Abstract

A tetra-n-butoxy zinc phthalocyanine (n-BuO)4ZnPc has been synthesized in a single step, starting from commercial precursors, and easily purified. The molecule can be solution processed to form an effective and inexpensive hole transport layer for organic and perovskite solar cells. These appealing features are suggested by the results of a series of chemical, optical, and voltammetric characterizations of the molecule, supported by the results of ab initio simulations. Preliminary measurements of (n-BuO)4ZnPc-methylammonium lead triiodide perovskite-based devices confirm such suggestion and indicate that the interface between the photoactive layer and the hole transporting layer is characterized by hole-extracting and electron-blocking properties, potentially competitive with those of other standards de facto in the field of organic hole transport materials, like the expensive Spiro-OMeTAD.


Hole transporting layers (HTLs) have demonstrated to be fundamental to achieve high efficiency and long-term durability in new-generation hybrid organic photovoltaic devices. They must provide a good chemical stability, a high hole mobility, and an ideal matching between their HOMO levels and the valence band maximum of the active layer, promoting an efficient collection of photogenerated holes at the electrode. In particular, they have been extensively employed in perovskite solar cells (PSCs) [13], contributing to boost their impressive evolution in the last ten years from an initial 3.8% [4] to a 22.1% [5] certified efficiency. More specifically, the presence of a HTL increases the extraction rate of photogenerated holes from the perovskite film and improves their transport to the back contact metal electrode. This, in turn, minimizes recombination losses and charge-induced decomposition of the active layer, thus increasing the overall PSC performance. Several small molecules, polymers, and inorganic materials have been studied and tested for this purpose [6]. 2,2,7,7-Tetrakis(N,N-di-p-methoxyphenilamine)-9,9-spirobifluorene (Spiro-OMeTAD) and poly(triarylamine) (PTAA) are among the most effective organic hole transporting materials (HTMs), yielding highly efficient PSCs [7, 8]. However, their synthesis requires complex multistep procedures resulting in a significant increase of the device price. Phthalocyanines (Pcs) and similar macrocyclic compounds are well-known p-type organic semiconductors and have found applications as hole transporters in organic photovoltaics (OPV) [9, 10] and in organic field effect transistors (OFETs) [11] due to their high chemical stability, large conjugated π-system, and tunable optical/electrochemical properties through chemical functionalization. Such well-known features promoted their use as HTLs in PSCs, and several examples have been reported for vacuum-evaporated and spin-coated layers, depending on their molecular structure [1215]. Recently, an impressive 17.5% efficiency scored by tetra-5-hexyl-2,2-bisthiophene zinc phthalocyanine in a mixed-ion perovskite solar cell has been reported by Cho et al. [16], due to the effect of a favorable packing of the molecule on its transport properties.

Aimed at pushing further down the fabrication cost, a key feature of OPV and PSC technologies, we developed a solution-processable HTL with the target molecule prepared in a one-step synthesis from easily available, cheap precursors. The molecule is a symmetric tetra-n-butoxy-substituted zinc phthalocyanine, hereinafter named (n-BuO)4ZnPc.

Its electron-donating alkoxy peripheral substituents have been chosen to accomplish two main tasks: (i) tuning the (n-BuO)4ZnPc electronic properties, thus optimizing the alignment of HOMO-LUMO levels to the perovskite band edges, and (ii) assuring the (n-BuO)4ZnPc processability within a device by formation of spin-coated homogeneous films with favorable molecular packing.

(n-BuO)4ZnPc has been synthesized with a 53% yield in a one-step reaction, using commercially available 4-n-butoxyphthalonitrile and ZnCl2 as templating salt, as shown in Scheme 1.

The crude has been purified by filtration on silica gel using dichloromethane and THF as eluents, and the resulting purple crystalline solid has been characterized by different techniques. UV-Vis spectra have been measured in transmission mode both in solution (solid black line) and as spin-coated film (solid red line), as shown in Figure 1(a).

