Fully Conjugated Porphyrin Glass: Collective Light-Harvesting Antenna for Near-Infrared Fluorescence beyond 1 μm.

Expanded π-systems with a narrow highest occupied molecular orbital-lowest unoccupied molecular orbital band gap encounter deactivation of excitons due to the "energy gap law" and undesired aggregation. This dilemma generally thwarts the near-infrared (NIR) luminescence of organic π-systems. A sophisticated cofacially stacked π-system is known to involve exponentially tailed disorder, which displays exceptionally red-shifted fluorescence even as only a marginal emission component. Enhancement of the tail-state fluorescence might be advantageous to achieve NIR photoluminescence with an expected collective light-harvesting antenna effect as follows: (i) efficient light-harvesting capacity due to intense electronic absorption, (ii) a long-distance exciton migration into the tail state based on a high spatial density of the chromophore site, and (iii) substantial transmission of NIR emission to circumvent the inner filter effect. Suppression of aggregation-induced quenching of fluorescence could realize collective light-harvesting antenna for NIR-luminescence materials. This study discloses an enhanced tail-state NIR fluorescence of a self-standing porphyrin film at 1138 nm with a moderate quantum efficiency based on a fully π-conjugated porphyrin that adopts an amorphous form, called "porphyrin glass".

Introduction

There is a growing fascination with near-infrared (NIR) photoluminescent materials aiming at potential applications in a wide range of fields along with the availability of the semiconductor, indium gallium arsenide (InGaAs), to detect NIR light longer than 1 μm, namely, in the shortwave infrared (SWIR) region.[1−3] A narrow highest occupied molecular orbital–lowest unoccupied molecular orbital band gap is indispensable to NIR photoluminescence. Among them, fully π-conjugated low-band-gap polymers have hitherto been a pivotal subject in organic photovoltaic applications because NIR light-harvesting capacity has a crucial role in photovoltaic efficiency.[4−7] However, narrow energy gaps are intrinsically prone to fast deactivation, known as the “energy gap law”.[8,9] Additionally, the expansion of π-conjugated systems is usually accompanied by a drawback arising from the propensity for π-stacking. These bottlenecks make the design of bright NIR photoluminescent chromophores and polymers highly challenging. Actually, lowering the band gap of π-conjugated polymers is not always effective in shifting the emission wavelength to the red unlike the absorption band; poor emission efficiency of some narrow-band-gap polymers apparently violates Kasha’s law,[10−13] although that defines the fluorescence as radiative decay from the lowest optical absorption band.[14] To overcome the intrinsic inefficiency of NIR/SWIR emission, high excitation efficiency should ensure the brightness of the emission, considering the fact that brightness of emission is defined as the product of light absorption efficiency and emission quantum yield. In this consequence, a strongly light-absorbing material is a possible candidate to yield bright NIR/SWIR emission, even if the emission quantum yield remains inefficient.

A greater absorption band enhances not only efficient light-harvesting capacity but also the inner filter effect in which fluorescence is usually reabsorbed, thus reducing the emission intensity. A large Stokes shift of the emission should increase only the light-harvesting antenna effect by circumventing the inner filter effect. It is known that the structural distribution of chromophore assemblies is the sum of Gaussian statistics and tail states, such as exponential tail[15,16] or heavy-tailed Lévy distributions (Scheme and Figure ).[17−21] Although the tail states display only a marginal fluorescence component, enhanced tail-state fluorescence might be a rational strategy to obtain effective NIR/SWIR emission. We envisioned that bulk chromophore aggregates could evolve the emission wavelength up to the lowest edge of the electronic absorption band through enhanced exciton funneling efficiency. Accordingly, only the tail-state fluorescence could be substantially obtained. From this point of view, bulk chromophore assemblies may be intriguing as NIR/SWIR luminescent materials.

Scheme 1

Synthesis of 2

Figure 1

Schematic representation of the density of states and the corresponding fluorescence spectrum as the sum of Gaussian statistics (black solid lines) and exponential tails (red dotted line) (A) and the space coordinate of the occupied density of states (B).

