Pt(II) Complexes with Tetradentate C^N*N^C Luminophores: From Supramolecular Interactions to Temperature-Sensing Materials with Memory and Optical Readouts

Abstract

A series of four regioisomeric Pt(II) complexes (PtL^a-n and PtL^b-n) bearing tetradentate luminophores as dianionic ligands were synthesized. Hence, both classes of cyclometallating chelators were decorated with three n-hexyl (n = 6) or n-dodecyl (n = 12) chains. The new compounds were unambiguously characterized by means of multiple NMR spectroscopies and mass spectrometry. Steady-state and time-resolved photoluminescence spectroscopy as well quantum chemical calculations show that the effect of the regioisomerism on the emission colour and on the deactivation rate constants can be correlated with the participation of the Pt atom on the excited state. The thermal properties of the complexes were studied by DSC, POM and temperature-dependent steady-state photoluminescence spectroscopy. Three of the four complexes (PtL^a-12, PtL^b-6 and PtL^b-12) present an intriguing thermochromism resulting from the responsive metal–metal interactions involving adjacent monomeric units. Each material has different transition temperatures and memory capabilities, which can be tuned at the intermolecular level. Hence, dipole–dipole interactions between the luminophores and disruption of the crystalline packing by the alkyl groups are responsible for the final properties of the resulting materials.

1. Introduction

Luminescent Pt(II) complexes have been widely studied as optical sensors for both physical (mechanic [1,2,3,4], temperature [5,6,7], polarity [6]) and chemical (vapours [2,8,9], ions [10,11], small molecules [12], molecular oxygen [5,13,14]) stimuli. Moreover, several patents on their use in optoelectronic devices and as sensors have been applied for [15,16,17,18]. The phosphorescence of Pt(II) complexes, which is widely discussed in the bibliographic literature, generally stems from metal-perturbed ligand-centred triplet states (³MP-LC), where the emissive species possesses a mixed ³LC (ligand-centred, i.e., π-π*) and ³MLCT (metal-to-ligand charge transfer, i.e., d-π*) character [19,20,21,22]. There are numerous examples of Pt(II) complexes with both bidentate and tridentate ligands that are flexible enough to allow for geometrical distortion out of the coordination plane, which upon photoexcitation facilitates non-radiative decay pathways through conical intersections with the ground state [23,24]. Hence, the use of tetradentate ligands can increase the rigidity and, in consequence, enhance the photoluminescence efficiency of the complexes [25,26]. Moreover, cyclometallation can be further used to enhance the ligand-field splitting by the introduction of strong σ-donors, raising the antibonding d_x²_-y² orbital to higher energies and, in consequence, destabilizing dissociative states (including d-d* and π-d* configurations), which in turn become less thermally accessible [26,27,28,29,30].

Several Pt(II)-based species with sensing applications rely on their capability to form aggregates, both in solution and in the solid state. The formation of Pt(II)-based aggregates is possible due to their planar coordination geometry and is mainly driven by van der Waals interactions. When the distance between the Pt atoms is approximately 3.5 Å or shorter (i.e., below the sum of their van der Waals radii), coupling between the d_z² orbital lobes protruding out of the coordination plane becomes feasible. The resulting aggregates show a red-shifted emission if compared with the monomeric species, which arises from triplet metal–metal-to-ligand charge transfer (³MMLCT) states with a certain degree of excimeric character [14,21,31,32]. Hence, we have now prepared a series of Pt(II) complexes where the modification in the length and position of peripheral alkyl chains allows us to control the photophysical properties, the characteristics of the solid phase and the transition temperatures of the resulting solids. These concepts can lead to temperature-sensing materials with memory based on Pt-Pt interactions and providing optical readouts.

The complexes presented in this work have been developed as a natural progression of our previous research [4]. The herein-reported replacement of the methoxyphenyl group at the bridging N-atom by alkyl chains aims at promoting the aggregation of the complexes. This stacking was partially hampered in our previous complexes, due to the 90° rotation of the phenyl ring with respect to the luminophoric plane [4,14]. In addition, modulation of the length and position of the alkyl chains affects the packing and supramolecular interactions between the molecular units in the solid state, leading to changes in the thermal properties of the complexes.

