Naphthalimide-Modified Tridentate Platinum(II) Complexes: Synthesis, Characterization, and Application in Singlet Oxygen Generation

Abstract

Singlet oxygen (¹O₂), representing an important reactive oxygen species, has promising applications in biomedical, material, and environmental sciences. Photosensitized production of ¹O₂ using organic dyes is highly desirable and the exploration of highly efficient photosensitizers has received considerable attention. Herein, two tridentate Pt(II) complexes, i.e., cationic 1(PF₆) and neutral 2, modified with the ethynylnaphthalimide chromophore, were designed and prepared for the application in ¹O₂ generation. Spectroscopic studies and computational results suggest that 1(PF₆) and 2 display the lowest-energy absorption bands centered at 435–465 nm with the molar extinction coefficients of 0.6–3.2 × 10⁴ M⁻¹ cm⁻¹, originating from the singlet ligand-to-ligand charge transfer (¹LLCT) and a mixture of ¹LLCT and singlet ligand-centered (LC) transitions, respectively. Moreover, they show similar phosphorescence at 620–640 nm assigned to the Pt-perturbed triplet LC emission of the ethynylnaphthalimide moiety. Thanks to the relatively long phosphorescence lifetimes, these complexes exhibit O₂-dependent phosphorescence intensities with good reversibility and stability. They are able to behave as efficient triplet photosensitizers to promote the ¹O₂ generation with high quantum yields (84–89%). This work indicates that the combination of an organic chromophore with Pt(II) complexes provides an effective method to obtain photosensitizers for ¹O₂ generation.

1. Introduction

Singlet oxygen (¹O₂), the lowest excited state of the dioxygen molecule, represents an important reactive oxygen species and has attracted great interest in biomedical, material, and environmental sciences due to its extensive applications in photodynamic therapy (PDT), fine chemical synthesis, wastewater treatment, etc. [1,2,3,4]. Up until now, a number of chemical and photochemical methods have been developed for the in situ generation of ¹O₂. Among them, photosensitized production of ¹O₂ using various dyes (also called photosensitizers) is more prevalent [5,6,7]. As briefly illustrated in Figure 1a, upon photoexcitation, the photosensitizer (PS) is first converted to its lowest singlet excited state (S₁) followed by rapid transition to the lowest triplet excited state (T₁) via fast intersystem crossing (ISC). If effective collision of PS with surrounding O₂ takes place at this moment, the ground triplet-state O₂ (³Σ_g) is subsequently excited to generate ¹O₂ (¹∆_g) through energy transfer (EnT) from the T₁ of PS to the ³Σ_g state of O₂ [8,9,10]. For an ideal PS, several characteristics should be met, including strong absorption, high ISC efficiency and long triplet excited-state lifetime (τ_T), good photostability, etc. [11,12].

Over the past few decades, several kinds of classical PSs have been explored and adequately investigated, such as aromatic hydrocarbons, quinones, porphyrins, phthalocyanines, borondipyrromethenes, and transition metal complexes [13,14,15,16,17,18,19,20]. Among them, transition metal complexes, especially Ru(II), Ir(III), Pt(II), and Au(I) complexes, exhibit advantages including large ISC rates benefiting from the strong spin-orbit coupling (SOC), rich coordination modes, and excellent structural modification abilities, endowing them with potential photosensitizing function [21,22,23,24]. In view of the larger SOC constant (ζ = 4481 cm⁻¹) of the Pt atom relative to other transition metals and the fascinating photophysical properties of related complexes [25,26,27], Pt(II) complexes become one of the most promising PSs. A certain number of Pt(II) complexes have been reported to display high quantum yields of ¹O₂ generation (Φ_∆). Nonetheless, some drawbacks, e.g., short absorption wavelengths (≤450 nm), low molar absorptivities (≤10⁴ M⁻¹ cm⁻¹), and short triplet excited-state lifetimes, are found in these complexes. Therefore, continuous studies on these complexes to further improve their photosensitizing performances are necessary [28,29,30,31,32,33].

