Softening the Donor-Set: From [Cu(P^P)(N^N)][PF₆] to [Cu(P^P)(N^S)][PF₆]

Abstract

We report the synthesis and characterization of [Cu(P^P)(N^S)][PF₆] complexes with P^P = bis(2-(diphenylphosphino)phenyl) ether (POP) or 4,5-bis(diphenylphosphino)-9,9- dimethylxanthene (xantphos) and N^S = 2-(iso-propylthio)pyridine (iPrSpy) or 2-(tert-butylthio)pyridine (tBuSpy). The single crystal structures of [Cu(POP)(iPrSPy)][PF₆] and [Cu(POP)(tBuSPy)][PF₆] have been determined and confirm a distorted tetrahedral copper(I) centre and chelating P^P and N^S ligands in each complex. Variable temperature (VT) ¹H and ³¹P{¹H} NMR spectroscopy reveals dynamic behavior with motion of the POP backbone in [Cu(POP)(iPrSPy)][PF₆] and [Cu(POP)(tBuSPy)][PF₆] frozen out at 238 K. VT NMR spectroscopic data including EXSY peaks in the ROESY spectrum of [Cu(xantphos)(tBuSPy)][PF₆] at 198 K reveal that two conformers exist in an approximate ratio of 5:1. Replacing bpy by the N^S ligands shifts the Cu⁺/Cu²⁺ oxidation to a higher potential. The copper(I) compounds are weak emitters in the solid state with PLQY values of <2%. These values are similar to those for [Cu(POP)(bpy)][PF₆] and [Cu(xantphos)(bpy)][PF₆] in the solid state.

1. Introduction

Around forty years ago, McMillin and co-workers reported the photoluminescence of the metal-to-ligand charge transfer (MLCT) state of copper(I) complexes containing both diimine (2,2′-bipyridine, bpy or 1,10-phenanthroline, phen) and PPh₃ or chelating bis(phosphane) ligands [1,2]. Since then, the field has extended to applications of [Cu(P^P)(N^N)]⁺ complexes (N^N = diimine chelates and P^P = bis(phosphane) [3,4] and to the use of copper(I) complexes containing N-heterocyclic carbenes [5,6] as emitters in light-emitting electrochemical cells (LECs). Many [Cu(P^P)(N^N)]⁺ compounds exhibit thermally activated delayed fluorescence (TADF), meaning that the energy gap between the singlet and triplet excited states is small enough to permit repopulation of the singlet from the triplet state under ambient conditions. This leads to indirect harvesting of triplet-state fluorescence, thereby improving LEC performance [7,8,9,10,11,12]. In [Cu(P^P)(N^N)]⁺ cations, the lowest unoccupied molecular orbital (LUMO) is localized on the N^N ligand and the emissive properties of these compounds are typically tuned by structural and electronic manipulation of the N^N domain [3,4,13]. The highest occupied molecular orbital (HOMO) of [Cu(P^P)(N^N)]⁺ is largely centered on copper with small contributions from phosphorus, and structural variation within the P^P domain is expected to affect steric rather than electronic properties [14], and the P–Cu–P bite angle is critical in understanding the detailed structure and properties of a [Cu(P^P)(N^N)]⁺ complex [15]. The most commonly encountered P^P ligands are 4,5-bis(diphenylphosphino)-9, 9-dimethylxanthene (xantphos) and bis(2-(diphenylphosphino)phenyl)ether (POP) (Scheme 1), both of which are commercially available.

Most investigations of copper(I) complexes for applications in LECs continue to focus on [Cu(P^P)(N^N)]⁺ compounds, and we were interested to note that relatively little attention has been paid to copper(I) complexes containing PR₃ or bis(phosphane) ligands in combination with hard-soft chelating N^S donors. A number of [Cu(PPh₃)₂(N^S)] and related compounds, in which H(N^S) is a thiosemicarbazone, have been described [16,17,18,19,20], and there are several reports of thiocyanato-bridged dicopper(I) compounds, for example [Cu₂(PPh₃)₂(2-Mepy)₂(μ-NCS)₂] [21], which present a distorted tetahedral CuP₂NS coordination sphere. Copper(I) complexes incorporating heterocyclic thioamide and phosphane ligands have also been described [22,23], and are shown to exhibit broad emission bands in the range λ^em_max = 490–495 nm (λ_exc = 270–291 nm) [21]. [Cu₂(dppdtbpf)₂(μ-NCS)₂] (dppdtbpf = 1-diphenylphosphino-1′-di-tert-butylphosphinoferrocene) exhibits a broad emission with λ^em_max ≈ 500 nm assigned to a metal-centered transition [24]. However, detailed studies of the emission behavior of [Cu(PR₃)₂(N^S)]⁺ complexes are, to the best of our knowledge, absent from the literature. We were therefore motivated to investigate a series of compounds that combined simple N^S chelates with the POP and xantphos ligands.

