Dual Emissive Zn(II) Naphthalocyanines: Synthesis, Structural and Photophysical Characterization with Theory-Supported Insights towards Soluble Coordination Compounds with Visible and Near-Infrared Emission

Metal phthalocyaninates and their higher homologues are recognized as deep-red luminophores emitting from their lowest excited singlet state. Herein, we report on the design, synthesis, and in-depth characterization of a new class of dual-emissive (visible and NIR) metal naphthalocyaninates. A 4-N,N-dimethylaminophen-4-yl-substituted naphthalocyaninato zinc(II) complex (Zn-NMe2Nc) and the derived water-soluble coordination compound (Zn-NMe3Nc) exhibit a near-infrared fluorescence from the lowest ligand-centered state, along with a unique push–pull-supported luminescence in the visible region of the electromagnetic spectrum. An unprecedentedly broad structural (2D-NMR spectroscopy and mass spectrometry) as well as photophysical characterization (steady-state state and time-resolved photoluminescence spectroscopy) is presented. The unique dual emission was assigned to two independent sets of singlet states related to the intrinsic Q-band of the macrocycle and to the push–pull substituents in the molecular periphery, respectively, as predicted by TD-DFT calculations. In general, the elusive chemical aspects of these macrocyclic compounds are addressed, involving both reaction conditions, thorough purification, and in-depth characterization. Besides the fundamental aspects that are investigated herein, the photoacoustic properties were exemplarily examined using phantom gels to assess their tomographic imaging capabilities. Finally, the robust luminescence in the visible range arising from the push–pull character of the peripheral moieties demonstrated a notable independence from aggregation and was exemplarily implemented for optical imaging (FLIM) through time-resolved multiphoton micro(spectro)scopy.


Introduction
Over the past decade, there has been a rising interest in the design and synthesis of compounds exhibiting dual emission (DE) [1].While most organic molecules typically manifest a single characteristic fluorescence, certain distinctive luminophores, wherein electron donor and electron acceptor moieties are connected by a single bond, may exhibit a dual emission [2].The phenomenon of dual fluorescence (DF) was initially observed for 4-(dimethylamino)benzonitrile (DMABN) by Lippert et al. [3].It arises from two distinct conformations of the same molecule in the first excited singlet state (S 1 ), specifically, the locally excited state (LE) and the intramolecular charge-transfer state (ICT) [4].Subsequently, several other systems were designed incorporating a strong electron-donating or -withdrawing group [5][6][7][8][9][10][11]. Owing to their nonlinear optical properties, these compounds have been extensively explored as part of the development of organic materials, serving various applications including their use as electrooptical switches, chemical sensors, and fluorescent markers [12][13][14][15].
Generally, due to their broad application range as functional materials for semiconductors, gas sensors, nonlinear optical limiters, liquid crystals and as sensitizers for photodynamic therapy (PDT) in cancer treatments, among many others, metal phthalocyanines (Pcs) are counted among the most extensively investigated dyes in history, besides others such as cyanine colorants and BODIPY derivatives [16][17][18].Metal naphthalocyanines (Ncs) constitute expanded analogues of their Pc counterparts that possess a linearly annulated benzene ring in the periphery of the macrocyclic core that leads to an about 100 nm red shift of the absorption band in the red if compared with Pcs.This shift, known as the rule of "100 nm", results from the destabilization of the HOMO (highest occupied molecular orbital) and the associated decrease in the HOMO-LUMO (lowest unoccupied molecular orbital) gap [19].Thus, Ncs are able to act as versatile tetradentate chelators for various metals and show a strong absorption in the near-infrared region (750-900 nm) with high molar absorption coefficients.The intense absorption and fluorescence bands of naphthalocyanines in the near-infrared offers interesting opportunities such as the fabrication of high-performance NIR photodiodes and bioimaging [20,21], and the complexation of openshell or late-transition element cations can add further functionalities related to spin-spin or spin-orbit coupling, respectively.
The structural flexibility of Ncs has been extensively illustrated through a diverse range of metal complexes and a wide array of substituents that can be attached to the periphery of the core or as axial ligands [19,[22][23][24][25][26][27].This chemical modification induces specific alterations in the electronic structure of the macrocyclic core, providing precise control over the physicochemical characteristics.However, despite their ability to absorb in the far red/near-infrared region, Ncs have been relatively neglected, mainly because of their intricate synthesis, with difficult purification, poor solubility, and a tendency to form inactive aggregates, owing to strong van der Waals interactions upon stacking [28,29].As a consequence of their high tendency to form aggregates, a drop of fluorescence and singlet oxygen quantum yields is observed [30,31].There have been several structural variations introduced to overcome the solubility and aggregation issues of (na)phthalocyanines, including the insertion of bulky substituents in the peripheral positions of the macrocycles [25,[32][33][34][35], the addition of axial ligands to the central atom [28,30,[36][37][38][39], or the encapsulation of the dye in colloidal particles [40,41].Furthermore, peripheral functional groups in Ncs macrocycles may be utilized to further tailor intra-and intermolecular interactions as well as optical, photochemical, and electrochemical properties; particularly, aromatic groups increase intermolecular π-π interactions [19,34].However, all these derivatives still exhibit a strong tendency to aggregate in aqueous media, which can significantly reduce their performance due to self-quenching [42].In the case of Pcs, these issues were solved by the introduction of several hydrophilic and amphiphilic groups (e.g., carboxylates, sulfonates, phosphonates, PEG chains) [43][44][45][46][47][48], as well as bulky axial ligands (e.g., −OSiMe 2 (CH 2 ) 3 NMe 2 and −OCH(CH 2 NMe 2 ) 2 ) [49][50][51][52][53] in the periphery of the macrocycle or at the central atom, respectively.Among these strategies, positively charged macrocycles are of particular interest, as they might target highly vulnerable intracellular sites and produce effective DNA damage in the context of phototherapy [42].Moreover, positively charged Pcs have been successfully employed for the photoinactivation of both Gram-negative and Gram-positive bacteria [54][55][56].However, in the case of water-soluble Ncs, the majority of the reported exemplars are based on silicon(IV)-based coordination compounds, mainly due to the presence of two axial positions that can be readily modified to reduce aggregation while improving solubility in aqueous media [27,57].While peripherally substituted water-soluble Pcs have been widely discussed in the literature [30,58], their Ncs counterparts with aque-ous solubility have remained vastly unexplored [29,59].Despite exhibiting NIR absorption and emission, Ncs exhibiting dual fluorescence (vis/NIR) have not been reported so far.
With this background in mind, we herein report on an unprecedented design pattern for single-component yet dual-emissive zinc(II) naphthalocyaninates, including a detailed synthesis as well as a structural and photophysical characterization.In order to suppress the intermolecular interactions and to improve their solubility, we aimed at the tailored substitution on the γ-position of the macrocyclic ligand to introduce eight 4-N,Ndimethylaminophen-4-yl moieties, which were further quaternized with methyl iodide to enhance their hydrophilicity.Moreover, substitution with 4-N,N-dimethylaminophen-4yl moieties opens up new possibilities for push-pull-based two-photon excitability and aggregation-independent luminescence in the visible region of the electromagnetic spectrum.To assess the charge-transfer nature of the excited state that was provided by the dimethylamino substitutents, an analogous 4-methoxyphen-4-yl-substituted zinc(II) naphthalocyaninate was synthesized.The assignment of the emissive states was carried out by TD-DFT calculations while rationalizing the electronic transitions dominating the absorption and emission spectra of these complexes.Thus, we anticipate that this strategy for preparing dual-emissive (visible and NIR) naphthalocyanines will pave the way for future innovation and accelerated developments in a broad range of application fields, including optoelectronics, energy conversion, and as contrast agents that are able to provide two orthogonal readouts for multiscale-multimodal imaging.