The solution spectrum in acetonitrile is dominated by the typical phthalocyanine Q-band absorption, with its maximum peak at 675 nm and a shoulder peak at 610 nm, which was also observed in gas-phase measurements of unsubstituted ZnPc [17]. The latter should therefore be considered as part of the vibrational envelope of the transition rather than a fingerprint of H-type (cofacial) aggregation of the molecules. As expected, the same Q-band is much broader in the case of the (n-BuO)4ZnPc film and composed by two redshifted bands centered at 692 and 737 nm. This behavior is compatible with an “edge-to-edge” macrocycle arrangement typical of J-type (coplanar) aggregation [18, 19]. The steady-state emission spectrum of a (n-BuO)4ZnPc solution in acetonitrile, recorded at  nm, shows a maximum centered at 686 nm, with a small Stokes shift of 11 nm. The related time-correlated single-photon counting measurements, shown in Figure 1(b), were performed at 686 nm, and the resulting decay curve was monoexponentially fitted, providing a  ns. An optical band gap of 1.82 eV was estimated from the interception of the normalized absorption and emission spectra. This result, combined with cyclovoltammetric measurements, allowed to estimate the reduction potentials of the (n-BuO)4ZnPc/(n-BuO)4ZnPc+ and (n-BuO)4ZnPc/(n-BuO)4ZnPc couples indicated as and , respectively. The optical and electrochemical characterizations of the compound are summarized in Table 1.

To evaluate the processability of (n-BuO)4ZnPc in terms of quality, homogeneity, and roughness of its films, atomic force microscopy topographic measurements have been performed. Molecular films have been prepared by spin coating, using 20 mg/ml of toluene solutions on nonconductive glass plates. Large-area scans of 5 μm × 5 μm surfaces, an example is reported in Figure 2, indicated a flat and uniform film in the explored regions, suggesting that the molecule can be successfully layered in a planar device. The root mean square roughness of the reported film was found to be 4.94 nm, with an average height of 4.76 nm and a surface skewness of 0.24 nm. The presence of rare overlying crystallites up to 288 nm (143 nm in Figure 2) has sometimes been detected.

Optical and electrochemical measurements are in agreement with ab initio results based on density functional theory (DFT) [20] and are compared to simulations of Spiro-OMeTAD and tetra-tert-butyl zinc phthalocyanine (hereinafter referred to as (t-Bu)4ZnPc) properties, chosen as internal standards. The theoretical results are summarized in Table 2 and support the idea that the n-butoxy substituents have a combined structural and electronic effect on the properties of (n-BuO)4ZnPc molecular films.

The peripheral n-butoxy substituents induce an extension of the spatial distribution of the (n-BuO)4ZnPc HOMO orbital responsible for the transport of holes to the back electrode, as opposed to the t-Bu substituents of the similar (t-Bu)4ZnPc molecule (see Figure 3). Such a resonant effect is accompanied with the strong electron-donating behavior of O-R groups that stabilizes positive-charge carriers and induces a 0.25 eV rising of the HOMO with respect to the weaker alkyl substituents. This indication is confirmed by an analysis of the charge displacement induced by a hole injected in (n-BuO)4ZnPc (Figure 3), which underlines the role of the peripheral n-butoxy substituents. Theoretical estimates of reduction potentials in acetonitrile [21] are displayed in Figure 4, left side, and offer further comparison between the properties of (n-BuO)4ZnPc and Spiro-OMeTAD. A calculated value of −5.23 eV (vs vacuum) in the case of (n-BuO)4ZnPc approaches the −4.96 eV value estimated in the case of Spiro-OMeTAD, whereas a corresponding reduction potential of −3.31 eV for the value is suitable for blocking free electrons that diffuse in the active perovskite layer. These results confirm that (n-BuO)4ZnPc is a suitable hole transport material for organic and perovskite solar cells. In particular, as shown in Figure 4, right side, in a methylammonium lead triiodide-based solar cell, the −5.18 eV oxidation potential of the phthalocyanine allows the hole transport from the photoactive layer from the perovskite/HTL interface to the gold electrode, while the reduction potential, higher than that of the perovskite, prevents the electrons to diffuse back and recombine on it.

