Paulina
Ratajczyk
a,
Andrzej
Katrusiak
*a,
Krzysztof A.
Bogdanowicz
b,
Wojciech
Przybył
b,
Piotr
Krysiak
b,
Anna
Kwak
b and
Agnieszka
Iwan
*b
aFaculty of Chemistry, Adam Mickiewicz University, Uniwersytetu Poznańskiego 8, 61-614 Poznań, Poland. E-mail: katran@amu.edu.pl
bMilitary Institute of Engineer Technology, Obornicka 136, 50-961 Wroclaw, Poland. E-mail: iwan@witi.wroc.pl
First published on 14th February 2022
Mechanical and thermal strains in organic photovoltaic layers deposited on poly(ethylene terephthalate) (PET) and glass substrates and the resulting effect on the bandgap energy have been determined for the organic charge-transfer polymer PTB7. Its strong bathochromic shift of the absorption spectra under pressure also depends on the layer preparation method. The effects of different solvents and PTB7 concentrations on agglomerate formation were also studied. A perfect linear dependence of the concentration vs. absorbance was observed up to 1.81 × 10−7 M for chlorobenzene or o-dichlorobenzene solution. The thermal image monitoring of the current response at an increasing potential showed a linear current-to-voltage tendency, indicating that the conductive behaviour of the polymer depends on the substrate used in PET/ITO/PTB7/Ag/ITO/PET and glass/ITO/PTB7/Ag/ITO/glass devices. The PTB7 layer on the PET/ITO substrate displays superior mechanical properties suitable for flexible photovoltaic panels exposed to strain and deformations.
At present, organic/polymer solar cells are being widely investigated in a few main directions: (i) the application of new polymers, copolymers or small compounds in the active layer of PV devices,2–7 (ii) the application of graphene, graphene oxide, their chemical modifications, carbon nanotubes, fullerenes and nonfullerenes, and their derivatives either as an active layer or as an interlayer in PV devices,8–14 (iii) the application of new polymers or graphene as flexible substrates in PV devices15–17 and (iv) the application of new polymers as hole transporting layers (HTLs), instead of poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS), in PV devices.18,19
To construct the active layer in polymer photovoltaic devices, poly(3-hexylthiophene) (P3HT) and poly[N-9′′-heptadecanyl-2,7-carbazole-alt-5,5-(4′,7′-di-2-thienyl-2′,1′,3′-benzothiadiazole)] (PCDTBT) have been widely investigated as donor materials. Theoretical studies from 2008 predicted “realistically achievable” power conversion efficiency (PCE) values of ca. 15% for fully optimized donor–acceptor bulk heterojunction (BHJ) organic solar cells.20 The basic questions to be addressed centre on the relationships between the chemical and physical compositions of interlayer materials and their energetic and electrical properties on nanometre length scales, often in environments with the complexity of the working solar cell platform.
Most recently, various strategies have been used for optimizing the organic/polymeric solar cells, e.g.: (i) in order to manipulate the molecular interactions of the binary blend, environmentally friendly solvents were used as the third component of the active layer;21 (ii) efficient deposition techniques for perovskites and charge transport layers such as doctor-blade coating, slot-die coating, screen printing, and spray deposition strategies contributed to the significant development of large-area perovskite solar modules;22 (iii) small-molecule donor–acceptor mixtures were used as stable device architectures without intermediate electrode layers;23 (iv) a broad family of organoboron small molecular donors/acceptors and organoboron polymer donors/acceptors became an important class of organic photovoltaic materials;5 (v) the stability of organic solar cells was significantly increased via material design, device engineering of the active layers, employing an inverted geometry, optimizing the buffer layers, and using stable electrodes and encapsulation materials;7 (vi) several printing techniques improved the production of large-area flexible transparent conductive films (TCFs);24 (vii) novel nanomaterials were designed and developed, including carbon nanotubes, graphene and their composites, which are now the main materials for flexible and transparent thin-film electrodes;17,25,26 and (viii) radioactive sources were used to generate photonic light in scintillators as converters of ionizing radiation to electricity in photovoltaic cells.27