Design, Synthesis, Purification, and Structural Characterization
The peripheral functionalization was achieved by means of a Suzuki-Miyaura coupling reaction between 6,7-dibromo-2,3-dicyanonaphthalene (1) and the corresponding boronic acids, following our previously reported methodology [60].This substitution pattern was chosen in accordance with our recent report showing that the push-pull character and perpendicular-to-plane rigidified arrangement of the two phenyl moieties can provide strong luminescence in aggregates as well as in solution.The detailed synthetic procedure is shown in Section 3 (vide infra).The substituted naphthalonitriles NMe 2 and OMe were converted into the corresponding naphthalocyaninato zinc(II) complexes Zn-OMeNc and Zn-NMe 2 Nc by refluxing them in isoamyl alcohol (i-amOH) in the presence of 1,8-diazabicycloundec-7-ene (DBU) in catalytic amounts (Figure 1).
The reaction was performed under inert conditions in order to prevent the oxidation of the macrocycle by atmospheric oxygen at high temperatures.The obtained products contained some fluorescent impurities (most likely low-molecular-weight condensates preceding the cyclotetramerization); attempts to use column chromatography with silica or alumina as stationary phases in combination with classical organic solvent gradients failed, as the polar components of the reaction mixture strongly bind to both materials.Reverse-phase silica was also explored for purification; also in this case, it was observed that the product strongly binds to the stationary phase.Thus, the crude reaction mixture was subjected to multiple Soxhlet extractions to remove bulk impurities.The obtained products were further purified by size exclusion chromatography (Sephadex LH-20, Amersham Pharmacia Biotech AB, Uppsala, Sweden) in N,N-dimethylformamide (DMF).Finally, the water-soluble octa-cationic zinc(II) naphthalocyaninate Zn-NMe 3 Nc was obtained by treating Zn-NMe 2 Nc with an excess of methyl iodide in DMF (Figure 2).Besides UVvis absorption spectroscopy, the Ncs were structurally characterized by means of mass spectrometry and NMR spectroscopy.Due to their intrinsic aggregation tendency, attaining well-resolved 1 H-and 13 C-NMR spectra proved to be particularly challenging.Nonetheless, we were able to obtain precise 2D-NMR spectra (including an unambiguous assignment for all peaks).The reaction was performed under inert conditions in order to prevent the oxidation of the macrocycle by atmospheric oxygen at high temperatures.The obtained products contained some fluorescent impurities (most likely low-molecular-weight condensates preceding the cyclotetramerization); attempts to use column chromatography with silica or alumina as stationary phases in combination with classical organic solvent gradients failed, as the polar components of the reaction mixture strongly bind to both materials.Reverse-phase silica was also explored for purification; also in this case, it was observed that the product strongly binds to the stationary phase.Thus, the crude reaction mixture was subjected to multiple Soxhlet extractions to remove bulk impurities.The obtained products were further purified by size exclusion chromatography (Sephadex LH-20, Amersham Pharmacia Biotech AB, Uppsala, Sweden) in N,N-dimethylformamide (DMF).Finally, the water-soluble octa-cationic zinc(II) naphthalocyaninate Zn-NMe3Nc was obtained by treating Zn-NMe2Nc with an excess of methyl iodide in DMF (Figure 2).Besides UV-vis absorption spectroscopy, the Ncs were structurally characterized by means of mass spectrometry and NMR spectroscopy.Due to their intrinsic aggregation tendency, attaining well-resolved 1 H-and 13 C-NMR spectra proved to be particularly challenging.Nonetheless, we were able to obtain precise 2D-NMR spectra (including an unambiguous assignment for all peaks).In order to obtain a better resolution in the 1 H-NMR spectra of Zn-OMeNc, Zn-NMe2Nc, and Zn-NMe3Nc, DMSO was used as the solvent in NMR spectroscopy, as it suppresses aggregation caused by stacking.Due to the de-shielding effect of the porphyrazinato core, a clear downfield shift of Hα protons compared to Hβ was observed [34].The interpretation of all signals in the 1 H-NMR spectra of the Ncs Zn-OMeNc, Zn-NMe2Nc, and Zn-NMe3Nc were enabled by 1 H-1 H-COSY, 1 H-13 C-HSQC, 1 H-13 C-HMBC, and 1 H-1 H-ROESY spectroscopies, where all signals were unambiguously assigned (also for the naphthalonitriles, see SI, Figures S1-S26).From the 1 H-NMR spectra of Zn-OMeNc and Zn-NMe2Nc, it was clear that these two coordination compounds aggregate at the required concentrations, whereas the eight positive charges in the periphery of Zn-NMe3Nc suppressed aggregation and yielded well-resolved 1 H-NMR spectra (Figure S21).Hence, the proton signals from Zn-NMe3Nc were better resolved than those from Zn-OMeNc and Zn-NMe2Nc.In the 1 H-1 H-ROESY spectra, cross-peaks were observed between the signals of Hα and Hβ protons and between the protons of the phenyl groups and signals corresponding to Hβ (see SI, Figures S14, S20, and S26).From 1 H-1 H-COSY, additional groups of cross-peaks were found between the protons of the phenyl groups In order to obtain a better resolution in the 1 H-NMR spectra of Zn-OMeNc, Zn-NMe 2 Nc, and Zn-NMe 3 Nc, DMSO was used as the solvent in NMR spectroscopy, as it suppresses aggregation caused by stacking.Due to the de-shielding effect of the porphyrazinato core, a clear downfield shift of H α protons compared to H β was observed [34].The interpretation of all signals in the 1 H-NMR spectra of the Ncs Zn-OMeNc, Zn-NMe 2 Nc, and Zn-NMe 3 Nc were enabled by 1 H-1 H-COSY, 1 H-13 C-HSQC, 1 H-13 C-HMBC, and 1 H-1 H-ROESY spectroscopies, where all signals were unambiguously assigned (also for the naphthalonitriles, see SI, Figures S1-S26).From the 1 H-NMR spectra of Zn-OMeNc and Zn-NMe 2 Nc, it was clear that these two coordination compounds aggregate at the required concentrations, whereas the eight positive charges in the periphery of Zn-NMe 3 Nc suppressed aggregation and yielded well-resolved 1 H-NMR spectra (Figure S21).Hence, the proton signals from Zn-NMe 3 Nc were better resolved than those from Zn-OMeNc and Zn-NMe 2 Nc.In the 1 H-1 H-ROESY spectra, cross-peaks were observed between the signals of H α and H β protons and between the protons of the phenyl groups and signals corresponding to H β (see SI, Figures S14, S20 and S26).From 1 H-1 H-COSY, additional groups of cross-peaks were found between the protons of the phenyl groups (see SI, Figures S11, S17 and S23).The mass spectra of Zn-NMe 3 Nc were obtained by using ESI-EM, whereas for Zn-OMeNc and Zn-NMe 2 Nc, MALDI-TOF had to be employed.The MS demonstrated a good agreement between the expected and observed peaks (see SI, Figures S27-S31).