2. Results and Discussion

2.1. Preliminary Theoretical Investigation

Before embarking on a synthetic study, we examined the ground state electronic structure of the model [Cu(POP)(MeSPy)]⁺ cation shown in Figure 1. After geometry optimization, DFT calculations revealed a similar partitioning of orbital character in the HOMO and LUMO, as has been shown for [Cu(N^N)(P^P]⁺ (see Introduction). The LUMO of [Cu(POP)(MeSPy)]⁺ is localized on the N^S domain and largely on the pyridine ring (Figure 1a), while the HOMO displays dominant copper character with smaller contributions from the ligands (Figure 1b). These results suggested that the LUMO energy of [Cu(P^P)(N^S)]⁺ complexes may be modified by structural modification of the N^S domain.

2.2. Synthesis of Ligands and Copper(I) Complexes

For an initial investigation, we chose the N^S ligands iPrSPy and tBuSPy shown in Scheme 1. They were prepared by following literature procedures and the ¹H and ¹³C{¹H} NMR spectra (Figures S1 and S2) were in agreement with those reported [25,26]. The heteroleptic [Cu(P^P)(N^S)][PF₆] complexes were obtained by addition of the N^S ligand to a CH₂Cl₂ solution containing a 1:1 mixture of [Cu(MeCN)₄][PF₆] and the P^P ligand. The [Cu(P^P)(N^S)][PF₆] compounds were isolated as colourless solids in 41% to 67% yield. The base peak in the positive mode electrospray mass spectrum of each compound arose from the [M–PF₆]⁺ ion (see Materials and Methods Section).

2.3. Structural Characterizations

Colourless single crystals of [Cu(POP)(iPrSPy)][PF₆] and [Cu(POP)(tBuSPy)][PF₆] were grown by diffusion of Et₂O into acetone solutions of the complexes. Both compounds crystallize in the monoclinic P2₁/c space group. Figure 2 and Figure S3 depict the cations in the two compounds. In [Cu(POP)(iPrSPy)][PF₆] (Figure S3), the phenyl ring with C40 was disordered and was modelled over two sites with 65% and 35% occupancies. A disorder involving the iPrSPy ligand was modelled over two equal-occupancy sites; the S atom was common to both ligand positions. The crystal structures confirm the κ²N,S-binding modes of iPrSPy and tBuSPy and the distorted tetrahedral geometry of copper(I). Selected bond parameters are given in the captions to Figure 2 and Figure S3 and are unexceptional. The unit cell dimensions for the two compounds are comparable (see Materials and Methods Section) and the similarity between the ligand conformations in the [Cu(POP)(iPrSPy)]⁺ and [Cu(POP)(tBuSPy)]⁺ cations is seen in Figure 3. Each cation is chiral and both enantiomers are present in the unit cell. No intracation π-stacking of aromatic rings is observed in either [Cu(POP)(iPrSPy)]⁺ or [Cu(POP)(tBuSPy)]⁺. This contrasts with the numerous examples of structurally characterized salts of [Cu(POP)(N^N)]⁺ and [Cu(xantphos)(N^N)]⁺ complexes that exhibit intracation π-stacking [27]. It is noteworthy that the PM3 optimized geometry of the model [Cu(POP)(MeSPy)]⁺ cation, shown in Figure 1, also exhibited no π-stacked pairs of aromatic rings.

