Panchromatic light funneling through the synergy in hexabenzocoronene–(metallo)porphyrin–fullerene assemblies to realize the separation of charges

Here, we present a novel butadiyne-linked HBC-ethynyl-porphyrin dimer, which exhibits in the ground state strong absorption cross sections throughout the UV and visible ranges of the solar spectrum. In short, a unidirectional flow of excited state energy from the HBC termini to the (metallo)porphyrin focal points enables concentrating light at the latter. Control over excitonic interactions within, for example, the electron-donating porphyrin dimers was realized by complexation of bidentate ligands to set up panchromatic absorption that extends all the way into the near-infrared range. The bidentate binding motif was then exploited to create a supramolecular electron donor–acceptor assembly based on a HBC-ethynyl-porphyrin dimer and an electron accepting bis(aminoalkyl)-substituted fullerene. Of great relevance is the fact that charge separation from the photoexcited HBC-ethynyl-porphyrin dimer to the bis(aminoalkyl)-substituted fullerene is activated not only upon photoexciting the HBCs in the UV as well as the (metallo)porphyrins in the visible but also in the NIR. Implicit is the synergetic interplay of energy and charge transfer in a photosynthetic mimicking manner. The dimer and bis-HBC-ethynyl-porphyrin monomers, which serve as references, were probed by means of steady-state as well as time-resolved optical spectroscopies, including global target analyses of the time-resolved transient absorption data.


Introduction
Exploiting the full spectrum of solar radiation for light harvesting applications remains a challenging task. In particular, architectures based on molecular chromophores are oen limited to discrete absorption bands. This pitfall has been overcome in nature by means of self-assembled superstructures of chlorophylls, which made porphyrins one of the most important classes of molecular entities in photosynthesis. Nowadays, they are also important model compounds in elds as diverse as catalysis, molecular electronics, sensing, and, last but not least, electron-transfer applications. [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19] Their strong absorption throughout the visible part of the solar spectrum makes them useful for solar-energy applications. Much work has been devoted to rening and tuning these features. Ethynyl substitution, for example, of the porphyrin core in the mesoand/or b-positions has been shown to shi their absorption bands and alter the corresponding relative intensities. [20][21][22][23][24][25][26] Porphyrin (homo/hetero) dimers linked by p-conjugated oligomeric bridges have been studied extensively, not only with regard to their spectroscopic properties, but also to their propensity to build supramolecular architectures and mediate charge transport between different porphyrins. [27][28][29][30][31][32][33][34][35][36][37] Another way of tailoring porphyrins is to couple polycyclic aromatic hydrocarbons, PAHs, directly to the porphyrin core (e.g. tetraphenylporphyrins, TPPs). Likewise, the incorporation of peripheral ethynyl groups allows for tuning ground-and excited-state interactions. 25,26,38,39 Hexa-perihexabenzo-coronenes (HBCs) have attracted a great deal of attention because of their electronic and optoelectronic properties. Aer a synthesis for soluble HBCs was published by Müllen and coworkers in the late 1990s, 40,41 HBCs have recently found use as graphene model systems. 42,43 Several reports have described the use of conjugates of HBC with perylenediimides and/or porphyrins for photochemical applications albeit with the use of rotationally exible bonds. [44][45][46][47][48][49][50] One of the detrimental consequences is a rather moderate electronic coupling between the individual building blocks and, in turn, an ineffective ow of charges across the entire conjugate in addition to structural exibility, conformational freedom, etc.