UV-Vis Absorption, Steady-State, and Time-Resolved Photoluminescence Spectroscopies
In agreement with the characteristic spectroscopic features of metal (na)phthalocyaninates, the UV-vis absorption spectra of the herein reported complexes feature a Soret-band (or B-band) in the UV range and Q-band in the NIR region [25].Figure 3 depicts the absorption spectra of Zn-OMeNc, Zn-NMe 2 Nc, and Zn-NMe 3 Nc in DMSO and in water.A red shift of the Q-band was observed when going from Zn-OMeNc (794 nm) to Zn-NMe 2 Nc (800 nm), implying that the energy gap between the HOMO and the LUMO is reduced when introducing a stronger electron π-donor (i.e., replacement of methoxy by dimethylamino moieties) [61,62].Consistently, in the case of Zn-NMe 3 Nc, a blue shift was observed (781 nm) in DMSO, due to the reduced electron-donating capacity of the cationic trimethylammonium groups (Table 1).Nonetheless, the eight positive charges of Zn-NMe3Nc suppress aggregation in DMSO, and the relative intensity of the Q-band vs. the B-band fall in the expected range for metal (na)phthalocyaninates [25,65].However, the absorption spectrum of Zn-NMe3Nc in water showed clear differences when compared to its spectrum in DMSO.In water, Zn-NMe3Nc still exhibits H-aggregation, as evidenced by the presence of two nonvibrational shoulders in the Q-band region (Figure 3d).The lower energy (red-shifted)  As shown in Figure 3a,b, the Q-band appears to be less intense than the B-band, which is surprising when considering that the latter is usually weaker in metal (na)phthalocyaninates.This observation can be attributed to the strong absorption in the blue and in the UV related to the eight methoxy-or dimethylamino-substituted moieties, which actually exhibit a sizeable push-pull character (in fact, this blue-shifted band is drastically reduced upon quaternization, as seen by the comparison of Zn-NMe 2 Nc with its derivative Zn-NMe 3 Nc; vide infra regarding the additional influence of aggregation phenomena on the relative intensities in the red and in the blue region of the spectrum).Similar trends were previously reported for Zn(II) phthalocyanines [63,64].The formation of red-shifted aggregates plays a role in the case of Zn-OMeNc, in agreement with the broadened signals that are observed in NMR spectroscopy (see SI, Figure S9); hence, the low molar absorption coefficient (ε) at λ max = 794 nm (ε = 4.4 × 10 4 M −1 cm −1 ) and the rather weak absorption band at around 876 nm point towards J-aggregation.Similarly, in the case of Zn-NMe 2 Nc, the intensity of the Q-band appears to be reduced if compared with the B-band, an observation that is reinforced by aggregation; therefore, a relatively low molar absorption coefficient at λ max = 800 nm (ε = 5.3 × 10 4 M −1 cm −1 ) is measured.Typically, the values for Q-bands in metal (na)phthalocyaninates are of the order of 10 5 M −1 cm −1 .Thus, by comparison of the three compounds presented herein, it is clear that the relatively high intensity of the B-band is prominently enhanced in the cases of Zn-OMeNc and Zn-NMe 2 Nc, due to the charge-transfer absorption of the eight aromatic electron-donating moieties (methoxy-or dimethylaminophenyl, respectively), which overlap with the B-band and disappear upon quaternization (vide infra); in addition, aggregation compromises the intensity of the Q-band.
Nonetheless, the eight positive charges of Zn-NMe 3 Nc suppress aggregation in DMSO, and the relative intensity of the Q-band vs. the B-band fall in the expected range for metal (na)phthalocyaninates [25,65].However, the absorption spectrum of Zn-NMe 3 Nc in water showed clear differences when compared to its spectrum in DMSO.In water, Zn-NMe 3 Nc still exhibits H-aggregation, as evidenced by the presence of two non-vibrational shoulders in the Q-band region (Figure 3d).The lower energy (red-shifted) band at about 784 nm can be attributed to the monomeric species, whereas the higher energy (blue-shifted) band at 732 nm suggests the presence of H-aggregates, as it was previously reported for comparable (na)phthalocyanines [29][30][31]66].The formation of H-aggregates in water is more evident when comparing the molar absorption coefficients (ε) of Zn-NMe 3 Nc in DMSO vs. water.In DMSO, the H-aggregates are disrupted, and Zn-NMe 3 Nc shows a characteristically high ε value at λ max = 781 nm (ε = 5.5 × 10 5 M −1 cm −1 ).This contrasts with the spectrum in water, where Zn-NMe 3 Nc peaks at λ max = 732 nm (ε = 2.2 × 10 5 M −1 cm −1 ), and λ max = 784 nm (ε = 1.8 × 10 5 M −1 cm −1 ) (Table 1).In water, H-aggregation reduces the Q-band intensity with respect to the B-band, and the molar absorption coefficient in the red drops by a factor of roughly two, which points toward dimerization.Despite the clear aggregation phenomena discussed above, the Lambert-Beer law was obeyed within the experimental range for all these compounds (see SI, Figures S35-S47) [67].In general, the molar absorption coefficients were obtained from the slope of the absorbance versus the concentration of the samples (the latter were determined by total reflection X-ray fluorescence (TXRF) addressing the concentration of Zn).
The UV-vis absorption, steady-state photoluminescence emission and excitation spectra of Zn-OMeNc, Zn-NMe 2 Nc, and Zn-NMe 3 Nc were measured in DMSO (Figure 4).In the case of Zn-NMe 3 Nc in water, a clear narrowing of the excitation spectrum was observed compared with the absorption spectrum, as only the emissive monomers were addressed, whereas all species (even aggregates) in a solution contribute to the broadened absorption.The proximity of the Q-band-related emission with respect to the Q-band excitation and absorption for Zn-OMeNc, Zn-NMe 2 Nc, and Zn-NMe 3 Nc in DMSO indicate that the nuclear configurations of the ground and excited states are comparable and therefore practically not affected by the excitation wavelength.
Int. J. Mol.Sci.2024, 25, x FOR PEER REVIEW 8 of 28 determined by total reflection X-ray fluorescence (TXRF) addressing the concentration of Zn).
The UV-vis absorption, steady-state photoluminescence emission and excitation spectra of Zn-OMeNc, Zn-NMe2Nc, and Zn-NMe3Nc were measured in DMSO (Figure 4).In the case of Zn-NMe3Nc in water, a clear narrowing of the excitation spectrum was observed compared with the absorption spectrum, as only the emissive monomers were addressed, whereas all species (even aggregates) in a solution contribute to the broadened absorption.The proximity of the Q-band-related emission with respect to the Q-band excitation and absorption for Zn-OMeNc, Zn-NMe2Nc, and Zn-NMe3Nc in DMSO indicate that the nuclear configurations of the ground and excited states are comparable and therefore practically not affected by the excitation wavelength.In agreement with previous reports on metal (na)phthalocyaninates, the photoluminescence emission spectra of Zn-OMeNc, Zn-NMe2Nc, and Zn-NMe3Nc mirror the excitation spectra with a typically small Stokes shift (Figure 4), if referring to the red-shifted Q-band.As can be observed in the full-range emission spectra (recorded while exciting in the UV), the compounds feature a surprisingly prominent dual emission, which will be discussed below (vide infra).This phenomenon was initially noticed while measuring the emission spectra of Zn-NMe2Nc in DMSO at different excitation wavelengths (λex).After varying λex, a broad emission band with a λem = 560 nm (λex = 335 nm) and a sharp nearinfrared emission with λem = 811 nm (λex = 750 nm) was observed (Figure 5c,d, Table 2).In In agreement with previous reports on metal (na)phthalocyaninates, the photoluminescence emission spectra of Zn-OMeNc, Zn-NMe 2 Nc, and Zn-NMe 3 Nc mirror the excitation spectra with a typically small Stokes shift (Figure 4), if referring to the red-shifted Q-band.As can be observed in the full-range emission spectra (recorded while exciting in the UV), the compounds feature a surprisingly prominent dual emission, which will be discussed below (vide infra).This phenomenon was initially noticed while measuring the emission spectra of Zn-NMe 2 Nc in DMSO at different excitation wavelengths (λ ex ).After varying λ ex , a broad emission band with a λ em = 560 nm (λ ex = 335 nm) and a sharp near-infrared emission with λ em = 811 nm (λ ex = 750 nm) was observed (Figure 5c,d, Table 2).In order to understand this rather unique dual emission, the analogous Zn-OMeNc was also designed, prepared, and studied.Interestingly, Zn-OMeNc also exhibited two clear emission bands, with a λ em = 508 nm (λ ex = 335 nm) and λ em = 804 nm (λ ex = 750 nm) in DMSO (Figure 5c,d, Table 2).It is important to mention that when using λ ex = 335 nm and screening the emission from the blue to the NIR region to acquire the full-range fluorescence spectra, both bands are visible in the case of Zn-OMeNc and Zn-NMe 2 Nc (Figure 5a,b, Table 2).
Int. J. Mol.Sci.2024, 25, x FOR PEER REVIEW 9 of 28 order to understand this rather unique dual emission, the analogous Zn-OMeNc was also designed, prepared, and studied.Interestingly, Zn-OMeNc also exhibited two clear emission bands, with a λem = 508 nm (λex = 335 nm) and λem = 804 nm (λex = 750 nm) in DMSO (Figure 5c,d, Table 2).It is important to mention that when using λex = 335 nm and screening the emission from the blue to the NIR region to acquire the full-range fluorescence spectra, both bands are visible in the case of Zn-OMeNc and Zn-NMe2Nc (Figure 5a,b, Table 2).A similar trend was observed for Zn-NMe 3 Nc when measured in water and in DMSO.In DMSO, Zn-NMe 3 Nc exhibited two emission bands, with a λ em = 550 nm (λ ex = 335 nm) and λ em = 801 nm (λ ex = 750 nm) (Figure 6c,d, Table 2).In contrast, the emission bands measured in water were observed at λ em = 432 nm (λ ex = 335 nm) and λ em = 804 nm (λ ex = 750 nm) (Figure 6c,d, Table 2).When a full-range emission spectrum for Zn-NMe 3 Nc in DMSO was measured (from blue to the NIR region using λ ex = 335 nm), a relatively faint emission band with λ em = 550 nm and a high-intensity luminescence with λ em = 800 nm were observed (Figure 6a).This can be attributed to the reduction in the push-pull character upon quaternization of the peripheral 4-N,N-dimethylaminophen-4-yl groups and the monomeric nature of Zn-NMe 3 Nc in DMSO, yielding a comparably strong NIR fluorescence and a weak visible emission.Thus, it is inferred that the monomers of Zn-NMe 3 Nc in DMSO yield an intense NIR emission and a comparatively weak luminescence between 400 nm and 500 nm.In contrast, the full-range emission spectrum of Zn-NMe 3 Nc in water (from blue to the NIR region using λ ex = 335 nm) exhibited a substantially stronger emission band, peaking at λ em = 400 nm, when compared with the NIR fluorescence band at λ em = 801 nm (Figure 6b); this can be understood when taking into account that in water, H-aggregation suppresses the fluorescence in the NIR, and a comparatively strong residual emission remains in the visible range.The dual emission of Zn-OMeNc, Zn-NMe 2 Nc, and Zn-NMe 3 Nc was interpreted and assigned by TD-DFT calculations (vide infra).
Table 2. Photophysical properties of the compounds (λ max and τ).P1 and P2 are the peaks in the visible and NIR regions in the emission spectra, respectively.For the bi-exponential photoluminescence decays, the amplitude-weighted average lifetimes (τ av_amp ) are shown.The τ values are rounded to match the significant figures of the uncertainties; raw time-resolved photoluminescence decays and fitting parameters are shown in the SI, Figures S48-S55.

