Synthesis and Photovoltaic Performance of β-Amino-Substituted Porphyrin Derivatives

Abstract

New β-amino-substituted porphyrin derivatives bearing carboxy groups were synthesized and their performance as sensitizers in dye-sensitized solar cells (DSSC) was evaluated. The new compounds were obtained in good yields (63–74%) through nucleophilic aromatic substitution reactions with 3-sulfanyl- and 4-sulfanylbenzoic acids. Although the electrochemical studies indicated suitable HOMO and LUMO energy levels for use in DSSC, the devices fabricated with these compounds revealed a low power conversion efficiency (PCE) that is primarily due to the low open-circuit voltage (V_oc) and short-circuit current density (J_sc) values.

1. Introduction

Fossil fuels such as coal, oil, and natural gas are still the main energy source for industrial and social human activities. However, because they are finite and produce unwanted consequences, namely pollution and climate change, it is becoming important to develop technologies based on renewable and non-polluting energy sources. Sunlight is an endless source of clean and free energy and, thus, harvesting solar energy is an obvious approach for dealing with energy problems. Dye-sensitized solar cells (DSSC) have attracted significant interest for the conversion of sunlight into electricity because they are easily fabricated, have low costs, and are environmentally friendly [1,2]. In general, sensitizers used in DSSC are divided into three main groups: ruthenium complexes [3,4,5], metal-free organic dyes [6,7,8,9], and porphyrin/phthalocyanine dyes [10,11,12,13,14].

One of the most successful DSSC described in the literature uses ruthenium-based dyes, with a power conversion efficiency (PCE) of about 12% [15,16] but, unfortunately, ruthenium is not an earth-abundant element and is expensive [17,18,19]. In order to solve this problem, several researchers have been focused on the replacement of ruthenium complexes by organic dyes that can achieve similar (or better) PCE. Porphyrin derivatives are an effective alternative to ruthenium sensitizers, as they exhibit a series of photophysical properties that are appropriate for DSSC, such as broad absorption in the visible region, and favorable cell kinetics for electron injection and charge recombination [20]. That explains why porphyrins have been so extensively investigated as dyes in DSSC during the last decade [11,13,20,21,22]. In 2014, the performance of porphyrin-based DSSC (PCE = 13%) [23] exceeded that of the ruthenium-based DSSC.

Porphyrins may be functionalized at the meso- or β-pyrrolic positions with substituents adequate for anchoring to TiO₂. Typically, porphyrins used in DSSC present meso-carboxyphenyl groups or meso-carboxyalkynylphenyl groups while β-substituted porphyrins usually have conjugated alkenyl systems bearing a terminal carboxylic acid [24,25].

In this paper, we report the synthesis of porphyrins bearing β-amino substituents containing a terminal benzoic acid unit. Some of the new porphyrins bear fused rings. The photophysical and electrochemical properties of the new compounds and their performance as dyes in DSSC are also discussed.

2. Results and Discussion

2.1. Synthesis

The synthesis of the porphyrin derivatives bearing carboxy groups 5a–c and 6a,b (Figure 1) required the previous preparation of (2-amino-5,10,15,20-tetraphenylporphyrinato)copper(II) (1) (Scheme 1). This compound was prepared by nitration of (5,10,15,20-tetraphenylporphyrinato)copper(II) followed by reduction of the 2-nitro group under standard Sn/HCl conditions [26]. The reaction of the 2-aminoporphyrin 1 with hexafluorobenzene was carried out in dimethylformamide (DMF) at 80 °C, for 5 h, and using K₂CO₃ as base. The TLC of the reaction mixture revealed the total consumption of the starting porphyrin and the formation of the N-pentafluorophenylamino derivative 2a as the major product (92% yield) accompanied by a minor amount of the bis(pentafluorophenyl)aminoporphyrin 3 (0.8% yield). The formation of compound 3 was not a surprise. In fact, we had already reported the synthesis of a similar porphyrin derivative bearing a bis(pentafluorophenyl)amino group at the meso position [27]. The structures of compounds 2a and 3 were confirmed by their absorption spectra (ESI, Figures S1 and S4) and mass spectra (ESI, Figures S2, S3 and S5) that reveal a peak at m/z = 857 for 2a and at m/z = 1022 for 3, corresponding to the protonated molecular ion [M + H]⁺ and to the molecular ion M⁺˙ of the proposed structures, respectively.

The demetalation of porphyrin 2a with a mixture of H₂SO₄/CH₂Cl₂ afforded the free base 2b in quantitative yield. The structure of 2b was confirmed by ¹H, ¹⁹F, and ¹³C NMR, UV–Vis and by MS (ESI, Figures S8–S15). Its ¹H NMR spectrum (ESI, Figure S6) presented the expected signals in the aromatic region due to the resonances of six β-pyrrolic protons (δ 8.60–8.90 ppm), and of the meso-phenyl groups plus the β-pyrrolic proton at position three (δ 7.60–8.30 ppm). Additionally, the spectrum showed a singlet at δ 5.95 ppm due to the resonance of one NH proton, which is in accordance with the occurrence of a mono-substitution. The ¹⁹F NMR spectrum (ESI, Figure S7) shows three signals at δ −143.9, −159.3, and −159.6 to −160.4 ppm, with relative intensities 2:2:1, in the form of doublet, triplet, and multiplet, respectively, which can be assigned, respectively, to the ortho-, meta-, and para-fluorine atoms.