Our approach to achieving NIR photoluminescent materials involves imparting a glassy nature to zinc porphyrin by accommodation with two 3,4,5-tri((S)-3,7-dimethyloctyloxy)phenyl groups at the meso-positions, called “porphyrin glass” in this study.[22] Molecular glasses or amorphous molecular materials have introduced an excellent morphological strategy in organic optoelectronic devices based on small molecules and polymers.[23−26] The porphyrin glass is highly resistant to crystallization and is tolerant of expansion of the π-system through supramolecular and covalent polymerization.[27,28] In principle, the amorphous or glass phase is kinetically constrained from crystallization and incorporates a free volume in the frozen liquid by losing macroscopic fluidity. Thus, the free volume allows some extent of microscopic conformational fluctuation in the photoexcited state, leading to excimer formation. Consequently, the glassy butadiynylene-bridged porphyrin dimer formed the excimer and exhibited solid-state NIR photoluminescence at approximately 970 nm with submicrosecond time constants,[22] and further expansion of stacked π-systems through supramolecular polymerization achieved SWIR photoluminescence at approximately 1 μm with microsecond decay time constants.[27] In contrast to a thermodynamically preferable supramolecular polymer structure due to the reversible backbone formation, the covalently linked polymer backbone likely entraps kinetically preferable states. Encouraged by the fact that no aggregation-caused fluorescence quench occurred in porphyrin glass, we explored the NIR/SWIR fluorescence properties of the covalent polymer of the porphyrin glass, expecting the increased probability of the formation of the tail states in bulk.

Oligomeric conjugated porphyrin arrays demonstrate excellent NIR fluorescence in organic solutions,[29,30] where the meso-ethynylene conjugation is a privileged motif of excellent π-electronic features, as established by Anderson et al.[31−34] and Therien et al.[29,35−39] To date, π-conjugation engineering of meso-ethynylene porphyrin arrays has been extended to fully conjugated polymers.[40−52] However, the porphyrin-conjugated polymers possess a drawback arising from large π-systems, such as undesired aggregation. Featuring an intrinsically glass-forming porphyrin unit, the “molecular wire effect” could be effective in incorporating a light-harvesting antenna function,[53−56] in a manner similar to the supramolecular polymer of porphyrin glass.[27] Our present objective focuses on a covalently linked polymer glass with a fully π-conjugated backbone.

One of the most efficient π-conjugations of aryl-bridged ethynyl-conjugated porphyrin dimers has been achieved by a possible resonance between a benzenoid–acetylene and quinoidal–cumulenic conjugation. This approach was proposed by Anderson and co-workers using a 9,10-anthracene-bridged ethynyl-conjugated porphyrin dimer for the first time[57−59] and has been reported for other aryl-bridges such as thiophene,[57−59] pentacene,[60] benzobisthiadiazole,[60] dimethylthiadiazoloquinoxaline,[61] and dithienometallole.[62] However, further extension to an alternating anthracene–porphyrin polymer has not been conducted. Along with the study of porphyrin glass, we recently reported an alternating anthracene–porphyrin polymer 1 (Mn = 54 000, Mw/Mn = 4.2, Tg = 94 °C) that formed a metal-lustrous self-standing film, “porphyrin foil” (Figure ).[28] Polymer 1 exhibited a remarkably red-shifted Q band at 904 nm, together with a high refractive index, as an outcome of the considerable persistence length of the π-conjugated backbone.[63,64] Compared with a newly synthesized model compound 2, this study discusses an SWIR fluorescent system based on the collective light-harvesting antenna effect of the porphyrin foil.

Figure 2

Chemical structure of 1 in a benzenoid–acetylene form and a quinoid–cumulene form (A), and photographs of the toluene solution in a cuvette (B) and the porphyrin foil (C) under ambient light. Chemical structure of a model compound 2 (D).