2. Synthesis

The synthesis of the ligands (Scheme 1) started with the reaction between 2,6-dibromopyridine and 2-amino-6-bromopyridine promoted by NaH to give the precursor 1. The intermediate 1 was reacted with 1-bromohexane or 1-bromododecane to yield the precursors 2-6 and 2-12, respectively. A further Suzuki–Miyaura cross-coupling with an excess of the corresponding hydroxyphenylboronic acid allowed us to prepare the intermediates 3a-n and 3b-n (n = 6 or 12). The phenolic species were not isolated and the ligands L^a-n and L^b-n were finally obtained by Williamson’s ether synthesis using the corresponding alkylbromide [4]. The Pt(II) complexes were prepared by a standard cycloplatination reaction with K₂PtCl₄ in acetic acid at reflux with yields between 30 and 44%, which are in agreement with previously reported syntheses [4,14]. The comparable yields obtained for both regioisomers suggest that, in the case of PtL^b-n complexes, only one of the three potential cyclometallation products is formed. This result may be attributed to the steric hindrance imposed by the alkyl chains during the cycloplatination reaction. The complexes show a good solubility in polar organic solvents, such as DCM and THF, mostly due to the presence of the peripheral alkyl chains [4]. All the precursors and complexes were unambiguously characterized by means of 2D-NMR (including a full assignment of ¹H and ¹³C signals) as well as by high-resolution mass spectrometry (Figures S1–S61). Further experimental details can be found in the ESI.

3. Photophysics in Diluted Conditions

The photophysical characterization of the complexes was initially performed in diluted DCM solutions (complex concentration c ≈ 10⁻⁵ M) at room temperature. The absorption bands in the region between 235 and 300 nm can be assigned to transitions into ¹LC states, while the bands in the region between 350 and 450 nm are related to mixed singlet states with variable degrees of LC and MLCT character, as was previously reported for comparable complexes (Figure 1 left, Figure S62 and Table S1) [14,21,33,34,35]. Interestingly, the PtL^b-n complexes present a stronger absorption band in the ¹MLCT (≈420 nm) region if compared with the PtL^a-n complexes. This can be attributed to the higher contribution of the alkoxy group in para position (with respect to the Pt atom) to the HOMO than in case of the moiety in meta position, and therefore to a higher destabilization of the HOMO (as previously observed for Ir(III) complexes and [36,37] in phenylpyridine Pt(II)-based complexes [38]).

The emission spectra of the PtL^a-n complexes (Figure 1 right, Figure S63) show a maximum at 517 nm with vibrational shoulders at 553 nm, which arise mainly from ³MP-LC states [14,33,34,39]. The spectra of PtL^b-n are red-shifted with a maximum at 555 nm (Figure 1 right, Figure S64). The red shift in the emission can be also ascribed to a destabilization of the HOMO and consequently to a reduction in the gap between the excited and ground state [37,38]. To support and interpret the experimental results, (TD-)DFT calculations were performed for the methoxy-substituted complexes (PtL^a-1 and PtL^b-1) as model compounds. After the geometry optimization of the ground state (S₀) and the lowest triplet state (T₁) was achieved, the emission maxima were calculated (Figure S77) [40,41,42,43,44,45,46,47].

A detailed characterization of the emissive state can be achieved with the aid of a correlated electron-hole pair analysis using the software package for Theoretical Density, Orbital Relaxation and Exciton analysis (TheoDORE) [48]. The result of the decomposition into contributions from MLCT, LMCT (ligand-to-metal charge transfer), LC and MC is presented in Figure 2. The higher MLCT character in the excited state of PtL^b-n, particularly if compared with PtL^a-n (25.0% vs 21.1%, respectively), may be responsible for the lack of vibronic progression in PtL^b-n complexes, due the stronger geometrical distortion in the excited state. The distortion of the excited state (ΔQ) can be estimated by the so-called Huang–Rhys factor (S) according to Equation (1), where I_0-0 is the emission intensity of the 0-0 transition and I_1-0 the one corresponding to the first vibronic peak (

Table 1. Huang–Rhys factor in fluid solution at r.t. and in glassy matrices at 77 K for the four complexes.

Complex	Huang–Rhys Factor S
	Fluid Solution at r.t.	Glassy Matrix at 77 K
PtLa-6	0.61	0.57
PtLa-12	0.66	0.46
PtLb-6	0.83	0.61
PtLb-12	0.80	0.54

) [26]. Consistently, S is higher for PtL^b-n than for PtL^a-n complexes.

$$S=\frac{I_{0-1}}{I_{0-0}}$$

(1)

The photophysical properties were also studied in diluted glassy matrices of 2-methyltetrahydrofuran (complex concentration, c ≈ 10⁻⁵ M) at 77 K. All the complexes show a blue shift in the emission spectra of approximately 10 nm, if compared with the solutions at r.t. This is a consequence of the weaker charge-transfer stabilization associated with the limited solvent reorientation in the rigid matrices, thereby decreasing the ³MLCT character of the emissive state [49,50]. Another consequence of the restricted solvent reorientation is the reduced density of solvent-related roto-vibrational states, thus leading to a well-defined vibrational progression [51]. Interestingly, the reduction in S is more significant for PtL^b-n than for PtL^a-n when the fluid solution is changed to the glassy matrix, which is in agreement with a higher loss of MLCT character from the lack of solvent reorientation.

