Mixed perovskite (MAPbI_3-xCl_x) solar cells using light-emitting conjugated polymer DMP end-capped MDMO-PPV as a hole transport material

1 Introduction

In the previous decade, investigation pursuits on hybrid organic–inorganic lead halide-based perovskite solar cells have greatly increased, creating devices with power conversion efficiencies exceeding 23 %, which is pentafold as high as the first cells revealed in 2009 (Kojima et al., 2009; Yang and You, 2017). These devices utilize a lead halide perovskite as the primary light-absorbing layer with excellent optoelectronic characteristics, including a superior absorption coefficient, extended charge carrier diffusion length, small exciton binding energy, and bandgap tunable (Chen et al., 2015; Kim et al., 2012; Leijtens et al., 2014; Noh et al., 2013; Stranks et al., 2013; Wehrenfennig et al., 2014; Zheng et al., 2015). Using simple solution-based processing techniques, they are feasible and can be synthesized at a low cost. Methylammonium lead iodide (CH₃NH₃PbI₃ or MAPI) is the utmost often utilized active layer among the numerous perovskite materials available and is relatively easy to produce (Leguy et al., 2015; Tvingstedt et al., 2014). Researchers have been looking into mixed cations Formamidinium (FA), Cesium (Cs), and other halides, particularly Cl- and Br-based mixed perovskites, because of their sensitivity to degradation under environmental conditions, particularly that of temperature and humidity (Miyasaka, 2021; Tan et al., 2018). Studies have generally found smooth surfaces and high surface coverage in perovskite (CH₃NH₃PbI₃) films made from lead acetate-based precursors (Huang et al., 2016; Zhang et al., 2015). Through single-step spin coating process, the rapid elimination of methylammonium acetate (CH₃NH₃Ac) as a byproduct was achieved and it resulted in the creation of smooth, pinhole-free perovskite films with miniscule grain sizes. Rapid crystallization causes huge defects and high trap densities in the perovskite layer, which manifest as recombination centers and traps at the grain borders, lowering the photovoltaic characteristics of such solar harvester devices (Qing et al., 2015; Shao et al., 2014).

High-efficiency solar cells based on perovskites mostly employ a sophisticated environmental control comparable to that of semiconductor fabrication labs. Despite decades of advancement, we now know that not all countries can make their own semiconductor fabrication labs due to the many commercial and industrial prerequisites. There would be much less strain on productivity on a large scale if the strict environmental requirements were reduced. Furthermore, if we mitigate the defects and improve the stability of Perovskite SC, solar cells, even with PCE of 10–14 % and durability of 7–10 years, could become commercially viable and cheap solar cells and could avoid oligopoly. Furthermore, the fabrication of integrated devices such as OLEDs, storage batteries (supercapacitors), sensors, and solar cells would become much simpler by using multipurpose materials for different purposes at different layers of the integrated devices.

There are many mitigation strategies to improve the stability of perovskite solar cells, such as adding interconnecting materials such as polymers, defect management, humidity repulse additive incorporation, and adding excessive PbI₂. One of the reinforcement pathways is the control, reduction, or management of the number of grain boundaries, i.e., increasing the grain size, which is one technique to limit the repercussions of trap states and reduce hysteresis. Larger grains have a greater charge carrier mobility, an extended charge carrier lifetime, and a greater open-circuit voltage, all of which contribute to improved device efficiencies (Bai et al., 2017; Yuan et al., 2016). Furthermore, grain boundaries allow moisture to permeate through the perovskite films, resulting in deterioration (Rong et al., 2017). Applying such techniques as engineered doping, casting by temperature, and annealing by solvent, larger grains with uniform distribution can be formed, making perovskite films more resistant to moisture deterioration. Xu et al used NH₄H₂PO₂ (ammonium hypophosphite) as an additive to enhance grain size up to the µm range. Addition of additive, such as Phenyl-C61-butyric acid methyl ester (PCBM), could enhance the working of solar cells (Xu et al., 2018). In addition to PCBM, substituting iodide ions in MAPI with chloride ions decreases surplus iodine in the perovskite films, lowering trap state density and suppressing centers of recombination. Furthermore, several studies have used small organic molecules as a passivation layer to reduce non-radiative recombination sites and surface defects (Liu et al., 2018). Hence, mitigation and repurposing (like using OLED material in the solar cell) strategies are important tools to improve the viability of perovskite solar cell technology.