Results and Discussion

NIR Emission Behaviors in Solution

Polymer 1 exhibited an exceptionally long Q band at 904 nm, whose intensity was unusually stronger than that of the Soret band at 438–543 nm (Figure A). In principle, the Q band of porphyrin is optically only partially allowed, and therefore the Q band is usually much weaker than the Soret band.[31] To the best of our knowledge, an unusually intense Q band was not reported except for a few exceptions: butadiyne-linked octaethylporphyrin array[40−42] and the meso-ethynylene-conjugated porphyrin array with an alternating pyrido[3,4-b]pyrazine-zinc porphyrin sequence[45] other than a supramolecular polymer of the porphyrin glass in our previous study.[27] The intense and red-shifted Q band should be an outcome of the increased π-conjugation.

Figure 3

(A) Electronic absorption spectra of 1 at 298 K in toluene (red line) and spectrometric titration with pyridine (gray and green lines). (B) Fluorescence spectra of 1 in toluene under argon-saturated conditions obtained by excitation at 450 nm in the absence and presence of pyridine (red and green, respectively), together with normalized excitation spectra monitored at 800 nm (gray line) and 1000 nm (black line) in the absence of pyridine.

Extended π-system often causes undesired aggregation even under dilute conditions. The addition of an axial ligand is effective in the dissociation of stacked zinc porphyrins. The addition of pyridine to 1 gave rise to a blue shift of the Q band from 904 to 885 nm with remarkable sharpening, although the emission maxima shifted to the red (Figure A). The observed spectral change was unusual, considering that axial coordination typically resulted in a red-shift of the Q band. Actually, a titration experiment of model compound 2 with pyridine resulted in a small red-shift of both the Soret bands at approximately 440 and 530 nm, of the Q bands from 690 to 710 nm, and fluorescence maxima (Figure ), where the electronic structures were relevant to the reported homologue.[57] The spectral change through several pseudo-isosbestic and isoemissive points and the estimated binding strength indicated that 2 was substantially free from aggregation under the experimental conditions. Comparison with the electronic structure of 2 emphasizes the exceptionally sharp Q band of 1 in the presence of pyridine together with the extended π-conjugation. The band width of the Q band, which mirrors the torsional conformations around the ethynylene linkage, was reminiscent of that of the conjugated porphyrin array with all of the units fully planar,[32,34] suggesting a narrow conformational distribution. Considering the very low torsional barrier around the ethynylene linkage,[57] the 3,4,5-tri((S)-3,7-dimethyloctyloxy)phenyl groups may enhance the geometrical complementarity of π-stacked interactions, similar to the π-stacked porphyrin arrays even in the presence of pyridine.[65−67] We consider that a higher molecular weight component of polymer 1 could provide a π-stacked and/or entangled structure. We discuss the electronic properties of polymer 1 in the absence of pyridine to directly compare the photoelectronic properties in solution with those in a pyridine-free amorphous phase.

Figure 4

Spectrometric titration of 2 ([2]0 = 1.4 × 10–4 M, red line) with pyridine (gray lines up to 1 × 102 equivalent, and 1 × 104 equivalent, green line) in electronic absorption and fluorescence spectra at 25 °C in toluene (A). Stepwise binding constants were estimated as K1 = 4.1 × 104 M–1 and K2 = 8.3 × 103 M–1 based on global titration analyses (B).

Photoluminescence properties of polymer 1 were different from those of Kasha’s law[14] at a glance. Polymer 1 showed a relatively strong emission at 827 nm, together with a weak emission at 944 nm (Figure B), whose emission lifetime was ranged in the time scale of fluorescence. It should be noticed that the principal emission at 827 nm was attributed to the energy level at 1.53 eV (809 nm as the midpoint of the deconvoluted absorption maximum and emission maximum), higher than the lowest excited state at 1.34 eV (929 nm as the midpoint of the deconvoluted absorption maximum and the longer emission maximum) (Figure ). The fluorescence maximum at 827 nm was much longer than that of 2 at 743 nm in toluene, indicating that the energy level of the fluorescent species of 1 is lower than that of the π-conjugated domain of a porphyrin–anthracene–porphyrin unit.