In order to address the issues mentioned above, one straightforward approach is attaching a suitable light-harvesting chromophore with strong absorptivity to the metal component though a π-conjugated bond [11,33,34,35,36]. For instance, Zhao and coworkers prepared bidentate (N^N)-Pt(II) complex 3 with two ethynylnaphthalimide ligands, displaying a prolonged τ_T of up to 64 μs (in N₂-saturated solution) and higher Φ_∆ of 89% in CH₃CN relative to those (τ_T = 0.98 μs; Φ_∆ = 38%) of a Pt(II)-ethynylbenzene analogue [35]. Similarly, they reported two tridentate (N^N^C)- and (N^N*C)-Pt(II) complexes, 4 and 5, modified with an ethynylnaphthalimide ligand to present a moderate Φ_∆ of 57% and 65% in CH₃CN, respectively [36]. Herein, two tridentate Pt(II) complexes, cationic 1(PF₆) and neutral 2 distinguished by an N^N^N and cyclometalated N^N^C coordination mode, respectively, are presented. Both of them contain an appended ethynylnaphthalimide moiety as an additional chromophore to modulate their photophysical properties. Spectroscopic measurements demonstrate that 1(PF₆) and 2 show strong visible absorptions extending to ca. 550 nm and improved phosphorescence lifetimes. Meanwhile, their distinct responses to triplet oxygen (³O₂) accompanied by the dramatic decrease in the phosphorescence intensities and highly efficient in situ generation (Φ_∆ = 84–89%) of ¹O₂ via EnT are clearly detected and fully discussed.

2. Results and Discussion

2.1. Synthesis and X-ray Single-Crystal Diffraction Analysis

As outlined in Scheme 1, two Pt(II) complexes, 1(PF₆) and 2, were prepared with the one-step ancillary ligand exchange reaction of 2-dodecyl-6-ethynyl-1H-benzo[de]isoquinoline-1,3(2H)-dione (NI) with the cationic precursor [(tpy)PtCl]Cl (tpy represents 2,2′:6′,2″-terpyridine) and the neutral precursor (pbpy)PtCl (pbpy denotes 6-phenyl-2,2′-bipyridine), with the assistance of KOH and a catalytic amount of CuI in a 76% and 88% yield, respectively. Their molecular structures were fully characterized with proton nuclear magnetic resonance (¹H NMR) spectroscopy, high-resolution mass spectrometry (HRMS), and an elemental analysis (details are given in the Section 3). The introduction of the long alkyl chain (n-C₁₂H₂₅) into the NI ligand and the resulting complexes can significantly improve their solubilities in common organic solvents, and thus facilitate the separation and purification of these materials.

A single crystal of 2 suitable for the X-ray diffraction analysis was successfully obtained with slow diffusion of ethyl ether into the solution of 2 in CH₂Cl₂. The relevant crystallographic data are summarized in Table S1 and the molecular structure and stacking model are shown in Figure 2 and Figure S1. The planar Pt component and the ethynylnaphthalimide fragment are linked together with a large dihedral angle of ca. 81° via a Pt-C σ-bond with a length of 1.97 Å (Pt1-C70). This means that these two components are almost perpendicular to each other. The Pt-C σ-bond of Pt with the cyclometalating ligand (Pt1-C10) has a slightly longer length of 2.063 Å. The two Pt-N bonds have a length of 2.000 Å (Pt1-N8; with the middle pyridine of the tridentate ligand) and 2.062 Å (Pt1-N20; with the side pyridine), respectively (Figure 2a). These bond lengths are consistent with those of known tridentate Pt(II)-acetylene analogues [37,38]. Additionally, complex 2 crystallizes in a triclinic P$\overline{1}$ space group with two molecules being included in one unit cell. In the crystal packing, a type of chain-like supramolecular polymer is formed via the alternate π–π stackings of two intermolecular naphthalimide moieties with a distance of 3.47 Å and two cyclometalated Pt components with a distance of 3.305 Å, respectively. No effective metallophilic interaction is present in the crystal structure as the shortest detected intermolecular Pt···Pt distance is about 5.11 Å, which is much longer than the van der Waals contact distance of 3.5 Å of two Pt atoms (Figure 2b and Figure S1).