Compound/ Solvent
Zn-OMeNc/DMSO 508/804 Zn-NMe 2 Nc/DMSO 560/811 Zn-NMe 3 Nc/DMSO 550/801 The dual fluorescence can be explained by two sets of singlet states (with orthogonal charge-transfer vs. main-ligand-centered character).The near infrared fluorescence originates from the S 1/2 → S 0 transition, and the visible emission stems from the S 3 → S 0 transition.The doubly degenerate S 1/2 state has a π-π * character, whereas the S 3 states have a charge-transfer (n-π * ) character involving the electron-rich methoxy-or dimethylaminosubstituted phenyl moieties.The energy gap between S 1/2 and S 3 states is relatively large; the lack of excited state geometry distortion (i.e., parallel potential hypersurfaces or "nested states") hamper the non-adiabatic crossover, leading to relatively slow internal conversion processes.Hence, in the case of Zn-OMeNc, Zn-NMe 2 Nc, and Zn-NMe 3 Nc, both an emission in the visible region and an NIR fluorescence are observed.In the case of Zn-NMe 3 Nc in water, the S 1/2 state is quenched due to aggregation, the intensity of the NIR emission is low, and the push-pull character is reduced due to quaternization.
Taking into account that the relative intensities of the two emission bands remain invariant when comparing independently synthesized batches, and considering the thorough purification that was carried out with no traces of impurities in the NMR spectra, we can confidently exclude spurious emission from side products or other contamination (photodecomposition is ruled out, as the absorption and emission spectra do not significantly vary during the photophysical characterization).Moreover, TD-DFT calculations support the assignment (vide infra).Taking into account that the relative intensities of the two emission bands remain invariant when comparing independently synthesized batches, and considering the thorough purification that was carried out with no traces of impurities in the NMR spectra, we can confidently exclude spurious emission from side products or other contamination (photodecomposition is ruled out, as the absorption and emission spectra do not significantly vary during the photophysical characterization).Moreover, TD-DFT calculations support the assignment (vide infra).

TD-DFT Calculations
In order to assign the transitions that were observed in the absorption and fluorescence spectra, we optimized the structures of Zn-OMeNc, Zn-NMe2Nc, and Zn-NMe3Nc in their electronic ground states (for computational details, see Section 3.4).All three complexes exhibited a highly planar core, which was marginally influenced by the peripheral substituents (see Figure 7).At the optimized ground state geometries, the UV-vis absorption spectra were calculated using TD-DFT/B3LYP.

