1 Introduction

In 2015, the United Nations (UN) set the 2030 agenda for sustainable development to achieve the 17 Sustainable Development Goals (SDGs). With regard to SDG 7 “Affordable and Clean Energy”, it is necessary to search for new, clean, sustainable, and modern energy sources in a way for balancing economic and social needs [1]. It has been estimated that the solar energy delivered to the Earth is 3 × 1024 Joules per year, representing about 10,000 times the world's energy intake [2]. Hence, solar energy can be considered a renewable, clean and inexhaustible energy source. Therefore, a low-cost method for converting solar light into electrical energy is recently gaining popularity [3, 4].

Dye-sensitized solar cells (DSSCs), the third-generation of photovoltaic cells, base their functionality on the photo-excitation of dyes, in which electron transfer occurs miming photosynthesis processes naturally carried out by plants [5]. Common dyes include synthetic dyes [6, 7] and natural dyes [2]. According to the International Energy Agency (IEA), for improving the device efficiency and decreasing their production cost [3], the application of natural resources has attracted much attention. Natural resources application aims to achieve sustainability and reliability, supporting at the same time an eco-friendly environment [8].

Natural pigments have been considered a promising alternative for DSSCs production due to their low cost, large quantity, and environmentally friendly characteristics [9,10,11,12,13]. Furthermore, natural dyes are usually non-toxic and fully biodegradable [14].

Currently, the natural pigments used for the device production, such as carotenoids, betalains, anthocyanins, and chlorophyll, are mainly of plant origin, obtained from fruits, flowers, leaves, seeds, roots and wood [15,16,17,18,19,20].

The implementation of natural dyes for achieving solar energy conversion has been largely investigated, resulting in a cheap and simple approach based on their physical and chemical processing, avoiding at the same time any hazardous waste by-products [21,22,23,24,25,26,27].

Natural pigments can also be produced by microorganisms mainly as secondary metabolites allowing them to colonize different ecological niches to ensure their survival [28]. Pigments confer important adaptive roles by functioning primarily as photo-protective agents for the cell [29]. Moreover, it is known that some of them can be involved in the cellular respiration process, particularly within the electron transport chain [30], or even act as intermediary mediating agents in redox reactions when cells interact with their surrounding environment [31]. These unique physical and chemical microbial pigments properties can be extrapolated from biological systems to emerging technologies aimed at electricity generation. In this sense, microbial pigments have been managed to be the main character in developing microbial fuel cells and more recently, in dye-sensitized solar cells [28]. As the development of photovoltaic devices progresses, microbial pigments have been shown to have several substantial advantages over plant-derived pigments, such as the capability to be produced on a large scale through rapid fermentation processes that can be carried out throughout the year, less water use, along with their excellent stability and solubility [3, 32].

The microorganism’s ability to fast-growing in inexpensive media, the opportunity to modify the pigment production according to the medium and the fermentation conditions [33], the easy downstream processing, and their constant cultivation independent of the seasonal variations are all advantages of microbes over plants as a source of pigment production [34]. Besides, there is an increasing concern that large-scale production of plants intended for the extraction of their pigments could cause severe problems to the ecosystem, including deforestation and infringement on the diversity of local species; therefore, the practice is not sustainable [35]. Thus, in recent years, it has been proposed that microbial pigments could be an excellent alternative for these devices development [35].

To date, a small group of microorganisms has been investigated as an alternative source for natural pigments [25, 36] for DSSCs application. Among these microorganisms are included bacteria, such as Chryseobacterium spp. [37], Hymenobacter spp., [35, 37] Streptomyces fildesensis [2], Escherichia coli [38, 39], Rhodopseudomonas spp. [40], Serratia marcescens [28]; microalgae such as Porphyridium cruentum [41], Chlorella spp. [42, 43], Haematococcus pluvialis [44], Chlamydomonas reinhardtii [45], and Scenedesmus obliquus [46], and fungi Monascus spp. [47, 48].

Hence, given the several advantages of microorganisms over plants, it is recommended to continue screening new molecules that have not been previously studied as sensitizers and could potentially generate high photocurrent efficiencies becoming a new sustainable option for meeting future energy requirements.

In this sense, it is important to encourage the development of this promising technology through the generation of alternatives for the construction of these devices in order to expand the existing range of materials of microbial origin that can be used as photosensitizers.