As a proof of concept, several CH3NH3PbI3 PSCs with an active area of 0.04 cm2 have been prepared using a two-step deposition protocol: a widely employed device architecture with a mesoporous TiO2 electron transporting layer (ETL) and a tert-butylpyridine- (TBP-), lithium bis(trifluoromethanesulfonyl)imide- (LiTFSI-), and tris(2-(1H-pyrazol-1-yl)-4-tert-butylpyridine)-cobalt(III) tris(bis(trifluoromethylsulfonyl)imide))-doped HTL has been used. Spiro-OMeTAD control cells have also been prepared and measured as benchmarks. The obtained data have been subjected to the Peirce’s criterion [22] to discard anomalous results. The phthalocyanine scored average efficiencies of 6.06% ± 0.74 (forward bias) and 9.00% ± 0.21 (reverse bias), while 6.75% ± 0.19 (forward bias) and 8.61% ± 0.60 (reverse bias) are the average values obtained using Spiro-OMeTAD. These values, along with the averaged , , and , are reported in Table 3.

Open-circuit voltages average around 0.91 V for both the HTMs, with acceptable standard deviations. Values around 1.0 V are reported in several articles for CH3NH3PbI3-based solar devices [23, 24]; our 10% discrepancy may be due to several factors such as a large reverse saturation current occurring because of a high recombination rate at the TiO2/perovskite and HTL/perovskite interfaces and to the presence of pinholes in the charge transport layers that create small short circuits with the metal contacts. and the fill factor have lower values than the state-of-the-art and high standard deviations, indicating that the examined batches are subjected to some degree of structural inhomogeneity likely ascribable to the quality of the perovskite layer in terms of crystallinity and evenness.

The current density-voltage () curves and the principal photovoltaic parameters of the best n-(BuO)4ZnPc- and Spiro-OMeTAD-based cells are reported in Figures 5(a) and 5(b), respectively.

Both devices show a remarkable hysteresis with consistent performance discrepancies depending on the scan direction, mostly ascribable to differences in and fill factors. No investigations have yet been performed to rationalize this phenomenon in this specific case; but according to the literature, it can be caused by a number of factors such as scan direction and rate, cell preconditioning, and device architecture [25, 26]. As frequently reported, these cells yielded a higher PCE in reverse bias scans rather than in forward bias scans, likely due to the uprising of dynamic trapping and detrapping processes of charge carriers at the perovskite/HTL and perovskite/ETL interfaces and to ion migration that alters the interfacial energy barriers and consequently modifies their height for the carrier collection [27]. The external quantum efficiency (EQE) spectra, shown in Figure 6, have been recorded between 300 and 850 nm.

The photocurrent onset is around 760 nm in both cases, which is in agreement with the 1.55 eV band gap of CH3NH3PbI3. In the Spiro-based device, we observe a maximum efficient charge extraction at 570 nm (EQE = 63%), while the phthalocyanine-based cell peaks at 540 nm (EQE = 58%).

For both Spiro-OMeTAD and (n-BuO)4ZnPc, the integrated photocurrent is lower than the averaged one measured in the JV curves. This effect can occur in many different photovoltaic devices and may be due to several factors [28]. In particular, a significant role may be played by a difference in intensity of the incident light in EQE and JV measurement setups (in orders of magnitude). This points to the sublinear dependence of photocurrent in both cells [29] and need for further optimization of current collection and in particular the cell contact system (including contact/conducting layer interfaces). These preliminary evidences suggest that there is wide room for performance improvements focusing the attention on the optimization of the cell construction, which was beyond our initial goal. Nevertheless, in a nonoptimized device, our phthalocyanine-based HTL is capable of forming an interface with the photoactive layer characterized by hole-extracting and electron-blocking properties. These promising results can be further improved with a careful study of the layer deposition parameters and of the dopant ratio to enhance its hole transport properties.