Conjugated polymers (CPs) are chemically inert, stable, and hydrophobic and can therefore protect perovskite SCs against humidity. CP has been extensively studied for its ability to produce ASE (amplified spontaneous emission) at different wavelengths of visible spectra (Aljaafreh et al., 2021a, Aljaafreh et al., 2021b), OLEDs (AlSalhi et al., 2011), and sensors (Li and Liu, 2010). Thiophene based CPs are specially made for solar cell applications but have occasionally also been used in laser applications (AlSalhi et al., 2018). The best polymer for solar cell applications is PEDOT: PSS due to its high electron conductivity (up to 1.8x10-3 sv/cm⁻³). It is often used to replace the FTO and ITO layers in flexible solar cells; however, it has yet to or will never completely replace ITO and FTO without some performance compromises. At the same time, light-emitting CPs have also been utilized as hole conductive materials in perovskite solar cells. Masi et al fabricated a nanocomposite of CP MEH-PPV/MAPbI₃ perovskite that is dynamically used in solar cells. The polymer is utilized to control the inclusion of a small and uniform perovskite domain, leading to the fabrication of a smooth film using a single-step technique (Masi et al., 2015). Liren. Z et al. designed and synthesized various types of planar conjugated polymers that were primarily composed of bis(alkyloxy)benzene and/or bis(dithiophene) units. These polymeric materials act as hole transport layers in perovskite solar cells and can effectively act as a moisture barrier, preventing the perovskite from interacting directly with moisture (Zhang et al., 2019). The solid-state NMR (ss NMR) spectroscopy provides nanoscale insight into the blending nature of semi-ordered and amorphous regions; the remarkable insight provided by ssNMR about organic semiconductors and metal halide perovskites has been summarized by Seifrid et al. (2020), and Dahlman et al. (2021). Atomic-scale interaction between conjugated materials PM6 and Y6 in a bulk heterojunction (BHJ) and framework for understanding morphological differences and their impact on PCEs (Luginbuhl et al., 2022). The CP MDMO-PPV end-DMP produces laser at 560 nm and 605 nm (Prasad et al., 2020). It can be used to produce OLED and photoreactors. However, it is rarely used in solar cells and has not been used in combination with MAPbI_3-xCl_x as an active layer.

In this context, we report solar cells fabricated with red-emitting conjugated polymer MDMO-PPV as a hole transport layer, along with highly stable chlorine iodine mixed halide perovskite. The heterojunction solar cell (PEROVSKITE + CP)-based solar cell did not function well due to the low penetration of perovskite + CP into TiO₂ mesopores. When the thickness of the m-TiO₂ was adjusted to match the penetration of perovskite + CP, we were able to achieve better efficiency of our HJSC, and it remained stable for more than 4 weeks. To the best of our knowledge, this is the first report of a perovskite solar cell with the conjugated copolymer MDMO-PPV as a hole transport layer with a mesoporous TiO₂ architecture.

2 Materials and methods

The chemicals and solvents used in the experiment were acquired from commercial providers and used as received unless otherwise stated. Methylamine (33 wt%, absolute ethanol, titanium isopropoxide (TIP), hydroiodic acid (57 wt%, aqueous,), PbI₂, PbCl₂ were acquired from Sigma Aldrich, and diethyl ether VWR Chemicals, and were used in the preparation of methylammonium iodide (MAI). Fluoride-doped indium-doped tin oxide (FTO)-coated glass (15 O/cm²) was purchased from Osilla, Sheffield, UK. The CP poly[2-methoxy-5-(3,7-dimethyloctyloxy)-1,4-phenylene-vinylene] (MDMO-PPV) was procured from an American dye source,. The crystal structure is typical of mixed halide perovskites, as depicted in Fig. 1a, and also illustrates the molecular structure of MDMO-PPV. TiO₂ nanoparticles were synthesized using a hydrothermal process reported earlier (Oh et al., 2009).