Talaromyces spp., previously classified as Penicillium spp. [49], are filamentous fungi which secrete large amounts of Monascus-like red pigments, without producing any known mycotoxins.

Pigments produced by this fungus have shown a good stability in a wide pH range maintaining the red color mainly at neutral conditions [50,51,52,53]. This behavior makes these materials particularly promising for different industrial applications, such as textile [54], cosmetics [55], and food industries [56]. On the other hand, in our knowledge, no previous studies have been carried out on the application of these pigments as photosensitizer. This work represents the first study focused on investigating Monascus-like azaphilone red pigments produced by Talaromyces atroroseus GH2 for evaluating their suitability for dye-sensitized solar cells implementation.

2 Materials and methods

2.1 Microorganism and culture medium composition

Talaromyces atroroseus GH2 (Department of Food Science and Technology strain repository, Autonomous University of Coahuila, Saltillo, Mexico) was used for pigment production through immobilized biomass fermentation. Fungal spores were stored at − 20 °C in skimmed milk and glycerol solution.

For spore preparation, the T. atroroseus GH2 strain was grown for 120 h at 30 °C in 250 mL Erlenmeyer flasks containing 50 mL of Potato Dextrose Agar medium.

After the incubation time, a sterile aqueous solution of Tween 20 (0.01% v/v) was used for obtaining a spores suspension by washing the cultures. T. atroroseus GH2 spore suspension (1 × 105 spores/mL) was used as inoculum for 250 mL Erlenmeyer flasks containing 50 mL of Potato Dextrose Broth medium. The flasks were incubated at 30 °C for 72 h in an orbital shaker (Inova 94, New Brunswick Scientific, Edison, NJ, USA) at 200 rpm. The resulting mycelial suspension was used as inoculum for pigment production by immobilized biomass fermentation.

According to Ruiz-Sánchez et al. [57], Czapek–Dox-modified (CDM) medium was used for pigment production. CDM consisted of (g/L) D-xylose (15.0), NaNO3 (3.0), MgSO4·7H2O (0.5), FeSO4·7H2O (0.1), K2HPO4 (1.0), KCl (1.0), and ethanol (20.0). Initial pH was adjusted to 5 using HCl solution before sterilization through 0.22 μm sterile membranes (Millipore, Billerica, MA, USA).

2.2 Fermentation and pigment extraction

As previously reported by Ruiz-Sánchez et al. [57], pigment production was carried out using immobilized biomass in 125 mL Erlenmeyer flasks containing nylon sponges (Delicate Duty) support with a volume of 1 cm3.

Before starting the fermentation process, Erlenmeyer flasks containing the supports were sterilized. Later, 25 mL of sterile CDM medium was added aseptically. Flasks were inoculated with a mycelial suspension (10% v/v) of T. atroroseus GH2 and incubated at 30 °C in an orbital shaker at 200 rpm. After 168 h of incubation, the medium containing the extracellular pigment was recovered and samples were lyophilized. Extracellular pigment recovery was performed according to Morales-Oyervides et al. [58]. Briefly, samples were centrifuged at 6236 g for 20 min at 4 °C (Sorvall, Primo R Biofuge Centrifugation Thermo, USA) and the extracellular pigmented extracts were filtered (Cellulose filter, 0.45 μm. Millipore, USA), frozen (− 20 °C), lyophilized and stored at 4 °C [58].

2.3 Pigment extract characterization

Pigment characterization was carried out in accordance with the methodology reported by Venkatachalam et al. with some modifications [56]. The lyophilized powder (100 mg) was dissolved in methanol (1 mL) and subsequently filtered through PTFE syringe filter (0.45 μm), followed by a secondary filtration (0.20 μm PTFE syringe filter). The resulting crude filtrates were stored at 4 °C in a vial pending HPLC analysis.

HPLC-DAD-ESI-MS analyses were conducted utilizing a Shimadzu Prominence UFLC XR system, and the acquired data were subjected to analysis using the Shimadzu Lab Solution Software (Shimadzu, Kyoto, Japan). Chromatographic separation was performed on a Kinetix C18 column (100 × 2.1 mm; 1.7 μm, Phenomenex, Torrance, CA, USA). Eluent A comprised water, while eluent B consisted of acetonitrile; both phases were acidified with formic acid (1%). The chromatographic conditions included a flow rate of 0.2 mL/min, a sample injection volume of 1 μL, and an oven temperature set to 30 °C.