TD-DFT Calculations
In order to assign the transitions that were observed in the absorption and fluorescence spectra, we optimized the structures of Zn-OMeNc, Zn-NMe 2 Nc, and Zn-NMe 3 Nc in their electronic ground states (for computational details, see Section 3.4).All three complexes exhibited a highly planar core, which was marginally influenced by the peripheral substituents (see Figure 7).At the optimized ground state geometries, the UV-vis absorption spectra were calculated using TD-DFT/B3LYP.As can be seen from Figure 8, the theoretical spectra generally agree well with the experimental data.Due to the degeneracy of the LUMO and LUMO + 1 orbitals, the S1 and S2 states are also doubly degenerate, as they can be described by HOMO → LUMO and HOMO → LUMO+1 excitations, respectively.The band corresponding to the S0 → S1/2 transition-located at 805 nm, 756 nm, and 740 nm for Zn-NMe2Nc, Zn-OMeNc, and Zn-NMe3Nc, respectively-shows a progressive blue shift with decreasing electron donor ability, in good agreement with the experimental data (and also qualitatively with respect to the relative intensities of the B-and Q-bands).The high intensity of this NIR maximum can be explained by the fact that it represents a local excitation within the 2,3-naphthalocyaninato ligand (the black part of the structure depicted in Figure 1, see Figure 9 for a visualization of the relevant frontier orbitals).As can be seen from Figure 8, the theoretical spectra generally agree well with the experimental data.Due to the degeneracy of the LUMO and LUMO + 1 orbitals, the S 1 and S 2 states are also doubly degenerate, as they can be described by HOMO → LUMO and HOMO → LUMO+1 excitations, respectively.The band corresponding to the S 0 → S 1/2 transition-located at 805 nm, 756 nm, and 740 nm for Zn-NMe 2 Nc, Zn-OMeNc, and Zn-NMe 3 Nc, respectively-shows a progressive blue shift with decreasing electron donor ability, in good agreement with the experimental data (and also qualitatively with respect to the relative intensities of the B-and Q-bands).The high intensity of this NIR maximum can be explained by the fact that it represents a local excitation within the 2,3naphthalocyaninato ligand (the black part of the structure depicted in Figure 1, see Figure 9 for a visualization of the relevant frontier orbitals).As can be seen from Figure 8, the theoretical spectra generally agree well with the experimental data.Due to the degeneracy of the LUMO and LUMO + 1 orbitals, the S1 and S2 states are also doubly degenerate, as they can be described by HOMO → LUMO and HOMO → LUMO+1 excitations, respectively.The band corresponding to the S0 → S1/2 transition-located at 805 nm, 756 nm, and 740 nm for Zn-NMe2Nc, Zn-OMeNc, and Zn-NMe3Nc, respectively-shows a progressive blue shift with decreasing electron donor ability, in good agreement with the experimental data (and also qualitatively with respect to the relative intensities of the B-and Q-bands).The high intensity of this NIR maximum can be explained by the fact that it represents a local excitation within the 2,3-naphthalocyaninato ligand (the black part of the structure depicted in Figure 1, see Figure 9 for a visualization of the relevant frontier orbitals).In addition, we calculated the 0-0 fluorescence wavelengths from the S1/2 and S3 states by optimizing the respective excited state geometries and subtracting the optimized ground state energy.While the S1/2 → S0 de-excitation involved a local LUMO/LUMO + 1 → HOMO configuration change (Figure 9), which was reasonably well described by all functionals that were tested herein, the S3 → S0 de-excitation turned out to be much more problematic, as it involved a pronounced charge-transfer character for the relevant excited state-at least for Zn-OMeNc and Zn-NMe2Nc (see Figure 10).For Zn-NMe2Nc, geometry optimization with the B3LYP functional leads to a collapse of the S3 state onto the S1/2 state, yielding spurious emission wavelengths.The long-range corrected functional CAM-B3LYP, on the other hand, significantly overestimated the S3 energy (see SI, Table S1).In fact, the PBE0 functional emerged as the best compromise for all three compounds, considering both emission lines in the visible and the NIR (see Figure 11).It is able to reproduce the position of the experimental 0-0 peaks accurately, including the slight blue shift when going from Zn-NMe2Nc to Zn-NMe3Nc via Zn-OMeNc.For Zn-OMeNc, the S3 peak was also described very well, while for Zn-NMe2Nc, the agreement with the experiment was particularly poor, with the PBE0 peak being red-shifted by about 150 nm.In addition, we calculated the 0-0 fluorescence wavelengths from the S 1/2 and S 3 states by optimizing the respective excited state geometries and subtracting the optimized ground state energy.While the S 1/2 → S 0 de-excitation involved a local LUMO/LUMO + 1 → HOMO configuration change (Figure 9), which was reasonably well described by all functionals that were tested herein, the S 3 → S 0 de-excitation turned out to be much more problematic, as it involved a pronounced charge-transfer character for the relevant excited state-at least for Zn-OMeNc and Zn-NMe 2 Nc (see Figure 10).For Zn-NMe 2 Nc, geometry optimization with the B3LYP functional leads to a collapse of the S 3 state onto the S 1/2 state, yielding spurious emission wavelengths.The long-range corrected functional CAM-B3LYP, on the other hand, significantly overestimated the S 3 energy (see SI, Table S1).In fact, the PBE0 functional emerged as the best compromise for all three compounds, considering both emission lines in the visible and the NIR (see Figure 11).It is able to reproduce the position of the experimental 0-0 peaks accurately, including the slight blue shift when going from Zn-NMe 2 Nc to Zn-NMe 3 Nc via Zn-OMeNc.For Zn-OMeNc, the S 3 peak was also described very well, while for Zn-NMe 2 Nc, the agreement with the experiment was particularly poor, with the PBE0 peak being red-shifted by about 150 nm.
Presumably, this is due to the stronger charge-transfer character in the case of Zn-NMe 2 Nc, with a notable amount of charge density being located at the NMe 2 substituent in the HOMO-1.In addition, the S 3 state contains a 7% contribution of the local HOMO → LUMO + 1 excitation.The use of the M06-2X functional leads to a considerable improvement for Zn-NMe 2 Nc (see Figure 11), but not for the other two compounds.PBE0 predicts the S 3 state to be governed by a HOMO-1 → LUMO excitation in the case of Zn-NMe 2 Nc and Zn-OMeNc, whereas for Zn-NMe 3 Nc, the S 3 state is dominated by a HOMO → LUMO + 2 excitation (Figure 10).Presumably, this is due to the stronger charge-transfer character in the case of Zn-NMe2Nc, with a notable amount of charge density being located at the NMe2 substituent in the HOMO-1.In addition, the S3 state contains a 7% contribution of the local HOMO → LUMO + 1 excitation.The use of the M06-2X functional leads to a considerable  Presumably, this is due to the stronger charge-transfer character in the case of Zn-NMe2Nc, with a notable amount of charge density being located at the NMe2 substituent in the HOMO-1.In addition, the S3 state contains a 7% contribution of the local HOMO → LUMO + 1 excitation.The use of the M06-2X functional leads to a considerable

MSOT Imaging in Gel Phantoms
In recent studies, it has been demonstrated that naphthalocyaninato complexes (Ncs) hold significant potential as photoacoustic contrast agents [27,68,69].In comparison with other organic compounds, such as cyanine dyes and nanoparticles, naphthalocyanines exhibit higher molar absorption coefficients in the NIR region, featuring sharp maxima between 770 nm and 850 nm [16].The well-defined NIR absorption peaks mentioned above contribute to more precise spectral unmixing, distinguishing them from other exogenous contrast agents [70].As a result, Zn-NMe 2 Nc and Zn-NMe 3 Nc were employed as representative examples for generating photoacoustic (PA) images utilizing multi-spectral optoacoustic tomography (MSOT).These measurements were conducted within tissuemimicking gel phantoms, as depicted in Figure 12  predicts the S3 state to be governed by a HOMO-1 → LUMO excitation in the case of Zn-NMe2Nc and Zn-OMeNc, whereas for Zn-NMe3Nc, the S3 state is dominated by a HOMO → LUMO + 2 excitation (Figure 10).

MSOT Imaging in Gel Phantoms
In recent studies, it has been demonstrated that naphthalocyaninato complexes (Ncs) hold significant potential as photoacoustic contrast agents [27,68,69].In comparison with other organic compounds, such as cyanine dyes and nanoparticles, naphthalocyanines exhibit higher molar absorption coefficients in the NIR region, featuring sharp maxima between 770 nm and 850 nm [16].The well-defined NIR absorption peaks mentioned above contribute to more precise spectral unmixing, distinguishing them from other exogenous contrast agents [70].As a result, Zn-NMe2Nc and Zn-NMe3Nc were employed as representative examples for generating photoacoustic (PA) images utilizing multi-spectral optoacoustic tomography (MSOT).These measurements were conducted within tissuemimicking gel phantoms, as depicted in Figure 12 [69].The experiments were carried out at different concentrations, and the obtained photoacoustic profiles qualitatively agree with the absorption spectra regarding the wavelengths that induce the highest photoacoustic response.The phantom studies were performed by dissolving Zn-NMe2Nc in DMSO, Zn-NMe3Nc in DMSO, and Zn-NMe3Nc in water, respectively.Among these, Zn-NMe3Nc in DMSO provided the highest photoacoustic signal even at a low concentration (1 µM), with Zn-NMe2Nc in DMSO and Zn-NMe3Nc in water showing similar yet slightly lower PA signals (Figure 13).The observed difference in signal intensity of Zn-NMe3Nc in DMSO and water can be attributed to the monomerization in DMSO and aggregation in water.Also, when comparing the photoacoustic (Figure 13) and absorption spectra (Figure 3c) of Zn-NMe3Nc in DMSO at λmax = 781 nm, the relative intensities of the vibronic features appear distorted.For the photoacoustic spectra of Zn-NMe3Nc in DMSO, the inner filter effect limits the excitation at the main maxima, even at very low concentrations (1 µM), while causing a saturation plateau between λmax = 781 nm and λmax = 700 nm.As Zn-NMe3Nc in DMSO provides the best photoacoustic signals at a low concentration, it is clear that the suppression of aggregation enhances the photoacoustic output.In the future, this could be further improved with vanadyl-or siliconnaphthalocyanines, where the axial ligands could prevent aggregation [30,69,71]; the insertion of vanadyl centers could further shift the absorption maxima to longer wavelengths in the infrared region [72][73][74].13).The observed difference in signal intensity of Zn-NMe 3 Nc in DMSO and water can be attributed to the monomerization in DMSO and aggregation in water.Also, when comparing the photoacoustic (Figure 13) and absorption spectra (Figure 3c) of Zn-NMe 3 Nc in DMSO at λ max = 781 nm, the relative intensities of the vibronic features appear distorted.For the photoacoustic spectra of Zn-NMe 3 Nc in DMSO, the inner filter effect limits the excitation at the main maxima, even at very low concentrations (1 µM), while causing a saturation plateau between λ max = 781 nm and λ max = 700 nm.As Zn-NMe 3 Nc in DMSO provides the best photoacoustic signals at a low concentration, it is clear that the suppression of aggregation enhances the photoacoustic output.In the future, this could be further improved with vanadyl-or silicon-naphthalocyanines, where the axial ligands could prevent aggregation [30,69,71]; the insertion of vanadyl centers could further shift the absorption maxima to longer wavelengths in the infrared region [72][73][74].