Close

Fig. 1a

The crystal structure of unit drawing using VESTA software methylammonium mixed halide perovskite [35].

Export to PPT

CPs have poor solubility in polar solvents. However, MDMO-PPV has two oxygens with 2 acceptor sites each; hence, it has good solubility in moderately polar solvents such as chlorobenzene (CB) and chloroform (CHF). We used CHF to dissolve MDMO-PPV. When the perovskite precursors in DMF and MDMO-PPV in CHF were mixed at 70 °C, they were completely miscible.

The absorption spectra were recorded using a PerkinElmer Lambda – 950, and the fluorescence spectra were obtained using a PerkinElmer LS 55. A nanosecond Ti:Sapphire laser was used for the ablation of FTO to provide isolation. We used Field Emission Scanning Electron Microscope (FE-SEM) was utilize to achieve high resolution images of SCs samples. The current voltage characteristics of the solar cells are performed using a FYTRONIX SM-9000 solar simulator and QE-9000 quantum efficiency system (Fig. 1(b)).

Close

Fig. 1b

Molecular Structure of CP MDMO-PPV.

Export to PPT

2.1

2.1 Computational studies of CP

The donor's LUMO and HOMO were determined to be −1.43 eV and −4.52 eV, respectively, as presented in Fig. 2(a). Fig. 2 (b) depicts the UV–VIS spectrum of MDMO-PPV (n = 3) calculated using the TD-DFT B3LYP method with a basis set of 6–31G*. The absorption spectrum exhibits a single prominent peak at a wavelength of 441.2 nm with an oscillator strength of 2.4; the oscillator strengths at 360 and 380 nm are insignificant, with oscillator strength values of 0.0038 and 0.0213, respectively. Nevertheless, experimental evidence indicates that these three oscillators are singlets and can be attributed to S0-0, S0-1, and S0-2. For n = 3, the donor's energy gap was approximately 3.09 eV. As illustrated in Fig. 2 (c), this CP exhibits a transition to a negative cotton effect from a positive cotton effect in its simulated ECD spectrum. This is necessary for the MDMO-PPV to be polarizable. Fig. 2 (d) [inset of Fig 2b] shows the acceptor sites of CP.

Close

Fig. 2a

Simulated molecular orbitals structure (furrier densities), (2b) US-VIS and (2c) ECD structure of MDMO-PPV [n = 3] using TD-DFT B3LYP method with 6–31G basis set, and (2d) Acceptor sites of MDMO-PPV [inset of Fig 2b].

Export to PPT

Fig. 2(e) shows the estimated bandgap which is plotted against the inverse of repetitive monomer units, and the extrapolation up to n, which tends to infinity (n → ∞) molecules. The intercept of the 1/n graph was 2.31 eV, which was compared to the experimental value of 2.25 eV as determined by the intersection of the absorption and fluorescence spectra method (yellow square). The calculated and experimental energy gaps differ by 0.06 eV, indicating that the simulated HOMO – LUMO structures and locus of the energy levels are extremely reliable. This is critical for comprehending the charge separation and interaction at the interface between the MDMO-PPV and perovskite.

Close

Fig. 2e

The estimated bandgap using oligomer model simulation up to n = 4 plotted against the inverse of repetitive monomer units, and extrapolation up to n tends to infinity (n → ∞) molecular and compared with experimentally estimated using the intersection of absorption and fluorescence spectra method (yellow square).

