Supramolecular architectures featuring the antenna effect in solid state DSSCs

Abstract

Hybrid artificial antenna systems, with implementation of nature's basic concept of self-organization of chromophores, attract vast interest due to their potential application in solar cells. Herein, we report a new supramolecular system comprising two porphyrins, a free base and its zinc analogue, bearing diphenylalanine units. The first is grafted on the TiO₂ surface and plays the role of an initiator in the self-assembly of the second, inducing the formation of multi-molecular self-assembled spherical nanostructures. The achieved assemblies display an antenna effect and are used as biomimetic chromophore systems for dye-sensitized solar cells (DSSCs). This self-assembled antenna system gives a proof of concept that multichromophoric supramolecular assemblies based on hydrogen bonds can be of practical use for photovoltaic applications.

---

Introduction

Panchromatic light absorption is the primary function of any solar energy conversion system to efficiently transform the broad band sunlight energy source into electricity (photovoltaic cell) or into chemical potential (artificial photosynthesis). Plants, algae and purple bacteria provide an inspiring strategy to maximize light collection by the antenna effect as illustrated by the function of the light harvesting complexes.^1–3 The latter function is based on the initial light collection by multiple dyes featuring distinct absorption spectra, which is then followed by directional energy transfer to a single and terminal energy acceptor.

Hence a myriad of chromophores (bacteriochlorophyll and carotene derivatives) mostly held by hydrogen interactions and coordination bonds are elegantly organized around the photosynthetic reaction center inside the light harvesting antennas. Nature's capability to capture sunlight via supramolecular assemblies is not necessary to be followed likewise, in order to create efficient light harvesting units. Moreover, synthesizing artificial antenna mimics using large polypeptides is expensive and inapplicable. For that reason, self-assemblies of short peptides such as diphenylalanine (L-Phe–L-Phe, FF) have been intensively studied in recent years.^4–7

Formation of well-organized nanostructures with distinct architectures due to specific molecular recognition is promoted by non-covalent interactions. Combining the unique properties of FF motifs with various chromophores such as porphyrins and corroles, via covalent attachment or non-covalent interactions, has proven efficient in order to achieve more complex well-ordered nano-structures.^8–12 The enhancement of the light harvesting cross-section by the antenna effect was previously reported in dye sensitized solar cells (DSSCs) in their liquid^13–17 and solid-state form,^18–22 with antenna molecules that are distinct from the sensitizer. The antenna, whose role is to fuel the sensitizer by energy transfer with the incoming sunlight, was generally covalently linked to the sensitizer.^23–25 We have embarked in a program whose objective is to design an antenna for DSSCs through molecular self-assembly similarly as in biological systems. So far, coordination bonds by axial ligation of a heterocyclic amine onto a zinc porphyrin are the most used strategy to tether an antenna to the sensitizer by metallo-supramolecular interaction.^26–29 However, hydrogen bonds are certainly the most ubiquitous noncovalent interaction found in natural systems as they control the structures of proteins and nucleic bases. In this context, we have contemplated the possibility to use H bonds to assemble antenna molecules around the sensitizers in DSSCs. Indeed, self-organization from correctly informed complementary building blocks to generate ordered and complex molecular structures represents a powerful strategy to design functional materials with tailored properties. Towards this goal, we recently developed a system where diphenylalanine–porphyrin conjugates spontaneously self-assemble forming well-defined supramolecular nano-spherical architectures.^9,30 More precisely, we have considered the possibility to use such a peptide-based porphyrin as a sensitizer for TiO₂ (compound FF-H₂P-COOH), which would then serve after immobilization on TiO₂ as a seed to initiate the formation of a multi-molecular self-assembled antenna with the corresponding zinc porphyrin subunit (compound FF-ZnP) (Fig. 1). Indeed, energy transfer from zinc porphyrin to free base porphyrin is a well-established process.³¹ In this study, we report on the construction of such self-assembled molecular systems on the TiO₂ surface and their photovoltaic characterization in liquid and solid state DSSCs. Exploiting a fully solid-state device was expected to maintain the supramolecular assembly once formed better than in the presence of a liquid electrolyte as in classical DSSCs. The higher performance of the solar cell incorporating a supramolecular antenna gives a proof of concept of the practical use of self-assembled supramolecular systems for solar energy conversion.

Fig. 1 Schematic representation of the supramolecular architectures formed on the TiO₂ surface. Light blue cartoons (discs) symbolize H₂P-COOH anchored on the TiO₂ surface and purple discs represent ZnP spherical assemblies whereas FF-dipeptide is pointed with blue and the Boc moiety with green.

Results and discussion

Synthetic procedures

The synthesis of porphyrin–diphenylalanine derivatives FF-ZnP (2), FF-H₂P-COOH (6) and FF-ZnP-COOH (8) is presented in Scheme 1. Compound 1 was prepared according to the literature, after the covalent attachment of the Boc-protected diphenylalanine in an amino-substituted porphyrin.⁹

Scheme 1 Synthesis route for the porphyrin–diphenylalanine compounds FF-ZnP (2), FF-H₂P-COOH (6) and FF-ZnP-COOH (8).

Compound FF-ZnP was formed after metallation with zinc acetate of the free-base porphyrin 1. Porphyrin 3 was used as the starting material for the preparation of the two other derivatives (FF-H₂P-COOH (6) and FF-ZnP-COOH (8)) and it was synthesized following published procedures.³² Peptide bond formation between the amino group of porphyrin 3 and the free carboxylic acid of the Boc-protected diphenylalanine 4 gave compound 5 in high yield (89%). The methyl-ester group of porphyrin 5 was hydrolyzed by potassium hydroxide producing compound FF-H₂P-COOH. Finally, the free base porphyrin 5 was metallated by zinc acetate, followed by hydrolysis of the methyl-ester group to afford porphyrin FF-ZnP-COOH in high yield (90%). All new compounds were fully characterized by NMR experiments (¹H and ¹³C) and MALDI-TOF mass spectrometry.