The excited state of Pt(II) complexes can be deactivated in fluid solution by ³O₂-mediated diffusional quenching, which is reflected in the decrease (by a factor > 50) in the k_nr after Ar purging of the samples. The PtL^a-n series shows longer lifetimes (Figures S65–S76) and higher quantum yields than their PtL^b-n isomers (Table 2); a similar effect was previously reported for Ir(III) complexes [37].

When the complexes are studied in dilute fluid solutions, different conformations are thermally accessible and possess comparable excited state characters yet different deactivation rates; therefore, multiexponential decays can be observed in some cases. This effect is also noticeable in glassy matrices, yielding multiexponential decays as previously reported for other Pt(II) complexes [33,51,52]. Compared to the analogous complex with a methoxyphenyl group (which has been previously reported by our group [4]), PL^a-6 presents somewhat higher k_r and k_nr rates. This can be attributed to the reduced steric strain associated with the n-alkyl substitution, as opposed to a phenyl moiety at the bridging N-atom. The k_r and k_nr values of PtL^a-n and PtL^b-n show two interesting trends. The k_r are in the same range for both luminophores; however, PtL^b-n complexes present k_nr values between 2 and 3 times higher than the PtL^a-n compounds. In this case, the faster non-radiative intersystem crossing rate constant between the T₁ and S₀ states can be ascribed to the higher geometrical distortion in the excited state of PtL^b-n, if compared with the PtL^a-n complexes. It would be expected that a lower MLCT character of the excited state leads to a higher k_r at 77 K than at r.t. However, the four complexes show a counterintuitive tendency, with higher k_r at lower temperatures. This was already observed for other Pt(II) complexes with tridentate luminophores, and (TD-)DFT calculations at a higher level of sophistication (including a temperature-dependent Boltzmann analysis of the radiative rate constants) will be needed to elucidate the nature of this observation [53].

Table 2. Photophysical parameters of the complexes.

	PtLa-6				PtLa-12				PtLb-6				PtLb-12
	τav a (µs)	ΦL b	kr c/104 (s−1)	knr d/104 (s−1)	τav a (µs)	ΦL b	kr c/104 (s−1)	knr d/104 (s−1)	τav a (µs)	ΦL b	kr c/104 (s−1)	knr d/104 (s−1)	τav a (µs)	ΦL b	kr c/104 (s−1)	knr d/104 (s−1)
Air-equilibrated DCM solution (r.t.)	0.203 ± 0.001 e	<0.02	<10	480 < knr < 490	0.174 ± 0.001 e	<0.02	<12	560 < knr < 570	0.163 ± 0.001 e	<0.02	<12	600 < knr < 610	0.209 ± 0.001 f	<0.02	<10	470 < knr < 480
Ar-purged DCM solution (r.t.)	11.22 ± 0.02 e	0.49 ± 0.02	4.4 ± 0.2	4.5 ± 0.2	10.60 ± 0.02 f	0.57 ± 0.02	5.4 ± 0.2	4.1 ± 0.2	6.102 ± 0.003 e	0.26 ± 0.02	4.3 ± 0.3	12.1 ± 0.3	8.55 ± 0.03 f	0.27 ± 0.02	3.2 ± 0.2	8.5 ± 0.3
2-MeTHF glassy matrix (77 K)	16.8 ± 0.1 e	0.99 ± 0.01	6.0 ± 0.2	<6	11.4 ± 0.8 f	0.99 ± 0.01	8.8 ± 0.8	<9	7.60 ± 0.03 f	0.99 ± 0.01	13.2 ± 0.3	<13	8.7 ± 0.1 f	0.99 ± 0.01	11.6 ± 0.4	<11