Considering that the extension of the electronic π-system of a porphyrin macrocycle results, in general, in the red-shift of the Q bands (or the appearance of new absorption bands), we decided to synthesize porphyrin derivative 4 bearing a fused ring system. The formation of 4 involved the oxidative cyclization of the 2-aminoporphyrin 2b following a procedure reported by our group [28]. The desired compound 4 was isolated in 53% yield and its structure was confirmed by ¹H, ¹⁹F, and ¹³C NMR, UV–Vis and MS (ESI, Figures S12–S17). The mass spectrum (ESI, Figure S16) showed a peak at m/z = 794 corresponding to the protonated molecular ion [M + H]⁺. In the ¹H NMR spectrum (ESI, Figure S12), the absence of the singlet at δ 5.95 ppm, corresponding to the resonance of the NH proton observed in 2b, and the presence of two multiplets at δ 8.15–8.30 ppm, corresponding to the resonances of six ortho-Ph-H protons only, are clear evidence that the oxidative cyclization occurred.

Knowing that pentafluorophenyl groups react easily with nucleophiles [27], and particularly with thiols [29], by nucleophilic aromatic substitution reactions, and considering that dyes to be applied in DSSC should contain an anchoring group to bind to the surface of titanium dioxide, we performed the reaction of porphyrins 2b and 4 with 4-sulfanyl- and 3-sulfanylbenzoic acid. The structures of the resulting porphyrins 5a,b and 6a,b are shown in Figure 1. The nucleophilic aromatic substitution took place in DMF in the presence of K₂CO₃ at room temperature. The porphyrin derivatives 5a,b and 6a,b were isolated in good yields (63–74%) and their structures were confirmed by ¹H, ¹⁹F, and ¹³C NMR, UV–Vis and MS (ESI, Figures S18–S41). The main evidence that the nucleophilic aromatic substitution occurred was observed in the ¹⁹F NMR spectra. For example, the spectrum of 5a showed two main signals, in the form of doublets, integrating for two protons each, at δ −132.1 and −149.0 ppm (ESI, Figure S19).

In order to evaluate the influence of the porphyrin metalation on the efficiency of DSSC devices, it was decided to prepare the Zn(II) complex 5c. The metalation was carried out by adding zinc(II) acetate to a solution of 5b in chloroform and methanol. After crystallization from dichloromethane/methanol, compound 5c was obtained in 98% yield. The structure of the Zn(II) complex 5c was confirmed by its ¹⁹F NMR, UV–Vis and MS spectra (ESI, Figures S42–S44). It was not possible to obtain the ¹H NMR spectrum of this compound due to their very low solubility (probably due to the formation of aggregates or ordered polymeric structures, resulting from intermolecular Zn–amine interactions) [30,31].

2.2. Spectroscopic Properties

The absorption and emission spectra of porphyrin derivatives 2–6 in chloroform are shown in Figure 2, Figure 3 and Figure 4 and summarized in

Table 1. Absorption and emission data of compounds 5a–c and 6a,b in chloroform.

Compounds	λmax,abs (nm) a	log ε (M−1 cm−1)	λmax,em (nm) b	Stokes Shift (cm−1) c
5a	424	5.26	655 719	94
	523	4.31
	554	3.89
	594	3.86
	651	3.48
5b	415	5.03	657	164
	424	5.03719
	522	4.13
	550	3.79
	594	3.72
	650	3.35
5c	430	4.74	608	—
	558	3.73
	592 sh	3.22	658
	629 sh	2.87
6a	425	4.46	661	69
	440	4.48
	544	3.52
	584	3.74	727
	605	3.59
	658	3.59
6b	424	4.43	661	69
	444	4.50
	544	3.46
	585	3.71	728
	605	3.54
	658	3.52

^a 10⁻⁵–10⁻⁶ mol L⁻¹ in CHCl₃. ^b Excitation at maximum absorption. ^c Q4 band.

for compounds 5a–c and 6a,b. As expected, all free-base porphyrins (2b, 4, 5a, 5b, 6a, 6b) exhibit absorption spectra composed of an intense Soret band (415–444 nm) and four less intense Q bands (ca. 500–658 nm). In turn, metalated derivatives 2a, 3, and 5c show fewer Q bands as a result of a change in symmetry and orbital degeneracy upon complexation [32]. Metalation leads to a higher planarization of the porphyrin, with increased delocalization and red shift of the absorption bands, as observed when comparing 2a with 2b, and 5c with 5a and 5b. Leaving behind the synthetic intermediate porphyrins 2–4, a more detailed analysis of the spectroscopic properties of the final dyes 5–6 shows no significant difference between the meta- and para- substitution of the carboxylic acid group in the fluorinated benzene ring. More significant is the rigidification imposed by ring fusion when transitioning from porphyrins 5 to porphyrins 6, as can be noticed by the different colors of both types of porphyrins (orange-red for 5; greenish for 6 when adsorbed on TiO₂). The more conjugated ring-fused porphyrins 6 present red-shifted bands when compared to 5 (16–20 nm in the Soret band, 7–8 nm in the Q4 band), with significant splitting of the Soret band. The emission spectra reflect corresponding red-shifts in the Soret and Q bands. The smaller Stokes shift of porphyrins 6 relative to 5 reflect the higher rigidification in the cyclized porphyrins. In the case of the metalated derivative 5c, a smaller red-shift could also be expected relative to 5b due to planarization, but the lowest-energy Q band is too broad for calculating an accurate absorption maximum.