In this work, we present the design of a new family of covalently-linked, ethynyl-bridged (metallo)porphyrin-HBC conjugates in their free-base (1) and/or metallated with Cu (1-Cu) or Zn (1-Zn) forms, which were complemented by an HBCterminated, butadiyne-bridged zinc-porphyrin (ZnP) dimer 2-Zn. Of great importance for our molecular design is the choice of ethynyl bridges as they foster the panchromatic absorption throughout the UV and visible regions of the solar spectrum all the way into the near-infrared region. Going beyond this aspect, ethynyl bridges enable control over the electronic couplings between the HBCs and the (metallo)porphyrins, on one hand, and the ow of energy/charges, on the other hand. An incentive for placing the HBCs at the termini and the (metallo)porphyrins at the focal points is to set up a gradient along which a unidirectional ow of energy supports the pooling of light. Ultimately, the efficient utilization of the concentrated light is realized by coordinating, for example, an electron-accepting C 60 to the (metallo) porphyrin dimer. As such, the synergy of the aforementioned characteristics enables the perfect blueprint of mimicking photosynthesis, namely the basic principles of light-harvesting arrays and reaction centers. The absorptive and emissive properties of all conjugates in their ground and excited states were characterized thoroughly by an arsenal of steady-state and timeresolved optical spectroscopies. The insight gathered for 1, 1-Cu, and 1-Zn helped interpret the excited-state reactivity of 2-Zn coordinated to 3: the sequence of a unidirectional ow of energy followed by that of charges. Experimental results are complemented by molecular modeling to address the coordinationinduced changes in the excited-state characteristics (Fig. 1).
For details of the synthetic procedures and characterization of intermediate products, please refer to the ESI. †
Inspection of the spectra reveals two main contributions, the HBC-centered absorption features with a major peak at 360/ 361 nm and two minor peaks at 346 and 391/392 nm, and the (metallo)porphyrin-centered absorption features between 400 and 800 nm. Signicant differences are found relative to, for example, well-known zinc tetraphenylporphyrin (ZnTPP). Ref-Zn exhibits a split Soret-band absorption that is red-shied relative to ZnTPP in THF. The most intense Q-band absorption is found at 636 nm with two smaller bands/shoulders at 584 and 620 nm, whereas ZnTPP displays two distinct Q-band absorptions at 550 and 595 nm.
Both the Soret-and Q-band absorptions in 1-Zn are further red-shied as well as broadened and result in an asymmetric Soret-band with a 461 nm maximum and an asymmetric Q-band with a 675 nm maximum. Similar changes are discernable for 1-Cu and 1, that is, red-shied absorptions and lower number of broadened Q-bands. A major contribution to these effects stems  Top: Extinction coefficients of the investigated systems and references in THF (Note: individual spectra set off by 10 5 M À1 cm À1 for clarity); Bottom: Fluorescence spectra of the investigated systems and reference HBC(tBu) 6 in THF normalized to the respective emission maximum; excitation at 350 nm for 1-Cu, 1-Zn, 2-Zn, HBC(tBu) 6 (second order diffraction peak at 700 nm), 360 nm for 1.
from the presence of ethynyl groups in the meso-positions of the porphyrins. On the one hand, the extended p-system results in smaller band gaps and, on the other hand, the symmetry is lowered from D 4h to D 2h for all metalloporphyrins.
Additional interactions with the HBC substituents induce further red-shis, as well as broadening and changing intensities of the spectral features. Previous work on similar di-and tetra-substituted porphyrins and porphyrin-dimers, bearing aryl-ethynyl substituents, revealed similar effects. 22,25,[51][52][53][54] The absorption spectrum of dimeric 2-Zn also is inuenced by the factors discussed above. In addition, both the Soret-band (463 and 495 nm) and the Q-band (683 and 737 nm) absorptions are split into two, suggesting electronic coupling of the two porphyrin moieties: placing two or more porphyrins in close proximity and dened geometry, as in 2-Zn, is known to cause excitonic splitting. However, for butadiyne-linked porphyrin dimers such as 2-Zn, the spectral changes have also been described in terms of an extended conjugated p-system across the two porphyrins and the bridge that depends on the dihedral angle between the porphyrin subunits. 36,52,53,[55][56][57][58][59] The same trends, i.e., red-shied (relative to ZnTPP) and single uorescence maxima, were found for the (metallo) porphyrin-centered features upon 450 nm photoexcitation. Only 1-Cu does not uoresce because of the open-shell character of the central metal ion. HBC photoexcitation at 350 nm ( Fig. 2, bottom) shows that the HBC-centered uorescence is quenched and replaced by the (metallo)porphyrin-centered uorescence. The underlying energy transfer occurs with quantum yields close to unity, as also found previously in similar HBC-porphyrin systems. [46][47][48][49] Independent conrmation of the postulated energy transfer came from excitation spectra of the (metallo)porphyrin-centered uorescence. A perfect match to the absorption spectra includes the HBC and (metallo) porphyrin ngerprints. The uorescence quantum yields of the uorescent 1, 1-Zn and 2-Zn were evaluated by analogously using ZnTPP (F Fl ¼ 0.04) as uorescence standard - Table 2.