Since we have stressed the importance of lowering the fabrication costs of novel hole transport materials to foresee their future commercialization, we propose a cost-per-gram analysis as a function of the synthetic complexity of (n-BuO)4ZnPc and compare the obtained results with the current price of Spiro-OMeTAD. We point out that moving from lab to fab requires a complete cost analysis that must include parameters like energy, facility maintenance and personnel costs, taxes, and other charges that we have not taken into account. The total material cost (in euros) for our hole transport material has been estimated according to a paper published in 2013 by Osedach et al. [30]. For the sake of simplicity, Sigma-Aldrich has been chosen as the only supplier for reagents, solvents, and purification materials with the exception of 4-n-butoxyphthalonitrile that has been quoted from TCI chemicals. In all cases, the bulkiest batches available on the online catalogs have been considered—the list of which is reported in table I in the supplementary materials. The costs of our synthesis have been estimated assuming linearity when scaling the procedure from grams to kilograms, calculating the required amounts of chemicals to produce 1 kg of (n-BuO)4ZnPc in a 53% yield (see the synthesis flowchart in Figure S3 in the supplementary materials) and multiplying them by the respective prices. The resulting costs for reagents, solvents, and purification materials are reported in Table 4, given as euro per gram of the synthesized product.

Given that the experimental protocol can be further optimized and that solvents and reagents can be partially recovered and recycled, two total costs are provided. The “worst-case scenario” (wcs) accounts that any unreacted reagent and all the purification materials are lost as wastes. In the “best-case scenario” (bcs), the purification costs are not included because the chromatographic filtration is replaced by a vacuum treatment (10–3 mmHg at 250°C) that completely avoids the use of silica gel and solvents. Even if we did not include them in our analysis, it is worth to mention that diminishing the amount of wastes is beneficial for reducing their disposal costs.

Spiro-OMeTAD is commercially available and has been quoted from three suppliers as reported in table II in the supplementary materials. With respect to the less-expensive batch, supplied by Sigma-Aldrich at 242.40 €/g, our molecule is 2.8 times cheaper in the worst-case scenario and 3.3 times cheaper in the best-case scenario. Since a large-scale production requires kilograms of input reagents and 4-n-butoxyphthalonitrile is sold in 5 g batches at the most, we requested a bulk quotation for a batch of 1 kg, reducing its price from 42.40 €/g to 17.50 €/g. The recalculated synthesis costs, given as euro per gram of synthesized product, are summarized in Table 5.

In this way, (n-BuO)4ZnPc results to be 7.9 times cheaper than Spiro in the best-case scenario, and given the limited number of suppliers that we have taken into consideration and the possibility to request bulk quotations of all the other materials, we are confident that its fabrication costs can be further lowered.

In conclusion, we have synthesized and characterized an easily processable, symmetrically tetrasubstituted zinc phthalocyanine, which is a cheap and promising hole transporting material for new-generation photovoltaics. We have also provided a first proof of concept assessment of its implementation in perovskite solar cells as proof of concept, obtaining promising results even in nonoptimized devices. We suggest therefore that improvements in the cell construction as well as the optimization of the molecular layer can promote (n-BuO)4ZnPc as a possible alternative to commercially available organic p-type semiconductors thanks to its straightforward, potentially scalable, and less-expensive synthetic procedure.