2.3. Electrochemical Characterization

The electrochemical properties of compounds 5 and 6 were studied using differential pulse voltammetry (DPV) aiming to evaluate the suitability of these porphyrins as sensitizers for DSSC. The electrochemical data and the energy-level diagram are summarized in

Table 2. Electrochemical properties of porphyrin derivatives 5 and 6 obtained from differential pulse voltammetry measurements in CH₂Cl₂ solution with a dye concentration of 0.5 mM and 0.1 M TBAPF₆ at scan rates of 10, 20, and 30 mV s⁻¹.

Dye	Eox SCE (V)	HOMO vs. Vacuum (eV)	Ered SCE (V)	LUMO vs. Vacuum (eV)	Band Gap (eV)
5a	0.819	−5.26	−1.063	−3.38	1.88
5b	0.761	−5.20	−1.151	−3.29	1.91
5c	0.649	−5.09	−1.210	−3.23	1.86
6a	0.638	−5.08	−1.068	−3.37	1.71
6b	0.677	−5.12	−0.995	−3.45	1.67

and Figure 5, respectively. From the obtained onsets of the oxidation and reduction peaks, the values of the HOMO and LUMO energies were calculated with the following equation: E [eV] = −(E_onset (V vs. SCE) + 4.44) [33]. The onset values were considered the intersection points between the tangent lines of the rising current and the baseline current (see Supplementary Materials, Figures S45–S49). Additionally, the HOMO/LUMO potentials of porphyrins 5 and 6 were compared with those of the electrolyte (I⁻/I₃⁻ redox pair) and the conducting band (CB) of TiO₂. The highest occupied molecular orbital energy levels, located between −5.26 and −5.08 eV, are more negative than the I⁻/I₃⁻ redox couple potential (−4.60 eV) [34], which ensures efficient regeneration of the oxidized dye. The lowest unoccupied molecular orbital energy levels, located between −3.45 and −3.23 eV, are higher than the CB of TiO₂ (−4.0 eV) [34], indicating that electron injection from the excited state of the dyes to the CB of TiO₂ surface is thermodynamically permitted.

Porphyrins with fused rings (6a and 6b) have lower band gap values due to the extension of the conjugated system, corroborating the observed red-shifts when transitioning from porphyrins 5a and 5b to 6a and 6b. This rigidification also results in lower oxidation potentials when compared to the non-rigidified analogs, suggesting an increased ability to donate electrons. It was expected that this would lead to an improvement in photocurrent generation when compared with 5a and 5b, but no correlation was observed. The Zn(II) complex 5c has higher HOMO and LUMO levels and a narrower bandgap when compared with 5b.

2.4. Fabrication of DSSC Using the Porphyrin Derivatives 5a–c and 6a,b

The preparation of the DSSC devices with each porphyrin 5a–c and 6a,b is detailed in the experimental section. Upon adsorption onto a thin film of TiO₂ deposited on the surface of an FTO-coated glass slide, all compounds resulted in highly colored and homogeneous films (Figure 6).

The photovoltaic performance of the resulting DSSC devices was compared with that obtained from the dye N719, which was selected as reference. The results were obtained through I–V measurements under 100 mW cm⁻² simulated AM1.5G illumination, using 0.8 M LiI and 0.05 M I₂ in acetonitrile/pentanenitrile (85:15, % v/v) as electrolyte. The results summarized in Figure 7 and

Table 3. Photovoltaic performance of the porphyrin derivatives 5 and 6 and N719 under 100 mW cm⁻² simulated AM 1.5 G illumination, using 0.8 M LiI and 0.05 M I₂ in acetonitrile/pentanenitrile (85:15, % v/v) as electrolyte.

Compound	Voc (mV)	Jsc (mA cm−2)	FF	PCE (%)
5a	213 ± 4	0.30 ± 0.02	0.41 ± 0.08	0.03 ± 0.01
5b	206 ± 25	0.23 ± 0.03	0.43 ± 0.12	0.02 ± 0.01
5c	361 ± 7	1.8 ± 0.1	0.56 ± 0.06	0.37 ± 0.06
6a	185 ± 5	0.17 ± 0.03	0.42 ± 0.03	0.013 ± 0.003
6b	280 ± 8	0.35 ± 0.01	0.49 ± 0.05	0.049 ± 0.003
N719	427 ± 12	15.3 ± 0.5	0.57 ± 0.01	3.70 ± 0.09

show that all porphyrin derivatives exhibit low efficiency with PCE values varying between 0.4% and 10% of the value obtained for N719.

The low performance of these dyes is primarily due to their low open-circuit voltage (V_oc) and the short-circuit current density (J_sc) values (Table 3). This outcome can be attributed to the common occurrence of aggregation in this family of compounds and/or to an inefficient electron-injection from the excited molecule to the semiconductor.