Absorption spectroscopy

The electronic absorption spectra of compounds FF-ZnP and FF-H₂P-COOH in dichloromethane are depicted in Fig. 2. In the case of FF-ZnP, the spectral features observed correspond to a Soret band at 420 nm and three vibronically resolved Q-bands at 548, 587 and 630 nm. The latter are attributed to π–π* electronic transitions corresponding respectively to the population of the second (S2) and the first (S1) singlet excited states that are located within the porphyrin framework. Porphyrin FF-H₂P-COOH exhibits typical signatures of free-base porphyrins, featuring an intense Soret band at 420 nm and four Q-bands at 515, 550, 591 and 648 nm.

Fig. 2 UV-Vis absorption spectra of compounds FF-ZnP and FF-H₂P-COOH in CH₂Cl₂ and expansion of the Q bands.

The TiO₂ electrodes were initially surface-modified after the attachment of the free base porphyrin FF-H₂P-COOH through the carboxylic acid group, by soaking the electrode in a solution of FF-H₂P-COOH in a dichloromethane/ethanol mixture (1/1) overnight. Then, the functionalized electrodes were immersed in a solution of FF-ZnP in a CH₂Cl₂/heptane mixture (2/3). The UV-vis absorption spectrum of the TiO₂ film coated with FF-H₂P-COOH alone shows the same pattern compared to the corresponding spectrum in solution, while after dipping in a solution of zinc porphyrin FF-ZnP the intensity of the Q bands at 554 and 596 nm was significantly increased and a new broad peak around 630 nm was also observed (Fig. 3).

Fig. 3 Normalized absorption spectra of compound FF-H₂P-COOH adsorbed on TiO₂ (black line) and after the attachment of porphyrin FF-ZnP (red line) (Left). Pictures of the electrodes coated with FF-H₂P-COOH alone and with the assembly FF-H₂P-COOH + FF-ZnP (Right).

These results support the formation of the supramolecular assemblies on the TiO₂ surface since the spectrum of the electrode corresponds to the superposition of the spectra of free base FF-H₂P-COOH and the zinc porphyrin FF-ZnP. Moreover, the color change of the electrodes after the addition of FF-ZnP indicates the effective immobilization of zinc porphyrin on the TiO₂ surface (Fig. 3). If we assume that the extinction coefficients of these porphyrins have not changed upon formation of the supramolecular organization, based on the increase absorbance on the FF-ZnP Q-bands, we can estimate that there is about one zinc porphyrin which appended each FF-H₂P-COOH that attached to TiO₂.

Emission spectroscopy

Compounds FF-ZnP and FF-H₂P-COOH were studied by steady state emission spectroscopy in CH₂Cl₂ at ambient temperature (Fig. 4). When irradiated at 550 nm, porphyrin FF-ZnP showed emission which is typical of a zinc metallated porphyrin chromophore with two maxima at 598 and 646 nm. Derivative FF-H₂P-COOH exhibits the characteristic fluorescence of a free-base porphyrin with emission maxima at 654 and 716 nm. The emission spectrum of free base porphyrin FF-H₂P-COOH has a significant overlap with the absorption spectrum of zinc porphyrin FF-ZnP (Fig. S1†). Moreover, the determination of the zero–zero excitation energy (E₀₀) of free base porphyrin FF-H₂P-COOH (E₀₀ = 1.9 eV) and zinc porphyrin FF-ZnP (E₀₀ = 2.1 eV) can be calculated from the wavelength at the interception of the absorption and emission spectra. This supports that the efficient Forster energy transfer process from the zinc porphyrin FF-ZnP to the free-base porphyrin FF-H₂P-COOH can occur since the spectral overlap integral and the driving force are both significant.

Fig. 4 Normalized emission spectra of derivatives FF-H₂P-COOH and FF-ZnP in CH₂Cl₂ after excitation at 550 and 515 nm, respectively.

Electrochemical study

The electrochemical properties of the three porphyrin conjugates FF-ZnP, FF-H₂P-COOH and FF-ZnP-COOH were investigated by cyclic and square wave voltammetry in tetrahydrofuran (THF). The electrochemical redox data are summarized in Table 1 and the voltammograms are displayed in Fig. S2 and S3 (see ESI†). Compound FF-ZnP exhibits two reversible oxidation processes at 1.02 V and 1.32 V and two reversible reductions, the first at −1.31 V and the second at −1.77 V. Derivative FF-H₂P-COOH in the anodic region shows two processes at 1.27 V and 1.51 V, both stemming from the oxidation of the central porphyrin ring. In the cathodic region, the voltammograms feature two reversible reduction processes at −1.16 V and −1.55 V. In the case of FF-ZnP-COOH, the typical data for meso-tetra-aryl substituted zinc porphyrins were observed, specifically two reversible oxidations at 1.03 V and 1.37 V and two reversible reductions at −1.40 V and −1.55 V. Overall, free base porphyrin FF-H₂P-COOH is oxidized at higher potentials compared to zinc porphyrins FF-ZnP and FF-ZnP-COOH. From these observations, we can assume that the remaining hole on FF-H₂P-COOH after electron injection into TiO₂ can be possibly transferred on FF-ZnP, which can play the role of a secondary electron donor and slow down charge recombination. Conversely, in the cathodic region compounds, FF-ZnP and FF-ZnP-COOH displayed more cathodic reduction potentials compared to FF-H₂P-COOH.