2.4. NMR Spectroscopic Characterization and Solution Dynamics

The solution NMR spectra of the compounds were recorded in acetone-d₆. The complexes are kinetically stable in this solvent with respect to ligand redistribution, at least for the period of data collection. Atom labelling for the NMR assignments is given in Scheme 1 and Scheme 2. The solution ³¹P{¹H} NMR spectrum of each [Cu(P^P)(N^S)][PF₆] compound at 298 K (Figure S4) showed a broadened singlet arising from POP or xantphos (δ −12.6 ppm for each POP-containing compound, and δ −13.1 and −14.4 ppm for [Cu(xantphos)(iPrSPy)][PF₆] and [Cu(xantphos)(tBuSPy)][PF₆], respectively) in addition to a septet from the [PF₆]⁻ ion. At 298 K, the solution ¹H NMR spectra (Figures S5–S8) contained sharp signals for the N^S ligands, with broadened signals from POP or xantphos being indicative of dynamic behavior. In addition, the coordinated sulfur atom (which is a stereogenic center) can undergo inversion. As in the crystallographically determined structure of the cations shown in Figure 3, the two P atoms in these [Cu(P^P)(N^S)]⁺ complexes would be inequivalent if inversion at sulfur were frozen out. Since this was never observed in the low temperature ³¹P{¹H} NMR spectra (see below), we conclude that inversion at the coordinated-S atom is a low energy process in all four complexes.

Over the temperature range 298–238 K, the ¹H NMR spectra of [Cu(POP)(iPrSPy)][PF₆] and [Cu(POP)(tBuSPy)][PF₆] remain essentially unchanged in the alkyl region as shown for [Cu(POP)(tBuSPy)][PF₆] in Figure S9. Figure 4 and Figure S10 show the aromatic regions of the variable temperature spectra of [Cu(POP)(iPrSPy)][PF₆] and [Cu(POP)(tBuSPy)][PF₆], respectively, and reveal that signals assigned to the protons of the phenyl rings, which are broad at 298 K, are partly resolved into two sets (labelled D and D’) at 238 K. Exchange (EXSY) cross-peaks are observed in the ROESY spectrum at 238 K between signals for the pairs of protons H^D2/H^D’2 and H^D3/H^D’3, as shown for [Cu(POP)(tBuSPy)][PF₆] in Figure 5. A low intensity NOESY cross-peak is observed in the ROESY spectrum between protons H^D2 and H^A6, and this is displayed for [Cu(POP)(tBuSPy)][PF₆] in Figure S11. There is no corresponding NOESY correlation between H^A6 and H^D’2, indicating that the D rings point up towards to N^S ligand while the D’ rings are directed away (see Figure 3 and Scheme 2). For a molecular weight of around 700, NOESY peaks in a ROESY spectrum are expected to have a close to zero intensity [28]. In summary, dynamic processes in [Cu(POP)(iPrSPy)][PF₆] and [Cu(POP)(tBuSPy)][PF₆] involve motion of the POP backbone that is frozen out at 238 K, while inversion at the coordinated sulfur atom is still rapid on the NMR timescale at 238 K.

The xantphos ligand has less flexibility than POP, because of the insertion of the CMe₂ bridge across the back of the ligand (Scheme 1). However, we have previously detailed dynamic processes involving the inversion of the xanthene unit in coordinated xantphos, which leads to the interconversion of, for example, conformers of [Cu(xantphos)(Phbpy)]⁺ [29] and of [Cu(xantphos)(1-Pyrbpy)]⁺ [30], (Phbpy = 6-phenyl-2,2′-bipyridine, 1-Pyrbpy = 6-(1-pyrenyl)-2,2′-bipyridine). For [Cu(xantphos)(tBuSPy)][PF₆], the broad ³¹P{¹H} NMR signal at 298 K from the xantphos ligand (δ −14.4 ppm) is replaced at 198 K by two signals at δ −16.8 and −10.4 ppm with relative integrals of 5:1 (Figure S12). This integral ratio is the same as that observed for signals arising from two species present in the ¹H NMR spectrum of [Cu(xantphos)(tBuSPy)][PF₆] at 198 K (Figure 6 for the aromatic region and Figure S13 for the alkyl region). COSY, ROESY, HMQC, and HMBC spectra recorded at 198 K were used to assign the major peaks in the spectrum, and EXSY peaks in the ROESY spectrum at 198 K were used to correlate the major and minor species. The EXSY peaks for [Cu(xantphos)(tBuSPy)][PF₆] are instructive. Figure 7 shows the aromatic region and reveals the very different environments of proton H^A6 of the pyridine ring (Scheme 2) in the two species. The EXSY peaks shown in Figure 8 verify the exchange between the tert-butyl groups of the two species, as well as the exchange between the inner and outer pointing methyl groups of the xantphos CMe₂ units in the two species. This can only occur if the xanthene unit undergoes the inversion, as shown in Scheme 3, for one pair of possible conformers. In summary for [Cu(xantphos)(tBuSPy)][PF₆], two species are present in solution at 198 K, assigned to conformers which interconvert at higher temperatures through inversion of the xanthene “bowl”; fast inversion at the coordinated sulfur atom occurs on the NMR timescale at 198 K.