2.2. Steady-State Absorption and Emission Spectroscopies

UV-Vis absorption and steady-state emission properties of 1(PF₆) and 2 were first investigated in 1,2-dichloroethane (DCE) with a concentration of 5 × 10⁻⁵ M (Figure 3 and

Table 1. Photophysical parameters ¹.

	Compound	1(PF6)		2		NI
Parameter		DCE	CH3CN	DCE	CH3CN	DCE
λabs (nm)/ε(104 M−1 cm−1)		465/0.6 389/2.4 348/2.1 286/3.1	435/1.3 385/2.6 343/2.1 282/3.4	437/3.2 412/2.9 335/2.0 282/4.6	440/1.7 405/1.3 330/0.9 279/2.1	366/2.0 350/2.2
λem (nm)		620	618	640	636	400
τ (μs) air/N2		0.6/4.8	0.27/3.2	0.72/7.0	0.24/1.65	3 × 10−4/--
Φ 2 (%) air/N2		0.3/1.9	--/0.4	0.3/1.9	0.1/0.9	8.1/--
kr 3 (103 s−1) air/N2		5.0/4.0	--/1.2	4.2/2.7	4.2/5.4	2.7 × 105/--
knr 4 (105 s−1) air/N2		16.6/2.0	--/3.1	13.8/1.4	41.6/6.0	30.6 × 103/--
kq 5 (108 M−1 s−1)		9.1	18	7.8	19	--
ΦΔ 6 (%)		--	89	--	84	--
PT,O2 7 (%)		--	92	--	85	--
fT,Δ 8 (%)		--	97	--	99	--

¹ Monitored in specific solution with the concentration of 5 × 10⁻⁵ mol L⁻¹. The mark of “--” denotes “No determined”. ² Absolute quantum yield. ³ k_r = Φ/τ. ⁴ k_nr = (1 − Φ)/τ. ⁵ Phosphorescence quenching rate constant by air (containing about 21% O₂). ⁶ Quantum yield of ¹O₂ generation with respect to that of [Ru(bpy)₃]Cl₂ in CH₃CN (Φ_Δ = 57%). ⁷ The proportion of triplet excited state of PS quenched with ³O₂. ⁸ The fraction of the triplet state of PS quenched with ³O₂, which leads to ¹O₂ generation.

). These two complexes display a series of intense absorption bands in the monitored spectral region of 250–600 nm with the molar extinction coefficients (ε_max) of 0.6–3.1 × 10⁴ and 2.0–4.6 × 10⁴ M⁻¹ cm⁻¹, respectively (Figure 3a,c). With reference to previous reports [37,38,39,40], the high-energy vibronic-structured absorptions of 1(PF₆) and 2 with the absorption maximum (λ_abs) below 380 nm are mainly ascribed to the ligand-centered (¹LC) ¹π→π* transitions of the NI moiety, tpy and pbpy ligands, while the broad absorptions in the range of 380–550 nm can be assigned to an admixture of ¹LC/charge-transfer (¹CT) transitions. In comparison with precursors [(tpy)PtCl]Cl and (pbpy)PtCl, complexes 1(PF₆) and 2 show red-shifted absorptions with distinctly enhanced molar absorptivities in the lower-energy region, which are also more intense than those of related Pt(II)-phenylacetylide complexes. This reflects the effective electronic interaction between the Pt component and the NI moiety, leading to the appearance of new CT absorptions. The specific assignments with more detailed information will be further discussed later with the aid of time-dependent density functional theory (TD-DFT) calculations.