Figure 5

Possible Jablonski diagram, combined with the electronic absorption spectrum deconvoluted with Gaussian-curve fitting (yellow) and the fluorescence spectrum (green) in toluene.

A similar fluorescence behavior is known for low-band-gap polymers, such as poly{benzo[1,2-b:4,5-b′]dithiophene-2,6-diyl}{3-fluorothieno[3,4-b]thiophenediyl}, called “PYB7” it has one of the most excellent properties for photovoltaic applications, exhibiting fluorescence at a shorter wavelength than the longest absorption band. This effect is attributed to the folded structure.[6−10] To access the fluorescence properties, the excitation spectra (Figure B) were compared with the deconvoluted electronic absorption spectrum (Figure ). The distinctive difference of the longer emission at 944 nm from the shorter emission at 827 nm was the greater contribution of the excitation at 543 nm. Additionally, the emission efficiency at 944 nm was very low when the longest Q band at 909 nm was excited. Then, the entire observations may be interpreted in a manner similar to the case of PYB7; photoexcitation of polymer 1 produced two distinct states, that is, a localized exciton on an isolated shorter chain and a relatively delocalized exciton on the stacked or entangled longer chain. The lowest level at 1.34 eV may be prone to deactivation according to the energy gap law,[13,14] thus dissipating the exciton to reduce the emission efficiency at 944 nm. In solution, exciton funneling from the localized state to the delocalized state may be ineffective due to their large spatial separation between energy-donating and energy-accepting sites, where the Förster critical radius governs the distance of exciton migration.[67] Thus, the shorter emission band at 827 nm was relatively conspicuous. The rational mechanism for the fluorescence behaviors of 1 in toluene is proposed in Figure .

The above explanation for the complicated fluorescence behavior of polymer 1 assumes partial aggregation or entanglement of the polymer backbone depending on molecular weight. A spin-cast film of 2 prepared from toluene solution would provide aggregates, whose fluorescence behavior could provide fruitful insights into the fluorescence properties of polymer 1. The spin-cast film of 2 showed a split Q band at 680 and 770 nm and the finely structured Soret band at 443 and 537 nm due to exciton coupling (Figure ), unambiguously indicating partial aggregation. The emission of the spin-cast film at 958 nm showed remarkably large Stokes shift, suggesting excimer formation. The close similarity of the emission wavelength of the spin-cast film of 2 at 958 nm and the solution of 1 at 944 nm may indicate structural resemblance of the emission species.

Figure 6

Electronic absorption (black), fluorescence (red), and excitation (gray) spectra of a spin-cast film of 2 on a quartz substrate. Emission and excitation spectra were obtained by excitation at 443 nm and monitoring at 958 nm, respectively.

Solid-State SWIR Photoluminescence

In the second stage, our attention was focused on the porphyrin foil of 1, a bulk self-standing neat film approximately 160 μm in thickness (Figure C). In the porphyrin foil, 1 adopted a form of amorphous phase at room temperature.[28] The porphyrin foil exhibited emission at 1138 and 950 nm (Figure ). In this case, the considerable overlap of intense absorption of the porphyrin foil should distort the spectral shape in the wavelength region shorter than the Q band and the normal emission should partly emerge at 950 nm, although the SWIR emission longer than the Q band should also be transmitted. Thus, the spectral shape of the SWIR fluorescence of the porphyrin foil was not strictly the same depending on the experimental conditions (Figures , 8, and S3). The small difference in the excitation efficiency and/or the collective light-harvesting antenna efficiency also affected the intensity of the fluorescence at approximately 950 nm, although the wavelength of the SWIR fluorescence band was reproducible. More importantly, the lack of the emission at 809 nm suggested that the localized exciton was substantially converted to the delocalized exciton, and consequently the tail state. The NIR/SWIR emission of 1 was obtained in 0.26% of external quantum yield, which encompassed a moderate efficiency as the emission in this wavelength region. Although the quantum efficiency remained comparable to the conventional NIR-luminescence chromophores, the considerable light-harvesting capacity of the porphyrin foil led to significant brightness of the SWIR emission.