The excitation spectrum taken at the 640 nm maximum is strikingly similar to the absorption spectrum of Ref-Zn, while that of the 750 nm maximum agrees closely with that of 2-Zn. We rationalize these ndings on the basis of two different molecular transitions, the rst localized on the individual (metallo)porphyrins and the second, which is orthogonal to the rst, delocalized across the porphyrin-butadiyne-porphyrin system. These distinct states correspond to the dihedral angle between the two porphyrin, that is, monomer-like states at around 90 , and delocalized states at around 0 . This is in sound agreement with the literature on similar butadiynelinked ZnP dimer systems. 59,60 Both uorescent states are also populated via energy transfer from HBC, as inferred from contributions in the 300 to 400 nm range in both excitation spectra (Fig. 3).
In order to investigate the impact of symmetry, functionalization, dimerization and coplanarity on the absorption and emission properties, we turned to molecular modeling. A series of porphyrins of different degrees of functionalization was optimized with B3LYP and CAM-B3LYP density functionals utilizing the def2-TZVP basis set. Time-dependent density functional theory (TD-DFT) was used to investigate the nature of the excited states. Fig. S38 and S39 † show the S 0 to S 1 /S 2 Q-band vertical excitation energies and oscillator strengths thus calculated for a series of D 4h and D 2h functionalized porphyrin Table 1 Peak positions with corresponding absorbances in parentheses for the HBC-porphyrin compounds and references (recorded in THF) a
Complexation with 1,10-DA yields the most pronounced spectral changes in the 2-Zn titration experiments - Fig. 5. In particular, in the range of the Q-band absorptions, that is, between 550 and 800 nm, the initial maxima at 580, 632 (shoulder), 674, and 724 nm, shi to 591, 645 (shoulder), 700, and 761 nm, respectively. Hereby, the intensity of the longwavelength maximum increases strongly, while those at 674 and 724 nm decrease in intensity. In the 400 to 550 nm Soretband range, new transitions evolve at 452 and 502 nm. Isosbestic points develop at 451, 473, 492, 497, and 731 nm.   Importantly, the HBC-centered absorptions do not change throughout the titration assays. Our observations are consistent with excitonic splitting and indicate a coplanar or nearly coplanar arrangement between the two ZnPs in 2-Zn$1,10DA. B3LYP/def2-TZVP optimization of 2-Zn$1,10DA yields the same results - Fig. 6. Interestingly, for 2-Zn$1,7DA even at 30 equivalents the equilibrium is not reached. In contrast, the equilibrium is completed for 1,12DA at 2 equivalents (Fig. S23 †). An immediate consequence is a higher binding energy for 1,12DA and, in turn, a higher association constant.