The NI ligand shows a ¹LC emission band at 400 nm (Figure 3b). Complex (pbpy)PtCl displays a triplet LC (³LC) emission band at 560 nm (Figure 3d) [41], while complex [(tpy)PtCl]Cl is non-emissive at rt [42,43]. In comparison, complexes 1(PF₆) and 2 display a similar lower-energy emission with vibronic characteristics in DCE (5 × 10⁻⁵ M). Under the air-equilibrated condition, the emission of 1(PF₆) locates at 620 nm with an absolute quantum yield (Φ_air) of 0.3% and lifetime (τ_air) of 0.6 μs, and the emission of complex 2 is observed at 640 nm with Φ_air of 0.3% and τ_air of 0.72 μs. Meanwhile, at the N₂-saturated condition, both emissions exhibit a distinct intensity enhancement (Φ_N2 = 1.9%) and elongated lifetimes with τ_N2 of 4.8 and 7.0 μs, respectively (Table 1 and Table S2). These properties are superior to those of two Pt(II) precursors and reported Pt(II)-phenylacetylide analogues [37,38,39]. The highly O₂-dependent emissions indicate their phosphorescence features. In accordance with the previous works [36,40,44] and TD-DFT calculations, the phosphorescences of 1(PF₆) and 2 are considered to stem from the Pt-perturbed ³π→π*(³LC) transition of the NI fragment (see further discussions in following section). Furthermore, benefiting from the long lifetimes of triplet excited states and the good photostability of 1(PF₆) and 2, their phosphorescence emissions showed a good reversibility in response to the alternate saturation with N₂ and air (containing about 21% O₂). Under repeated treatment with N₂ and air for eight cycles, the phosphorescence intensity showed no apparent attenuation, indicative of their potentials for O₂ sensing (Figure S2).

Considering that the ¹O₂ generation experiment is often measured in CH₃CN [3,9], other than DCE, the comparative study of the spectroscopic properties of 1(PF₆) and 2 in these two solvents was subsequently conducted (Figure S3). For cationic 1(PF₆), the lowest energy absorption band displays a distinct hypsochromic shift from 465 nm in DCE to 437 nm in the more polar CH₃CN. The negative solvatochromic effect suggests that the dipole moment of 1(PF₆) in the ground state is larger with respect to that in the excited state and the ground-state molecules are better stabilized with solvation than the excited-state molecules [44,45]. In contrast, the absorptions of neutral 2 in DCE and CH₃CN display an insignificant spectral change. In a sense, the obvious solvent polarity-dependent absorption spectral shift of 1(PF₆) points to its more CT-contributed character of the lowest-lying singlet electronic transition relative to that of 2. Unlike the absorption spectra, the phosphorescence profiles of 1(PF₆) and 2 almost remain unchanged in these two solvents under air/N₂-satured conditions, indicative of their dominant ³LC transition characters as mentioned above.

2.3. Theoretical Calculations

To further elucidate the electronic structures of 1(PF₆) and 2, DFT and TD-DFT calculations were carried out. In order to save time and cost, the calculations were performed on the two corresponding model complexes, [Me-1]⁺ and Me-2, with a methyl group as the substituent on the nitrogen atom of the NI fragment. The crystallographic data of 2 were used to build the initial structures for the geometry optimization of [Me-1]⁺ and Me-2 at the ground state (S₀) and the lowest-lying triplet state (T₁). The DFT-optimized structures of Me-2 and [Me-1]⁺ at the ground state show smaller dihedral angles between NI and the Pt fragments with respect to that of 2 in the crystal state (Figure S4). The highest occupied molecular orbitals (HOMOs) are mainly related to the ethynylnaphthalimide moiety (89% for [Me-1]⁺ and 82% for Me-2) with a minor contribution of Pt (8.2% for [Me-1]⁺ and 13.6% for Me-2). The lowest unoccupied molecular orbital (LUMO) of [Me-1]⁺ is primarily localized on the tpy fragment (86%) with a minor distribution on Pt (5.6%), but that of Me-2 has almost equal contributions from pbpy (50%) and NI fragments (46%) with a little participation of Pt (4%) (Figure 4c,d). More detailed information on the frontier molecular orbitals of [Me-1]⁺ and Me-2 is displayed in Figures S5 and S6.

In order to obtain insight into the excited state characters of 1(PF₆) and 2, TD-DFT calculations were performed for the optimized S₀ structures of Me-2 and [Me-1]⁺, and the predicted vertical excitations from S₀ to various singlet (S_n) excited states are summarized in

Table 2. Selected TD-DFT-computed singlet vertical excitations from S₀ to S_n states based on optimized S₀ structure ^a.