Table 1 Electrochemical data of compounds FF-ZnP, FF-H₂P-COOH and FF-ZnP-COOH in tetrahydrofuran. All potentials are reported vs. SCE and FcH/FcH⁺ was used as the internal standard (E^Ox_1/2 is 0.60 V)

Compound	E ^Red21/2 (V)	E ^Red11/2 (V)	E ^Ox11/2 (V)	E ^Ox21/2 (V)
FF-ZnP	−1.77	−1.31	1.02	1.32
FF-H2P-COOH	−1.55	−1.16	1.27	1.51 (ir)
FF-ZnP-COOH	−1.85 (ir)	−1.40	1.03	1.37

---

These data enable us to confirm that the electron injection reaction from the singlet excited state of FF-ZnP is thermodynamically allowed since its oxidation potential (1.02 − 1.9 = −0.88 V vs. SCE) is more negative than the conduction band of TiO₂ (−0.7 V vs. SCE). In addition, since the oxidation potential of FF-ZnP is more positive than that of the redox couple (E(I₂˙⁻I⁻) = 0.55 V vs. SCE), there is also a significant driving force for the regeneration reaction (ΔG = −0.47 eV).

Scanning electron microscopy (SEM)

SEM was used in order to explore the structural features of the nanostructures which were either deposited or developed onto the TiO₂ electrodes. Macroporous substrates were imaged before any conjugate attachment (blank substrate) (Fig. 5a). First, FF-H₂P-COOH was immobilized on TiO₂ (Fig. 5b), and then, the substrate was dipped into a solution of FF-ZnP (Fig. 5c). Finally, in order to examine the ideal conditions of self-assembly on TiO₂ electrodes, we followed a different preparation protocol. With this methodology, previously prepared nanostructures of FF-ZnP were deposited on the FF-H₂P-COOH grafted TiO₂ surface (Fig. 5d). SEM results suggest that even though the macroporous substrate exhibits a rough surface, this feature did not affect the formation of the nanostructures. The adsorption of compound FF-H₂P-COOH also does not affect the structural characteristics of the substrate. The formation of spherical microstructures onto the surface of the electrode was observed after dipping in the solution of compound FF-ZnP; these microstructures appeared to be held together onto the substrate by a network of nanofibrils.

Fig. 5 SEM images of TiO₂ electrodes: (a) blank substrate, (b) after adsorption of compound FF-H₂P-COOH, (c) after dipping the FF-H₂P-COOH grafted TiO₂ substrate in a solution of derivative FF-ZnP, and (d) after depositing previously prepared nanostructures of compound FF-ZnP on the FF-H₂P-COOH grafted TiO₂ electrode.

Upon depositing, previously prepared nanostructures of compound FF-ZnP on the FF-H₂P-COOH functionalized substrate, conglomerates of nanospheres covered parts of the TiO₂ surface. Since in both cases we observed the formation of nano-spheres, for the next studies we used the dipping method because it's simpler and the shape of the spheres is more homogeneous.

Photovoltaic measurements

The free-base porphyrin FF-H₂P-COOH and the zinc porphyrin FF-ZnP-COOH were first used as sensitizers in liquid DSSCs with the classical redox couple iodide/triiodide in acetonitrile as the electrolyte. The metrics of the solar cells are gathered in Table 2 along with those of the reference dye N719, which were recorded under the same conditions.

Table 2 Characteristics of the DSSCs with the iodide/triiodide electrolyte recorded under AM1.5 G simulated sunlight (1000 W m⁻²)

Sensitizer	J sc (mA cm^−2)	V oc (mV)	ff (%)	PCE (%)
FF-H2P-COOH	2.07 ± 0.1	413 ± 10	72 ± 2	0.62 ± 0.3
FF-ZnP-COOH	2.46 ± 0.1	407 ± 10	68 ± 2	0.68 ± 0.3
N719	14.15 ± 0.2	734 ± 10	74 ± 2	7.64 ± 0.5

---

Interestingly, the zinc porphyrin FF-ZnP-COOH performs only slightly better than the free base FF-H₂P-COOH, justifying thus to be used as the sensitizer of TiO₂ for the fabrication of the supramolecular assembly. In that case, it can play the dual role of the electron injector into TiO₂ and the terminal energy acceptor from zinc porphyrin FF-ZnP. Then, the supramolecular assembly FF-H₂P-COOH + FF-ZnP, immobilized on the nanocrystalline TiO₂ electrode, was used as the photoanode to fabricate DSSCs.

Unsurprisingly, the supramolecular assembly maintained by hydrogen bonds was immediately disrupted once the electrolyte was introduced into the cell. The substitution of acetonitrile by other solvents (such as dichloromethane and chlorobenzene) was attempted, but they systematically led to the same result. Accordingly, the liquid electrolyte was replaced by the well-known solid-state hole transporting material (HTM) tetrakis(N,N-di-p-methoxyphenylamine)-9,9′-spirobifluorene (spiro-OMeTAD).^33,34 Classically, this HTM is infiltrated inside the photoanode by spin-coating of a concentrated solution in chlorobenzene (200 mg mL⁻¹) in the presence of polar additives (bis(trifluoromethane)sulfonimide lithium salt (Li-TSFI) and tert-butylpyridine).

Preliminary tests showed that these conditions are not compatible with the stability of the supramolecular assembly, this latter being immediately destroyed during the spin-coating cycle, before complete solvent evaporation. To address this issue, therefore we adopted a two-stage deposition sequence. First, the TiO₂ photoanode bearing the supramolecular assembly was infiltrated using a dilute solution of spiro-OMeTAD (10 mg mL⁻¹) in a mixture of n-heptane/chlorobenzene (60/40 volume ratio), in an attempt to form a first capping layer of the HTM above the TiO₂/FF-H₂P-COOH + FF-ZnP surface. The fraction of n-heptane in chlorobenzene was optimized to ensure the complete dissolution of the spiro-OMeTAD molecular glass. After drying for five minutes in ambient air, a concentrated solution of spiro-OMeTAD in pure chlorobenzene (200 mg mL⁻¹), either without any additives or doped with Li-TSFI and/or tert-butylpyridine, was then spin coated on this first layer. Among all the conditions tested, we observed that the two-step procedure was able to preserve the supramolecular assembly of H₂P-ZnP, as long as no tert-butylpyridine was involved during the deposition of the concentrated spiro-OMeTAD solution. In this case, the ratio of absorption bands at 518 and 554 nm, which reflects the molar ratio of FF-ZnPversus (FF-H₂P-COOH + FF-ZnP), decreases compared to a FF-H₂P-COOH-sensitized infiltrated in a similar manner but in the absence of the supramolecular assembly (Fig. 6).