Poly[[4,8-bis[(2-ethylhexyl)oxy]benzo[1,2-b:4,5-b′]dithio-phene-2,6-diyl][3-fluoro-2-[(2-ethylhexyl)oxy]carbonyl]thieno [3,4-b]thiophenediyl] (PTB7) (Fig. 1) presently gives some of the highest reported efficiencies (PCE = 8.92%) for PTB7:PC71BM BHJ solar cells in the presence of 3.0 v% 1,8-diiodooctane (DIO) and 1 wt% polystyrene (PS).28 For the inverted architecture indium tin oxide (ITO)/poly[(9,9-bis(3′-(N,N-dimethylamino)propyl)-2,7-fluorene)-alt-2,7-(9,9-dioctylfluorene)] (PFN)/PTB7:PC71BM/MoO3/Al, the power conversion efficiency was found to be 9.15% (for standard BHJ solar cell ITO/PEDOT:PSS/PTB7:PC71BM/PFN/Ca, PCE = 8.24%).29 However, this efficiency is still far from the theoretical value of PCE, which is equal to 15%. One of the proposed ways to achieve or even exceed the theoretical value of PCE is the construction of tandem perovskite cells with ITO/SnO2/CsPbI2Br/P3HT/MoO3/Ag/PFN-Br/PTB7-Th:IEICO-4F/MoO3/Ag architectures containing poly([2,6′-4,8-di(5-ethylhexylthienyl)benzo[1,2-b;3,3-b]dithiophene]{3-fluoro-2[(2-ethylhexyl)carbonyl]thieno[3,4-b]thiophenediyl}) (PTB7-Th) as a component of the hole transporting material (HTM) layer with PCE = 17.24%.30
Murray et al. described the ITO/HTL/PTB7:PC71BM/LiF/Al architecture for polymer solar cells, where graphene oxide (GO) or PEDOT:PSS was used as the HTL.31 Devices with GO exhibited a PCE equal to 7.5%, indicating a 5-fold enhancement in the thermal aging lifetime and a 20-fold enhancement in the humid ambient lifetime versus analogous PEDOT:PSS-based devices.
It should be emphasized that polymer solar cells based only on PTB7 are utilized in the inverted device with the ITO/ZnO/PTB7/MoO3/Ag architecture, and the PCE value is only 0.14%, indicating almost no photovoltaic performance.32 For solar cells with an active layer based on PTB7:PC71BM or PTB7:PC71BM:DIO, the PCE values are 4.74% and 7.41%, respectively.
In addition to modifying the solar cell architecture, scientists are studying the selected electrochemical parameters of PTB7:PC71BM in detail to understand the differences in behaviour of this mixture compared to P3HT:PCBM. For example, Park et al. discovered that, at the PTB7:PC71BM interface, the interface dipole and band bending were not observed due to their identical charge neutrality levels. On the other hand, a large interfacial dipole was observed at the P3HT:PC61BM interface. It was found that the photovoltaic energy gaps (EPVG) were 1.10 eV and 0.90 eV for PTB7:PC71BM and P3HT:PC61BM, respectively, inducing a larger open-circuit voltage in PTB7:PC71BM than that in P3HT:PC61BM.33 Jin et al. investigated the transient THz conductivities of PTB7 and P3HT blended with PC71BM using ultrafast THz spectroscopy.34 They demonstrated that the carrier–carrier interactions at higher density play a significant role in decreasing the charge carrier mobility in PTB7:PC71BM. Transient absorption spectroscopy of PTB7:PC71BM blends containing a broad range of acceptor contents (0.01–50% by weight)35 showed that with the increase in the polaron signal PC71BM concentrations of more than 10%, most polarons are generated within 200 fs, while for lower acceptor concentrations, the polaron signal increases gradually over ∼1 ps, with most polarons generated after 200 fs.
Detailed studies on the selected properties of PTB7 led to new solutions for improving the charge mobility and morphology of PTB7. For example, Savikhin et al. studied the effect of high-temperature annealing on the properties of PTB7 and PTB7:fullerene blends and revealed that annealing at a moderate temperature (260 °C) improves the PTB7 morphology and optoelectronic properties.36 Although high-temperature annealing (290 °C) also improves the morphology, it caused poorer optoelectronic properties as an effect of aliphatic side chain modification in PTB7, increased electronic disorder and decreased mobility. It was found that solar irradiation modifies the chemical structure of PTB7 and affects the device photostability.