Complexes	Electronic Transitions	E (eV)	λ (nm)	f	Major Transitions (Percentage)	Characters
[Me-1]+	S0→S1	2.64	470	0.2487	HOMO→LUMO (97%)	LNILtpyCT/MLtpyCT
	S0→S2	3.17	390	0.0114	HOMO→LUMO+1 (97%)	LNILtpyCT/MLtpyCT
	S0→S3	3.20	388	0.4426	HOMO→LUMO+2 (57%) HOMO−1→LUMO (40%)	LCNI MLtpyCT/LNILtpyCT
	S0→S4	3.28	378	0.1013	HOMO−1→LUMO (55%) HOMO→LUMO+2 (41%)	MLtpyCT/LNILtpyCT LCNI
Me-2	S0→S1	2.81	440	0.3582	HOMO→LUMO (93%)	LCNI/LNILbpyCT/MLbpyCT
	S0→S2	2.92	424	0.0019	HOMO−1→LUMO (71%) HOMO−1->LUMO+1 (24%)	ILpbpyCT/LbpyLNICT ILpbpyCT/LbpyLNICT
	S0→S3	3.03	408	0.4450	HOMO→LUMO+1 (84%)	LCNI/LNILbpyCT/MLbpyCT

^a Calculation method: PBE1PBE/6-311G*/SDD.

. The simulated singlet absorption spectra are shown in Figure 4a,b. The S₀→S₁ excitations were predicted to locate at 470 nm for [Me-1]⁺ with an oscillator strength (f) of 0.2487 and at 440 nm with f of 0.3582 for Me-2, respectively. These excitations are in good agreement with the observed lowest-energy absorption maximum of 465 nm for 1(PF₆) and 437 nm for 2, respectively. The S₀→S₁ excitations of [Me-1]⁺ and Me-2 are both dominated by the HOMO→LUMO electronic transitions and they can be interpreted as the major ligand-to-ligand charge-transfer (¹L_NIL_tpyCT) and mixed ¹LC_NI/¹L_NIL_bpyCT (bpy stands for the bipyridine segment of the pbpy ligand) transitions, respectively, which are in accordance with the observation of their solvent-dependent absorptions.

Furthermore, the spin density distributions of [Me-1]⁺ and Me-2 were calculated on the optimized T₁ geometries to probe the natures of the phosphorescence of 1(PF₆) and 2 (Figure 5). The results show that the spin densities are mainly distributed on the π-orbitals of the ethynylnaphthalimide moiety with a minor involvement of the d-orbital of Pt metal. On the basis of these results, we tentatively assign the low-energy phosphorescence of 1(PF₆) and 2 to the Pt-perturbed ³LC transition of the ethynylnaphthalimide segment. This assignment is consistent with the fact that the phosphorescence of complexes 1(PF₆) and 2 involves similar solvent polarity-independent vibronic structures and emission wavelength range.

2.4. Singlet Oxygen Generation

The small triplet radiative transition rate constants (k_r), even at the N₂-saturated condition, which is 4 × 10³ s⁻¹ for 1(PF₆) and 2.7 × 10³ s⁻¹ for 2, respectively, result in their low Φ_N2 of ca. 1.9% (Table 1). Moreover, Φ_N2 as well as the average τ are further diminished under the air-equilibrated condition along with the relatively large nonradiative transition rate constants (k_nr in the order of 10⁶ s⁻¹). The phosphorescence diminishment under the air-equilibrated condition can be rationalized with Equation (1) as a result of the emission quenching with ³O₂ [46].

$$\frac{{\mathsf{\tau }}_{\mathrm{N}2}}{{\mathsf{\tau }}_{\mathrm{a}\mathrm{i}\mathrm{r}}}=\frac{{\mathsf{\Phi }}_{\mathrm{N}2}}{{\mathsf{\Phi }}_{\mathrm{a}\mathrm{i}\mathrm{r}}}=1+k_{\mathrm{q}}{\mathsf{\tau }}_{\mathrm{N}2}\left[{\mathrm{O}}_2\right]$$