2.5. Electrochemistry

Cyclic voltammetry was used to investigate the electrochemistry of the copper(I) complexes. Each undergoes a quasi-reversible or irreversible oxidation, which is assigned to a Cu⁺/Cu²⁺ process (

Table 1. Cyclic voltammetry data for [Cu(P^P)(N^S)][PF₆] complexes in CH₂Cl₂ (1–5 × 10⁻³ M, vs. Fc⁺/Fc, [ⁿBu₄N][PF₆] as supporting electrolyte, scan rate = 0.1 V s⁻¹).

Compound	E1/2ox/V	(Epc–Epa/mV)
[Cu(POP)(iPrSPy)][PF6]	+0.86	220
[Cu(POP)(tBuSPy)][PF6]	+0.96	irreversible
[Cu(xantphos)(iPrSPy)][PF6]	+1.02	220
[Cu(xantphos)(tBuSPy)][PF6]	+0.92	240

and Figures S14 and S15). For the benchmark-compounds [Cu(POP)(bpy)][PF₆] and [Cu(xantphos)(bpy)][PF₆], the Cu⁺/Cu²⁺ process occurs at +0.72 and +0.76 V, respectively [13]. Thus, replacing bpy by the N^S ligands shifts the oxidation to a higher potential. This observation results from an interplay of two opposing effects. A change from bpy to the less π-accepting N^S ligands should make Cu⁺ easier to oxidize, whereas the soft S-donor favours the copper(I) oxidation state. If the forward CV scan is taken past about +1.1 V, a second (irreversible) oxidation process, assigned to phosphane oxidation, is observed. No reduction processes were observed in the cyclic voltammograms within the solvent accessible window.

2.6. Photophysical Properties

Figure 9 displays the solution absorption spectra of the [Cu(P^P)(N^S)][PF₆] complexes. For comparison, the absorption spectra of the N^S ligands are shown in Figure S16; the absorption spectra of POP and xantphos have previously been reported [31]. The compounds are colourless, and the intense high-energy absorption bands (

Table 2. Solution (CH₂Cl₂, 2.5 × 10⁻⁵ mol dm⁻³) absorption maxima and solid-state emission data for [Cu(P^P)(N^S)][PF₆].

Complex Cation	Solution Absorption	Solid-State Emission
	λabsmax/nm (ε/L mol−1 cm−1)	λemmax/nm (λexcitation/nm)	PLQY/% (λexcitation)
[Cu(iPrSPy)(POP)]+	248 (22,600), 280 (17,600)	470 (280)	<1% (280)
[Cu(tBuSPy)(POP)]+	250 (23,600), 275 (20,700)	490 (280)	<1% (280)
[Cu(iPrSPy)(xantphos)]+	269 (24,300)	475 (280)	<2% (280)
[Cu(tBuSPy)(xantphos)]+	272 (26,600)	490 (280)	<1% (280)

) are assigned to ligand-based, spin-allowed π*←π and π*←n transitions. The compounds are non-emissive in solution, and solid samples excited at 280 nm gave emission maxima between 470 and 490 nm (Table 2 and Figure S17), blue-shifted with respect to the reference compounds [Cu(POP)(bpy)][PF₆] and [Cu(xantphos)(bpy)][PF₆] (λ_em^max = 581 and 587 nm with λ_excitation = 365 nm [13]). The solid-state photoluminescence quantum yields (PLQY) are <2% (Table 2), which is significantly lower than for many [Cu(P^P)(N^N)][PF₆] compounds, for example [13,27,29,30]. However, it is important to note that PLQY values for [Cu(POP)(bpy)][PF₆] and [Cu(xantphos)(bpy)][PF₆] are also low in the solid state (3.0% and 1.7%, respectively [13]), and that enhancement of PLQY is only achieved by tuning the substitution pattern of the bpy ligand [13,27,29].

3. Materials and Methods

3.1. General

¹H, ¹³C{¹H} and ³¹P{¹H} NMR spectra were recorded on a Bruker Avance 500 spectrometer (Bruker BioSpin AG, Fällanden, Switzerland); spectra for the ligands were recorded at 295 K, spectra for the complexes were recorded at 238 K unless otherwise stated. ¹H and ¹³C NMR chemical shifts were referenced to the residual solvent peaks with respect to δ(TMS) = 0 ppm and ³¹P NMR chemical shifts and with respect to δ(85% aqueous H₃PO₄) = 0 ppm.