Red-shis were also found when HA or pyridine was added to either 1-Zn or 2-Zn, affording binding constants of 4 Â 10 5 M À1 for 1-Zn$HA and 7.9 Â 10 5 M À1 for 2-Zn$HA - Fig. S23. † The Qband absorptions shi in any of these cases by less than 25 nm and without any major changes in intensity. No appreciable changes are found in the Soret-band absorptions. Titrations with 1,7-DA or 1,12-DA are accompanied by similar, although less pronounced, changes in the Soret-and Q-band range. Based on the concentration dependence and the 760 nm intensity in combination with the aforementioned we conclude that 1,7-DA is too short and 1,12-DA is too long to induce any coplanar conformation. This conclusion is in sound agreement with the modeling. Due to chelating effects and the best exibility, 1,12-DA features the highest binding constant of 1.3 Â 10 8 M À1 , which renders it ideal for supramolecular assaysvide infra.
Fluorescence spectra mirror the changes seen in the absorption spectra (ESI, Fig. S24 †). In particular, the short-wavelength uorescence maxima of 2-Zn shi upon their excitation along with the longest-wavelength absorption features without, however, affecting the Stokes shis. Isosbestic points develop around 760 nm for 1,7-DA, 1,10-DA, and 1,12-DA, while reference titrations with HA lacked clear isosbestic points. Coordination in reference experiments with HA and 1-Zn results in red shis as large as 20 nm. Notably, addition of 1,10-DA to 2-Zn red-shis the maximum relative to 2-Zn$HA by another 15 nm to a total of 35 nm due to coplanarization of the two ZnPs and excitonic splitting. Overall, the optical band-gap is decreased by as much as 80 meV when comparing 2-Zn in the absence and in the presence of 1,10-DA - Table 3. Fitting the underlying binding isotherm of the uorescence experiments affords a binding constant of 1.7 Â 10 6 M À1 for 2-Zn$1,10DA - Fig. S24. † As a complement, the inuence of complexation with the different mono-and diamines was investigated by molecular modeling. Relaxed torsion scans, in which the angle between the two porphyrin planes was varied, yield rotation barriers of 0.6 kcal mol À1 (2.5 kJ mol À1 ) independent of further functionalization on the outer periphery (Fig. S29 †), in good agreement with existing literature. 59, 61 We next studied axial amine coordination to the Zn center with a reference system (ZnP$methylamine) in the gas phase (Fig. S30 & Table S1 †). The binding energy and the length of the Zn-N bond depend strongly on the size of the basis set, suggesting signicant basis set superposition error (BSSE) for def2-SVP, while the triple z def2-TZVP gives results within 1 kcal mol À1 of the largest basis set used. Furthermore, including empirical dispersion corrections and using range-separated functionals increases the binding energy signicantly. Our estimate of the binding energy for a monodentate coordination lies in the range of 10-18 kcal mol À1 . This was conrmed when studying the force-eld annealed and DFToptimized structures of 2-ZnDPP with all three diamines (Fig. S31, Tables S2 and S3 †). The binding energies for bidentate complex formation are always less than twice the monodentate binding energy because of reorganization effects within the diamines and porphyrin dimer. However, the chelating nature enables binding energies higher than 16 kcal mol À1 for 1,10-DA and 1,12-DA in the gas phase, which is 6 kcal mol À1 better stabilized than the monodentate coordination of 1,7-DA. Including solvation in toluene and THF with the PCM model reduces the binding energy to 13 and 10 kcal mol À1 , respectively (Fig. S32 †). As this is 14 times higher than the torsional barrier observed for 2-Zn, the hybrid geometry is determined by the binding motif and the length of the DA aliphatic chain. 2-Zn$1,10-DA exhibits both the conformation with the smallest dihedral angle between the ZnP subunits (8 B3LYP/TZVP, 0 with dispersion correction) and the largest rotation barrier. The minimum-energy conformation for 2-Zn$1,12-DA lies close to 28 (72 with dispersion correction included) ZnP-ZnP torsional angle because the alkyl chain is longer than the Zn-Zn distance (Fig. 7, S34-S36 and Table S5 †). Torsional barriers rise steeply on bidentate complexation with the diamines; from 0.6 kcal mol À1 for 2-Zn to 2.0 or 4.8 kcal mol À1 for complexes with 1,12-DA and 1,10-DA, respectively. 1,7-DA cannot coordinate to 2-Zn in a bidentate fashion without severe bending of the ZnP-butadiyne-ZnP scaffold (45 out of plane) and is, therefore, likely to bind to only one of the two ZnPs.