Export to PPT

3 Results and discussions

Optical absorption and photoluminescence emission spectroscopy on MAPbI_3-xCl_x were performed under ambient temperatures to understand the changes in the electronic structure of the mixed halide anion version with regard to the photon harvesting material, as depicted in Fig. 3. Fig. 3(a) represents the absorption spectra of a thin film perovskite with a thickness of 450 nm coated on a quartz substrate. The spectrum showed a bandgap transition at approximately 814 nm (1. 52 eV), which is typical of MAPbI_3-xCl_x. It also showed a small dip at approximately 483 nm (2.57 eV). The MAPbI_3-xCl_x absorption band edge onset is centered at 760 nm at 24.85 °C, and we estimate the corresponding band gap to be 1.63 eV (using a bad edge method), which is consistent with previous results (Zhang et al., 2016). The normalized fluorescence (red line) shows the peak at 781 nm and a large overlap between the absorption spectrum and has a Stokes shift of 20 nm.

Close

Fig. 3a

Normalized Absorption and fluorescence spectra of thin-film perovskite (MAPbI_3-xCl_x) with a thickness of 450 nm coated on a quartz substrate.

Export to PPT

Close

Fig. 3b

Absorption spectra of perovskite (MAPbI_3-xCl_x), MDMO-PPV, and the blend of thin films of thickness 450 ± 30 nm coated on the quartz substrate.

Export to PPT

The absorption spectrum of the MDMO-PPV absorption peak is approximately 488 nm for a thickness of 400 nm, which coincides with the dip in the absorption spectrum of MAPbI_3-xCl_x. as presented in Fig. 3(b). The dip at 483 nm has been observed in many reports previously (Mehdi et al., 2020; Ralaiarisoa et al., 2019), and its origin could be attributed to grain boundary interface scattering and defect states. For typical nanomaterials, absorption increases towards shorter wavelengths. Similarly, the mix shows the perfect combination between MDMO-PPV and the MAPbI_3-xCl_x blend. The mix showed that absorption spectra of the blend are constructively superposed upon each other and add up to the sum of the individual components, indicating the nonappearance of strong interactions in the ground state.

The designed solar cells have both layered and heterojunction architectures. In general, the idea is to make a stable perovskite heterojunction layer that is resistant to environmental factors by using a conjugated polymer as both an encapsulation and hole transport material. One of the most significant prerequisites for optimum device performance in solar cells is an energy cascade enabled by the energy gradient between the light absorber, hole transport layer, and electron transport layer. The solar cell geometry consists of a homojunction or heterojunction layer of materials with energy levels given in a generalized energy diagram, as shown in Fig. 4. The energy levels of the various materials were taken from previously established reports (Heo and Im, 2016; Veenstra et al., 2004; Yuan et al., 2018; Zhang et al., 2018). The mixed perovskite has a valence band and conduction bands at −5.43 eV and −3.9 eV, respectively. The CO MDMO-PPV has HOMO and LUMO energies of −5.3 and −3, respectively. The large energy gradient between perovskite and MDMO-PPV largely stops the migration of electrons and enables charge separation at the interfacial junction. Additionally, LiTSO could be used as a bridge between HJ and Au. This is done because the mixing of perovskite and MDMO-PPV could reduce the bandgap and increase the barrier for the hole to jump from the perovskite/MDMO-PPV hybrid structure. We did not vary the thickness of MDMO-PPV, LiTSO, and gold (Au) to limit the permutation and combinations of the devices in this comparison study. The main objective was to compare the performance of devices fabricated under low humidity and easily achievable lab environments. The contact layer thickness was maintained at 85 ± 15 nm, and gold had an energy level of −4.5 eV.

Close

Fig. 4

Energy level of perovskite (MAPbI_3-xCl_x), MDMO-PPV, and sprio-OMeTAD layers.

Export to PPT

The synthesized TiO₂ had high nanopore porosity, and the distribution showed that most of the particles were less than 100 nm in diameter, as shown in Fig. 5. The thickness of the m-TiO₂ layer varied from 400 nm to 800 nm. The inset of Fig. 5 shows the blend layer at the same magnification. The perovskite tends to form submicron and nanosized domains that are connected with MDMO-PPV polymers. The FESEM image also indicates that the blend film is smooth and continuous even at the very small region (nano level) or at the molecular level.

Close

Fig. 5

FESEM image of TiO₂ and inset is the image of the blend.