A comparison between the DSSC performance using the free-base dyes 5a and 5b allows us to conclude that the position of the carboxy group in the terminal benzoic acid unit has no significant effect (Table 2). This is in line with the very similar absorption spectra observed for both dyes, and also the observed similar bandgap. It could be that the position of the carboxy group would lead to a different orientation of the dye on the surface of the anode, as observed in other porphyrin systems [35], but the similar poor efficiency of the DSSC cells of 5a and 5b suggests that no such effect is present. Comparing 5b and 5c, the presence of Zn(II) in 5c leads to a noticeable efficiency enhancement with an increase in the PCE from 0.02 to 0.37%. This improvement can be attributed to an increase of approximately 160 mV in V_oc (1.8-fold increase) and of 1.6 mA cm⁻² in the J_sc value (7.8-fold increase). Large enhancements in DSSC efficiency upon Zn metalation in porphyrins have been attributed to a mediator role of the metal ion in the injection step [36]. The closed-shell nature of the Zn(II) ion with empty coordination sites can allow for a rapid injection of photoexcited porphyrin electrons into the TiO₂ conduction band, reflected in the high increase in J_sc in 5c when compared to 5b.

When comparing the PCE values obtained for the non-cyclized derivatives 5 with those of the corresponding rigidified structures 6, there is no apparent pattern for the rigidification effect that can be established. For instance, a decrease in the PCE was observed when comparing 5a with 6a (0.03% vs. 0.013%), while the opposite was observed for the other pair 5b and 6b (0.02% vs. 0.049%).

The overall low efficiencies of the DSSC devices built with these porphyrin derivatives may be associated with the presence of the electronegative fluorine atoms, which might lead to an inefficient electron-injection from the excited molecule to the semiconductor.

2.5. Effect of CDCA Addition on the DSSC Performance

To assess if the low efficiency results obtained for the DSSC are due to the formation of aggregates, studies with chenodeoxycholic acid (CDCA) were performed. CDCA is a recognized de-aggregating agent commonly employed in DSSC. The incorporation of CDCA as a co-adsorber in DSSC has been shown to enhance cell performance by influencing the photocurrent through increased charge collection and/or electron injection, as well as by improving open-circuit photovoltage through the suppression of charge recombination [37,38,39,40].

In this study, 50 mM of CDCA was added to the solution of 5a, as well as a few drops of methanol to facilitate the solubilization of CDCA. The outcome of this operation was a significant decrease in the adsorbed dye (Figure 8), suggesting a direct competition of CDCA for the anchoring sites on the TiO₂ surface.

The addition of CDCA had a positive impact on the performance of the resulting DSSC, as evident from the outcomes detailed in

Table 4. Comparison of the photovoltaic performance under 100 mW cm⁻² simulated AM 1.5G illumination for DSSC based on compound 5a adsorbed from 0.5 mM solutions in dichloromethane, and 5a plus 50 mM CDCA, using 0.8 M LiI and 0.05 M I₂ in acetonitrile/pentanenitrile (85:15, % v/v) as electrolyte.

Compound	Voc (mV)	Jsc (mA cm−2)	FF	PCE (%)
5a	213 ± 4	0.30 ± 0.02	0.41 ± 0.08	0.03 ± 0.01
5a+ CDCA	329 ± 7	0.53 ± 0.02	0.63 ± 0.01	0.11 ± 0.01

and in Figure 9. The incorporation of CDCA resulted in the improvement in all cell parameters, resulting in a final efficiency of 0.11%. This increment in photocurrent and photovoltage can be attributed to the de-aggregating effect, which will avoid quenching and, thus, facilitate charge injection, as well as to TiO₂ surface passivation by CDCA molecules, which translates into the clear increase verified in the FF, from 0.41 to 0.63, possibly indicating a decrease in dark current phenomena. The increase in PCE upon the addition of CDCA is, however, limited by the lower adsorption of the dye which led us to not extend this test to the other porphyrins.

3. Materials and Methods

3.1. Materials

All chemicals were used as provided. Solvents were used as received or distilled and dried using standard procedures according to literature procedures [41]. The 2-amino-5,10,15,20-tetraphenylporphyrinato)copper(II) (1) was obtained by nitration of 5,10,15,20-tetraphenylporphyrin (TPP), followed by reduction of the 2-nitro derivative, according to literature procedures [26].

3.2. Methods

¹H and ¹³C NMR spectra were recorded on a Bruker Avance 300 (at 300.13 and 75.47 MHz, respectively) or on a Bruker Avance 500 (at 500 and 125 MHz, respectively) spectrometers (Bruker, Wissembourg, France). ¹⁹F NMR spectra were obtained on a Bruker Avance 300 at 282 MHz. CDCl₃ or DMSO-d₆ were used as solvents with tetramethylsilane (TMS) as the internal reference. Chemical shifts are expressed in δ (ppm) and the coupling constants (J) in hertz (Hz). UV–Vis spectra were recorded on a Shimadzu UV-2501PC spectrophotometer (Shimatzu, Kyoto, Japan) using CHCl₃ as solvent. λ_max values are in nm and log ε values were calculated from molar absorptivity in M⁻¹cm⁻¹. The emission spectra were recorded on a Horiba Jobin-Yvon Fluoromax 4 spectrofluorimeter (Shimatzu, Kyoto, Japan) using CHCl₃ as solvent. Mass spectra (MS) were recorded using a Micromass Q-TOF-2TM mass spectrometer (Micromass, Manchester, UK) and CHCl₃ as solvent. High-resolution mass spectra (HRMS-ESI) were obtained in a Q-Exactive^® hybrid quadrupole Orbitrap^® mass spectrometer (Thermo Fisher Scientific, Bremen, Germany). The instrument was operated in positive mode, with a spray voltage at 3.0 kV, and interfaced with a HESI II ion source. The analyses were performed through direct infusion of the prepared solutions at a flow rate of 10 μL min⁻¹ into the ESI source. Spectra were analyzed using the acquisition software Xcalibur (ver. 4.0, Thermo Scientific, San Jose, CA, USA). Preparative thin layer chromatography was carried out on 20 cm × 20 cm glass plates coated with silica gel (1 mm thick). Column chromatography was performed using silica gel (Merck (Darmstadt, Germany), 35–70 mesh). Analytical TLC was carried out on precoated sheets with silica gel (Merck 60, 0.2 mm thick).