Fig. 6 Absorbance spectra of TiO₂ photoanodes sensitized by compound FF-H₂P-COOH only (black curve) and by the supramolecular assembly of compounds FF-H₂P-COOH + FF-ZnP (red curve), after infiltration by spiro-OMeTAD using a 2-step deposition method. Li-TFSI is used as the dopant in a Li-TFSI/spiro-OMeTAD weight ratio of 1/27.

Finally, solid-state DSSCs were obtained by depositing a thin film of gold on top of the infiltrated dye-sensitized photoanodes to collect the holes injected into the HTM. Photon-to-current conversion efficiency (IPCE) spectra and current density-to-voltage characteristics (J–V curves under standard solar simulation) of the solid-state cells are reported in Fig. 7, while the corresponding photovoltaic parameters are collected in Table 3.

Fig. 7 (a) IPCE spectra of solar cells based on FF-H₂P-COOH (black curve) and FF-H₂P-COOH + FF-ZnP (red curve). (b) Current density/voltage characteristics of solid-state DSSC based on spiro-OMeTAD hole transporter and sensitizers FF-H₂P-COOH (black curve) and FF-H₂P-COOH + FF-ZnP (red curve) under simulated solar emission (1000 W m⁻², AM1.5 G).

Table 3 Characteristics of the DSSCs with the solid HTM spiro-OMeTAD and recorded under AM1.5 G simulated sunlight (1000 W m⁻²)

Sensitizer	J sc (mA cm^−2)	V oc (mV)	ff (%)	PCE (%)
FF-H2P-COOH	1.39 ± 0.20	298 ± 15	27 ± 2	0.11 ± 0.02
FF-H2P-COOH + FF-ZnP	1.88 ± 0.20	280 ± 15	35 ± 2	0.19 ± 0.02

---

Clearly, the cell containing the supramolecular association FF-H₂P-COOH + FF-ZnP produces a significantly higher photocurrent density than the cell with only FF-H₂P-COOH (Table 3). This is the direct consequence of a higher light harvesting efficiency owing to the presence of the FF-ZnP supplementary light collectors, which play the role of an antenna and convey the excitation energy to the free base sensitizer FF-H₂P-COOHvia energy transfer. IPCE spectra of the aforementioned solid state DSSCs confirm this trend. The cell with the supramolecular assembly exhibits higher IPCE values compared to the cell containing only the FF-H₂P-COOH porphyrin. This is consistent with the information obtained from normalized IPCE spectra (Fig. S14†), where the second Q band is slightly increased after the assembly of the FF-ZnP, indicating the contribution of the second porphyrin moiety on the IPCE spectrum. These results are similar to those of previously reported donor–acceptor antenna incorporating systems.^35–37 However, in our system the IPCE profile is not similar to the solid state absorption spectrum (Fig. 3). This can be attributed to the partial decomposition of the supramolecular assemblies after the addition of the spiro-OMeTAD. The relatively low open-circuit voltage observed for both cells is related to the absence of tert-butylpyridine as the additive on the spiro-OMeTAD solution, which usually induces a band edge shift to the vacuum level by deprotonation of the TiO₂ surface and drastically inhibits the interfacial charge carrier recombination.³⁸

Preliminary transient photo-voltage measurements performed under open-circuit voltage as a function of incident light intensities (Fig. 8) show that the recombination kinetics are slower in the presence of the supramolecular assembly compared to those associated with FF-H₂P-COOH only, over the entire range of illumination conditions tested. Considering similar charge carrier mobilities in both the nanocrystalline TiO₂ network and the spiro-OMeTAD HTM for both devices, such observation is also consistent with a larger charge collection efficiency of the devices bearing the supramolecular assembly. However, the relatively lower open-circuit voltage evidenced in this case indicates that the supramolecular assembly probably drives charges closer to the interface. It could be also possible that the slower recombination kinetics in the presence of the supramolecular assembly are associated with the hole transfer from the oxidized free base porphyrin to the zinc porphyrin ad layer after photo-injection, leading to better device performance as well. This is also consistent with the significant improvement in FF.

Fig. 8 Charge recombination time as a function of steady-state bias light intensity measured by transient photo-voltage on the solid state DSSC based on FF-H₂P-COOH alone (black data) and on the FF-H₂P-COOH + FF-ZnP supramolecular assembly (red data). The data are obtained under open-circuit conditions.

Conclusions

In summary, we have established an elegant method for the construction of supramolecular solid state DSSCs with self-assembling nano-architectures featuring an antenna effect. Diphenylalanine–porphyrin conjugates FF-ZnP self-assemble and form spherical nanostructures on a FF-H₂P-COOH functionalized TiO₂ surface. From the absorption studies we estimated that the ratio between the FF-H₂P-COOH porphyrin and the FF-ZnP on the TiO₂ surface is almost 1 : 1. The formation of spherical nanostructures on the TiO₂ surface was verified from scanning electron microscopy studies. The functionalized TiO₂ electrodes were used for the preparation of solid-state DSSCs based on the molecular spiro-OMeTAD hole transporter. The photovoltaic measurements showed a significant improvement in the performance of the device containing the supramolecular association FF-H₂P-COOH + FF-ZnP. The corresponding power conversion efficiency recorded under standard solar emission is multiplied by a factor of two, indicating that the concept of a supramolecular antenna can find practical use in solid-state DSSC devices. It is worth mentioning that this is the first example where supramolecular interactions through hydrogen bonds are used for the development of solid state DSSCs. The strategy of incorporating an antenna by supramolecular interactions represents a viable approach to prepare better performing solid-state hybrid devices for photovoltaic applications.