Export to PPT

The device structure of a regular perovskite device Fig. 6 (a) (D1) contained the following layers: glass/FTO/m-TiO2/perovskite/LiTSO/Au. The perovskite layer was 460 nm and infiltrated into the m-TiO₂ layer. Apart from that, perovskite also formed a 100–90 nm thick layer on top of the perovskite layer LiTSO with a thickness of 120 nm and a contact layer of gold with a thickness of 80 nm. D1 showed a PCE of 7.68 % V_oc = 0.83, and J_sc was 14 mA. It is low compared to high-performance solar cells based on perovskite, but we have not employed sophisticated environmental conditions during fabrication. m-TiO₂-based perovskite solar cells are simple to fabricate; however, they are very sensitive to outside environments. The introduction of new processes, such as the addition of 1,8-diiodooctane (DIO), solvent vapor annealing (SVA) (Sun et al., 2016), and PANI, improves the performance of solar cells (Zheng et al., 2019). Similarly, the addition of CP could improve the stability of the SC, and hence, we employed CP MDMO-PPV.

Close

Fig. 6

(a-d) Device structure of solar cells and (e and f) FESEM of solar cells D2 and D3.

Export to PPT

The D2 device introduces MDMO-PPV as a hole conducting layer instead of LiTSO. The device of D2 is Glass/FTO/m-TiO2/perovskite/MDMO-PPV/Au, as shown in Fig. 6 (b). The performance is low again, with a PCE of 5.32 %, a V_oc of 0.836 V, with a lower open-circuit current J_sc of 13.25 mA and FF of 0.478. Fig. 6 (e) (D2) shows the solar cell structure and FESEM image. The TiO₂ layer was 415 ± 20 nm thick with a completely penetrated perovskite layer. The perovskite layer is topped with an MDMO-PPV layer with a thickness of 170 nm. D2 has good V_oc but a poor fill factor and thus performance; however, it has very high stability due to the presence of MDMO-PPV.

Next, we made HJSC with a blend layer (D3). The D3 was a set of solar cells will the same structure but with different thicknesses of the m-TiO₂. This type of devices has a structure with layers Glass/FTO/m-TiO₂/(Perovskite + MDMO-PPV)/Au, as shown in Fig. 6(c), which depicts one of the solar cells in D3 set. The penetration of blend inside m-TiO₂ was 250 ± 10 nm irrespective of the thickness of m-TiO₂, beyond m-TiO₂ thickness greater than 260 nm. The maximum efficiency was observed for m-TiO₂ thickness 300 nm. The thickness of the m-TiO₂ varied from 500 nm to 300 nm; the D3 devices have a PCE of 5.68 % to 7.43 % for higher and lower thicknesses, respectively. The FSEM image in Fig. 6(e) shows that the blended materials partially penetrated inside the m-TiO₂ by approximately 246 nm. The wide variation in PCE is obtained by changing the m-TiO₂ layer and by optimizing the thickness of MDMO-PPV. The reduction in the PCE could be attributed to two factors: (a) the perovskite + MDMO-PPV irregular contact made by the HJ and gold contact layer and (b) the blend not penetrating completely into m-TiO₂. The above issues could lead to increased recombination and lower charge separation. The other parameters of the HJSC (D3) type were a very low V_oc = 0.63–0.85 V, J_sc of 9.6 mA to 11.3 mA, and FF ranging from 0.5 to 0.68. The maximum efficiency in the D3 set of solar cells is given in the Fig. 7 (pink – star line), that with V_oc = 0.85 V, J_sc of 9.67 mA, FF of 0.68 and PCE of 7.438 %.

Close

Fig. 7

The J–V characteristics for the solar cell’s samples D1 – D4 types. (JV of D3 is maximum efficiency in the subset of D3 type solar cell).