3.3. Synthesis and Characterization

3.3.1. Synthesis of 2a

A solution of 1 (185 mg, 268 µmol), hexafluorobenzene (2 mL), and potassium carbonate (222 mg, 1.6 mmol, 6 equiv.) in dry DMF (4 mL) was stirred at 80 °C for 5 h under a nitrogen atmosphere. After cooling at room temperature, the reaction mixture was diluted with dichloromethane and washed with water. The organic phase was dried (Na₂SO₄) and the solvent was evaporated under reduced pressure. The residue was purified by column chromatography (silica gel) using dichloromethane/hexane (1:1) as the eluent. The fraction with higher R_f was identified as the bis(pentafluorophenyl)aminoporphyrin 3 (2.1 mg, 0.8% yield). A second fraction was identified as the porphyrin 2a (210 mg, 92% yield).

2a: UV–Vis (CHCl₃): λ_max (log ε) 420 (4.63), 545 (3.58), 587 (3.08). MS (ESI(+)): m/z 857.3 [M + H]⁺. HRMS-ESI(+): m/z calcd. for C₅₀H₂₉CuF₅N₅ [M + H]⁺ 857.1634; found 857.1631.

3: UV–Vis (CHCl₃): λ_max (log ε) 422 (5.34), 546 (4.18), 582 (3.70). MS (ESI): m/z 1022.3 M⁺˙.

3.3.2. Synthesis of 2b

The demetalation of 2a (50 mg) was carried out with 10% H₂SO₄ in CH₂Cl₂. After about 5 min at room temperature, the reaction mixture was neutralized with an aqueous K₂CO₃ solution and extracted with dichloromethane. The organic layer was dried over Na₂SO₄ and, after evaporation of the solvent under reduced pressure, the residue was crystallized from dichloromethane/methanol. Porphyrin 2b was obtained in quantitative yield.

¹H NMR (300 MHz, CDCl₃): δ −2.75 (s, 2H, NH), 5.95 (s, 1H, NH), 7.64–7.77 (m, 10H, m,p-Ph-H + 3-H), 7.82–7.84 (m, 3H, m,p-Ph-H), 8.14–8.23 (m, 8H, o-Ph-H), 8.62 (d, 1H, J = 4.8 Hz, β-H), 8.77–8.85 (m, 5H, β-H). ¹⁹F NMR (282 MHz, CDCl₃): −159.9 to −160.3 (m, 1F), −159.3 (t, 2F, J = 21.2 Hz), −143.9 (d, 2F, J = 21.2 Hz). ¹³C NMR (75 MHz, CDCl₃): 111.7, 116.8, 118.3, 120.2, 121.4, 126.1, 126.6, 126.7, 126.8, 126.9, 127.7, 127.8, 128.3, 128.6, 129.2, 132.8, 134.4, 134.5, 134.6, 134.7, 140.7, 142.0, 142.2, 142.5. δ. UV–Vis (CHCl₃): λ_max (log ε) 419 (5.07), 522 (4.13), 559 (3.62), 594 (3.69), 650 (3.37). MS (ESI): m/z 796.5 [M + H]⁺. HRMS-ESI(+): m/z calcd. for C₅₀H₃₁N₅F₅ [M + H]⁺ 796.2494; found 796.2498.

3.3.3. Synthesis of 4

A solution of 2b (20 mg, 25.2 µmol) in nitrobenzene (2 mL) was kept stirring under reflux for 72 h. The reaction mixture was poured on the top of a silica gel chromatography column, and the nitrobenzene was eluted with hexane. Then, the reaction product was eluted using a gradient of hexane/dichloromethane. The desired compound 4 was obtained in 53% yield (10.6 mg) after crystallization from dichloromethane and methanol.

¹H NMR (500 MHz, CDCl₃): δ −1.67 (s, 2H, NH), 7.50 (d, 1H, J = 7.9 Hz), 7.73–7.79 (m, 10H), 7.84 (s, 1H), 7.88 (t, 1H, J = 7.9 Hz), 8.15–8.19 (m, 4H), 8.26–8.28 (m, 2H), 8.66 (d, 1H, J = 4.6 Hz, β-H), 8.72 (d, 1H, J = 4.6 Hz, β-H), 8.74 (d, 1H, J = 4.9 Hz, β-H), 8.82 (d, 1H, J = 4.9 Hz, β-H), 8.90 (d, 1H, J = 4.7 Hz, β-H), 9.62 (d, 1H J = 7.9 Hz), 9.75 (d, 1H, J = 4.7 Hz, β-H). ¹⁹F NMR (282 MHz, CDCl₃): δ −155.1 to −155.4 (m, 2F), −147.0 (t, J = 21.0 Hz, 1F), −139.1 to −139.4 (m, 2F). ¹³C NMR (125 MHz, CDCl₃): δ 101.9, 109.7, 113.6, 117.7, 117.8, 119.1, 122.9, 123.2, 124.0, 124.5, 124.6, 126.77, 126.81, 127.0, 127.3, 127.5, 127.6, 127.7, 127.9, 128.5, 128.7, 133.1, 134.2, 134.5, 134.8, 134.9, 135.0, 135.8, 137.0, 138.4, 138.5, 142.1, 142.2, 142.3, 144.0, 146.7, 147.1, 147.7, 150.2, 152.2, 154.6, 156.0. UV–Vis (CHCl₃): λ_max (log ε) 421 (5.26), 440 (5.19), 545 (4.15), 584 (4.41), 604 (4.21), 658 (4.18). MS (ESI): m/z 794.4 [M + H]⁺. HRMS-ESI(+): m/z calcd. for C₅₀H₂₉N₅F₅ [M + H]⁺ 794.2338; found 794.2349.