Solution absorption and emission spectra were measured using an Agilent 8453 spectrophotometer (Agilent Technologies Inc., Santa Clara, CA, USA) and a Shimadzu RF-5301PC spectrofluorometer (Shimadzu Schweiz GmbH, Roemerstr., Switzerland), respectively. A Shimadzu LCMS-2020 instrument or a Bruker esquire 3000plus instrument was used to record electrospray ionization (ESI) mass spectra. Quantum yields (CH₂Cl₂ solution and powder) were measured using a Hamamatsu absolute photoluminescence (PL) quantum yield spectrometer C11347 Quantaurus-QY. Powder emission spectra were measured with a Hamamatsu Compact Fluorescence lifetime Spectrometer C11367 Quantaurus-Tau, using an LED light source with λ_exc = 280. Electrochemical measurements were carried out using a CH Instruments 900B potentiostat with [ⁿBu₄N][PF₆] (0.1 M) as the supporting electrolyte and at a scan rate of 0.1 V s⁻¹. The working electrode was glassy carbon, the reference electrode was a leakless Ag⁺/AgCl (eDAQ ET069-1) and the counter-electrode platinum wire. Final potentials were internally referenced with respect to the Fc/Fc⁺ couple (E_1/2 = 1.126 V with respect to Ag/AgCl at 298.15 K [32]).

POP and xantphos were purchased from Acros (Chemie Brunschwig AG, Basel, Switzerland) and Fluorochem (Hadfield, UK), respectively. [Cu(MeCN)₄][PF₆] was prepared by the published method [33]. The N^S ligands were prepared according to the literature and the NMR spectroscopic data matched with those reported [25,26].

3.2. [Cu(POP)(iPrSPy)][PF₆]

[Cu(MeCN)₄][PF₆] (0.23 g, 0.62 mmol) and POP (0.33 g, 0.62 mmol) were dissolved in CH₂Cl₂ (40 mL). The reaction mixture was stirred at room temperature for 2 h, then 2-((iso-propylthio)methyl)pyridine (0.10 g, 0.62 mmol) was added by syringe. Then the reaction mixture was stirred for 1 h and subsequently the solvent was removed under reduced pressure. The crude product was washed with Et₂O and crystallized by vapour diffusion of Et₂O into a CH₂Cl₂ solution of the compound. [Cu(POP)(iPrSPy)][PF₆] was obtained as colourless crystals (0.38 g, 0.41 mmol, 67%). ¹H NMR (500 MHz, acetone-d₆, 248 K) δ/ppm 8.13 (d, J = 5.1 Hz, 1H, H^A6), 8.01 (td, J = 7.7, 1.7 Hz, 1H, H^A4), 7.78 (d, J = 7.8 Hz, 1H, H^A3), 7.52–7.42 (m, 12, H^{D4+D’2+D’3+D’4}), 7.40 (m, 2H, H^C5), 7.33 (m, 4H, H^D3), 7.24 (m, 1H, H^A5), 7.14–7.00 (m, 4H, H^C4+C6), 6.98–6.84 (m, 4H, H^D2), 6.77 (m, 2H, H^C3), 4.13 (s, 2H, H^a), 2.57 (septet, J = 6.6 Hz, 1H, H^Pr-CH), 0.97 (d, J = 6.6 Hz, 5H, H^Pr-Me). ¹³C{¹H} NMR (126 MHz, acetone-d₆, 248 K) δ/ppm 158.3 (C^C1), 156.3 (C^A2), 149.8 (C^A6), 139.5 (C^A4), 134.7 (C^C3), 134.5/132.9/129.3 (C^D’2,3,4), 132.9 (C^C5), 132.7 (C^D2), 131.2 (C^D’1), 130.7 (C^D4), 130.0/125.7 (C^C6+C4), 129.5 (C^D3), 125.5 (C^D1), 125.1 (C^A3), 124.6 (C^A5), 120.8 (C^C2), 38.6 (C^a+Pr-CH), 21.9 (C^Pr-Me). ³¹P{¹H} NMR (202 MHz, acetone-d₆, 298 K) δ/ppm −12.6 (broad, FWHM = 170 Hz), −144.3 (septet, J_PF = 710 Hz). ESI-MS: m/z 768.2 [M − PF₆]⁺ (base peak, calc. 768.2). Found: C 58.97, H 4.73, N 1.61; and C₄₅H₄₁CuF₆NOP₃S requires C 59.11, H 4.52, N 1.53.