Kettle et al. studied the chemical degradation of PTB7 by X-ray photoelectron spectroscopy (XPS)37 and demonstrated that PTB7 seems to be essentially unstable when illuminated in air. XPS studies were used to confirm that the rapid photodegradation process occurs due to changes in the chemical structure of PTB7. The transient magnetic field effects of photoexcitation in PTB7 were studied by Huynh et al.38
Among the considerable number of techniques employed for studying solar cell properties in recent years, high-pressure techniques have been used to reveal the impressive potential of photovoltaics based on perovskites.39–46 Pressure is a fundamental thermodynamic variable that can induce a variety of structural changes, such as phase transitions,47 Jahn–Teller interplay with the lattice strain,48 and amorphization.49,50 Generally, high-pressure studies are ideal for understanding the structure–property relationship,51 as they provide fundamental insight into the structural, optoelectronic, and magnetic properties of materials and thus can guide the design of new photovoltaic materials with desired properties. An important feature of photovoltaic materials, generally used in the form of thin and thick layers, is the effect of strain on the absorption and current-generation efficiency. Strain can be generated in the layer-deposition process due to the layer–substrate interactions, as governed by the different multilayer thermal expansions in photovoltaic panels, the bending strain in flexible devices and other effects in the production process and specific applications in different environments. Furthermore, the operation of solar cells occurs with significant temperature changes that can amplify stress, which may completely alter the structure and optoelectronic properties of the materials, affecting their performance. Nowadays, strain engineering is increasingly used to tailor optoelectronic functionalities.52,53 To our knowledge, no pressure effects have been described for the absorption edge of organic photovoltaic materials. Therefore, we have undertaken this study on the photovoltaic polymer PTB7. Apart from the photovoltaic performance, we also continue to be interested in finding new materials for pressure sensors. Our present study includes the effects of solvents and additives on the morphological, optical, mechanical and electrical behaviours of the created PTB7 layer. We have shown that the PTB7-based panels endow flexible and adaptive multilayer systems offering sufficient performance of the bulk and interfaces for their operation under the strain required in flexible, wearable and exposed products under extreme conditions. Such a performance can result not only from high stretchability, but also from autonomous self-healing capabilities.
To establish the possible agglomerate formation depending on the type of solvent, we investigated a wide range of concentrations (Fig. 3), from 5.3 × 10−9 to 2.56 × 10−7 M (6.67 × 10−4 to 3.22 × 10−2 mg mL−1), within the limit of 2.1 absorbance units. Following the changes in spectra, the ratios of the integrated intensities of the bands at 621 and 680 nm were compared. We found that the normalized signal for 680 nm reflected a change in the proportion from 1.21:1/1.19:1 to nearly 1.10:1/1.06:1 for chlorobenzene/o-chlorobenzene. According to Zhang et al., this direction of change is due to the reduced polymer ordered aggregation in solution that occurs upon increasing the concentration.32 In our study, disordered aggregation was observed for concentrations above 1.81 × 10−7 M (2.28 × 10−2 mg mL−1), which was evident for the solution in o-DCB. Thus, an almost perfect linear dependence of the concentration vs. absorbance was observed up to 1.81 × 10−7 M (2.28 × 10−2 mg mL−1) for both solvents, with only a small discrepancy for the CB solution. This shows that the dipole-moment change from 1.54 D (CB) to 2.14 D (o-DCB) influences the in-solution aggregation ordering of the PTB7 polymer, resulting in a stronger disorder at higher concentrations. This relationship indicates that the dipole moment of solvent molecules has a similar effect on aggregation as the addition of DIO to the CB solution.32
(i) The PTB7 layer deposited by uniaxial compression: after placing a small amount of the sample on the diamond culet, it was strongly pressed from above by a thick glass plate until it fused into a uniform layer (Fig. 4d).
Fig. 4 High-pressure absorbance spectra of PTB7 measured for (a) a solution of PTB7 in dichloromethane (the sample was liquid up to 1.25 GPa, when dichloromethane freezes at 296 K),61,62 (b) a PTB7 layer on one culet, formed upon evaporation of dichloromethane, (c) the same layer as in (b), but on two diamond culets, and (d) a layer pressed on one culet. |
(ii) The PTB7 layer deposited by solvent evaporation: a drop of the PTB7 solution in dichloromethane, chlorobenzene or 1,2-dichlorobenzene was placed on the culet, and the layer was formed after solvent evaporation (Fig. 4b, c and 5).