The pronounced geometrical changes on inclusion of dispersion corrections have two main reasons; the low torsional barrier of 0.6 kcal mol À1 of the 2-Zn dimer and the attractive dispersive interaction of the aliphatic backbone of the DA with the butadiyne bridge. Benchmark calculations (Fig. S36 and   Table S5 †) show that all methods without dispersion correction underestimate this interaction signicantly, while empirical dispersion corrections overestimate it. Secondly, the s-p interaction of À0.66 kcal mol À1 for the two possible binding motifs (CCSD(T) reference calculation) is only 0.06 kcal mol À1 higher than the torsional barrier, so that slight deviations from this reference value lead to large geometrical changes. Finally, we used AM1 semiempirical direct moleculardynamics simulations to study the movement around the Zn-Zn bond for 2-Zn and its complexes with 1,10-DA and 1,12-DA (Fig. S37 †). Both diamines block the rotation efficiently, restricting the oscillations to the range of À90 to 90 at 150 K and 300 K aer 100 ps equilibration time. The harmonic oscillator model gives a 50% increased stiffness for 2-Zn$1,10-DA compared to 2-Zn$1,12-DA (Table S6 †). This is further corroborated by the increased width of the Gaussian t in the statistical analysis.
The results let us conclude that 1,10-DA is best suited for stabilizing 2-Zn in a planar arrangement, while the backbone of 1,7-DA is too short to chelate 2-Zn. In contrast, 1,12-DA is too long for an efficient planarization, although it leads to the best stabilizationin line with the titration experiments (ESI, Fig. S23 and S24 †).
Binding constants on the order of 10 8 M À1 encouraged us to test C 60 -derivative 3 bearing two pentylamine side chains to coordinate to 2-Zn and to afford electron donor-acceptor 2-Zn$3 with near panchromatic absorptions from the UV to the NIR. We opted for the strongest binding rather than the highest degree of coplanarity. The rather poor solubility of 3 necessitated partial protonation by adding 0.5 eq. of triuoroacetic acid. 63 Importantly, the absorption spectrum of 2-Zn$3 is a good match of that found for 2-Zn$1,12-DA - Fig. 8. Next, 2-Zn$3 was probed by time-resolved transient absorption spectroscopy.

Time-resolved transient absorption spectroscopy
The excited state properties of the HBC-(metallo)porphyrins 1, 1-Zn, 1-Cu, and 2-Zn were probed in pump-probe experiments with ultrashort 387 or 450 nm laser pulses in the absence of molecular oxygen. Additional experiments were performed with 676 nm excitation for 1-Zn, with 505, 676, and 775 nm excitations for 2-Zn, and 775 nm excitation for 2-Zn$3.
The excited state deactivation of all HBC-porphyrin-conjugates presented is extremely complex in comparison to (metallated) TPP derivatives. The splitting of states induced by the lowered symmetry and additional conjugation with the ethynyl-HBC-substituents, as discussed above, introduces additional energy levels/states (Fig. 4), which are transiently populated en route to the ground state. Detailed descriptions of the time constants of these states and their differential absorption features can be found in the ESI, Fig. S1-S21. † For the monomeric HBC-porphyrin-conjugates 1, 1-Cu, and 1-Zn, Soret-band excitation gives rise to deactivation through ve states, which we assign as four singlet excited states and the respective lowest triplet excited state. An example of transient spectra and the species associated spectra of the states involved in the excited state deactivation is given for 1 in Fig. 9. 2-Zn deactivates via ve singlet excited states and its lowest triplet excited statecompare Fig. 4. Choosing a signicantly lower excitation energy leads to the population of fewer states in the excited state deactivation (ESI, Fig. S14-S21 †).