3.3. [Cu(xantphos)(iPrSPy)][PF₆]

[Cu(MeCN)₄][PF₆] (0.23 g, 0.62 mmol) and xantphos (0.36 g, 0.62 mmol) were dissolved in CH₂Cl₂ (40 mL). The reaction mixture was stirred at room temperature for 2 h, then 2-((iso-propylthio)methyl)pyridine (0.10 mg, 0.62 mmol) was added by syringe. Then, the reaction mixture was stirred for 1 h and then the solvent was removed under reduced pressure. The crude product was washed with Et₂O to yield [Cu(xantphos)(iPrSPy)][PF₆] as a colourless solid (0.32 g, 0.35 mmol, 57%). ¹H NMR (500 MHz, acetone-d₆, 238 K) Major species (see text): δ/ppm 8.11–7.98 (m, 2H, H^A6+A4), 7.88 (d, J = 7.7 Hz, 2H, H^C5), 7.73 (d, J = 7.8 Hz, 1H, H^A3), 7.53–7.39 (m, 12H, H^{D4+D’2+D’3+D’4}), 7.38–7.27 (m, 7H, H^A5+C4+D3), 6.84 (m, 4H, H^D2), 6.74 (m, 2H, H^C3), 3.74 (s, 2H, H^a), 2.60 (m, 1H, H^iPr-H), 1.95 (s, 3H H^xantphos-Me), 1.40 (s, 3H H^xantphos-Me), 0.89–0.86 (m, 6H, H^Pr-Me). ¹³C{¹H} NMR (126 MHz, acetone-d₆, 238 K) δ/ppm 156.8 (C^A2), 155.1 (C^C1), 149.8 (C^A6), 140.1 (C^A4), 134.4 (C^C6), 134.3/130.9/129.9 (C^{D4+D’2+D’3+D’4}), 132.9 (C^D2), 132.8 (C^D1), 132.0 (C^C3), 131.4 (C^D’1), 129.8 (C^D3), 129.0 (C^C5), 125.8 (C^A3+C4), 125.7 (C^A5), 120.1 (C^C2), 39.0 (C^a), 38.5 (C^Pr-CH), 36.4 (C^{xantphos bridge}), 32.0/24.9 (C^xantphos-Me), and 21.8 (C^Pr-Me). ³¹P{¹H} NMR (202 MHz, acetone-d₆, 298 K) δ/ppm −13.1 (broad, FWHM = 200 Hz), −144.3 (septet, J_PF = 710 Hz). ESI-MS: m/z 808.2 [M − PF₆]⁺ (base peak, calc. 808.2). Found: C 61.06, H 4.89, N 1.48; C₄₈H₄₅CuF₆NOP₃S requires C 60.41, H 4.75, N 1.47.

3.4. [Cu(POP)(tBuSPy)][PF₆]

[Cu(MeCN)₄][PF₆] (0.20 g, 0.62 mmol) and POP (0.30 g, 0.62 mmol) were dissolved in CH₂Cl₂ (40 mL). The reaction mixture was stirred at room temperature for 2 h, then 2-((tert-butylthio)methyl)pyridine (0.10 g, 0.55 mmol) was added by syringe. Then the reaction mixture was stirred for 1h and subsequently the solvent was removed under reduced pressure. The crude product was washed with Et₂O and crystallized by vapour diffusion of Et₂O into a CH₂Cl₂ solution of the product. [Cu(POP)(tBuSPy)][PF₆] was obtained as colourless crystals (0.21 g, 0.23 mmol, 41%). ¹H NMR (500 MHz, acetone-d₆, 238 K) δ/ppm 8.15 (d, J = 5.0 Hz, 1H, H^A6), 8.00 (td, J = 7.7, 1.5 Hz, 1H, H^A4), 7.81 (d, J = 8.0 Hz, 1H, H^A3), 7.52–7.42 (m, 10H, H^{C5+D4+D’3+D’4}), 7.42–7.37 (m, 4H, H^D’2), 7.34 (m, 4H, H^D3), 7.24–7.17 (m, 1H, H^A5), 7.16–7.06 (m, 4H, H^C4+C6), 6.97–6.86 (m, 4H, H^D2), 6.79–6.70 (m, 2H, H^C3), 4.25 (s, 2H, H^a), 1.01 (s, 9H, H^t-Bu). ¹³C{¹H} NMR (126 MHz, acetone-d₆, 238 K) δ/ppm 158.8 (C^C1), 157.4 (C^A2), 149.9 (C^A6), 139.6 (C^A4), 135.0 (C^C3), 134.7 (C^D’1), 134.5 (C^D’2), 133.2 (C^C5+D1), 133.1 (C^D’3+D’4), 132.9 (C^D2), 129.9 (C^D4), 129.6 (C^D3), 125.8 (C^C6/C4), 121.0 (C^C6/C4), 125.1 (C^A3), 124.5 (C^A5), 124.1 (C^C2), 47.8 (C^Bu-Cq), 37.7 (C^a), 29.5 (C^t-Bu) ³¹P{¹H} NMR (202 MHz, acetone-d₆, 298 K) δ/ppm −12.6 (broad, FWHM = 170 Hz), −144.3 (septet, J_PF = 710 Hz). ESI-MS: m/z 782.2 [M − PF₆]⁺ (base peak, calc. 782.2). Found: C 59.43, H 4.76, N 1.70; C₄₆H₄₃CuF₆NOP₃S requires C 59.51, H 4.67, N 1.53.