3.3.4. General Procedure for 5a,b and 6a,b

Porphyrins 2b or 4 (50 mg) were added to solutions of 3-sulfanylbenzoic acid or 4-sulfanylbenzoic acid (14.5 mg, 94 µmol, 1.5 eq.) in DMF (4 mL) and K₂CO₃ (36 mg, 260 µmol, 4 eq.). The mixtures were stirred at room temperature for 2 h or 8 h. Compounds 5a,b and 6a,b were obtained after precipitation from dichloromethane/methanol. 5a: 36.8 mg, 63% yield; 5b: 43.2 mg, 74% yield; 6a: 43.3 mg, 74% yield; 6b: 40.3 mg, 69% yield.

5a: ¹H NMR (300 MHz, DMSO-d₆, 40 °C): δ −2.81 (s, 2H, NH), 7.31 (d, 2H, J = 8.4 Hz), 7.62–7.70 (m, 3H), 7.78–7.85 (m, 10H), 7.92 (d, 2H, J = 8.4 Hz), 8.09–8.12 (m, 2H), 8.18–8.22 (m, 6H), 8.29 (s, 1H, NH), 8.62 (d, 1H, J = 4.9 Hz, β-H), 8.73–8.85 (m, 5H, β-H). ¹⁹F NMR (282 MHz, DMSO-d₆): −132.1 (d, 2F, J = 19.0 Hz), −149.0 (d, 2F, J = 19.0 Hz). ¹³C NMR (125 MHz, DMSO-d₆): δ 118.9, 119.5, 119.8, 120.8, 121.3, 126.6, 127.5, 127.6, 127.7, 128.6, 130.8, 133.7, 133.9, 134.5, 134.65, 134.72, 139.9, 141.3, 141.7, 141.9, 162.8, 167.5. UV–Vis (CHCl₃): λ_max (log ε) 424 (5.26), 523 (4.31), 554 (3.89), 594 (3.86), 651 (3.48). MS (ESI): m/z 930.4 [M + H]⁺. HRMS-ESI(+): m/z calcd. for C₅₇H₃₆F₄N₅O₂S [M + H]⁺ 930.2556; found 930.2559.

5b: ¹H NMR (300 MHz, DMSO-d₆, 40 °C): δ −2.82 (s, 2H, NH), 6.58 (s, 1H, NH), 7.40–7.45 (m, 2H), 7.56–7.66 (m, 4H), 7.78–7.86 (m, 11H), 8.06 (d, 2H, J = 6.7 Hz), 8.17–8.22 (m, 6H), 8.29 (s, 1H), 8.59 (d, 1H, J = 4.9 Hz, β-H), 8.73–8.85 (m, 5H, β-H). ¹⁹F NMR (282 MHz, DMSO-d₆): −132.4 (d, 2F, J = 19.4 Hz), −149.4 (d, 2F, J = 19.4 Hz). ¹³C NMR (125 MHz, DMSO-d₆): δ 118.9, 119.5, 120.5, 121.2, 127.5, 127.6, 128.3, 128.6, 128.9, 129.7, 130.8, 133.6, 134.5, 134.6, 134.7, 139.8, 141.3, 141.7, 141.9, 142.0, 143.2, 146.1, 146.7, 146.8, 162.4, 168.4, 186.6. UV–Vis (CHCl₃): λ_max (log ε) 415 (5.03), 424 (5.03), 522 (4.13), 550 (3.79), 594 (3.72), 650 (3.35). MS (ESI): m/z 930.4 [M + H]⁺. HRMS-ESI(+): m/z calcd. for C₅₇H₃₆O₂N₅F₄S [M + H]⁺ 930.2520; found 930.2530.

