2 O was orientationally disordered and was modelled over two positions with fractional occupancies of 0.70 and 0.30. The phenyl ring in ligand 3 was also orientationally disordered over two positions, and the rings had to be refined isotropically and with rigid body restraints.

Conclusions
We have prepared and characterized two new bpy ligands, 2 and 3, which contain  6 ] . 0.5Et 2 O, and this confirmed the expected distorted tetrahedral copper(I) coordination environment. The 6-substituent is oriented so that the αand β-CH 2 units reside in the xanthene 'bowl' of the xantphos ligand, and the conformation of the chain is such that is desymmetrizes the structure. This has implications for the interpretation of the solution NMR spectra of the five complexes, and analysis of the 2D spectra provides evidence for different combinations of possible dynamic processes operating in different compounds. Each copper(I) complex exhibits a broad MLCT absorption band with λ max in the range 381-384 nm, and excitation into this band results in a very weak, orange emission in solution. In the solid state, the heteroleptic complexes exhibit emission maxima between 542 nm and 555 nm, and PLQY values range from 13% for [Cu(POP)(2)][PF 6 ] to 28% for [Cu(xantphos)(1)][PF 6 ]. These quantum yields are not significantly lower than that of [Cu(xantphos)(6-Mebpy)][PF 6 ] and the decay lifetimes of the new compounds are also similar to that of the analogous 6-Mebpy containing derivative. These results demonstrate that going from a 6-methyl to longer-chain substituent is not unfavourable in terms of the photophysical properties.  Sample Availability: Samples of the compounds are not available from the authors.