3.5. [Cu(xantphos)(tBuSPy)][PF₆]

[Cu(MeCN)₄][PF₆] (0.23 g, 0.62 mmol) and xantphos (0.32 g, 0.62 mmol) were dissolved in CH₂Cl₂ (40 mL). The reaction mixture was stirred at room temperature for 2 h, then 2-((tert-butylthio)methyl)pyridine (0.10 g, 0.55 mmol) was added by syringe. Then the reaction mixture was stirred for 1 h and subsequently the solvent was removed under reduced pressure. The crude product was washed with Et₂O to give [Cu(xantphos)(tBuSPy)][PF₆] as a colourless solid (0.28 g, 0.28 mmol, 51%). ¹H NMR (500 MHz, acetone-d₆, 198 K) major species (see text, integrals of some signals are affected by minor species and are not stated): δ/ppm 8.31 (d, J = 5.1 Hz, 1H, H^A6), 8.05 (t, J = 7.3 Hz, 1H, H^A4), 7.94 (d, J = 7.6 Hz, 2H, H^C5), 7.74–7.62 (m, H^A3+D’4), 7.60–7.49 (m, H^D’2+D’3), 7.49–7.42 (m, H^A5+D4), 7.39 (m, 2H, H^C4), 7.34 (m, 4H, H^D3), 6.88 (m, 2H, H^C3), 6.80–6.73 (m, 4H, H^D2), 3.40 (s, 2H, H^a), 2.00 (s, 3H, H^xantphos-Me), 1.27 (s, 3H, H^xantphos-Me), 0.83 (s, 9H, H^t-Bu). ¹³C{¹H} NMR (126 MHz, acetone-d₆, 198 K) δ/ppm 157.7 (C^A2), 154.7 (C^C1), 149.5 (C^A6), 139.7 (C^A4), 134.5 (C^D’4), 133.9 (C^C6), 132.3 (C^C3), 132.1 (C^D2), 131.4 (C^D’2/D’3), 130.5 (C^D4) 130.4 (C^D1), 129.6 (C^D’2/D’3), 129.4 (C^D3), 129.0 (C^C5), 125.4 (C^C4), 124.9 (C^D’1), 124.8 (C^A5), 124.5 (C^A3), 119.6 (C^C2), 47.9 (C^Bu-Cq), 36.9 (C^a), 35.9 (C^{xantphos-bridge}), 33.0 (H^xantphos-Me), 23.8 (H^xantphos-Me), 28.7 (C^t-Bu). ³¹P{¹H} NMR (202 MHz, acetone-d₆, 298 K) δ/ppm −14.4 (broad, FWHM = 175 Hz), −144.3 (septet, J_PF = 710 Hz). ESI-MS: m/z 822.2 [M − PF₆]⁺ (base peak, calc. 822.2). Found: C 60.82, H 4.95, N 1.55; C₄₉H₄₇CuF₆NOP₃S requires C 60.77, H 4.89, N 1.45.

3.6. Crystallography

Data were collected on a Bruker Kappa Apex2 diffractometer (Bruker Biospin AG, Fällanden, Switzerland) with data reduction, solution and refinement using the programs APEX [34] and CRYSTALS [35]. Structural analysis was carried out using Mercury v. 3.5.1 [36,37]. In [Cu(POP)(iPrSPy)][PF₆], the disordered rings were refined as rigid bodies, and the disordered pyridine ring had to be refined isotropically.