1. Experimental and Computational Details

1.1. (n-BuO)4ZnPc Synthesis and Characterization

ZnCl2 and all the solvents (RPE grade) were purchased from Carlo Erba except for N,N-dimethylethanolamine (99.5%) which was purchased from Aldrich. Acetonitrile was distilled over calcium hydride. 4-n-Butoxyphthalonitrile (95%) was purchased from Tokyo Chemical Industry Co. (TCI). 1,5-Diazabicyclo[5.4.0]undec-5-ene (DBU) (≥99%) was purchased from Fluka. Reactions were monitored by thin-layer chromatography (TLC) employing a polyester layer coated with 250 mm F254 silica gel. The crude purification was performed by filtration using silica gel Carlo Erba Reactifs SDS 60A C.C. 35–70 mm and dichloromethane and THF as eluents. To perform the fluorescence measurements, a further purification has been performed by size exclusion chromatography using Bio-Beads S-X Resin (Bio-Rad) as the stationary phase and THF as the mobile phase. 1H NMR spectra were recorded in THF d8 on a Bruker AVANCE 600 NMR spectrometer operating at a proton frequency of 600.13 MHz; chemical shifts (δ) are given in ppm relative to the residual solvent peaks of the deuterated solvent (3.58, 1.73 ppm). Infrared spectra were recorded on a Shimadzu FT-IR prestige-21 spectrometer using an attenuated total reflectance (ATR) unit. All the UV-Vis measurements were performed in direct transmission mode. Absorption spectra of 10−5 M solution in acetonitrile were recorded on a PerkinElmer LAMBDA 950 UV-Vis/NIR spectrophotometer, while transparent sensitized films have been analyzed using a Cary 5000 UV-Vis-NIR spectrophotometer (Agilent Technologies). MALDI-TOF spectra have been recorded at Toscana Life Sciences facility in a MALDI-TOF/TOF Ultraflex III (Bruker) using α-cyano-4-hydroxycinnamic acid (HCCA) as matrix. Isotopic pattern simulations have been performed by Bruker Daltonics program. Steady-state fluorescence spectra were recorded on a 10−6 M solution in acetonitrile with a Jobin Yvon Fluorolog3 spectrofluorometer, in the range 620–850 nm, using 5 nm grids for excitation and 5 nm for emission ( nm). The corresponding excitation spectrum has been collected in the range 350–700 nm, under an excitation wavelength of 750 nm, with 10 nm grids. No filters have been used. Time-resolved fluorescence measurements were carried out by a time-correlated single-photon counting (TCSPC) system (HORIBA Jobin Yvon), using a 405 nm pulsed laser diode and collecting the emission decay at the corresponding maximum emission wavelength (686 nm). The fluorescence decay profile was analyzed through decay analysis software (DAS6a HORIBA Scientific). The curve has been fitted with a monoexponential decay, with a final time decay of 2.4 ns. The quality of the fit was checked by examining the residual distribution having a value of 1.4. Voltammograms were recorded at 25°C with a potentiostat-galvanostat Metrohm PGStat 204 in a conventional three-electrode cell; a platinum disk (~1 mm) was used as working electrode together with a platinum wire as auxiliary electrode. The reference electrode was Ag/AgNO3 (0.01 M) in distilled acetonitrile (ACN) (Aldrich) ( vs NHE = 0.548 V), and the Fc+/Fc (ferrocenium/ferrocene) couple was used as external standard. The sample solutions were ~10−4 M in distilled anhydrous solvents, and dry tetra(n-butyl)ammonium tetrafluoroborate (TBATFB) (Aldrich) was used as the supporting electrolyte at 0.1 M. The solutions were previously purged for 10 minutes with nitrogen, and all measurements were performed under nitrogen. Cyclic voltammetry (CV) investigations were carried out at scan rates typically ranging from 0.05 to 2 Vs−1. AFM images have been acquired in contact mode using a Keysight (formerly Agilent/MI) 5500 AFM with AppNano HYDRA6R-200N probes in air at room temperature and analyzed with the freely downloadable software WSxM 5.0 develop 8.4 December2016 [31].