(iii) The PTB7 layer deposited by the solution under high pressures: the solution of PTB7 in dichloromethane was loaded into the DAC chamber, and the layer was formed as a precipitate after increasing the pressure (Fig. 4a).
When measuring the absorbance spectrum of PTB7 dissolved in dichloromethane above the freezing pressure of dichloromethane (above 1.25 GPa), a clear jump-like increase was observed in the absorbance (Fig. 4a). This jump is associated with the precipitation of small grains of PTB7 extruded from dichloromethane upon crystallization. The crystallization of the solvent in the DAC chamber is clearly visible through the microscope, but this process significantly increases the scattering of the probing-beam light on the tiny grains of PTB7 dispersed in the sample. We also observed this effect for other freezing solutions.
The UV-vis spectra measured in the DAC for PTB7 layers prepared using different methods (i–iii) display similar strong bathochromic pressure effects (Fig. 6). The absorption edges up to 5 GPa are shifted by approximately 25 nm GPa−1 in the range 700–870 nm. It is a characteristic of all spectra series that, in the initial increased pressure ranges up to approximately 1 GPa, the absorption edge shift is linearly dependent on pressure. Then, above 1 GPa, dλ/dp gradually decreases, and above 4 GPa, dλ/dp is several times smaller compared to that in the range from 0.1 MPa to 1 GPa.
Fig. 6 The absorption edge as a function of pressure, depending on the PTB7 preparation method. As described in the plot, the different colours refer to the data in Fig. 4 and 5. |
This finding correlates with the tendency of PTB7 to form aggregates, which are known to cause red absorption.54,55 In a photoluminescence study, Hedley et al. revealed that small nanoaggregates of polymer chains can enhance the red absorption range above 600 nm, which they connected to microaggregation.56
Scanning electron microscopy (SEM) images were recorded to inspect the homogeneity of the prepared layers: (i) PTB7 powder compressed into the culet (Fig. S2a in the ESI†) and (ii) a layer formed by evaporation of dichloromethane (Fig. S2b in the ESI†). The SEM images allowed us to determine that the thickness of the layer formed by pressing the powdered product into the diamond was approximately 15 μm. The PTB7 layers formed via solvent evaporation were much thinner. The SEM micrographs show that the surface of the PTB7 layers formed by powder compression is homogenous and uniform, whereas the layers formed by dichloromethane evaporation consist of small patches with gaps and cracks that cause various patterns of defects in the deposited film.
X-ray powder diffraction (XRD) measurements (Fig. S3 in the ESI†) were used to confirm that both samples (i) prepared by pressing PTB7 and (ii) obtained by evaporating the solvent were amorphous.
The heat generated by the electric current in a through-plane thin layer was registered through thermal imaging. The experiment includes thermographic images that were each registered at the end of 3 minute intervals, when a stable state was reached for a constant current and temperature. To assess the electrical and thermal behaviours of PTB7 compounds, the change in the thermal signature was registered while applying an external potential. Two devices were constructed with the following configurations: (a) glass/ITO/PTB7/Ag/ITO/glass and (b) PET/ITO/PTB7/Ag/ITO/PET. These configurations differ based on the glass and PET substrates. In both devices, the thin PTB7 layers display nearly linear current-to-voltage dependences (Fig. 7), which confirms that the polymer is an electric conductor. The electric resistance of glass-supported PTB7 was 178.5 Ω and that of PET-supported PTB7 was 119.1 Ω. The difference is related to the pixeled glass support giving a longer conductive pathway than in the case of the PET/ITO support. However, for PET-supported PTB7, the lower resistance resulted in a larger current, causing a higher temperature, which in turn led to a drastic decline in the current above 6 V/136 °C. Nevertheless, high temperatures observed in the thermal images suggest the progressive degradation of the device, although at a lower pace.