Experimental section

Materials

All commercially available compounds were used as received unless otherwise noted. Boc-protected diphenylalanine 4 was purchased from Bachem. Porphyrins 1 (ref. 9) and 3 (ref. 32) were prepared according to the literature procedures.

NMR spectra were recorded on Bruker AVANCE III-500 MHz and Bruker DPX-300 MHz spectrometers using solutions in deuterated solvents and the solvent peak was chosen as the internal standard.

High-resolution mass spectra were obtained on a Bruker UltrafleXtreme MALDI-TOF spectrometer using trans-2-[3-(4-tert-butylphenyl)-2-methyl-2-propenylidene]malononitrile (DCTB) as the matrix.

Absorption spectra were recorded on a Shimadzu UV-1700 PharmaSpec instrument. Steady-state emission spectra were obtained using a JASCO FP-6500 fluorescence spectrophotometer. Cyclic and square wave voltammetry experiments were carried out at room temperature using an AutoLab PGSTAT20 potentiostat. All measurements were carried out in THF in the presence of tetrabutylammoniumtetrafluoroborate (0.1 M) as the supporting electrolyte. A three-electrode cell setup was used with a platinum working electrode, a saturated calomel (SCE) reference electrode, and a platinum wire as the counter electrode.

Preparation of porphyrin FF-ZnP (2)

A mixture of porphyrin 1 (80 mg, 0.078 mmol) and zinc acetate dihydrate (200 mg, 0.91 mmol) in CH₂Cl₂ (30 mL) and MeOH (8 mL) was stirred at room temperature overnight. The solvent was removed in a rotary evaporator and the desired compound was isolated by silica column chromatography CH₂Cl₂–EtOH (100–2) to obtain porphyrin FF-ZnP (2) as a purple solid (79 mg, 93%). ¹H NMR (CDCl₃, 500 MHz): δ 8.96 (s, 8H), 8.24 (m, 6H), 8.11 (d, J = 8.0 Hz, 2H), 7.97 (s, 1H), 7.75 (m, 9H), 7.68 (br s, 2H), 7.27 (m, 3H), 7.22 (m, 3H), 6.90 (br s, 2H), 6.85 (br s, 2H), 6.07 (d, J = 8.0 Hz, 1H), 4.36 (br s, 1H), 3.94 (br s, 1H), 3.72 (br s, 1H), 2.97 (m, 1H), 2.64 (m, 1H), 2.50 (m, 2H), 1.29 (s, 9H). ¹³C NMR (CDCl₃, 75 MHz): δ 170.7, 168.5, 155.8, 150.4, 143.1, 139.0, 137.0, 136.0, 135.7, 135.0, 134.7, 132.1, 129.3, 129.2, 129.03, 128.98, 127.6, 127.3, 126.7, 121.2, 120.8, 118.3, 81.2, 56.0, 53.7, 36.9, 28.2. UV/vis (CH₂Cl₂) λ_max, nm (ε, mM⁻¹ cm⁻¹) 420 (543.8), 548 (21.8), 587 (5.6). HRMS (MALDI-TOF): m/z calcd for C₆₇H₅₆N₇O₄Zn: 1086.3685 [M + H]⁺; found: 1086.3694.

Preparation of porphyrin 5

Boc-protected di-phenylalanine 4 (18 mg, 0.044 mmol) was dissolved in CH₂Cl₂ (2 mL) and the solution was cooled to 0 °C in an ice bath. N,N′-Dicyclohexylcarbodiimide (DCC; 11 mg, 0.053 mmol) and 1-hydroxybenzotriazole hydrate (HOBt; 7 mg, 0.052 mmol) were added and the stirring was continued for 30 min at 0 °C. Then porphyrin 3 (33 mg, 0.043 mmol) was added and the resulting mixture was stirred at 8 °C for 48 h. After addition of CH₂Cl₂ the solution was washed once with water, and the organic layer was dried with Na₂SO₄, filtered and concentrated. The title compound 5 was isolated by column chromatography (silica gel, CH₂Cl₂–EtOH, 100–1) as a purple solid (46 mg, 89%). ¹H NMR (CDCl₃, 500 MHz): δ 8.82 (d, J = 4.7 Hz, 2H), 8.74 (d, J = 4.7 Hz, 2H), 8.70 (m, 4H), 8.60 (s, 1H), 8.43 (d, J = 8.8 Hz, 2H), 8.31 (d, J = 8.0 Hz, 2H), 8.15 (d, J = 8.5 Hz, 2H), 7.98 (d, J = 7.0 Hz, 2H), 7.36 (m, 6H), 7.29 (s, 4H), 7.25 (m, 2H), 7.14 (m, 2H), 6.46 (m, 1H), 5.00 (m, 1H), 4.88 (br s, 1H), 4.39 (m, 1H), 4.11 (s, 3H), 3.52 (m, 1H), 3.11 (m, 3H), 2.63 (s, 6H), 1.84 (s, 12H), 1.34 (s, 9H), −2.63 (s, 2H). ¹³C NMR (CDCl₃, 125 MHz): δ 171.3, 169.2, 167.5, 156.2, 147.1, 139.5, 138.5, 137.9, 137.6, 136.3, 135.9, 135.0, 134.7, 130.3, 129.6, 129.5, 129.4, 129.2, 129.1, 128.0, 127.9, 127.7, 127.5, 119.5, 118.7, 118.5, 117.8, 81.5, 56.7, 54.2, 52.6, 37.5, 37.2, 28.3, 21.8, 21.6. UV/vis (CH₂Cl₂) λ_max, nm (ε, mM⁻¹ cm⁻¹) 420 (427.5), 515 (18.4), 550 (8.5), 591 (5.8), 647 (5.0). HRMS (MALDI-TOF): m/z calcd for C₇₅H₇₂N₇O₆: 1166.5544 [M + H]⁺; found: 1166.5549.