Export to PPT

We aim to improve the efficiency of perovskite/MDMO-PPV solar cells using the following method. We first optimized the thickness of m-TiO₂ to 350 nm, and thus, the blended materials diffuse almost completely onto the m-TiO₂ layer. Thus, we introduce a layer of LiTSO with a PCE of 7.74 %, FF of 0.7 and V_oc = 0.89 V and J_sc of 12.07 mA (D4). The FSEM image shows a LiSTO base solar cell with layers of 78 ± 15 nm, 350 ± 10 nm, and 105 ± 15 nm of compact TiO₂, m-TiO₂ (MDMO-PPV/perovskite), and LiSTO, as shown in Fig. 6 (f).

Fig. 7 shows the JV curves of solar cell types D1, D2, and D4. The design of D3 involves a change in TiO₂ thickness. The depicted JV curve of D3 belongs to the maximum efficiency solar cell with an optimized thickness of m-TiO₂ (300 nm).

The solar cells were randomly numbered and shuffled, and tested at another laboratory in a single-blind study. Additionally, the incident photon to charge carrier efficiency (IPCE) was measured at another laboratory. Fig. 8 shows the IPCE of devices D1 to D4. The J_sc calculated using IPCE was 13.89 mA for D1, 13.29 mA for D2, and 9.72 to 11.4 mA for D3 (group of different HJSC with different m-TiO₂ thicknesses), and for the last solar cell D4, the values were 11.52 mA.

Close

Fig. 8

IPCE spectra of the solar cells for all D samples.

Export to PPT

To check the stability of the designed solar cells, the devices were connected to a 1 kΩ resistor as a load and exposed to sunlight and ambient atmosphere for a long time. The PEC of the solar cell was measured at weekly intervals. D1 dropped 22 % of its initial performance, D2 lost 13 %, and D3 showed 11 % lower performance after 4 weeks. The improved stability could be attributed to the molecular-level interaction between MDMO-PPV and perovskite. The plausible expiations are derived from the earlier insight provided by G. N. M. Reddy group (Krishna et al., 2021; Raval et al., 2022). The reduction in performance of D3-type solar cells with large m-TiO₂ could be due to the exposure of the microdomain of perovskite, and the uninfused m-TiO₂ layer could attract H₂O. Finally, the performance of D4-type HJSC with the LiSTO layer dropped by only 9 %. The solar cell devices of each type (D1 to D4) were fabricated and characterized at two different labs. The performance characteristics of different types of solar cells are tabulated in Table 1.

4 Conclusions

In this work, we designed a series of solar cells with mixed halide methylammonium perovskite active layers and a CP MDMO-PPV as an HTM in layered and heterojunction architecture. The layered solar cells (FTO/TiO₂/perovskite/MDMO-PPV/Au) have a high short circuit current density (J_sc) of 13.295 mA but a very poor fill factor (FF) of 0.47 and PCE of 5.32 %. The heterojunction solar cell (HJSC) made with perovskite, and MDMO-PPV mix showed a PCE of 5.68 % for the same thickness of m-TiO₂. The FESEM showed that in HJSC, the perovskite and CP were mixed well, but they did not penetrate deeply into the TiO₂ mesoporous layer, which hinders electron transportation and thus causes efficiency loss. Upon optimization of the thickness of m-TiO₂, the HJSC reached a PCE of 7.438 %. The added layer of LiTSO on top of the blend layer improved the PCE to 7.74 %, which is comparable to the performance of the standard perovskite solar cell fabricated under the same environmental and operational conditions. The usage of a conjugated polymer as a hole transport layer minimizes reliance on separate components while also increasing the stability of the solar cells. It could also pave the way for a simple, minimalistic solution for integrated, flexible power electronics that include perovskite/CP solar cells, an organic light-emitting diode, and a graphene supercapacitor.

CRediT authorship contribution statement

Saradh Prasad: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Data curation, Writing – original draft, Visualization. Mamduh J. Aljaafreh: Software, Validation, Formal analysis, Investigation, Data curation, Writing – original draft, Visualization. Abeer Alshammari: Formal analysis, Investigation, Visualization. Mona A.S. Almutairi: . Jagannathan Madhavan: Methodology, Writing – review & editing, Supervision. Mohamad S. AlSalhi: Conceptualization, Methodology, Validation, Resources, Writing – review & editing, Supervision, Project administration.