Fig. 7 Electric current (black) and temperature (red) as a function of applied voltage for devices composed of (a) glass/ITO/PTB7/Ag/ITO/glass and (b) PET/ITO/PTB7/Ag/ITO/PET. |
The thermal response, culminating at approximately 53 °C for glass devices and at 136 °C for PET-supported PTB7 devices, shows an approximate logarithmic dependence on the voltage. For the flexible support, a higher temperature affects the organic layers and reduces their performance. According to the thermal images (Fig. 8), the edges of the device are most affected and overheated compared to the other regions. Defects located along the edges are most likely the cause of overheating and subsequent degradation of the PTB7 layers. In the present form, the PET-supported device performs well up to approximately 5 V.
Fig. 9 (a) Stress at break and (b) elongation at break for PET/ITO/PTB7 single layer (samples no. 1–4) and double layer (no. 5–9) devices. |
The correlation between the stretching force and elongation (Fig. 10) drastically differs for the single- and double-coated PET/ITO samples in the multilayer substrate devices. For all samples, the tests revealed both elastic and plastic behaviours, with the transition reaching approximately 2 mm elongation (approximately 0.8% of the initial length) at a force of approx. 135 N (79 MPa), consistent with the elastic properties of the PET/ITO support. After reaching the transition, the PTB7 single-coated samples became fully plastic and elongated until they broke with a negligible increase in the applied force. The plastic deformations of the samples with double-layer PTB7 clearly differed from those with the single layer. The force–elongation dependence for double-layer PTB7 deviated from the typical plastic behaviour due to the effect of the approximately 100 N force enhancement that occurred before reaching the break point. This shows that the double layer of PTB7 adds an elastic component to the typical plastic behaviour, increasing the value of the applied force and the value of the stress at the breaking point. The PTB7 double layer reduced the breaking-point elongation to an average value of 29.3 mm (samples 5–9) compared to an average value of 37 mm for the PTB7 single-layer samples 2 and 4. This change is connected to the higher values of the stretching force and the breaking point for the PTB7 double-layer samples (on average 250 N) compared to the PTB7 single-layer samples (150 N). However, the PTB7 single-layer samples 1 and 3 broke at the lowest elongation values of 19 and 25 mm, respectively. In the case of the 5th sample, a slight deviation from the normal behaviour was attributed to the imperfection of the PET/ITO support probably caused by the cutting process.
Fig. 10 Stretching force values as a function of elongation for PET/ITO/PTB7 single layer (samples 1–4) and double layer (5–9) devices. |
The basic transmission UV-vis spectra of PTB7 in solution were acquired using an A360 UV–Vis spectrophotometer (AOR Instruments, Shanghai, China) with an interval of 0.2 nm and medium scan speed in the range of 200 nm to 800 nm.
Scanning electron microscopy (SEM) images for the two layers, (i) PTB7 powder compressed into the culet and (ii) formed by evaporation of dichloromethane, were investigated using a Zeiss EVO 40 scanning electron microscope.
The PTB7 sample and the sample after evaporation of 1,2-dichlororobenzene were characterized under ambient conditions by X-ray powder diffraction (XRD). A Bruker AXS D8 diffractometer equipped with a Johansson monochromator selecting the Cu Kα1 line was used (λ = 1.54060 Å). Silicon single-crystal plates were used as the sample cuvette to reduce the background.
The thermal behaviour was observed upon the application of a potential using a thermographic camera (VIGOcam v50, VIGO System S. A., Ożarów Mazowiecki, Poland) and a multichannel potentiostat–galvanostat (PGStat Autolab M101, Metrohm, Barendrecht, Nederland) as described in detail elsewhere.60 The samples were prepared through the spin casting method with the PTB7 solution (10 mg mL−1 in chlorobenzene) on ITO-coated PET or the glass support (900 rpm, 30 s). Additionally, to improve the electric contact, a layer of silver paste was used to bind the electrode separated from the sample.
The mechanical properties (tensile strain) were investigated using an Instron 33R4469 (Instron, Norwood, MA, USA) testing machine with a load cell of 5 kN, and the results were registered using the Bluehill 3.0 software (Instron, Norwood, MA, USA). The PTB7 solution (10 mg mL−1 in chlorobenzene) was spun cast (900 rpm, 30 s) on a 5 × 5 cm2 ITO-coated PET support. For mechanical tests, 5 × 1 cm2 plates were used.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/d1ma01066b |
This journal is © The Royal Society of Chemistry 2022 |