2.1.

Mesoporous Structured Solar Cell

Mesoporous structured solar cell (MSSC) is a third-generation device with low fabrication cost. ${\mathrm{TiO}}_2$ is the most common mesoporous layer in them. Though ${\mathrm{TiO}}_2$ nanoparticles are the most commonly used in the mesoporous layer, success reports are available while employing ${\mathrm{TiO}}_2$ nanosheets,⁶³ nanorods, and nanofibers. MSSC consists of a single-phase triple-conducting oxide (TCO), blocking layer, mesoporous ${\mathrm{TiO}}_2$ or ${\mathrm{Al}}_2{\mathrm{O}}_3$ scaffold, perovskite absorber, hole transport layer (HTL), and a metal electrode in its construction (Fig. 8).⁶⁴ Usually, MSSC fabrication done by a two-step deposition method. The processing temperature is in the range of 150°C. However, highly efficient MSSC devices have a high sintering temperature of 500°C.⁶⁵ During high-temperature sintering, the mesoporous layer allows rapid extraction of photo-induced electrons, which shorten the electron transport length. This shortening of length makes MSSC an efficient light harvester. However, they suffer from the various disadvantage of having low open-circuit voltage ($V_{\mathrm{oc}}$) and lower light absorbance at wavelengths greater than 700 nm. Mesoporous ${\mathrm{Al}}_2{\mathrm{O}}_3$, which acts as an insulator, is found to be more stable than ${\mathrm{TiO}}_2$, under continuous illumination. ${\mathrm{TiO}}_2$ introduces nonstoichiometry defects, such as titanium interstitials and oxygen vacancies, which causes trap sites.⁶⁶ Encapsulation of layers can prevent moisture ingress and degradation. Etgar et al.⁶³ recently reported that spin-casting a methylammonium lead iodide (${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3$) solution on a 400-nm thick ${\mathrm{TiO}}_2$ nanosheet would form a mesoscopic $({\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3)/{\mathrm{TiO}}_2$ heterojunction. Evaporating a gold film on top of ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3$ creates a back contact. It provided PCE of 5.5%, short circuit photocurrent $J_{\mathrm{sc}}=16.1\mathrm{mA}/{\mathrm{cm}}^2$, and open-circuit photovoltage $V_{\mathrm{oc}}=0.631\mathrm{V}$ under Am 1.5 solar light of $1000\mathrm{W}/\mathrm{m}$.⁶³

Fig. 8

Schematic illustrating of mesoporous PSCs. The mesoporous oxide can be an electron transporting semiconductors, such as ${\mathrm{TiO}}_2$, ZnO, or an insulating layer, such as ${\mathrm{Al}}_2{\mathrm{O}}_3$ or ${\mathrm{ZrO}}_2$ (MSSC). Reprinted with permission from Ref. 64. © 2012 by the American Association for the Advancement of Science.

2.2.

Planar Structure

Film structure has a simple construction as it does not have a mesoporous scaffold or electronic layer in it, facilitating ease in fabrication. Figure 9 shows the architecture of n-i-p and p-i-n variants with different location of the charge transport layer, which is their primary difference.⁶⁷ A simple planar junction designed between perovskite and HTM layer, without scaffolding, corresponds to a simple planar thin-film cell. The major problem in the fabrication of this structure is to avoid contact between HTM and blocking layer. Vapor deposition and dual-source vapor method excel in achieving this required feature. Docampo et al.⁶⁸ applied ${\mathrm{NiO}}_x$ and ${\mathrm{VO}}_x$ in planar heterojunction solar cells with ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_{3\text{-}\mathrm{x}}{\mathrm{Cl}}_{\mathrm{x}}$ as photoactive material. However, these ${\mathrm{NiO}}_x$ and ${\mathrm{VO}}_x$ interlayers resulted in low photovoltaic performance due to poor coverage on the substrate by perovskite film.⁶⁸ Alternatively, UV treatment was reported to enhance surface wetting properties with increased PCE by 7% on these materials. HCl improves crystallization and coverage of ${\mathrm{PbI}}_2$ in thin film, which subsequently improves the morphology achieving a PCE of 14.8% for perovskite thin film.⁶⁹ Planar heterojunction additionally assists in evaluating and testing of the new materials and their effect, for solar cell fabrication. ${\mathrm{SnO}}_2$ was also experimented to replace ${\mathrm{TiO}}_2$ as electron-selective contact. ${\mathrm{SnO}}_2$ was spin coated and sintered at a much lower temperature of 200°C than ${\mathrm{TiO}}_2$. Its fabrication included the solvent vapor annealing, which crystallizes the residual ${\mathrm{PbI}}_2$ and adds it at the interface of ${\mathrm{SnO}}_2/{\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3$. PSC devices based on ${\mathrm{SnO}}_2$ displayed high stability of holding the PCE of 13% till 700 h.⁷⁰