Synthesis of (n-BuO)4ZnPc (Mixture of Regioisomers). 200 mg of 4-n-butoxyphthalonitrile (1.0 mmol, 4 eq), 34 mg of zinc chloride (0.250 mmol, 1 eq), and a catalytic amount of 1,5-diazabiciclo[5.4.0]undec-5-ene (DBU) (1%mol, 1.7 μl) have been dissolved in 1 ml of N,N-dimethylethanolamine in a round bottom flask. The mixture has been refluxed for 15 hours and then cooled down, diluted with water, and filtered to obtain a solid which has been washed with water until the pH is neutral and the washings are colorless. The desired molecule has been purified by filtration on silica gel using dichloromethane and THF as eluents, affording 114 mg (yield 53%) of the target molecule as a shiny purple crystalline solid. 1H NMR (THF-d8, ppm), δ: 1.21–1.25 (m, 12H; CH3), 1.80–1.88 (m, 8H; CH2), 2.08–2.16 (m, 8H; CH2), 4.50 (m, 8H; OCH2), 7.45–7.55 (m, 4H; macrocycle), 8.39–8.49 (m, 4H; macrocycle), 8.79–8.90 (m, 4H; macrocycle); IR (cm−1): 2954–2870 (cluster), 1604, 1236, 1089, 1047, 742, UV-Vis (acetonitrile, nm,): 675, 354, MALDI-TOF (m/z) [M+H+] 866.3, elemental analysis: calcd = C (66.54%), N (12.93%), H (5.58%), found: C (66.07), N (13.05%), H (5.70%).

1.2. Device Fabrication

All the chemicals used for the device fabrication were purchased from Aldrich except methylamine (40% in methanol) which was purchased from Tokyo Chemical Industry Co. (TCI). TiO2 paste was purchased from Dyesol. Methylammonium iodide (MAI) was synthesized by stirring 55.6 ml of methylamine (40% in methanol) with 60 ml of hydroiodic acid (57 wt% in water) in a 250 ml round bottom flask at 0°C for 2 h. The solvent was then evaporated at 50°C for 1 h using a rotary evaporator; the resulting solid was collected and stirred with ethanol for 30 minutes. It was then filtered and repeatedly washed with diethyl ether to obtain a white solid that has been dried at 60°C in a vacuum oven for 24 h.

The fluorine-doped tin oxide-coated SnO2:F (FTO) conducting glass (15, Pilkington) substrates have been cut in 2.5 × 1.5 cm size. The substrates were then etched with Zn/HCl and cleaned properly. A hole-blocking layer was spin coated on FTO using a solution of titanium diisopropoxide bis(acetylacetonate) (TiDIP) (75% in 2-propanol) in ethanol and then annealed at 450°C for 30 minutes. A mesoporous titania layer was then deposited by spin coating using TiO2 paste (DSL-90T TiO2 NPs paste) diluted in ethanol (0.1 g of paste in 0.32 g of absolute ethanol) and calcinated at 500°C for 30 minutes. The TiO2-deposited substrates were then treated with TiCl4 for 30 minutes at 70°C, then rinsed, and annealed at 500°C for 30 minutes. The perovskite has been deposited using a two-step method. In the first step, 1 M solution of PbI2 in a DMF : DMSO mixture (85 : 15 ratio) was spin coated on the substrates and annealed at 70°C for 45 min. The PbI2-coated films were then immersed in a 10 mg/ml solution of MAI in 2-propanol for 30 seconds, resulting in the formation of dense black-brown MAPbI3. The cells were further annealed at 90°C for 30 min. 40 μl of chlorobenzene solutions of (n-BuO)4ZnPc (40 mg/ml) and Spiro-OMeTAD (72.3 mg/ml) was deposited by spin coating. 26.3 μl of Bis(trifluoromethane)sulfonimide lithium salt in acetonitrile (520 mg/ml), 29 μl of tris(2-(1H-pyrazol-1-yl)-4-tert-butylpyridine)-cobalt(III) tris(bis(trifluoromethylsulfonyl)imide)) in acetonitrile (300 mg/ml), and 19.2 μl of 4-tert-butilpyridine were added as dopants. Finally, 70 nm-thick gold-back contacts have been evaporated under pressure of 5 × 10−6 Torr. The active area was 0.09 cm2, and a 0.04 cm2 screen mask has been used during the measurements.