ESI-MS was conducted under the same parameters as the above-mentioned work [56]. Identification of azaphilones was accomplished by comparing retention times, DAD, and MS spectra obtained from the analyzed samples with those documented in the literature.

2.4 Cell’s fabrication procedure and working principles

A typical DSSC assembles four elements: working electrode (WE), sensitizer (dye), redox-mediator (electrolyte), and counter-electrode (CE).

For the working electrode preparation, a conductive fluorine-doped tin oxide (FTO) of 2.5 × 2.5 cm (SnO2: F; 2.2 mm TCO (transparent conductive oxide)) [10, 35] glass was cleaned with water, soap, acetone, and ethanol in an ultrasonic bath. UV-ozone treatment was applied for about 18 min to the glasses for removing organic matter and improving the hydrophilicity of substrates and TiO2 deposition [59]. Successively, the electrodes were immersed for 30 min at 70 °C in a solution of TiCl4 (844 μL in 100 mL of distilled H2O. Since the reaction is exothermic, the solution was obtained in an immersed beaker in an ice bath). This method called “blocking layer” increases the resistance of the interaction between the FTO substrate and the TiO2 film and blocks the charge recombination process between the electrons present on the semiconductor conduction band and the I2 in the electrolyte.

The substrates, cleaned with water and ethanol, were dried in the muffle at 80 °C and then the photo-anode was prepared by depositing a film of the TiO2 paste (Dyesol 18NRT, particle size of about 10–15 nm) on the FTO-conducting glass using the screen-printing technique [60] which allows us to obtain TiO2 films with precisely defined areas and constant thickness. Factors, such as the size of the TiO2 nanoparticles, the crystalline purity of the same, and the subsequent sintering steps, are extremely important for the efficiency of the cell [61]. The deposition on substrates to realize WE was made by an electric flat screen Printer, AT-45PA by ATMA using modules and frames with suitable geometry.

The device photo-anode thickness was measured using a DektakXT profilometer (Bruker) equipped with a diamond-tipped stylus (radius of 2 μm) and selecting a vertical scan range of 524 μm with an 8.0 nm resolution, a programmed scan length of 6000 μm, and a stylus force of 1 mg.

For this experiment, a double deposition was carried out in order to obtain a mesoscopic oxide film of around 9 µm thick and transparent with an active area of approximately 0.181 cm2. The obtained photo-anodes were placed in a box on absorbent paper soaked with ethanol and then dried on a heating plate at 125 °C for 6 min.

The anode was finally thermally treated in an air atmosphere using a temperature gradient program with five levels at 125 °C (6 min), 325 °C (5 min), 375 °C (5 min), 450 °C (15 min), and 500 °C (15 min). After cooling, it was immersed in a 40 mM TiCl4 solution in deionized water at 70 °C for 30 min and then annealed at 500 °C for 30 min in order to create the upper blocking layer.

Finally, the cooled anodes were soaked in the dye solution (pigment in ethanol at different pHs) and the dye sensitization was carried out at room temperature for one night; excess dye was removed by rinsing with ethanol and dried in an oven. The pH of the electrolyte solutions was measured by a Waterproof pHTestr 10 (Eutech instruments—Oakton) calibrated up to three points (pH 3, 7, and 10) using certified single-use pouch NIST buffer set standards, dipped in the electrolyte solutions.

Two different home-made electrolytes were used:

  • J8* (LiI 0.1 M, I2 0.05 M, MPII 0.6 M, TBP 0.5 M in AN:VN 70:30) [62];

  • AS8* (LiI 0.8 M, I2 0.05 M in AN:VN 85:15) [10] where MPII is 1-methyl-3-propyl imidazolium iodide, TBP is 4-tert-butyl-pyridine, AN is acetonitrile and VN is valeronitrile.

For the counter-electrode preparation, a layer of platinum paste (Dyesol PT-1) thinner than 1 µm was printed on TCO glass and then sintered at 500 °C for 30 min. The two electrodes were sealed using a 25-µm-thick Surlyn frame and a thermal press for the device realization (Fig. 1). Drops of electrolyte solution were introduced into the device by the vacuum backfilling technique through a shaped hole drilled (1.5 mm diameter) in the back of the counter-electrode. Then, the hole was sealed with 3 M tape; for long-term stability test Surlyn and a glass coverslip were used as sealing materials.