No HBC-centered features, namely singlet excited-state maxima at 560 and 620 nm as well as triplet excited-state maxima at 500 Fig. 7 B3LYP/def2-SVP calculations gas phase relaxed potential energy scans for 2-ZnPDPP (red), complexed with 1,10 DA (greytwo global minima conformers, blacklowest energy structure, allowing rotation around C-C bonds in DA) and 1,12 DA (cyantwo global minima conformers, bluelowest energy structure, allowing rotation around C-C bonds in DA). and 540 nm, were found for any of the conjugates on either 387 nm HBC or 450 nm (metallo)porphyrin excitation (compare ESI, Fig.  S22 †). We infer a unidirectional HBC to (metallo)porphyrin energy transfer faster than the time resolution of our experimental set-up from the steady-state uorescence experiments, giving a lower rate limit of >10 12 s À1 . Interestingly, in the free-base compound 1, the Zn-monomer 1-Zn and the Zn-dimer 2-Zn systems we observe a marked delay of the stimulated uorescence features (Fig. 10). The uorescence signals reach their maximum intensity aer 600 ps for 1, and approximately 400 ps for 1-Zn and 2-Zn. Transient maxima in the NIR region that exhibit the same formation kinetics are found in all cases (also for 1-Cu). These ndings suggest a cascade of deactivation through the split excited states identied by molecular modeling (see above), before populating the lowest, emissive singlet excited state.
For butadiyne-linked ZnP dimers, whose structures resemble 2-Zn, it has been demonstrated that intramolecular rotation of the porphyrins relative to the butadiyne axis occurs at room temperature (DG 298 K z 2 kJ mol À1 ), 61 so that a broad distribution of porphyrin-porphyrin dihedral angles is present.
Overall, the room-temperature rotation has been linked both to a rise in uorescence with a time constant of approximately 100 ps and to changes in transient absorption with approximately 200 ps. 60 Variations in temperature, 59-61 solvent viscosity, 60 and excitation wavelength 60 affect the distribution of the dimer conformations (perpendicular or coplanar). Not only the 2-Zn dimer displays delayed formation of uorescence and related transient absorption features, but also 1, 1-Cu, and 1-Zn monomers. For example, marked rise times in uorescence and shis in the transient features are linked to excitation dependence. Changes in 1, 1-Cu, and 1-Zn monomers are, however, attributed to the electronic structure of the ethynyl-extended porphyrin cores and their reduced symmetry/increased number of energetically non-degenerate states. 22,64 Regarding 2-Zn$3, 775 nm pulses give rise to unambiguous evidence for the 2-Zn-centered excited-state formation, from which charge separation and charge recombination affords the intermediate 2-Znc + $3c À charge-separated state - Fig. 11. For example, short laser pulses generate the excited state of 2-Zn in the form of ground state bleaching of the Soret-and Q-band absorptions in the 400 to 500 nm and 700 to 800 nm ranges, respectively. The latter are accompanied by a maximum in the range between 1100 to 1300 nm. A number of differences are discernable relative to pristine 2-Zn: rst, in the 400 to 500 nm range, the most intense bleaching evolves at 505 nm, accompanied by minor features at 475 and 455 nm, rather than at 465 nm, which is followed by a weaker 500 nm minimum. Second, a set of two similarly strong minima at 685 and 740 nm is replaced by a single minimum at 765 nm. Third, the 1145 nm maximum is red-shied to 1160 nm. The presence of C 60 in 2-Zn$3 induces a charge separation from the excited state of 2-Zn rather than the intrinsic decay seen in the experiments without C 60 - Fig. S15-S21. † Evidence for the charge separation stems from the NIR ngerprint absorptions of the one-electron reduced form of C 60 and the one electron-oxidized form of 2-Zn. One-electron oxidized butadiyne-bridged ZnP-dimers such as 2-Zn are known to exhibit intense NIR absorption bands, 65 with an intense maximum around 1000 nm, whereas the one-  electron reduced form of C 60 exhibits a characteristic absorption around 1020 nm. 66 The newly developing feature around 1000 nm is a superimposition of contributions from 2-Znc + and C 