Figure 7

Emission–excitation contour map of the porphyrin foil. Excitation spectrum monitored at 1000 nm (black line), absorption spectrum obtained by spectroscopic ellipsometry (gray line, adapted from ref (28)) (left panel), and emission spectra obtained by excitation at 470 and 800 nm (upper panel). It should be noticed that no emission at 827 nm was found.

Figure 8

Possible Jablonski diagram, combined with the electronic absorption spectrum in toluene (yellow) and the fluorescence spectrum of porphyrin foil (red).

The spectral shape showed no excitation wavelength dependence (Figure ), unlike the fluorescence properties in toluene. This fact provides direct evidence that the energy conversion efficiency to the lowest level was unity irrespective of the initial photoexcited level. In the porphyrin foil, the close proximity of the fully conjugated porphyrin arrays should be appropriate for enhancing interchromophore Coulombic coupling for efficient exciton funneling to the lowest energy trap.[67] Although the inner filter effect should also involve trivial energy transfer through emission/re-excitation processes, the contribution should be insignificant considering substantial fluorescence quantum yield.

On the contrary, the spin-cast film of 1 exhibited very faint fluorescence, whose SWIR component was only marginal (Figure S4; details described in the Supporting Information). The collective light-harvesting antenna effect shifted the emission wavelength up to the lowest edge of the electronic absorption band and produced SWIR luminescence.

The emission lifetime measurements of the porphyrin foil revealed that the emission at 1138 nm decayed within a subnanosecond time scale (Figure S3), elucidating that the SWIR emission was definitively in the time range of fluorescence, in contrast with the unusually long emission lifetime of the solid-state NIR/SWIR photoluminescence of previous porphyrin glasses.[23,27] Therefore, the SWIR fluorescence was likely assigned to the radiative process from the self-trapped exciton at the tail state. The interpretation is most pertinent to the SWIR fluorescence of the porphyrin foil, as represented in Figure .

The comparison of the synchrotron wide-angle X-ray diffraction (WAXD) profile of the porphyrin foil and glazing-incidence (GI) WAXD profile of the spin-cast film may corroborate the above fluorescence behaviors (Figure A). The whole WAXD peaks of the porphyrin foil were broader than the GIWAXD peaks, indicating that the crystalline domains of polymer 1 in the bulk were smaller or more disordered than those in the spin-cast film. The intensity of the primary GIWAXD peak at d = ca. 30 Å depended on the azimuthal angle (Figure B). On the basis of the geometry-optimized model, the d-spacing of ca. 30 Å may be assignable to the rodlike polymer stacking (Figure A). At an azimuthal angle of 90° along the q axis, the peak intensity showed the maximum (Figure C), indicating that the chain axis of the rodlike polymer chains was parallel to the surface of the silicon wafer and they stack on one another perpendicular to the surface. Our previous study revealed that the 3,4,5-tri((S)-3,7-dimethyloctyloxy)phenyl groups segregate to the outer surface and reduce the surface free energy of the porphyrin foil.[28] Judging from these results, the polymer chains confined by the surface differentiated from the bulk.

Figure 9

WAXD profile of the porphyrin foil (red line, adapted from ref (28)) and GIWAXD profile of the spin-cast film of the porphyrin array on a silicon wafer (blue line) after background correction (A), where representative peaks are shown as d-spacing (d = 2π/q). The observed GIWAXD pattern (B) and azimuthal-angle dependence of the intensity of the primary GIWAXD peak at the magnitude of the scattering vector, q, of approximately 2.1 nm–1 (C).