Fig. 1
figure 1

Images of lyophilized extract sample (a), the solution in ethanol at different pH values (b), the photo-anodes obtained after immersion of the TiO2 substrate in the dye (c) the assembled devices (d), respectively

Full size image

Device’s behavior was investigated in relation to the electrolyte and to pH which was fixed by adding some drops of a HCl (1.0 M) or NaOH (1.0 M) solution to the electrolyte in order to reach the required pH. The pH values evaluated were 3, 7, and 10.

2.5 Devices performance characterization

UV–Vis measurements were performed by a Perkin Elmer Lambda 25 Uv–Vis spectrophotometer, in the region between 190 and 1200 nm.

The devices performance depends on a series of parameters, such as power conversion efficiency (PCE, %), short-circuit current (JSC, mA/cm 2), open-circuit voltage (VOC, V), maximum power output [Pmax], and fill factor (FF), which are influenced by various factors such as the reactions that are triggered by the elements of DSSCs [63]. A digital Keithley 236 multimeter connected to a PC and controlled by a homemade program was used to obtain the current–voltage (I–V) curves for the constructed devices. Simulated sunlight irradiation was provided by a LOT-Oriel solar simulator (Model LS0100-1000, 300 W Xe-Arc lamp, powered by LSN251 power supply equipped with AM 1.5 filter, 100 mW/cm2). Incident irradiance was measured with a Si-based pyranometer (Pyris 03).

The parameter used to evaluate the efficiency (PCE) of a PV device is defined as follows:

$$ {\text{PCE}} = P_{\max } /P_{{\text{in}}} $$
$$ {\text{PCE}} = {\text{ FF }}J_{{\text{sc}}} V_{{\text{oc}}} /P_{{\text{in}}} $$

where Pin is the radiation power incident on the cell, Jsc is short-circuit current density at zero voltage, Voc is the open-circuit voltage at zero current density and FF is the fill factor [64].

Electrochemical Impedance Spectroscopy (EIS) provides an explanation of the electrical and electrochemical processes occurring in the device.

By analyzing these impedance characteristics, it can be deduced the efficiency of electron transfer, the rate of recombination, and the overall performance of the DSSC. This information is crucial for optimizing cell design and material selection [61].

The DSSC cells were connected to an Autolab Potentiostat/Galvanostat (Metrohm) equipped with a frequency response analyzer (FRA). EIS experiments were conducted in the frequency range 100 kHz–100 mHz at 0.2 V, with a potential pulse amplitude of 0.01 Vrms.

3 Results and discussion

The study was carried out on DSSCs based on a natural dye, Monascus-like azaphilone pigments produced by T. atroroseus GH2, characterized mainly by single carboxylic groups [65].

3.1 Ultraviolet–visible spectrophotometry and pigment characterization

In the context of DSSC application, the selection of a dye with absorption capabilities in the visible spectrum is crucial [66]. The optical spectrum analysis of the pigments extract, diluted in ethanol at different pH values (Fig. 2), revealed two distinct absorbance peaks at 430 nm and 500 nm. These peaks are attributed to the presence of various colored compounds within the pigmented extract.

Fig. 2
figure 2

Absorbance spectra of solutions of pigment in ethanol at different pH values

Full size image

Talaromyces strains are recognized producers of Monascus-like azaphilone colorants, including yellow (at 400 nm), orange (at 450 nm), and red (at 500 nm) [67]. The resulting colorant mixture varies depending on the conditions of the process and the composition of the media [58].

A total of 22 different azaphilone-type pigments were detected in this study in Talaromyces atroroseus and the identified compounds, both intra-cellular (IC) and extracellular-cellular (EC) pigments, are presented in Table S1 (Supplementary material) which includes tentatively identified compounds, their chromatographic retention time, DAD and MS data. Characterization results indicated the presence of exclusively orange and red pigments. Figure 3 shows the chemical structure PP-O, which was the most abundant compound among the identified ones, which bear a carboxylic acid group—known to be important for the DSSCs functioning. Within the red pigments, specific molecules were identified, namely Rubropunctamine derivatives (N-methionine-rubropunctamine and N-tryptophan-rubropunctamine) and Monascorubramine derivatives (N-threonine-monascorubramine and N-valine-monascorubramine). Meanwhile, the orange pigments encompass monascorubrin and its derivatives (PP-O, PPO-Serine, and PPO-Glutamic acid). Notably, PPO emerged as the predominant pigment.