^a Amplitude-weighted average lifetime. ^b Photoluminescence quantum yield. ^c Average radiative rate constant $k_{\mathrm{r}}=\frac{{\Phi }_{\mathrm{L}}}{{\tau }_{\mathrm{a}\mathrm{v}}}$; ${∆k}_{\mathrm{r}}=\left(\frac{{∆\Phi }_{\mathrm{L}}}{{\tau }_{\mathrm{a}\mathrm{v}}}\right)+\left(\frac{{\Phi }_{\mathrm{L}}}{{{\tau }_{\mathrm{a}\mathrm{v}}}^2}∆{\tau }_{\mathrm{a}\mathrm{v}}\right)$ [54]. ^d Average radiationless deactivation rate constant $k_{\mathrm{n}\mathrm{r}}=\frac{1-{\Phi }_{\mathrm{L}}}{{\tau }_{\mathrm{a}\mathrm{v}}}$; ${∆k}_{nr}=\left(\frac{{∆\Phi }_L}{{\tau }_{av}}\right)+\left(\frac{∆{\tau }_{av}}{{{\tau }_{av}}^2}\right)+\left(\frac{{\Phi }_L}{{{\tau }_{av}}^2}∆{\tau }_{av}\right)$ [54]. ^e Monoexponential decay. ^f Biexponential decay.

Table 2. Photophysical parameters of the complexes.

	PtLa-6				PtLa-12				PtLb-6				PtLb-12
	τav a (µs)	ΦL b	kr c/104 (s−1)	knr d/104 (s−1)	τav a (µs)	ΦL b	kr c/104 (s−1)	knr d/104 (s−1)	τav a (µs)	ΦL b	kr c/104 (s−1)	knr d/104 (s−1)	τav a (µs)	ΦL b	kr c/104 (s−1)	knr d/104 (s−1)
Air-equilibrated DCM solution (r.t.)	0.203 ± 0.001 e	<0.02	<10	480 < knr < 490	0.174 ± 0.001 e	<0.02	<12	560 < knr < 570	0.163 ± 0.001 e	<0.02	<12	600 < knr < 610	0.209 ± 0.001 f	<0.02	<10	470 < knr < 480
Ar-purged DCM solution (r.t.)	11.22 ± 0.02 e	0.49 ± 0.02	4.4 ± 0.2	4.5 ± 0.2	10.60 ± 0.02 f	0.57 ± 0.02	5.4 ± 0.2	4.1 ± 0.2	6.102 ± 0.003 e	0.26 ± 0.02	4.3 ± 0.3	12.1 ± 0.3	8.55 ± 0.03 f	0.27 ± 0.02	3.2 ± 0.2	8.5 ± 0.3
2-MeTHF glassy matrix (77 K)	16.8 ± 0.1 e	0.99 ± 0.01	6.0 ± 0.2	<6	11.4 ± 0.8 f	0.99 ± 0.01	8.8 ± 0.8	<9	7.60 ± 0.03 f	0.99 ± 0.01	13.2 ± 0.3	<13	8.7 ± 0.1 f	0.99 ± 0.01	11.6 ± 0.4	<11

^a Amplitude-weighted average lifetime. ^b Photoluminescence quantum yield. ^c Average radiative rate constant $k_{\mathrm{r}}=\frac{{\Phi }_{\mathrm{L}}}{{\tau }_{\mathrm{a}\mathrm{v}}}$; ${∆k}_{\mathrm{r}}=\left(\frac{{∆\Phi }_{\mathrm{L}}}{{\tau }_{\mathrm{a}\mathrm{v}}}\right)+\left(\frac{{\Phi }_{\mathrm{L}}}{{{\tau }_{\mathrm{a}\mathrm{v}}}^2}∆{\tau }_{\mathrm{a}\mathrm{v}}\right)$ [54]. ^d Average radiationless deactivation rate constant $k_{\mathrm{n}\mathrm{r}}=\frac{1-{\Phi }_{\mathrm{L}}}{{\tau }_{\mathrm{a}\mathrm{v}}}$; ${∆k}_{nr}=\left(\frac{{∆\Phi }_L}{{\tau }_{av}}\right)+\left(\frac{∆{\tau }_{av}}{{{\tau }_{av}}^2}\right)+\left(\frac{{\Phi }_L}{{{\tau }_{av}}^2}∆{\tau }_{av}\right)$ [54]. ^e Monoexponential decay. ^f Biexponential decay.

4. Thermochromic Properties

The thermochromic properties of the compounds were studied by differential scanning calorimetry (DSC), polarized optic microscopy (POM) and temperature-dependent steady-state luminescence spectroscopy.

In DSC experiments (

Table 3. Transition temperatures and enthalpies of the complexes.