Figure 10

Geometry-optimized models of the bundle structure of π-stacked chains (blue and red chains) (A), and distorted single backbone (B) produced using an MM+ force field (HyperChem Ver. 8.0 software).

The population of the tail state should be minor and unobservable by GIWAXD measurements. In this context, the GIWAXD peaks observed the structures that acted as the light-harvesting antenna. We propose possible structures based on the geometry-optimized model for the WAXD peaks (Figure ). The anthracene bridging unit of 1 produces a space to stack with the porphyrin ring on the juxtaposed backbone, and the interdigitation of 3,4,5-tri((S)-3,7-dimethyloctyloxy)phenyl groups provides a π-stacked structure with a longitudinal displacement (Figure A). The structure is consistent with Bragg’s reflections for the interchain Zn···Zn distances of 10.3 and 8.7 Å as observed at the magnitude of the scattering vector, q, of 6.1 and 7.2 nm–1, respectively. The peak found at the d-spacing of 16.7 Å implied a nonlinearly ordered backbone with ca. 16.7 Å of the intrachain Zn···Zn spacing (Figure B) and overwhelmed the peak of the 18 Å of the intrachain Zn···Zn separation in the π-stacked backbone. Additionally, the broad peak at q = ca. 14 nm–1 was attributed to the halo peak for amorphous alkyl chains, being accompanied by small peaks corresponding to the interchain π-stacked spacing of 3.7 Å. On the contrary, in the GIWAXD pattern measured for the spin-cast film of 1, with rodlike polymer stacking having d = ca. 30 Å, some possible π-stacked structures were observed (Figure B). On the basis of the fact that the SWIR fluorescence quantum yield ranged in a moderate order, these structures did not quench the photoexcited singlet. A slipped-cofacial arrangement of π-stacked porphyrin rings (Figure A) could be rational for a long-distance exciton migration because of the close resemblance of the chromophore arrangement in the bacterial light-harvesting antenna complexes.[68−70]

Of particular importance, the amorphous halo peak at q = ca. 14 nm–1 in the porphyrin foil was more significant than that in the spin-cast film, suggesting that the polymer chain incorporated free volume more as relieving the surface confinement in a bulk. On the basis of this comparison, we interpret that the bulk porphyrin assembly raised the probability of the formation of the tail states.

To ensure photofunctionality under conditions of exceptionally high spatial density of porphyrin rings, an effective morphological strategy featuring porphyrin glass should play a key role in the suppression of undesired aggregation. Therefore, collectively, the light-harvesting antenna function of the porphyrin foil imparted the SWIR with fluorescent properties based on the following factors: (i) an intense electronic absorption band captured the excitation light, (ii) the high spatial density of porphyrin mediated the long-distance exciton migration, and (iii) the substantial transmission of NIR emission circumvented the inner filter effect. The bright SWIR fluorescent site was likely assigned to the tail state, where the plausible structure was the interchain π-stacked structure or the distorted backbone.

Conclusions

Considerable absorption by the porphyrin foil showed substantial light-harvesting capacity in the excitation light as well as reabsorption of the fluorescence. The collective light-harvesting capacity efficiently funneled the exciton into the tail state, despite its low population. The exceptional Stokes shift from the absorption band circumvented the inner filter effect and yielded the SWIR fluorescence at 1138 nm at a moderate efficiency. The porphyrin glasses in relatively thermodynamically preferable states, such as porphyrin dimer and supramolecular polymer, showed excimer-like emission with sub- or microsecond time constants, as shown in our previous study.[22,27] By contrast, the covalently linked backbone of the fully conjugated porphyrin glass kinetically entrapped the tail states, wherein the lack of undesired aggregation of the porphyrin rings is of particular importance to accomplish the collective light-harvesting antenna function. The present approach highlighted a new methodology to ensure the brightness of the SWIR emission with an inevitably low quantum efficiency.

Experimental Section

Materials

Synthesis of polymer 1 was described in our previous report.[28] Model compound 2 was newly synthesized.