Fig. 3
figure 3

PPO Chemical Structure

Full size image

It is noteworthy that Talaromyces metabolic pathways closely parallel to those of Monascus. Initial synthesis of orange pigments precedes their potential reduction to yellow pigments, and amination of the orange pigments leads to the formation of red pigments [68].

The synergy derived from a mixture of pigments may enhance light absorption, extend the spectral range, and improve overall efficiency, making it a promising avenue for advancing DSSC technology. The successful utilization of a pigment mixture has been previously documented in the application of DSSCs [69], reinforcing the practical viability of employing a combination of pigments for enhanced efficiency. Previous study reported similar results but using simultaneously two microbial pigments as sensitizers [69]. According to previous literature, there is a shift of about 20 nm in the studied pH range which does not strongly affect the color stability [67, 68].

3.2 DSSCs photoelectrochemical properties

Sets of samples were made using two types of electrolytes, J8* and As8* at different pH values in order to investigate the photovoltaic behavior of the cell in various conditions. In Table 1, the values of open-circuit voltage (VOC), short-circuit photocurrent density (JSC), fill factor (FF), and power conversion efficiency (PCE) are reported. These parameters illustrate an estimate of the photovoltaic performances of the developed device.

Table 1 Values of open-circuit voltage (VOC), short-circuit photocurrent density (JSC), fill factor (FF) and power conversion efficiency (PCE)
Full size table

The results obtained from the IV measurements showed a trend of Voc and Jsc that depends not only on the pH value but also on the type of electrolyte used. Average and standard deviation values are calculated over a set of 4 cells.

3.3 Additives and pH effect on devices performance

The electrolyte J8* comprises TBP and MPII in addition to the iodine/ iodide couple that serves for electron transfer to the dye.

MPII is often added as an additive to enhance ionic conductivity in dye-sensitized solar cells [70]. Their chemical structures are reported in Table 2.

Table 2 TBP and MPII chemical structures used to prepare the tested electrolytes
Full size table

The inclusion of specific pyridine-based compounds like TBP in the electrolyte of dye-sensitized solar cells has been reported able to enhance their performance by increasing the open-circuit voltage (Voc) [71]. This enhancement is primarily due to the negative shift of the TiO2 conduction band edge induced by the addition of TBP. Additionally, TBP's adsorption onto TiO2 results in the suppression of dark current, as it covers the free space on nanoparticles [72, 73]. Dark current is generated by the reduction of triiodide at the photo-anode, causing an electron flow in the opposite direction of the photocurrent. Regarding the pH effect on devices performances, the pH-dependent conduction band edge is described by the following equation (Eq. 1):

$$ {\text{Ecb }}\left( {{\text{pH}}} \right) \, = {\text{ Ecb }}\left( {{\text{pH }} = \, 0} \right) \, - \, 0.0{\text{59 pH}} $$
(1)

The value of Ecb (pH = 0) for anatase TiO2 (vs SCE) is − 0.4 V. In the presence of tert-butyl-pyridine, the conduction band (CB) is estimated to be 0.94 V (V vs SCE) [74].

Furthermore, TBP in the I3−/I− electrolyte solution reacts with I3−, reducing the iodine concentration by an order of magnitude (Eq. 2) [71]. This reduction has been experimentally demonstrated in devices sensitized by ruthenium compounds [62].

$$ {\text{I}}_{3} - \, + {\text{ TBP}} \to {\text{ TBP}} - {\text{I}}_{2} + {\text{ I}} - $$
(2)

The interaction between TBP in the electrolyte mediator couple (i.e., I3−/I− couple) is not significant when natural dyes are used as sensitizers and the addition in the device does not affect positively its performance. In fact, in our experiments, it has been observed that although an increase in the VOC can be detected (Fig. 4a), on the other hand, the addition of TBP (0.5 M) decreased the JSC value (Fig. 4b), as also reported in the literature data. However, it is necessary to specify that this behavior depends heavily on the chemical structure of the dye and by the interaction that therefore occurs between the film and the dye.