[Cu(POP)(iPrSPy)][PF₆]: C₄₅H₄₁CuF₆NOP₃S, M = 914.35, colourless block, monoclinic, space group P2₁/c, a = 14.0474(11), b = 17.4236(14), c = 18.3028(13) Å, β = 109.786(2)°, U = 4215.2(6) ų, Z = 4, D_c = 1.441 Mg m⁻³, μ(Cu-Kα) = 2.810 mm⁻¹, T = 123 K. Total 33404 reflections, 7650 unique, R_int = 0.028. Refinement of 7275 reflections (541 parameters) with I > 2σ (I) converged at final R₁ = 0.0743 (R₁ all data = 0.0762), wR₂ = 0.1724 (wR₂ all data = 0.1728), gof = 0.9944. CCDC 1871032.

[Cu(POP)(tBuSPy)][PF₆]: C₄₆H₄₀CuF₆NOP₃S, M = 928.37, colourless block, monoclinic, space group P2₁/c, a = 13.8102(4), b = 18.4985(6), c = 18.0974(6) Å, β = 107.7455(16)°, U = 4403.3(2) ų, Z = 4, D_c = 1.400 Mg m⁻³, μ(Cu-Kα) = 2.699 mm⁻¹, T = 123 K. Total 25698 reflections, 7939 unique, R_int = 0.030. Refinement of 6739 reflections (532 parameters) with I > 2σ (I) converged at final R₁ = 0.0530 (R₁ all data = 0.0617), wR₂ = 0.1350 (wR₂ all data = 0.1388), gof = 0.9911. CCDC 1871033.

3.7. Density Functional Theory (DFT) Calculations

Ground state DFT calculations were carried out using Spartan 16 (v. 2.0.10) [38] at the B3LYP level with a 6-31G* basis set in vacuum. Although the choice of atomic orbital basis set (6-311++G** basis set on all atoms, 6-311++G** on Cu and 6-31G* basis set on C, H and N, or 6-31G* basis set on all atoms) greatly influences the calculated absorption spectra of related copper(I) dyes, the MO characteristics are little affected [39], and therefore a 6-31G* basis set on all atoms was selected to optimize computer time. Initial geometry energy optimization was carried out at a semi-empirical (PM3) level.

4. Conclusions

We have prepared and characterized [Cu(POP)(iPrSPy)][PF₆], [Cu(POP)(tBuSPy)][PF₆], [Cu(xantphos)(iPrSPy)][PF₆] and [Cu(xantphos)(tBuSPy)][PF₆] with the aim of comparing their structures and spectroscopic properties with those of the benchmark compounds [Cu(POP)(bpy)][PF₆] and [Cu(xantphos)(bpy)][PF₆]. The single crystal structures of [Cu(POP)(iPrSPy)][PF₆] and [Cu(POP)(tBuSPy)][PF₆] confirm a distorted tetrahedral environment for copper(I) and the presence of P^P and N^S chelating ligands. Variable temperature ¹H and ³¹P{¹H} NMR spectroscopic investigations demonstrate that dynamic processes in [Cu(POP)(iPrSPy)][PF₆] and [Cu(POP)(tBuSPy)][PF₆] involve motion of the POP backbone which is frozen out at 238 K. For [Cu(xantphos)(tBuSPy)][PF₆], two conformers are present in acetone solution and at 198 K, signal integrals in both the ³¹P{¹H} and ¹H NMR spectra show that the ratio of these species is ~5:1. Exchange between the species has been confirmed through the observation of EXSY peaks in the ROESY spectrum. In all four compounds, inversion at the coordinated stereogenic sulfur atom is rapid on the NMR timescale over the temperature ranges studied. Replacing bpy by the N^S ligands shifts the Cu⁺/Cu²⁺ oxidation to higher potential. All the copper compounds are weak emitters in the solid state, but PLQY values of <2% are of the same order of magnitude as for [Cu(POP)(bpy)][PF₆] and [Cu(xantphos)(bpy)][PF₆]. Since enhancement of emission behaviour in [Cu(POP)(N^N)][PF₆] and [Cu(xantphos)(N^N)][PF₆] where N^N is a derivative of bpy is critically dependent upon ligand functionalization [13,27,29] we are motivated to further investigate [Cu(P^P)(N^S)][PF₆] complexes which show noteworthy kinetic stability with respect to ligand redistribution in acetone solutions.