60 c À . It is important that, within the context of charge separation, the growth at 1020 nm occurs simultaneously with the decay at, for example, 1160 nm (Fig. S25 †). Important is the lack of charge separation in reference experiments with 2-Zn$1,10-DA and 2-Zn$1,12-DA - Fig. S26 †due to the absence of electron accepting fullerenes. On a longer timescale, the triplet excited state of 2-Zn persists. Most notable is the 1240 nm maximum, which indicates a signicant red-shi relative to the 1210 nm maximum found for 2-Zn in the absence of C 60 . The triplet excited state of C 60 is, however, not populated en-route towards ground state recovery. We hypothesize that the red-shis seen in the absorption spectra of 2-Zn relative to Ref-Zn places its triplet excited state energy below that of C 60 -derivative 3. Such an inactive participation seems reasonable, considering that the involvement of either C 60 -or porphyrin centered triplet excited states in charge recombination processes depends on their energy relative to the charge separated state. [67][68][69] Similar to Fig. 12, a global t of the transient absorption data of 2-Zn$3 in chlorobenzene yields good results using ve species (ESI, Fig. S25 †). In accordance with the spectral changes described above, the rst and second species, with lifetimes of 0.4 and 8 ps, represent the split rst singlet excited state of 2-Znvide supra. Both undergo charge separation to afford the one-electron reduced form of C 60 and the one-electron oxidized form of 2-Zn. The latter, however, is seen as the third and fourth species. Notably, the exibility of the linkers that connect 2-Zn with C 60 is likely to lead to a distribution of center-to-center distances between the electron donor and acceptor. We interpret the corresponding lifetimes as upper and lower limits for the distribution of lifetimes with values of 380 and 950 ps. The h species is the triplet excited state of 2-Zn. It is populated from the second species and can be considered the intrinsic intersystem crossing, which competes with the charge separation. Its overall quantum yield is only about 10%.
This constitutes an intriguing electron donor-acceptor system, which exhibits absorption exploitable to drive a charge separation from the UV ($350 nm) to the NIR ($800 nm). [70][71][72] The charge-separated state formed accordingly is generated with high efficiency and lives up to 1 ns before relaxing to the ground state.

Conclusion
In a series of novel ethynyl-bridged HBC-porphyrin conjugates, we have chemically linked HBCs to porphyrins and have investigated their photophysical properties by means of steadystate and time-resolved spectroscopic methods. In interplay with molecular modeling, a rm basis for the interpretation of the ground-and excited-state properties of these systems is found. Particular focus was placed on the tuning of the excitonic couplings. Starting with the porphyrin core, adding just bare ethynyl groups leads to appreciable changes in the absorption and uorescence across the visible range. Next, linking HBCs enables an expansion of the resulting absorption to the UV and, in turn, is the basis for a unidirectional and unit efficient HBCto-porphyrin energy transfer. Additional control over the absorptive and emissive features is realized by complexing bidentate ligands including a bis(aminoalkyl)-substituted fullerene to the porphyrin dimer with binding constants in the range from 10 6 to 10 8 M À1 . An immediate consequence is a panchromatic absorption reaching from around 350 nm in the UV to 800 nm in the near infrared and beyond. A suitable ligand length facilitates both the locking of the dihedral angle between the two porphyrins at close to 0 and a photosynthetic sequence of HBC-to-porphyrin energy transfer and porphyrinto-C 60 charge transfer.

Conflicts of interest
There are no conicts to declare.  Fig. 12 Energy diagram to illustrate the deactivation of 2-Zn$3 upon 450 and 775 nm photoexcitation on the left and right hand side, respectively. Orange, green, and yellow reflect the second singlet excited state, the first singlet excited state, and the triplet excited state, while blue/red the charge-separated state.