Fig. 4
figure 4

Photovoltaic parameters comparison for samples with two different electrolytes used J8* e As8*; a Voc, b Jsc, c PCE at different pH values

Full size image

From the literature, it is known that using pyridine (Py) (0.25 M) as an electrolyte additive, the Voc increases but the short-circuit current (Jsc) is significantly reduced, resulting in a lower PCE (Fig. 4c). This decrease in performance is attributed to the basic nature of Py and TBP, as well as the pH sensitivity of natural pigments like anthocyanins, betalains, carotenoids, chlorophylls, and others. For example, in some studies [62], it is reported that the addition of basic compounds to anthocyanin dye anodes leads to partial desorption of dyes from TiO2, cation changes, and a red shift in absorption spectra, decreasing the molar extinction coefficient (ε) [74,75,76]. So when Py (0.25 M) is present, the devices pH reaches approximately 9.3, causing a red shift in the cyanin maximum wavelength peak and reducing the molar extinction coefficient by over 50%. Consequently, optical absorption cross section and Jsc are reduced. Even for betalains, previous investigations with TBP use resulted in a modest gain in photo-voltage and fill factor but a dramatic decrease in photocurrent [62]. This effect is likely due to the reduced activity of indicaxanthin under basic conditions. This phenomenon was not reported by betalain pigments for dye-sensitized solar cells [77] because authors used purified betanin and removed betaxanthin. Betanin is more robust and less prone to decompose following pH changes [78]. The influence of additive TBP in apocarotenoid-based DSSCs has been also investigated, and it was demonstrated that TBP-free electrolytes are the best for this kind of sensitizers [10]. The behaviors reported as examples can be applied to every natural dyes pH-sensitive. For these reasons, AS8*, which is basic free electrolyte, is proposed. This electrolyte which does not contain TBP in its formulation shows better performances when applied in natural dyes sensitized solar cells, than the conventional one, which contradicts the established idea that adding certain chemical species (e.g., TBP, etc.) to the redox electrolyte can improve the DSSCs photovoltaic performances by tuning the semiconductor/electrolyte interface and preventing unwanted recombination reactions [10]. This can be explained by considering the chemical nature of carotenoids with their pH-sensitive carboxylic and carboxylate groups. The addition of basic compounds, such as TBP (kb ≈ 9.5·10 − 9), to carotenoids-based anodes causes hydrolysis of the ester group and partial conversion of bixin to norbixin, resulting in a decrease in the photo-electrochemical cell performances. As observed for carotenoids, Monascus-like azaphilone pigments are characterized by functional groups based on OH or COOH moieties responsible for the anchoring to the semiconductor interface [20]. In neutral conditions, the injection capacity of the dye is greater than basic pH because the HOMO LUMO levels are approaching, and therefore it improves the receptive capacity of the TiO2 as showed in Scheme 1. Unfortunately, in acid condition, the stability is moderate, the coloration of the solution is less intense, and the adhesion of the dye to the surface of the mesoporous film is less effective [67], and this compromises the injection of the electron into conduction band of semiconductor.

Scheme 1.
scheme 1

HOMO (TiO2, grey) LUMO (dye, blue) levels at different pH conditions

Full size image

For all these reasons, DSSCs based on T. atroroseus GH2 pigments extract at pH 7 using As8* as electrolyte, showed the best IV characteristics (Fig. 5). As8* application as electrolyte allowed us to obtain a Jsc of 0.69 mA/cm2, a Voc of 0.27 V, a FF of 0.60, and a PCE of 0.11%. Comparing the results of the overall conversion efficiency with previous study using microbial dyes, it can be pointed out that in this study, PCE reached out the same percentage previously obtained when bacterioruberin was used as dye for DSSCs [69], whereas the implementation of Talaromyces extract allowed us to obtain an improved efficiency in comparison to the one obtained for DSSCs based on bacteriorhodopsin [69].

Fig. 5
figure 5

IV curves of DSSCs based on a Monascus-like azaphilone pigment produced by T. atroroseus GH2 as sensitizer using As8* as electrolyte at different pH values

Full size image

Electrochemical impedance spectroscopy (EIS) has been used to better understand the electrical and electrochemical mechanisms occurring in the device. A first qualitative analysis of Nyquist spectra allows us to distinguish between series or ohmic and charge transfer resistance. The high-frequency intercept of the semicircles on the x-axis of the Nyquist plots is associated with the ohmic resistance (more precisely reported as series resistance, Rs). The difference between the low-frequency intercept in the Nyquist plot and Rs is assumed as the charge transfer resistance (Rct).