1.3. Photovoltaic Characterization

Photovoltaic measurements were made on a Newport system, composed of an Oriel I–V test station with an Oriel Sol3A simulator. The solar simulator is class AAA for spectral performance, uniformity of irradiance, and temporal stability. The solar simulator is equipped with a 450 W xenon lamp. The output power is adjusted to match AM1.5 global sunlight (100 mW·cm2). The spectral match classifications are IEC60904-9 2007, JIC C 8912, and ASTM E927-05. I–V curves were obtained by applying an external bias to the cell and measuring the generated photocurrent with a Keithley model 2400 digital source meter. A 10 mV voltage step and a 40 ms photocurrent delay time were used. External quantum efficiency was calculated by measuring the of both the tested cell and a reference silicon cell at every wavelength, according to the following equation:

1.4. Computational Details

The structural, electronic, redox, and optical properties of (n-BuO)4ZnPc, (t-Bu)4ZnPc, and Spiro-OMeTAD have been investigated using ab initio simulations based on density functional theory (DFT). In detail, the calculations have been performed by using the ORCA suite of programs [20] in a localized-basis-set framework. Kohn-Sham orbitals have been expanded on a def2-TZVPP Gaussian-type basis set [32]. Fully decontracted def2-TZVPP/J has been also used as an auxiliary basis set for Coulomb fitting in a resolution-of-identity/chain-of-spheres (RIJCOSX) framework [33]. Molecular geometries have been fully optimized at the B3LYP level of theory [34, 35], including dispersion forces calculated by using the DFT-D3 approach [36]. Redox potentials of the molecules have been calculated by using the M06-2X functional [37], as discussed in detail elsewhere [21]. The molecules and their positive and negative ions have been embedded in an implicit CH3CN solvent using the conductor-like polarizable continuum model (CPCM) [38] and fully optimized. The corresponding redox potentials have been calculated as values between neutral and charged species.

Data Availability

The phthalocyanine photoelectrochemical characterization data and the photovoltaic characteristics of the best-performing phthalocyanine-based and Spiro-based perovskite solar cells, used to support the findings of this study, are included within the article. The complete IV measurements can be furnished upon request. Furthermore, the experimental protocols to reproduce the synthesis and prepare the photovoltaic devices, the 1H NMR chemical shifts, the IR main peaks, the UV-Vis absorption maxima, the MALDI-TOF molecular ion mass to charge ratio, the elemental analysis values of our molecule, and the computational details for the ab initio simulations are made available in the experimental section. The 1H NMR full spectrum, MALDI spectrum, and anodic voltammogram of (n-BuO)4ZnPc, used to confirm the molecular structure of our compound and support the findings of this study, are included in the supplementary materials.

Conflicts of Interest

The authors declare no competing financial interests.

Authors’ Contributions

Gloria Zanotti and Giuseppe Mattioli contributed equally to the work.

Acknowledgments

G.Z. is thankful to the Ente Nazionale Energia e Ambiente (ENEA) and to the Italian Ministry of Foreign Affairs for a visitor post-doc fellowship to Ben Gurion University of the Negev. The authors are thankful to Sara Notarantonio for her precious assistance in the experimental work and in the electrochemical measurements. The research was funded in part by Israel’s Ministry of National Infrastructures, Water and Energy Resources (grant no. 0399202/215-11-037) and by the Adelis Foundation.

Supplementary Materials

1H NMR spectrum, MALDI spectrum, and anodic voltammogram of (n-BuO)4ZnPc. Further details of the HTM cost analysis, consisting in two price tables and a synthesis flowchart. (Supplementary Materials)