Time-Resolved Multiphoton Micro(spectro)scopy
In order to demonstrate the ability to image structural features by time-resolved multiphoton micro(spectro)scopy, we employed Zn-NMe2Nc to stain 1 µm sized aminated polystyrene microparticles (PSMPs) by a swelling-diffusion processes [75], as these are well-established models for microstructured samples (Figure 14b).As depicted in Figure 14, the photoluminescence emission spectra of Zn-NMe2Nc-loaded PSMPs (Zn-NMe2Nc@PSMP) in water show a broad emission band in the green (λmax ≈ 470 nm), but no emission was observed in the near-infrared region (Figure 14a); this points towards aggregation-caused quenching of the fluorescence from the macrocycle, as it was previously reported for other (na)phthalocyanines [76].However, the distinct emission in the visible region, originating from the push-pull character of the excited side groups, can be utilized as an additional optical readout, despite the lack of intrinsic luminescence from the macrocycle.In fact, the push-pull side groups have been previously introduced in our recent work as robust luminophores, both in aggregated and in monomeric forms [60].The excited state lifetime of the stained microparticles in liquid water was determined, and a τav_amp = 5.1 ns was obtained (Figure S56).

Time-Resolved Multiphoton Micro(spectro)scopy
In order to demonstrate the ability to image structural features by time-resolved multiphoton micro(spectro)scopy, we employed Zn-NMe 2 Nc to stain 1 µm sized aminated polystyrene microparticles (PSMPs) by a swelling-diffusion processes [75], as these are wellestablished models for microstructured samples (Figure 14b).As depicted in Figure 14, the photoluminescence emission spectra of Zn-NMe 2 Nc-loaded PSMPs (Zn-NMe 2 Nc@PSMP) in water show a broad emission band in the green (λ max ≈ 470 nm), but no emission was observed in the near-infrared region (Figure 14a); this points towards aggregation-caused quenching of the fluorescence from the macrocycle, as it was previously reported for other (na)phthalocyanines [76].However, the distinct emission in the visible region, originating from the push-pull character of the excited side groups, can be utilized as an additional optical readout, despite the lack of intrinsic luminescence from the macrocycle.In fact, the push-pull side groups have been previously introduced in our recent work as robust luminophores, both in aggregated and in monomeric forms [60].The excited state lifetime of the stained microparticles in liquid water was determined, and a τ av_amp = 5.1 ns was obtained (Figure S56).
Further experiments were carried out to assess the photophysical properties of discrete Zn-NMe 2 Nc@PSMP.Briefly, Zn-NMe 2 Nc@PSMP were placed on a microscope slide and analyzed using single-photon excitation (SPE) to carry out fluorescence lifetime imaging microscopy (FLIM).As depicted in Figure 15, the particles were homogeneously stained with the push-pull luminophore, confirming the observation that is shown in Figure 14, while possessing a homogeneously distributed amplitude-weighted average lifetime of τ ≈ 3.9 ns (Figures 15 and S57).In addition, employing a spectrophotometer coupled by an optical fiber to the confocal microscope, the emission spectra of discrete Zn-NMe 2 Nc@PSMP entities were obtained as dry samples while consistently reproducing the results for Zn-NMe 2 Nc@PSMP that was suspended in liquid water (Figure 15).In contrast to single-photon excitation (SPE), two-photon excitation (TPE) laser scanning microscopy is often anticipated to increase cell survival and tissue penetration [77].Thus, we explored herein a way to increase the detection sensitivity due to an improved sample penetration employing a near-infrared TPE laser; we therefore performed comparable FLIM experiments using a high-intensity Ti:Sa oscillator as the two-photon excitation source (Figure 15), where comparable photoluminescence lifetime maps were obtained.In addition, the emission spectra of discrete Zn-NMe 2 Nc@PSMP were also measured by TPE.These experiments demonstrate that no significant differences are to be expected when using SPE and TPE (both in lifetimes (Figures S57 and S58) and in the emission spectra) for optical imaging, while providing a second readout for microscopic imaging.Further experiments were carried out to assess the photophysical properties of discrete Zn-NMe2Nc@PSMP.Briefly, Zn-NMe2Nc@PSMP were placed on a microscope slide and analyzed using single-photon excitation (SPE) to carry out fluorescence lifetime imaging microscopy (FLIM).As depicted in Figure 15, the particles were homogeneously stained with the push-pull luminophore, confirming the observation that is shown in Figure 14, while possessing a homogeneously distributed amplitude-weighted average lifetime of τ ≈ 3.9 ns (Figures 15 and S57).In addition, employing a spectrophotometer coupled by an optical fiber to the confocal microscope, the emission spectra of discrete Zn-NMe2Nc@PSMP entities were obtained as dry samples while consistently reproducing the results for Zn-NMe2Nc@PSMP that was suspended in liquid water (Figure 15).In contrast to single-photon excitation (SPE), two-photon excitation (TPE) laser scanning microscopy is often anticipated to increase cell survival and tissue penetration [77].Thus, we explored herein a way to increase the detection sensitivity due to an improved sample penetration employing a near-infrared TPE laser; we therefore performed comparable FLIM experiments using a high-intensity Ti:Sa oscillator as the two-photon excitation source (Figure 15), where comparable photoluminescence lifetime maps were obtained.In addition, the emission spectra of discrete Zn-NMe2Nc@PSMP were also measured by TPE.These experiments demonstrate that no significant differences are to be expected when using SPE and TPE (both in lifetimes (Figures S57 and S58) and in the emission spectra) for optical imaging, while providing a second readout for microscopic imaging.

Materials and Methods
All chemicals were purchased with the maximum quality available from Sigma Aldrich (Taufkirchen, Germany) or from TCI Chemicals (Eschborn, Germany) and used without further purification.Reactions were carried out using dried solvents (99.9% purity) under argon atmosphere.They were monitored by thin-layer chromatography (TLC), which was performed on 0.2 mm Macherey-Nagel ALUGRAM (Eupen, Belgium) precoated silica gel aluminum sheets.Spots were visualized by a UV handlamp (254 and 365