A typical Nyquist plot for a DSSC consists of one or more semicircles, each representing different processes in the cell. The diameter of the semicircle correlates with the resistance of a specific process. The first semicircle at high frequencies usually represents the charge transfer resistance at the counter-electrode. The semicircle in the intermediate frequency range often corresponds to the charge transfer at the dye-semiconductor interface and recombination processes. At low frequencies, a Warburg impedance, characterized by a sloped line, indicates diffusion processes within the DSSC [61].

The total resistance corresponding to the low-frequency intercept on the abscissa in the Nyquist plot (Rs + Rct) corresponds to the differential resistance of the IV curves.

In Fig. 6, the real part of the impedance (Z') is plotted on the x-axis and the imaginary part (Z'') on the y-axis. In the inset, the Bode diagram of the impedance phase is shown. A larger Z' value indicates higher resistance, and a larger Z'' value indicates higher capacitance or reactance in the cell.

Fig. 6
figure 6

Full EIS spectra (a), magnification at high frequencies (b) and Bode diagram of phase (c) of DSSCs based on a Monascus-like azaphilone pigment produced by T. atroroseus GH2 as sensitizer using As8* as electrolyte at different pH values

Full size image

The series resistance (Rs) of the three DSSC cells analyzed was very similar, with a value with a value between 1.4 and 1.7 Ω cm2. However, the charge transfer resistance (Ret) varied significantly between the cells, with values of 870 Ω cm2, 1105 Ω cm2, and 10,800 Ω cm2 for the cells with electrolytes at pH 3, 7, and 10, respectively. These results suggest that the Rct is a key factor influencing the performance of DSSC cells. The lower Rct of the cell with the pH 3 electrolyte can be attributed to the lower resistance at the electrode/electrolyte interface. The FTO/TiO2 surface carries charges that are highly influenced by the pH. At different pH values, the surface may exhibit variations in the density and distribution of surface charges. Under acidic conditions (pH 3), the surface may have a higher concentration of positive charges, influencing the overall electrostatic interactions with the dye molecules. Under pH 3 conditions, the increased positive charge on the surface could enhance the electrostatic attraction between the negatively charged dye molecules and the positively charged surface. This stronger attraction facilitates a more efficient and rapid dye adsorption process. The lower Rct observed in the cell with pH 3 electrolyte can be directly linked to the facilitated charge transfer at the electrode/electrolyte interface. The increased adsorption of dye molecules on the FTO/TiO2 surface, driven by favorable electrostatic interactions at pH 3, promotes a closer alignment of energy levels between the dye and the semiconductor. This alignment reduces the impedance to charge transfer, resulting in a lower charge transfer resistance (Rct). The favorable conditions at pH 3 contribute to improved electron injection from the dye to the TiO2 semiconductor, as well as more efficient electron collection at the electrode. This is reflected in the lower Rct value, indicating enhanced charge transport kinetics [79].

The equivalent electrical circuit, under illumination, includes an ohmic resistance (series resistance) connected in series with two components each consisting of a parallel between a resistance and a constant phase element (CPE). The series resistance reflects the ohmic phenomena whereas the R//CPE components are associated with the electrode–electrolyte interfacial properties.

In pH7 and pH10 cases, there are two overlapping semicircles in the Nyquist plots (Fig. 6) corresponding to two different relaxation times in the Bode plots (inset in the Fig. 6), represented by two overlapping peaks in the curves related to the phase shift. On the other hand, in the case of pH3, there is an additional relaxation time at low frequency. As a result of these considerations, the parameters of the equivalent electrical circuit were derived and are shown in Table 3.

Table 3 Equivalent electrical circuit parameters of DSSCs based on a Monascus-like azaphilone pigment produced by T. atroroseus GH2 as sensitizer using As8* as electrolyte at different pH values
Full size table

The photovoltaic performance parameters of DSSCs at different pH levels show a correlation with the impedance data previously discussed.

  • At pH 3: The lower charge transfer resistance (Rct) observed in the impedance data suggests more efficient electron transfer at the electrode/electrolyte interface. However, the photovoltaic performance is still low (PCE of 0.04%), which could be due to the lower open-circuit voltage (Voc) and fill factor (FF). The higher Jsc compared to pH 10 indicates better electron generation but overall low efficiency, potentially due to recombination processes not captured by the impedance alone.