(1)

wherein [O₂] represents the concentration of ³O₂ and is taken as 1.6 × 10⁻³ M in air-equilibrated DCE at rt [47] and τ_N2 and τ_air are phosphorescence lifetimes under N₂-saturated and air-equilibrated conditions, respectively. On the basis of Equation (1), the phosphorescence quenching rate constants (k_q) were calculated to be ca. 9.1 and 7.8 × 10⁸ M⁻¹ s⁻¹ for 1(PF₆) and 2, respectively, which are comparable with those of other known Pt analogues [28,48]. These k_q values are around 1/9 of the diffusion-controlled rate constant (k_diff), given that k_diff is taken to be 8.5 × 10⁹ M⁻¹ s⁻¹ in DCE at rt. This suggests that the phosphorescences of 1(PF₆) and 2 are quenched with ³O₂ exclusively through an EnT process [49]. The similar EnT phosphorescence quenching processes with ³O₂ are also believed to take place in CH₃CN ([³O₂] = 1.9 × 10⁻³ M) as the k_q of 1(PF₆) and 2 in CH₃CN (1.8 and 1.9 × 10⁹ M⁻¹ s⁻¹, respectively) are also around 1/9 of corresponding k_diff (1.9 × 10¹⁰ M⁻¹ s⁻¹). In addition, the excited-state energies of the T₁ state (E_T1,00) of 1(PF₆) and 2 were calculated with E_T1,00 = 1240/λ_emi to be ca. 2.0 and 1.9 eV, respectively, both of which are larger than the first (¹∆_g) and second (¹Σ_g) excited-state energies of ¹O₂ (ca. 1.6 and 1.0 eV, respectively) in non-aqueous solvents [50,51,52]. This suggests that the EnT processes from 1(PF₆) and 2 to ³O₂ are exothermic, leading to the effective generation of ¹O₂.

The generation of ¹O₂ was further quantified by integrating the NIR emission of ¹O₂ at 1270 nm in air-equilibrated CH₃CN of 1(PF₆) and 2 in comparison with that of [Ru(bpy)₃]Cl₂ as the reference compound (Figure 6 and Figure S7). By plotting the emission integrals of ¹O₂ (I_O2) as a function of the absorbances (A) at the excitation wavelength of 435 nm in a series of solutions of these compounds with variable concentrations, three linearly fitted curves with different slopes, S(I_O2/A), are obtained (Figure 6b). The quantum yields of the ¹O₂ generation of 1(PF₆) and 2 were thus estimated to be 89% and 84%, respectively, according to Equation (2), in which Φ_Δ,ref, I_O2,ref, A_ref, and S(I_O2,ref/A_ref) are the various parameters of the reference standard [Ru(bpy)₃]Cl₂ (Φ_Δ = 57% in CH₃CN) [49,50]. The determined Φ_Δ of these complexes are comparable with or even higher than those of known tridentate Pt(II)-arylacetylide analogues [36,53,54,55].

$$\frac{{\mathsf{\Phi }}_{\Delta }}{{\mathsf{\Phi }}_{\Delta ,\mathrm{r}\mathrm{e}\mathrm{f}}}=\frac{{\mathrm{I}}_{\mathrm{O}2}\times {\mathrm{A}}_{\mathrm{r}\mathrm{e}\mathrm{f}}}{{\mathrm{I}}_{\mathrm{O}2,\mathrm{r}\mathrm{e}\mathrm{f}}\times \mathrm{A}}=\frac{S({\mathrm{I}}_{\mathrm{O}2}/\mathrm{A})}{S({\mathrm{I}}_{\mathrm{O}2,\mathrm{r}\mathrm{e}\mathrm{f}}/{\mathrm{A}}_{\mathrm{r}\mathrm{e}\mathrm{f}})}$$

(2)

The good linear correlation between the integral of the ¹O₂ emission intensity and the absorbance at the excitation wavelength indicates that the triplet–triplet annihilation in these diluted solutions is negligible. Furthermore, Φ_Δ can be expressed with Equation (3):

$${\mathsf{\Phi }}_{\Delta }={\mathsf{\Phi }}_{\mathrm{T}}{\mathrm{P}}_{\mathrm{T},\mathrm{O}2}{f}_{\mathrm{T},\Delta }$$