Materials and Methods
All chemicals were purchased with the maximum quality available from Sigma Aldrich (Taufkirchen, Germany) or from TCI Chemicals (Eschborn, Germany) and used without further purification.Reactions were carried out using dried solvents (99.9% purity) under argon atmosphere.They were monitored by thin-layer chromatography (TLC), which was performed on 0.2 mm Macherey-Nagel ALUGRAM (Eupen, Belgium) precoated silica gel aluminum sheets.Spots were visualized by a UV handlamp (254 and 365 nm).Silica gel 60 (0.063-0.200 mm, Merck, Darmstadt, Germany) was used for column chromatography, if not otherwise stated.Fresh spectroscopic-grade solvents (Uvasol ® , Merck, Darmstadt, Germany) were utilized for the spectroscopic studies.
3.1.5.Synthesis of 3,4,12,13,21,22,30,31-Octakis(4-methoxyphenyl) Naphthalocyaninato Zinc(II) (Zn-OMeNc) Zn-OMeNc was prepared by treating OMe (0.1 g, 0.25 mmol) with Zn(OAc) 2 •2H 2 O (0.028 g, 0.127 mmol) and using DBU as the catalyst in 10 mL i-amOH.The obtained product was precipitated with a MeOH:H 2 O (10:1 v/v) mixture and filtered.The purification of the obtained products involved a sequential two-step approach.Initially, a Soxhlet extraction was employed using solvents of increasing polarity (from diethyl ether to hexane and ethyl acetate).Subsequently, the resulting product underwent a final purification using size exclusion chromatography with a resin (Sephadex-LH20) in N,N-dimethylformamide.The collected fractions were concentrated to obtain the Zn-OMeNc.Molecular formula: C 104 H 72 N 8 O 8 Zn (brown solid).Yield: 21% (90 mg, 0.05 mmol).Zn-NMe 2 Nc was prepared by treating NMe 2 (0.100 g, 0.24 mmol) with Zn(OAc) 2 •2H 2 O (0.026 g, 0.120 mmol) and using DBU as the catalyst in 10 mL i-amOH.The obtained product was precipitated with a MeOH:H 2 O (10:1 v/v) mixture and filtered.The purification of the obtained products involved a sequential two-step approach.Initially, a Soxhlet extraction was performed using solvents of increasing polarity (from diethyl ether to hexane and ethyl acetate).Subsequently, the resulting product underwent final purification using size exclusion chromatography with a resin (Sephadex-LH20) in N,N-dimethylformamide.Zn-NMe 3 Nc was prepared by treating Zn-NMe 2 Nc (0.030 g, 0.017 mmol) with methyl iodide (excess) in N,N-dimethylformamide for 3 days to quarternize the amino groups (Figure 2).Once the reaction was completed, the desired product was precipitated using acetone and then dried.The compound was further purified with a size exclusion resin (Sephadex-G25, Amersham Pharmacia Biotech AB, Uppsala, Sweden) using water as the eluent.The collected fractions were subjected to lyophilization to obtain Zn-NMe 3 Nc.Molecular formula: [C 120 H 120 N 16 Zn]I 8 (greenish brown solid).Yield: 87% (28 mg, 0.015 mmol). 1 µm sized aminated polystyrene microparticles (PSMPs) were bought from Micromod GmbH (Rostock, Germany).The highly hydrophobic complex was encapsulated in the 1 µm sized aminated polystyrene microparticles (PSMPs) via a simple one-step staining procedure as follows: First, 50 mg/mL of PS microparticles were diluted in deionized water to 2.5 mg/mL (1.2 mL).A 0.8 mM solution of the ZnNMe 2 Nc complex in tetrahydrofuran (THF) was prepared.Then, 200 µL of this solution in THF was added to the PSMP dispersion in water, and the sample was subsequently shaken for 40 min.Shrinkage of the particles was induced by adding 200 µL of deionized water to the dispersion, thereby encapsulating the complex in the microparticles.The particles were then centrifuged at 16.000 rpm for 5 min (Megafuge™ 16 from Thermo Scientific, Schwerte, Germany).The precipitate was washed three times with ethanol/water mixtures (volume ratios of 40/60, 30/70, and 20/80) and once with deionized water to remove the excess of dye that was adsorbed onto the particles' surface.Finally, the suspension containing the PSMPs loaded with the ZnNMe 2 Nc complex was diluted to 5 mg/mL (600 µL).

NMR Spectroscopy and Mass Spectrometry
NMR spectra were obtained at the Institut für Anorganische und Analytische Chemie (Universität Münster), using Bruker Avance Neo 500 and Bruker Avance III 400.All measurements were performed at room temperature unless otherwise specified.The 1 H-NMR and 13 C-NMR chemical shifts (δ) of the signals are given in parts per million and are referenced to the residual proton signal in the deuterated solvent DCM-d 2 ( 1 H: 5.32 ppm/ 13 C: 54.0 ppm) or DMSO-d 6 ( 1 H: 2.50 ppm/ 13 C: 39.52 ppm).The signal multiplicities are abbreviated as follows: s, singlet; d, doublet; t, triplet; q, quartet; br, broad; m, multiplet.
Steady-state excitation and emission spectra were recorded on a FluoTime 300 spectrometer from PicoQuant (Berlin, Germany), equipped with a 300 W ozone-free Xe lamp (250-900 nm), a 10 W Xe flash lamp (250-900 nm, pulse width < 10 µs) with repetition rates of 0.1-300 Hz, a singlegrating excitation monochromator (Czerny-Turner 2.7 nm/mm dispersion, 1200 grooves/mm, blazed at 300 nm), diode lasers (pulse width < 80 ps) operated by a computer-controlled laser driver PDL-820 (repetition rate up to 80 MHz, burst mode for slow and weak decays), two singlegrating emission monochromators (Czerny-Turner, selectable gratings blazed at 500 nm with 2.7 nm/mm dispersion and 1200 grooves/mm, or blazed at 1250 nm with 5.4 nm/mm dispersion and 600 grooves/mm), Glan-Thompson polarizers for excitation (Xe-lamps) and emission, a Peltier-thermostatized sample holder from Quantum Northwest (Washington, DC, USA) (−40-105 • C), and two detectors, namely, a PMA Hybrid 40 (transit time spread FWHM < 120 ps, 300-720 nm) and a R5509-42 NIR photomultiplier tube (transit time spread FWHM 1.5 ns, 300-1400 nm) with external cooling (−80 • C) from Hamamatsu (Shizuoka, Japan).Steady-state and fluorescence lifetimes were recorded in the TCSPC mode by a PicoHarp 300 (minimum base resolution 4 ps) from PicoQuant (Berlin, Germany).Emission and excitation spectra were corrected for source intensity (lamp and grating) by standard correction curves.Lifetime analysis was performed using the commercial EasyTau software (version 2.2) package from PicoQuant (Berlin, Germany).The quality of the fit was assessed by minimizing the reduced chi squared function (χ 2 ) and through visual inspection of the weighted residuals and their autocorrelation.At least three independent batches of the compounds were synthesized and measured, with the samples undergoing multiple measurements to ensure data consistency and to mitigate measurement errors.Across these different batches, the results were consistently identical, underscoring the reliability of our findings.

Computational Details
All density functional theory (DFT) and time-dependent density functional theory (TD-DFT) calculations were performed using the quantum chemistry package Gaussian 09 Rev. D.01 [80].Geometry optimizations in the ground state S 0 were carried out with DFT, while TD-DFT was employed to obtain the optimized S 1 and S 3 excited state structures.The solvent dimethyl sulfoxide (DMSO) was modelled by the polarizable continuum model (PCM) in an integral equation formalism framework [81] with atomic radii from the universal force field model (UFF) [82].In all cases, the spectral properties were calculated using the same functional as employed for geometry optimization.
UV-vis absorption spectra were calculated using the B3LYP functional [83], along with Grimme's D3 dispersion correction with Becke-Johnson damping (BJ) [84] and the SDD basis set, which applies an effective core potential for the Zn atom [85] and the D95 basis set for H, C, N, and O atoms [86].For each absorption spectrum, the 120 lowest excited singlet states were considered.A Lorentzian broadening with a half width of half maximum (HWHM) of 30 nm was used for each transition.
The emission wavelengths were obtained from the difference in energy between the optimized ground state and the corresponding excited state geometries (0-0).In addition to the B3LYP functional, emission spectra were also calculated with the CAM-B3LYP [87], PBE0 [88], and M06-2X [89] functionals due to the delicate balance between local and charge-transfer excitations.A Lorentzian broadening with a half width of half maximum (HWHM) of 20 nm was used for the S 1 → S 0 transition, while a value of 60 nm was applied for the S 3 → S 0 transition to match the experimental line shape.

MSOT Photoacoustic Imaging
Phantom photoacoustic imaging experiments were performed using a Multispectral Optoacoustic Tomography system (MSOT inVision 512-echo, iThera Medical, Munich, Germany).The phantom consists of a cylinder of 1.5% agarose with 2% of 20% soybean cream as scattering medium and 0.75% 2.5 mM Nigrosin, which was cast, including 2 straws as placeholders.Once hardened, these placeholders were removed and replaced with sealed straws, which had been filled with test solution and a control containing only the solvent.Multispectral photoacoustic imaging of phantoms containing the probe and control samples were performed tomographically for the length of the straw using an excitation wavelength ranging from 680 to 900 nm in steps of 5 nm.Images were loaded into an in-house developed image processing software, MEDgical (Version 0.9.9, EIMI, Münster, Germany).Spectra were exported and processed to create the presented plots.