  • At pH 7: The Jsc is the highest among the three, and the fill factor and PCE are improved (0.11%), which can be correlated with the moderate Rct. This suggests a balance between electron generation and transfer rates, and reduced recombination, leading to better overall cell performance.

  • At pH 10: Despite the highest Voc and FF, the Jsc is significantly lower, and the PCE drops to 0.06%. This aligns with the highest Rct seen in the impedance data, indicating substantial resistance to electron transfer, which severely limits the current generation capability of the cell, thus reducing the overall power conversion efficiency.

In conclusion, while the impedance data indicated how the electron transfer efficiency varies with pH, the complete photovoltaic performance also depends on other factors, such as the generation of electrons (Jsc) and the potential at which the cell operates (Voc). The Rct is a crucial factor but not the only determinant of the cell's performance (Fig. 7). Lower Rct helps but must be coupled with high Jsc and Voc to achieve higher PCE [61].

Fig. 7
figure 7

Equivalent electrical circuits of DSSCs based on a Monascus-like azaphilone pigment produced by T. atroroseus GH2 as sensitizer using As8* as electrolyte at a pH 3 and b at pH 7 and pH 10

Full size image

3.4 Device stability and bifaciality

The stability of a device over time depends on several factors, but the main problem is the electrolyte which evaporates easily and therefore it is important to properly seal the device to avoid these phenomena. For this reason, IV measurements were carried out over time to verify the stability and it has been observed that the efficiency values remain unchanged for over six months.

To further underline the importance of the use of this material as a dye for photovoltaic devices, measurements of the transmittance have been carried out on the photoanode made by depositing the TiO2 film on the TCO and sensitizing it with the dye (solution obtained at pH 7) as reported in Fig. 8. The result of this analysis is a transmittance value of about 80%, aspect of crucial importance because it allows the application of the device as a bifacial solar cell that collects light from both sides.

Fig. 8
figure 8

Transmittance spectra of the working electrode with and without dye showing a percentage of ± 80% suitable for use as a double-sided device

Full size image

However, the results obtained in this preliminary study are according to what has been pointed out for pigments extracted from Cortinarius spp. [80]. This behavior is due to the molecular structure, in particular, to the distribution of electrons into the heterocyclic structure and to the position of oxygen atoms and the sites where the photo-anode attaches. This configuration involves changing in the energy-level diagram of the ground state (HOMO) and of the excited state (LUMO) and consequently a reduction of the generated current flow [81]. Despite the microbial pigment extract investigated are poor in functional sites [68], in comparison with Cortinarius spp., the photovoltaic characteristics obtained were promising.

4 Conclusions

This study explored the potential of utilizing extracellular microbial pigment extracts obtained by T. atroroseus GH2 cultivation as suitable dyes for DSSCs implementation. Besides the already investigated industrial application of these pigments, the present research allows us to increase their area of application also on the energy field, being used for solar cell application.

The extract mainly comprising azaphilones pigments, where the most abundant compound was represented by PP-O, exhibited good stability at pH 7 using electrolyte TBP-free. Combining these conditions with the implementation of natural dyes allows us to obtain an eco-friendly and sustainable technology, considering also the easy downstream processing. Consequently, T. atroroseus GH2 can be considered a promising candidate for producing novel natural pigments for based solar modules for electric purposes, increasing the range of materials of microbial origin suitable as photosensitizers.

Further studies are required for setting up the best T. atroroseus GH2 cultivation conditions in order to direct the fermentation process towards specific colored dye production. Additionally, assessing the shelf life of the implemented devices is crucial for evaluating their long-term performance and practical applicability. Moreover, exploring different sealing procedures may enhance the photo-conversion efficiency values of the devices, contributing to improved overall performance and competitiveness of microbial pigment-based DSSCs.

The co-sensitization of TiO2 film using two or more bacterial dyes usually shows due to complementary absorption spectra, improved photovoltaic performance. In the future, devices will be tested combined dyes belonging to different molecules class to enhance the power conversion efficiency. In particular, Talaromyces atroroseus GH2 and Arthrobacter bussei extracts, mainly represented by azaphilone and 50 C carotenoid pigments respectively, will be evaluated together as suitable photosensitizer.