Synthesis of 2

The model compound 2 was newly synthesized as shown in Scheme . To a solution of precursor porphyrin 3 (0.15 g, 87 μmol)[65] in diisopropylamine (5 mL) degassed by successive freeze–pump–thaw cycles was added Pd(PPh3)4 (4.8 mg, 4.2 μmol), 9,10-diethynylanthracene (7.5 mg, 29 μmol), and CuI (1.4 mg, 7.4 μmol). The mixture was stirred at 60 °C under argon atmosphere for 15 h and then diluted with toluene to be washed with water. The organic layer separated was passed through a silica gel chromatography with toluene as the eluent. Target compound 2 was isolated by size-exclusion chromatography (toluene/pyridine, 95:5, v/v). 2 was obtained as a dark green substance in 5% yield (5.3 mg, 1.5 μmol). 1H NMR (600 MHz, CDCl3): δ 10.05 (d, J = 4.5 Hz, 4H; porphyrin-β), 9.71 (d, J = 4.5 Hz, 4H; porphyrin-β), 9.50 (dd, J = 6.3, 3.3 Hz, 4H; anthryl), 9.12 (d, J = 4.5 Hz, 4H; porphyrin-β), 8.99 (d, J = 4.5 Hz, 4H; porphyrin-β), 7.96 (dd, J = 6.3, 3.3 Hz, 4H; anthryl), 7.43 (s, 8H; meso-Ar), 4.45–4.05 (m, 24H; Ar–O–CH2−), 2.19–0.80 (m, 270H; alkyl). 13C NMR (151 MHz, CDCl3): δ 152.7, 152.2, 151.3, 150.5, 150.2, 138.2, 137.1, 133.3, 133.0, 131.3, 131.1, 128.0, 127.4, 123.3, 114.3, 72.1, 67.8, 39.5, 39.3, 37.7, 37.4, 36.6, 30.0, 29.7, 28.1, 27.9, 24.9, 24.7, 22.8, 22.7, 22.6, 19.8, 19.7, 19.1, 12.0. Matrix-assisted laser-dissociationionization time-of-flight (MALDI-TOF) mass spectra (dithranol): m/z calc for C224H326N8O12Si2Zn2: 3504.33; found 3504.56 [M]+.

Spectral Methods

NMR spectra were recorded on Bruker AVANCE III 600, and matrix-assisted laser-dissociationionization time-of-flight (MALDI-TOF) mass spectra were performed by Bruker Autoflex Speed. Electronic absorption spectra were recorded on a spectrophotometer (UV-1800, SHIMADZU) equipped with a Peltier thermoelectric temperature-controlling unit (TCC-240A, SHIMADZU). The NIR steady-state emission spectra were recorded on a Fluorolog 3 ps NIR spectroscopy system (HORIBA) and corrected for the response of the detector system. Emission lifetimes (τ) of the spin-cast film were measured by nano-LEDs (N-355, response time ≤1.2 ns, HORIBA, Kyoto) and a photomultiplier (R5108, response time ≤1.1 ns, Hamamatsu Photonics, Hamamatsu). Emission lifetimes were determined from the slope of logarithmic plots of the decay profiles. The absolute quantum measurement was performed by QE-5000 (Otsuka Electronics, Osaka) using an integrated sphere equipped with an excitation laser (450 nm). Time-resolved NIR fluorescence was performed by a microscopic UV/IR spectroscopic system (NX-FILM-T03, Tokyo Instruments, Inc.). GIWAXD experiments of the spin-cast film were performed using synchrotron radiation at the BL45XU beamline in SPring-8 (RIKEN SPring-8 Centre Hyogo, Japan) using a PILATUS 3X 2M (Dectris Ltd.) at 308.6 mm of the sample-to-detector distance.[71] Resonance Raman spectra were performed by a microscopic spectroscopic Raman spectrometer (HORIBA, LabRAM ARAMIS) by using 532 nm excitation wavelength (Nd:YVO4 laser), wherein the NIR luminescence did not overlap with Raman shift.