(3)

wherein Φ_T is the efficiency of triplet formation and estimated to be unity for 1(PF₆) and 2 in view of the strong SOC of the heavy Pt atom. In addition, P_T,O2 refers to the proportion of the triplet excited state quenched with ³O₂ and can be estimated to be 92% and 85% for 1(PF₆) and 2, respectively, with Equation (4).

$${\mathrm{P}}_{\mathrm{T},\mathrm{O}2}=1-\left({\mathsf{\tau }}_{\mathrm{A}\mathrm{i}\mathrm{r}}/{\mathsf{\tau }}_{\mathrm{N}2}\right)$$

(4)

The values of f_T,∆ in Equation (3), standing for the fraction of the triplet state quenched with ³O₂, which leads to the generation of ¹O₂, are calculated to be ca. 97% and 99% for 1(PF₆) and 2, respectively.

3. Experimental Section

3.1. General Information for Synthesis and Characterization

A Bruker Advance 400 MHz spectrometer was employed to measure the ¹H NMR spectra in the designated solvents. The ¹H NMR data are reported in ppm with respect to the residual protons of deuterated solvents. An Autoflex III matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometer was used to measure the high-resolution mass spectra. The elemental analysis was obtained on a Flash EA 1112 or Carlo Erba 1106 analyzer. Ligand NI and the precursor complexes [(tpy)PtCl]Cl and (pdpy)PtCl were prepared according to the previous reports [56,57].

Synthesis of 1(PF₆). Into a round-bottom flask containing the NI ligand (214 mg, 0.55 mmol, 1.1 equiv) and KOH (56 mg, 1.0 mmol, 2.0 equiv), 20 mL of CH₃OH was injected under a N₂ atmosphere. The obtained mixture was stirred at rt for 0.5 h. After that, [(tpy)PtCl]Cl (250 mg, 0.50 mmol, 1.0 equiv) and CuI (57 mg, 0.30 mmol, 0.60 equiv) were added and the mixture was bubbled with N₂ for 10 min. After stirring at rt overnight, the mixture was concentrated, followed by the addition of the proper amount of ethyl ether. The appeared precipitate was collected using filtration. The obtained solid was then dispersed in 20 mL of CH₃OH, followed by the addition of 20 mL of a saturated aq. KPF₆ solution. After stirring for an additional 3 h at rt, the precipitate was collected using filtration, and washed with H₂O, ethyl ether, and CH₂Cl₂ successively to give 210 mg of 1(PF₆) as a red solid in a 76% yield. ¹H NMR (400 MHz, DMSO-d₆): δ 9.15 (t, J = 4.8 Hz, 2H), 8.80 (t, J = 8.4 Hz, 1H), 8.50–8.70 (m, 5H), 8.47–8.53 (m, 3H), 8.38 (t, J = 7.6 Hz, 1H), 7.88–7.97 (m, 4H), 4.03 (t, J = 7.6 Hz, 2H), 1.64 (t, J = 7.6 Hz, 2H), 1.23–1.33 (m, 18H), 0.85 (t, J = 6.8 Hz, 3H). MALDI-HRMS calcd for [M − PF₆]⁺ C₄₁H₄₁N₄O₂Pt: 816.2877. Found: 816.2875. Anal. Calcd for C₄₁H₄₁F₆N₄O₂PPt·H₂O: C, 50.26; H, 4.42; N, 5.72. Found: C, 50.43; H, 4.26; N, 5.70.