Preparation of porphyrin FF-H₂P-COOH (6)

Porphyrin 5 (30 mg, 0.026 mmol) was dissolved in a solution of THF (10 mL), MeOH (4 mL) and H₂O (5 mL) and potassium hydroxide (0.27 g, 4.81 mmol) was added. The reaction mixture was stirred at room temperature overnight. After removing the organic solvents an aqueous solution of HCl (1 N) was added dropwise until the pH reached 6. A precipitate was formed, filtered and washed with water giving a purple solid (29 mg, 97%). ¹H NMR (CDCl₃, 300 MHz): δ 8.34 (m, 3H), 8.79 (d, J = 4.8 Hz, 2H), 8.73 (m, 4H), 8.57 (d, J = 8.1 Hz, 2H), 8.39 (d, J = 8.1 Hz, 2H), 8.17 (d, J = 8.3 Hz, 2H), 7.92 (d, J = 7.9 Hz, 2H), 7.38 (m, 8H), 7.29 (s, 4H), 7.18 (m, 1H), 7.10 (m, 1H), 6.66–6.44 (m, 1H), 5.24–4.96 (m, 1H), 5.06 (m, 1H), 4.45–4.21 (m, 1H), 3.51 (m, 1H), 3.14 (m, 3H), 2.63 (s, 6H), 1.85 (s, 12H), 1.45 (s, 3H), 1.36 (s, 6H), −2.62 (s, 2H). ¹³C NMR (CDCl₃, 125 MHz): δ 172.0, 171.5, 171.0, 169.3, 156.2, 147.9, 139.5, 138.5, 138.1, 138.0, 137.6, 136.4, 136.3, 135.9, 135.0, 134.8, 130.2, 129.7, 129.5, 129.4, 129.2, 129.14, 129.10, 128.8, 128.7, 127.9, 127.7, 127.6, 127.5, 119.6, 118.7, 118.6, 117.7, 81.4, 81.1, 57.6, 56.6, 54.6, 54.3, 37.8, 37.7, 37.3, 37.1, 29.6, 28.3, 21.8, 21.6. UV/vis (CH₂Cl₂) λ_max, nm (ε, mM⁻¹ cm⁻¹) 420 (411.2), 515 (18.0), 551 (8.2), 591 (5.7), 647 (4.6). HRMS (MALDI-TOF): m/z calcd for C₇₄H₇₀N₇O₆: 1152.5388 [M + H]⁺; found: 1152.5395.

Preparation of porphyrin 7

To a stirred solution of porphyrin 5 (37 mg, 0.030 mmol) in CH₂Cl₂ (18 mL) and MeOH (6 mL), zinc acetate dihydrate (125 mg, 0.57 mmol) was added. The reaction mixture was stirred at room temperature overnight. The solvent was removed in a rotary evaporator and the desired compound was isolated by silica column chromatography CH₂Cl₂–EtOH (100–1) to obtain porphyrin 7 as a purple solid (34 mg, 93%). ¹H NMR (CDCl₃, 500 MHz): δ 8.90 (d, J = 4.5 Hz, 2H), 8.82 (d, J = 4.5 Hz, 2H), 8.79 (m, 4H), 8.47 (br s, 1H), 8.40 (d, J = 8.2 Hz, 2H), 8.32 (d, J = 8.2 Hz, 2H), 8.13 (d, J = 8.4 Hz, 2H), 7.84 (d, J = 7.5 Hz, 2H), 7.35 (m, 6H), 7.28 (s, 4H), 7.21 (d, J = 6.9 Hz, 2H), 7.05 (br s, 2H), 6.37 (br s, 1H), 4.82 (br s, 1H), 4.77 (m, 1H), 4.30 (m, 1H), 4.09 (s, 3H), 3.32 (m, 1H), 3.03 (m, 3H), 2.63 (s, 6H), 1.83 (s, 12H), 1.31 (s, 9H). ¹³C NMR (CDCl₃, 75 MHz): δ 171.2, 169.0, 167.6, 156.1, 150.4, 150.2, 150.1, 149.6, 148.0, 139.4, 139.1, 138.8, 137.7, 137.3, 136.2, 135.9, 134.9, 134.6, 132.6, 132.0, 131.2, 131.0, 129.5, 129.4, 129.2, 129.1, 127.8, 127.7, 127.5, 120.3, 119.6, 118.8, 118.3, 81.4, 56.6, 54.1, 52.5, 37.5, 37.0, 28.2, 21.8, 21.6. UV/vis (CH₂Cl₂) λ_max, nm (ε, mM⁻¹ cm⁻¹) 421 (481.3), 549 (20.7), 588 (4.8). HRMS (MALDI-TOF): m/z calcd for C₇₅H₇₀N₇O₆Zn: 1228.4679 [M + H]⁺; found: 1228.4685.

Preparation of porphyrin FF-ZnP-COOH (8)

To a solution of porphyrin 7 (20 mg, 0.016 mmol) in THF (7 mL), MeOH (3 mL) and H₂O (4 mL) potassium hydroxide (0.18 g, 3.21 mmol) was added. The mixture was stirred at room temperature overnight. The organic solvents were removed in a rotary evaporator and an aqueous solution of HCl (1 N) was added dropwise until the pH reached 6. A precipitate was formed, filtered and washed with water giving a purple solid (18 mg, 92%). UV/vis (CH₂Cl₂) λ_max, nm (ε, mM⁻¹ cm⁻¹) 421 (448.6), 549 (20.1), 588 (4.6). HRMS (MALDI-TOF): m/z calcd for C₇₄H₆₈N₇O₆Zn: 1214.4523 [M + H]⁺; found: 1214.4531.