6a: ¹H NMR (300 MHz, CDCl₃): δ −1.63 (s, 2H, NH), 7.50 (d, 2H, J = 8.3 Hz), 7.73–7.75 (m, 4H), 7.78–7.85 (m, 8H), 8.15–8.20 (m, 8H), 8.25–8.29 (m, 2H), 8.66 (d, 1H, J = 4.6 Hz, β-H), 8.72 (d, 1H, J = 4.6 Hz, β-H), 8.78 (d, 1H, J = 5.0 Hz, β-H), 8.83 (d, 1H, J = 5.0 Hz, β-H), 8.90 (d, 1H, J = 4.8 Hz, β-H), 9.64 (dd, 1H, J = 7.9 and 1.0 Hz), 9.77 (d, 1H, J = 4.8 Hz, β-H). ¹⁹F NMR (282 MHz, CDCl₃): δ −137.8 (dd, J = 21.6 and 9.0 Hz, 2F), −125.5 (dd, J = 21.6 and 9.0 Hz, 2F). ¹³C NMR (125 MHz, CDCl₃): δ 102.1, 109.7, 113.8, 117.7, 117.8, 118.4, 122.8, 123.16, 123.24, 124.6, 126.77, 126.81, 127.0, 127.3, 127.6, 127.8, 127.9, 128.5, 128.66, 128.72, 129.1, 131.2, 131.3, 132.7, 133.2, 134.3, 134.5, 134.8, 134.9, 135.8, 138.8, 139.7, 142.0, 142.2, 142.3, 143.6, 162.1, 168.6. UV–Vis (CHCl₃): λ_max (log ε) 425 (4.46), 440 (4.48), 544 (3.52), 584 (3.74), 605 (3.59), 658 (3.59). MS (ESI): m/z 928.4 [M + H]⁺. HRMS-ESI(+): m/z calcd. for C₅₇H₃₄F₄N₅O₂S [M + H]⁺ 928.2364; found 928.2374.

6b: ¹H NMR (300 MHz, CDCl₃): δ −1.67 (s, 2H, NH), 7.52–7.56 (m, 2H), 7.72–7.78 (m, 9H), 7.85–7.87 (m, 2H), 7.92–7.96 (m, 2H), 8.14–8.19 (m, 6H), 8.24–8.27 (m, 3H), 8.33 (s, 1H), 8.65 (d, 1H, J = 4.6 Hz, β-H), 8.71 (d, 1H, J = 4.6 Hz, β-H), 8.74 (d, 1H, J = 5.1 Hz, β-H), 8.82 (d, 1H, J = 5.1 Hz, β-H), 8.88 (d, 1H, J = 4.7 Hz, β-H), 9.60 (d, 1H, J = 8.3 Hz), 9.74 (d, 1H, J = 4.7 Hz, β-H). ¹⁹F NMR (282 MHz, CDCl₃): δ −138.6 (dd, J = 18.4 and 8.6 Hz, 2F), −126.4 (dd, J = 18.4 and 8.6 Hz, 2F). ¹³C NMR (125 MHz, CDCl₃): δ 113.7, 113.9, 118.3, 123.2, 126.76, 126.80, 127.0, 127.3, 127.5, 127.6, 127.7, 127.8, 128.2, 128.48, 128.51, 128.6, 128.7, 129.1, 130.1, 132.7, 134.3, 134.8, 135.8, 136.2, 142.1. UV–Vis (CHCl₃): λ_max (log ε) 424 (4.43), 444 (4.50), 544 (3.46), 585 (3.71), 605 (3.54), 658 (3.52). MS (ESI): m/z 928.4 [M + H]⁺. HRMS-ESI(+): m/z calcd. for C₅₇H₃₄F₄N₅O₂S [M + H]⁺ 928.2364; found 928.2381.

3.3.5. Synthesis of 5c

A solution of 5b (10 mg, 10.8 µmol) and Zn(OAc)₂ (19.7 mg, 108 µmol, 10 equiv.) in a mixture of chloroform (3 mL) and methanol (1 mL) was heated at reflux for 15 min. After cooling, the solvents were removed under reduced pressure. The solid residue was dissolved in chloroform and the organic layer was washed with water and dried over Na₂SO₄. The Zn(II) porphyrin 5c was obtained in 98% yield (10.5 mg) after crystallization from dichloromethane and methanol.

¹⁹F NMR (282 MHz, CDCl₃): δ −150.0 (d, J = 20.9 Hz, 2F), −132.8 (d, J = 20.9 Hz, 2F). UV–Vis (CHCl₃): λ_max (log ε) 430 (4.74), 558 (3.73), 592 (3.22), 629 (2.87). MS (ESI): m/z 991.3 M⁺˙. HRMS-ESI(+): m/z calcd. for C₅₇H₃₄O₂N₅F₄SZn [M + H]⁺ 992.1656; found 992.1655.

3.4. Differential Pulse Voltammetry (DPV)

Differential pulse voltammetry (DPV) voltammograms were measured on a μAutolab Type III potentiostat/galvanostat (Metrohm Autolab B. V., Utrecht, The Netherlands), supervised by the GPES (General Purpose Electrochemical System) program version 4.9 (Eco-Chemie, B. V. Software, Utrecht, The Netherlands). The electrolytic cell, in which three electrodes can be placed together, had a volume capacity of 5 mL. A saturated calomel reference electrode (SCE, saturated KCl; Metrohm, Utrecht, The Netherlands) was used as the standard electrode. A glassy carbon electrode (f = 1.0 mm, BAS Inc., West Lafayette, IN, USA) was the chosen working electrode. The counter-electrode consisted of Pt wire. The working electrode was polished before use on 2–7/″ micro-cloth (Buehler) polishing pads using 1.0 and 0.3 mm alumina–water slurry (Buehler, Esslingen, Germany), then cleaned with water and ethanol. This desorption method was always repeated before carrying out electrochemical measurements. The electrolyte solution contained the porphyrin dye (0.5 mM) and the supporting electrolyte tetrabutylammonium hexafluorophosphate (0.1 M, TBAPF₆) dissolved in dry dichloromethane. The electrolyte solutions were degassed by purging N₂ before each measurement. The voltammograms were recorded at three different scan rates (10, 20, and 30 mV s⁻¹). The points of intersection between the tangent lines of the rising current and the baseline current were regarded as the onset values.