	Phase Transitions
n	6	12
PtLa-n	Cr 245 Idecomp	Cr 197 (22.3) I
PtLb-n	Cr 127 (30.8) I	Cr 132 (45.4) I

Temperatures given in °C; enthalpies in kJ·mol⁻¹ are indicated in parentheses. Cr: crystalline phase; I: isotropic liquid. Heating rates: 10 °C/s. Values are reported in the second heating cycle except for PL^a-6.

, Figures S78–S81), PtL^a-6 melts at 245 °C with decomposition. No changes below this temperature were observed in POM. PtL^a-12 shows, within a first cycle, a crystal-to-crystal transition at 74 °C, while melting at 197 °C and crystallizing again at 186 °C (Figure S82). Changes in the melting temperature within cycles for PtL^a-12 indicate a small decomposition during the phase transition. PtL^b-6 and PtL^b-12 both melt at around 130 °C, returning not to a crystalline phase, but rather to a glassy phase in a reversible way (Figures S83 and S84). In the case of the “open” series PtL^b-n, the increasing chain length produces an enhancement of the van der Waals interactions between the alkyl groups while causing a slight raise of the melting temperature and a significant increment of the melting enthalpy.

Interestingly, the transitions undergone by PtL^a-12 involve a change in the luminescence colour, with a decrease of the monomeric emission and the growth of an unstructured yet red-shifted band. This new luminescence can be ascribed to the formation of dimers and aggregates where the Pt···Pt distances are shorter than 3.5 Å. [3,14,34,55]. The assignment of this emission band was carried out by comparing the calculated emission spectra of the dimers with the experimental bands measured upon heating of the samples. The calculated emission spectra of PtL^a-1 and PtL^b-1 dimers are in very good agreement with the experimental data (Figure 3).

Powder X-ray diffractometry of PtL^a-12 was performed before and after heating the sample above the transition temperature. The diffraction pattern before the thermal treatment shows typical Bragg reflections of a polycrystalline sample. Upon heating, no Bragg reflections are observed while showing the formation of a mostly amorphous solid (Figure 4).

In the case of complexes PtL^b-6 and PtL^b-12, the monomeric emission intensities decrease with increasing temperature (Figures S86 and S87), as expected for the progressive population of roto-vibrational states while leading to faster non-radiative deactivation rates. On the other hand, PtL^a-12 shows a very stable emission intensity at 512 nm as a function of temperature below the transition temperature. In the range between 65 °C and 75 °C, the emission intensity at 512 nm decreases by roughly 80%.

The emission spectra of PtL^a-12, PtL^b-6 and PtL^b-12 were measured 48 h after melting (Figure 5). Notably, each complex exhibits a different behaviour. PtL^a-12 keeps the emission of the aggregates kinetically blocked and PtL^b-6 shows partial recovery, whereas for PtL^b-12 the original emission (i.e., before heating) is fully recovered. These observations can be explained in terms of the intermolecular interactions within the solids: the “closed” complexes PtL^a-n possess a higher dipolar moment than the “open” PtL^b-n series. Consequently, the interaction between the aromatic cores within the aggregates is more pronounced. Conversely, the longer alkyl moieties disrupt the interactions between the aromatic planes [56,57]. As a result of these combined effects, PtL^b-12 is able to return to a monomer-dominated phase, whereas PtL^b-6 achieves only a partial reversion. Ultimately, the more robust dipole interactions sustain the aggregates in the case of PtL^a-12.

5. Conclusions

Four new Pt(II) complexes with C^N*N^C ligands were synthesized and characterized. The position of the alkoxy groups has a crucial effect on the photophysical properties of the monomeric complexes by varying the metal participation on the excited state. The PtL^b-n complexes with an extended metal contribution display emission maxima that are red-shifted by approximately 45 nm while having higher k_nr values in comparison with their PtL^a-n isomers; this demonstrates that small changes in the ligand structure can significantly affect the photophysical properties. The alkyl chain length affects the capability of the complexes to form crystalline phases and their transition temperatures. Interestingly, the phase transitions facilitate the Pt-Pt coupling, which causes a red-shifted luminescence due to the formation of dimers. Depending on the substitution pattern and the alkyl chain length, the thermal properties of the materials can be tuned. Moreover, the balance between the different classes of intermolecular interactions affects the memory of thermal exposure above the transition temperatures. Hence, variable alkyl chain lengths can be exploited to design materials with different sensitivity ranges, in order to cover appropriated windows for their use as sensors with multiple optical readouts (i.e., based on photoluminescence wavelengths, intensities and lifetimes).