Preparation of the nanostructures on the TiO₂ electrodes

Two protocols were followed for the development of self-assembled nanostructures on the TiO₂ electrodes. The first protocol involves dipping the electrode in a 50% ethanol/50% CH₂Cl₂ solution of compound FF-H₂P-COOH (0.1 mM). The electrode was left in the solution overnight, washed with ethanol, dried under a nitrogen stream and dipped in a 40% CH₂Cl₂/60% heptane solution of compound FF-ZnP (0.1 mM) for 24 h. Afterwards, the electrodes were washed with heptane and left to dry in air. In the second protocol, the electrodes were dipped in a 50% ethanol/50% CH₂Cl₂ solution of compound FF-H₂P-COOH (0.1 mM) and remained in the solution overnight. The dyed electrodes were washed with ethanol and dried under a nitrogen stream. An aliquot of 100 μL from a stock solution of compound FF-ZnP (1 mM) in 30% CH₂Cl₂/70% heptane was deposited onto the electrodes and left to dry in air. SEM observation was conducted using a JEOL JSM-6390LV operating at 15–20 kV. All samples were dried overnight in air and coated with a 10 nm gold sputtering prior to observation.

Preparation of the solar cells

Solid-state dye-sensitized solar cells were fabricated on specifically etched FTO-coated glass substrates (square resistance around 15 Ω sq⁻¹) on which dense TiO₂ layers were initially sprayed at 450 °C from a titanium tetra-isopropoxide (TTIP) solution in an ethanol/acetylacetone mixture. Nanocrystalline TiO₂ layers were then deposited by spin-coating from a diluted TiO₂ paste (Dyesol, DSL 18 NR-T) to achieve around 2 μm thick films. Subsequent annealing up to 500 °C in air over 45 min was used in order to remove the organic additives and to achieve a percolated nanoparticle network (see our ref. 39 for complementary details). Following these steps, protocol 1 described above for the self-assembly of the porphyrin nanostructure on TiO₂ was applied before a specific two-step procedure was used for the infiltration by the spiro-OMeTAD HTM, to preserve the supramolecular self-assembly. First, a diluted spiro-OMeTAD solution in an n-heptane/chlorobenzene mixture (volume ratio of 60/40, HTM concentration of 10 mg mL⁻¹) was drop-cast on the porphyrin-treated TiO₂. Then, a concentrated spiro-OMeTAD solution (200 mg mL⁻¹ in pure chlorobenzene) was spin-coated on top of the substrates. In this latter case, bis(trifluoromethane)sulfonimide lithium salt (Li-TSFI) was added to the HTM solution prior to spin-coating using a 1/27 Li-TFSI/spiro-OMeTAD weight ratio. Finally, a 100 nm thick silver top electrode was evaporated under high vacuum through a shadow mask to define two independent devices per substrate, presenting an active area of 0.18 cm² each.

Photovoltaic measurements

Current density–voltage characteristics were measured under simulated solar emission provided by an AM1.5 G-filtered 1600 W Newport solar simulator. Spectral mismatch correction was applied to ensure an equivalent 100 mW cm⁻² irradiation on the tested cells. The electrical signal of the devices was recorded through a Keithley 2400 source-measurement unit.

Transient photo-voltage (TPV) decays in the μs to ms regime were measured under open-circuit conditions on full solar cells under continuous white light illumination provided by two white inorganic light emitting diodes (OSRAM) over the range 1–100 mW cm⁻². An additional light pulse was focused on the device active area using dedicated lenses, from a third LED controlled by a fast solid-state switch. The transient electrical signal at the electrodes of the devices was measured through a digital oscilloscope (Tektronics DPO 4032). All traces were fitted by mono-exponential decay functions, from which the recombination time was directly derived.

Acknowledgements

This work was supported by the European Commission's Seventh Framework Programme (FP7/2007–2013) under grant agreement no. 229927 (FP7-REGPOT-2008-1, Project BIOSOLENUTI) and Special Research Account of the University of Crete. The COST Action CM1202 PERSPECT-H₂O and Région pays de la Loire for LUMOMAT project are also acknowledged.