Synthesis of 2. Into a round-bottom flask containing the NI ligand (107 mg, 0.28 mmol, 1.1 equiv) and KOH (28 mg, 0.5 mmol, 2.0 equiv), 20 mL of CH₃OH was added under a N₂ atmosphere. The resulting mixture was stirred at rt for 0.5 h. After that, 20 mL of CH₂Cl₂ was injected, followed by the addition of (pbpy)PtCl (115 mg, 0.25 mmol, 1.0 equiv) and CuI (29 mg, 0.15 mmol, 0.60 equiv). The mixture was then bubbled with N₂ for 10 min, and stirred at rt overnight. Into the reaction system, 50 mL of H₂O was added and the mixture was extracted with CH₂Cl₂ (40 mL × 3). The obtained organic phase was combined. The crude product was purified with flash column chromatography on silica gel (eluent: CH₂Cl₂) to afford 167 mg of 2 as an orange solid in a 88% yield. ¹H NMR (400 MHz, CDCl₃): δ 9.26 (d, J = 5.2 Hz, 1H), 9.09 (dd, J = 8.4, 1.2 Hz, 1H), 8.59 (dd, J = 7.2, 1.2 Hz, 1H), 8.52 (dd, J = 8.0 Hz, 1H), 8.09 (td, J = 7.6, 1.6 Hz, 1H), 8.86–8.99 (m, 4H), 7.72 (dd, J = 8.0, 7.2 Hz, 1H), 7.65 (dd, J = 8.0, 1.2 Hz, 2H), 7.60 (dd, J = 7.6, 5.2 Hz, 1H), 7.43 (dd, J = 7.6, 1.2 Hz, 1H), 7.20 (td, J = 7.2, 1.2 Hz, 1H), 7.11 (td, J = 7.6, 1.2 Hz, 1H), 4.18 (t, J = 7.6 Hz, 2H), 1.71–1.79 (m, 2H), 1.26–1.45 (m, 18H), 0.88 (t, J = 6.8 Hz, 3H). MALDI-HRMS calcd for [M + H]⁺ C₄₂H₄₂N₃O₂Pt: 815.2925. Found: 815.2924. Anal. Calcd for C₄₂H₄₁N₃O₂Pt: C, 61.91; H, 5.07; N, 5.16. Found: C, 61.59; H, 5.05; N, 5.09.

3.2. Information on Other Physical Measurements

A TU-1810DSPC spectrometer of Beijing Purkinje General Instrument Co., Ltd., Beijing, China., was used to measure the absorption spectra at rt. An F-380 spectrofluorimeter of Tianjin Gang-dong Sci. & Tech. Development Co., Ltd., Tianjing, China., was used to measure the luminescence spectra. The excited-state lifetimes were obtained on Quantaurus-Tau Fluorescence lifetime spectrometer C11367 of Hamamatsu Photonics. The absolute emission quantum yields were determined on Hamamatsu Quantaurus-QY spectrometer C11347. Near-infrared ¹O₂ emission spectra were recorded on an FLS980 spectrometer from Edinburgh Instruments Co., Ltd. equipped with an NIR photomultiplier tube (NIR-PMT, R5509) from Hamamatsu Corporation using excitation at 435 nm with a Xe lamp.

X-ray single-crystal crystallography. A Rigaku Saturn 724 diffractometer was employed to obtain the X-ray diffraction data on a rotating anode (Mo Kα radiation, 0.71073 Å) at 173 K. The structure was analyzed using the direct method with SHELXS-9758 and refined on Olex 2.59. Crystallographic data are summarized in Table S1.

DFT and TD-DFT calculations. The Gaussian 09 package was employed for the DFT and TD-DFT calculations with the PBE1PBE exchange correlation functional [58]. The SDD basis set was used for Pt and 6-311G* for other atoms [26]. For all calculations, the solvent effect of 1,2-dichloroethane was taken into account using the self-consistent reaction field (SCRF) and solvent model based on density (SMD) [59].

4. Conclusions

In conclusion, two tridentate Pt(II) complexes, i.e., cationic 1(PF₆) and neutral 2, modified with an ethynylnaphthalimide chromophore, were successfully synthesized and characterized. These complexes possess strong UV-Vis absorptions and vibronically structured phosphorescence emissions at 620–640 nm in nonpolar DCE and polar CH₃CN. Theoretical calculations suggest that the lowest-energy absorption bands of 1(PF₆) and 2 are mainly associated with the ¹L_NIL_tpyCT and mixed ¹LC_NI/¹L_NIL_bpyCT transitions, respectively, while their low-energy phosphorescences are both associated with the ³LC transition of the ethynylnaphthalimide moiety. Benefiting from the long lifetimes of phosphorescence, these two complexes exhibit highly sensitive and reversible responses to O₂ with the high-yield generation (84–89%) of ¹O₂ via an EnT process. Thus, these complexes have potential for O₂ sensing and PDT in the future.