Fluorescence Lifetime Imaging Microscopy (FLIM) by Time-Resolved Multiphoton Micro(spectro)scopy
Fluorescence lifetime imaging microscopy (FLIM) was recorded on a fluorescence microscope IX 73 from Olympus (Shinjuku, Japan) with a complete confocal system and a laser-combining unit (LCU), an inverted microscope body, and a multichannel detection unit, namely a MicroTime 200 from PicoQuant (Berlin, Germany) equipped with diode lasers (providing adjustable output power and repetition rates up to 80 MHz inside a compact fiber couple unit with wavelengths between 375 and 900 nm).For beam diagnostics, a charge-coupled device (CCD) camera and a photodiode were available in the main optical unit (MOU) of the microscope.The MOU was equipped with two detectors, namely, a hybrid photomultiplier-based single photon counting module (PMA Hybrid 40, PicoQuant) and an SPAD-based photon-counting module (SPCM-AQR-14, Perkin-Elmer, Hopkinton, MA, USA).Different band-pass (BP) and low-pass (LP) filters were placed before these detectors on demand to acquire lifetime maps.Data acquisition was based on the unique time-tagged time-resolved (TTTR) measurement mode, where simultaneous data acquisition on two channels is possible.Data were processed and analyzed with the SymphoTime 64 (PicoQuant) software.In order to couple the MicroTime 200 and the FluoTime 300 instruments (both from PicoQuant, Berlin, Germany), a fiber coupler was employed.In this way, the spectrometer can be used to record either steady-state or time-resolved luminescence spectra and decays from a sample mounted on the microscope.Luminescence micrographs were acquired using the same microscope mentioned above, equipped with an X-CiteQ Lamp module (Excelitas Technologies, Waltham, MA, USA) as excitation source and a UI-5580SE (IDS) digital camera.Different band-pass (BP) and low-pass (LP) cubes were using accordingly.
Additionally, a two-photon (Mai Tai ® from SpectraPhysics, Darmstadt, Germany) Ti:Sapphire laser (Ti:Sa oscillator) with a pulse with <100 fs and tuning range of 690-1040 nm, connected to the MOU, can be used as excitation source.In order to reduce the repetition rate, the Ti:Sapphire laser was connected to a double pulse picker (A.P.E.® , Berlin, Germany).

Conclusions
In conclusion, we report on a set of novel Zn(II) naphthalocyaninates with a prominent dual fluorescence encompassing the visible and the NIR portion of the electromagnetic spectrum.A proper synthesis, purification, and full characterization of the peripherally octa-substituted zinc(II) naphthalocyaninato complexes Zn-OMeNc, Zn-NMe 2 Nc, and Zn-NMe 3 Nc was undertaken for the first time; hence, a detailed structural elucidation based on NMR studies is provided.The introduction of a push-pull system on the macrocycle enabled the integration of two orthogonal chromophores with unprecedented dual emission, which was assigned by means of TD-DFT calculations.Upon quaternization of

Figure 5 .Table 2 .Figure 5 .
Figure 5. Fluorescence in DMSO: (a) full-range emission spectrum of Zn-OMeNc (blue), λex = 335 nm; (b) full-range spectrum of Zn-NMe2Nc (red), λex = 335 nm; (c) normalized emission spectra of Zn-OMeNc (blue) and Zn-NMe2Nc (red) in the visible region, λex = 335 nm; (d) normalized emission spectra of Zn-OMeNc (blue) and Zn-NMe2Nc (red) in the near-infrared region, λex = 750 nm.In the full-range emission spectra (a,b), the bands may appear slightly different (if compared with (c,d), both in shape and intensity), due to the different instrumental settings and filters employed (for details, see Section 3.3).Table2.Photophysical properties of the compounds (λmax and τ).P1 and P2 are the peaks in the visible and NIR regions in the emission spectra, respectively.For the bi-exponential photoluminescence decays, the amplitude-weighted average lifetimes (τav_amp) are shown.The τ values are rounded to match the significant figures of the uncertainties; raw time-resolved photoluminescence decays and fitting parameters are shown in the SI, FiguresS48-S55.Compound/ Solvent

Figure 6 .
Figure 6.Fluorescence in DMSO and in water: (a) full-range emission spectrum of Zn-NMe3Nc (blue) in DMSO, λex = 335 nm; (b) full-range emission spectrum of Zn-NMe3Nc (red) in water, λex = 335 nm; (c) normalized emission spectra of Zn-NMe3Nc in DMSO (blue) and water (red) in the visible region, λex = 335 nm; (d) normalized emission spectra of Zn-NMe3Nc in DMSO (blue) and water (red) in the near-infrared region, λex = 750 nm.In the full-range emission spectra (a,b), the bands may appear slightly different (if compared with (c,d), both in shape and intensity) due to the different instrumental settings and filters employed (for details see Section 3.3).

Figure 6 .
Figure 6.Fluorescence in DMSO and in water: (a) full-range emission spectrum of Zn-NMe 3 Nc (blue) in DMSO, λ ex = 335 nm; (b) full-range emission spectrum of Zn-NMe 3 Nc (red) in water, λ ex = 335 nm; (c) normalized emission spectra of Zn-NMe 3 Nc in DMSO (blue) and water (red) in the visible region, λ ex = 335 nm; (d) normalized emission spectra of Zn-NMe 3 Nc in DMSO (blue)and water (red) in the near-infrared region, λ ex = 750 nm.In the full-range emission spectra (a,b), the bands may appear slightly different (if compared with (c,d), both in shape and intensity) due to the different instrumental settings and filters employed (for details see Section 3.3).
[69].The experiments were carried out at different concentrations, and the obtained photoacoustic profiles qualitatively agree with the absorption spectra regarding the wavelengths that induce the highest photoacoustic response.The phantom studies were performed by dissolving Zn-NMe 2 Nc in DMSO, Zn-NMe 3 Nc in DMSO, and Zn-NMe 3 Nc in water, respectively.

Figure 12 .
Figure 12.MSOT photoacoustic images of gel phantoms at different concentrations: (a) Zn-Nme 2 Nc in DMSO, (b) Zn-Nme 3 Nc in DMSO, and (c) Zn-NMe 3 Nc in water (λ ex = 745 nm).Among these, Zn-NMe 3 Nc in DMSO provided the highest photoacoustic signal even at a low concentration (1 µM), with Zn-NMe 2 Nc in DMSO and Zn-NMe 3 Nc in water showing similar yet slightly lower PA signals (Figure13).The observed difference in signal intensity of Zn-NMe 3 Nc in DMSO and water can be attributed to the monomerization in DMSO and aggregation in water.Also, when comparing the photoacoustic (Figure13) and absorption spectra (Figure3c) of Zn-NMe 3 Nc in DMSO at λ max = 781 nm, the relative intensities of the vibronic features appear distorted.For the photoacoustic spectra of Zn-NMe 3 Nc in DMSO, the inner filter effect limits the excitation at the main maxima, even at very low concentrations (1 µM), while causing a saturation plateau between λ max = 781 nm and λ max = 700 nm.As Zn-NMe 3 Nc in DMSO provides the best photoacoustic signals at a low concentration, it is clear that the suppression of aggregation enhances the photoacoustic output.In the future, this could be further improved with vanadyl-or silicon-naphthalocyanines, where the axial ligands could prevent aggregation[30,69,71]; the insertion of vanadyl centers could further shift the absorption maxima to longer wavelengths in the infrared region[72][73][74].

Table 1 .
Q-and B-bands observed in the absorption spectra.The corresponding molar absorption coefficient (ε) for each band is also listed in square brackets.