3.5. DSSC Fabrication and Photovoltaic Characterization

The conductive FTO-glass (TEC7, Greatcell Solar, Queanbeyan, Australia) employed for preparing transparent electrodes underwent meticulous cleaning with detergent followed by thorough rinsing with water and ethanol. For the anode preparation, the conductive glass plates (measuring 15 cm × 4 cm) were immersed in a TiCl₄/water solution (40 mM) at 70 °C for 30 min, then washed with water and ethanol before undergoing sintering at 500 °C for 30 min. This precise sequence is crucial for enhancing adherence of subsequent nanocrystalline layers and establishing a ‘blocking-layer’ to decrease charge recombination between electrons in the FTO and holes in the I⁻/I₃⁻ redox couple.

Subsequently, the TiO₂ nanocrystalline layers were deposited onto these pre-treated FTO plates through screen-printing with transparent titania paste (18NR-T, Greatcell Solar) using a polyester fiber frame with 43.80 mesh per cm². This dual-step process, involving coating and drying at 125 °C, was iterated twice. The TiO₂-coated plates underwent gradual heating up to 325 °C, followed by a temperature increase to 375 °C within 5 min, then further to 500 °C for sintering over 30 min, concluding with cooling to room temperature. A second treatment with the same TiCl₄/water solution (40 mM) was executed, following the aforementioned procedure, serving as an optimization step to augment surface roughness for improved dye adsorption, thereby positively influencing photocurrent generation under illumination.

Finally, a layer of reflective titania paste (WER2-O, Greatcell Solar) was screen-printed and sintered at 500 °C. This layer comprising anatase particles sized 150–200 nm functions as a ‘photon-trapping’ layer, further enhancing photocurrent. Each anode was precisely cut into rectangular pieces (measuring 2 cm × 1.5 cm) with a spot area of 0.196 cm² and a thickness of 15 μm. These prepared anodes were then immersed for 16 h in a 0.5 mM dye solution in dichloromethane at room temperature in darkness, followed by removal of excess dye via rinsing with the same solvent.

For investigating the impact of chenodeoxycholic acid (CDCA) addition on the photovoltaic properties, the adsorption process was repeated for each compound with a 0.5 mM dye solution prepared in dichloromethane with 50 mM CDCA.

Each counter-electrode comprised an FTO-glass plate (measuring 2 cm × 2 cm) with a 1.0 mm diameter hole drilled. These perforated substrates were meticulously cleaned with water and ethanol to eliminate residual glass powder and organic contaminants. Transparent Pt catalyst (PT1, Greatcell Solar) was deposited on the conductive face of the FTO-glass using a doctor blade technique. A strip of adhesive tape (3M Magic, Springfield, IL, USA) was applied to one edge of the glass plate to control film thickness and mask an electric contact strip. After uniform spreading of Pt paste on the substrate using a glass rod along the tape spacer, the adhesive tape strip was removed, and the glasses were heated at 550 °C for 30 min.

The photoanode and Pt counter-electrode were then assembled into a sandwich-type arrangement and sealed using a hot melt gasket made of Surlyn ionomer (Meltonix 1170-25, Solaronix SA, Aubonne, Switzerland) via a thermopress. The electrolyte, comprising the redox couple, I⁻/I₃⁻ (0.8 M LiI and 0.05 M I₂), dissolved in an acetonitrile/pentanenitrile (85:15, % v/v) mixture, was introduced into the cell via backfilling under vacuum through the hole drilled in the back of the cathode, which was subsequently sealed with adhesive tape. For each compound, a minimum of two cells were assembled under identical conditions, and the efficiencies were measured five times for each cell, resulting in a minimum of ten measurements per compound.

Current–Voltage curves were recorded using a digital Keithley SourceMeter multimeter (PVIV-1A) (Newport, M. T. Brandão, Porto, Portugal) connected to a PC. Simulated sunlight irradiation was provided by an Oriel solar simulator (Model LCS-100 Small Area Sol1A, 300 W Xe Arc lamp equipped with AM 1.5 filter, 100 mW/cm²) (Newport, M. T. Brandão). The thickness of the oxide film deposited on the photoanodes was measured using an Alpha-Step D600 Stylus Profiler (KLA-Tencor, Milpitas, CA, USA).

4. Conclusions

In this study, 2-amino-substituted porphyrin derivatives suitably functionalized with a carboxy group were prepared and evaluated as dyes in DSSC. The new compounds were successfully synthesized in two or three steps, in good overall yields, by nucleophilic aromatic substitution reactions with 3-sulfanyl- and 4-sulfanylbenzoic acid on porphyrins bearing a pentafluorophenyl group. Electrochemical studies revealed that the HOMO and LUMO energy levels are consistent with the requirements for effective electron transfer and dye regeneration, indicating that these dyes fit as suitable candidates for DSSC devices. However, the DSSC devices fabricated with these compounds revealed performances between 0.4% and 10% of the value obtained for N719. This low performance can be primarily attributed to their reduced open-circuit voltage and short-circuit current density values. One of the reasons for the reduced PCE values seems to arise from porphyrin aggregation phenomena, as demonstrated by the increased efficiencies in the presence of the disaggregating agent CDCA.