Notes and references

1. G. McDermott, S. M. Prince, A. A. Freer, A. M. Hawthornthwaite-Lawless, M. Z. Papiz, R. J. Cogdell and N. W. Isaacs, Nature, 1995, 374, 517–521.
2. T. Pullerits and V. Sundström, Acc. Chem. Res., 1996, 29, 381–389.
3. V. Sundström, T. Pullerits and R. van Grondelle, J. Phys. Chem. B, 1999, 103, 2327–2346.
4. T. O. Mason, D. Y. Chirgadze, A. Levin, L. Adler-Abramovich, E. Gazit, T. P. J. Knowles and A. K. Buell, ACS Nano, 2014, 8, 1243–1253.
5. M. Reches and E. Gazit, Science, 2003, 300, 625–627.
6. X. Yan, P. Zhu and J. Li, Chem. Soc. Rev., 2010, 39, 1877.
7. M. Reches and E. Gazit, Nat. Nanotechnol., 2006, 1, 195–200.
8. K. Karikis, E. Georgilis, G. Charalambidis, A. Petrou, O. Vakuliuk, T. Chatziioannou, I. Raptaki, S. Tsovola, I. Papakyriacou, A. Mitraki, D. T. Gryko and A. G. Coutsolelos, Chem.–Eur. J., 2016, 22, 11245–11252.
9. G. Charalambidis, E. Kasotakis, T. Lazarides, A. Mitraki and A. G. Coutsolelos, Chem.–Eur. J., 2011, 17, 7213–7219.
10. S. H. Kim, H. Park, M. S. Seo, M. Kubo, T. Ogura, J. Klajn, D. T. Gryko, J. S. Valentine and W. Nam, J. Am. Chem. Soc., 2010, 132, 14030–14032.
11. J. H. Kim, D. H. Nam, Y. W. Lee, Y. S. Nam and C. B. Park, Small, 2014, 10, 1272–1277.
12. J. H. Kim, M. Lee, J. S. Lee and C. B. Park, Angew. Chem., Int. Ed., 2012, 51, 517–520.
13. R. Amadelli, R. Argazzi, C. A. Bignozzi and F. Scandola, J. Am. Chem. Soc., 1990, 112, 7099–7103.
14. N. C. Jeong, H.-J. Son, C. Prasittichai, C. Y. Lee, R. A. Jensen, O. K. Farha and J. T. Hupp, J. Am. Chem. Soc., 2012, 134, 19820–19827.
15. B. E. Hardin, E. T. Hoke, P. B. Armstrong, J.-H. Yum, P. Comte, T. Torres, J. M. J. Frechet, M. K. Nazeeruddin, M. Grätzel and M. D. McGehee, Nat. Photonics, 2009, 3, 406–411.
16. C. Siegers, J. Hohl-Ebinger, B. Zimmermann, U. Würfel, R. Muelhaupt, A. Hinsch and R. Haag, ChemPhysChem, 2007, 8, 1548–1556.
17. J. Warnan, F. Buchet, Y. Pellegrin, E. Blart and F. Odobel, Org. Lett., 2011, 13, 3944–3947.
18. J.-H. Yum, B. E. Hardin, S.-J. Moon, E. Baranoff, F. Nueesch, M. D. McGehee, M. Grätzel and M. K. Nazeeruddin, Angew. Chem., Int. Ed., 2009, 48, 9277–9280.
19. B. Louahem M'Sabah, M. Boucharef, J. Warnan, Y. Pellegrin, E. Blart, B. Lucas, F. Odobel and J. Boucle, Phys. Chem. Chem. Phys., 2015, 17, 9910–9918.
20. S.-J. Moon, E. Baranoff, S. M. Zakeeruddin, C.-Y. Yeh, E. W.-G. Diau, M. Gratzel and K. Sivula, Chem. Commun., 2011, 47, 8244–8246.
21. N. Humphry-Baker, K. Driscoll, A. Rao, T. Torres, H. J. Snaith and R. H. Friend, Nano Lett., 2012, 12, 634–639.
22. K. Driscoll, J. Fang, N. Humphry-Baker, T. Torres, W. T. S. Huck, H. J. Snaith and R. H. Friend, Nano Lett., 2010, 10, 4981–4988.
23. F. Odobel, Y. Pellegrin and J. Warnan, Energy Environ. Sci., 2013, 6, 2041–2052.
24. T. Lazarides, G. Charalambidis, A. Vuillamy, M. Reglier, E. Klontzas, G. Froudakis, S. Kuhri, D. M. Guldi and A. G. Coutsolelos, Inorg. Chem., 2011, 50, 8926–8936.
25. S. Kuhri, G. Charalambidis, P. A. Angaridis, T. Lazarides, G. Pagona, N. Tagmatarchis, A. G. Coutsolelos and D. M. Guldi, Chemistry, 2014, 20, 2049–2057.
26. M. Morisue, S. Yamatsu, N. Haruta and Y. Kobuke, Chem.–Eur. J., 2005, 11, 5563–5574.
27. A. Nomoto and Y. Kobuke, Chem. Commun., 2002, 1104–1105.
28. J. Warnan, Y. Pellegrin, E. Blart and F. Odobel, Chem. Commun., 2012, 48, 675–677.
29. P. L. Marek, H. Hahn and T. S. Balaban, Energy Environ. Sci., 2011, 4, 2366–2378.
30. G. Charalambidis, E. Georgilis, M. K. Panda, C. E. Anson, A. K. Powell, S. Doyle, D. Moss, T. Jochum, P. N. Horton, S. J. Coles, M. Linares, D. Beljonne, J.-V. Naubron, J. Conradt, H. Kalt, A. Mitraki, A. G. Coutsolelos and T. S. Balaban, Nat. Commun., 2016, 7, 12657.
31. K. Pettersson, A. Kyrychenko, E. Rönnow, T. Ljungdahl, J. Mårtensson and B. Albinsson, J. Phys. Chem. A, 2006, 110, 310–318.
32. G. E. Zervaki, E. Papastamatakis, P. A. Angaridis, V. Nikolaou, M. Singh, R. Kurchania, T. N. Kitsopoulos, G. D. Sharma and A. G. Coutsolelos, Eur. J. Inorg. Chem., 2014, 2014, 1020–1033.
33. U. Bach, D. Lupo, P. Comte, J. E. Moser, F. Weissortel, J. Salbeck, H. Spreitzer and M. Gratzel, Nature, 1998, 395, 583–585.
34. B. Li, L. Wang, B. Kang, P. Wang and Y. Qiu, Sol. Energy Mater. Sol. Cells, 2006, 90, 549–573.
35. C. Y. Lee and J. T. Hupp, Langmuir, 2010, 26, 3760–3765.
36. M. Planells, L. Pellejà, P. Ballester and E. Palomares, Energy Environ. Sci., 2011, 4, 528–534.
37. Y. Wu, Q. Zhang, J.-C. Liu, R.-Z. Li and N.-Z. Jin, Org. Electron., 2016, 41, 60773–60779.
38. J. Krüger, R. Plass, L. Cevey, M. Piccirelli, M. Grätzel and U. Bach, Appl. Phys. Lett., 2001, 79, 2085–2087.
39. H. Melhem, P. Simon, L. Beouch, F. Goubard, M. Boucharef, C. Di Bin, Y. Leconte, B. Ratier, N. Herlin-Boime and J. Bouclé, Adv. Energy Mater., 2011, 1, 908–916.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6se00051g

---