Fig. 9

Schematic representation of two different configurations used PSC. (a) Standard architecture n-i-p and (b) inverted architecture p-i-n. Reprinted with permission from Ref. 67. © 2017 by Elsevier.

2.3.

HTL/ETL Free Cells

The fundamental task of any photovoltaic device is to create charge carriers, i.e., holes and electrons on the absorption of light energy. Further separating them at respective heterojunction and making them available at their respective electrodes. This separation depends on energy offsets versus exciton binding energy. It defines the formation of two primary interfaces, perovskite/electron transport layer (ETL) interface and perovskite/HTL interface. The architecture of a few PSCs does not have HTL or ETL layers. As these are not mandatory layers, numerous researchers have developed PSC without HTL^71,72 or ETL with good PCE. Perovskite layer has Au deposited over it while fabricating the HTL-free PSC. In this, perovskite layer conducts the duty of absorber and hole conductor by forming a heterojunction with ETL. Aharon et al.⁷¹ have reported a two-step deposition method to fabricate highly optimized HTL-free PSC with a PCE of 10.85%, FF of 68%, and $V_{\mathrm{oc}}$ of 0.84 V. Their performance is strongly dependent on the width of the depletion layer formed by ${\mathrm{TiO}}_2\text{-}{\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3$. Depletion layer supports charge separation and restricts back reaction from ${\mathrm{TiO}}_2$ to ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3$. The width of the depletion layer is evaluated using Mott Schottky analysis.⁷³ Figure 10 shows a schematic diagram of HTL-free PSC with their energy levels.⁷¹ Figure 11(b) shows ETL-free PSC in-which an insulator (${\mathrm{Al}}_2{\mathrm{O}}_3$) replaces ${\mathrm{TiO}}_2$ scaffold,⁶⁴ which proves that perovskite material could effectively transport electrons without an ETL. Production cost for HTL-less cells is high, leading to an increase in device cost. Further, the presence of the HTL layer complicates the mechanism followed for extraction.

Fig. 10

(a) Structure of hole conductor free ${\mathrm{TiO}}_2-{\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3$ perovskite-based solar cells. Arrow indicates the depletion region observed at ${\mathrm{TiO}}_2-{\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3$ junction. Reprinted with permission from Ref. 71. Copyright (2014) Royal Society of Chemistry. (b) Energy level diagram that shows charge separation process. The position of energy levels is taken from Ref. 63. © 2012 American Chemical Society.

Fig. 11

Structure of PSC: (a) HTL-free PSC and (b) ETL-free PSC.

2.4.

Tandem Cells

Tandem cells consist of two formats, i.e., four-terminal and two-terminal devices. Figure 12 shows the arrangement of four-terminal tandem device, which has two-terminal devices/cells. They are either mechanically stacked or monolithically integrated through tunneling junction. Baillie and his coworkers⁷⁴ with copper indium gallium selenide, Asadpour et al.⁷⁵ with a highly efficient a-Si:H/c-Si heterojunction bottom cell and Mailoa et al. using the c-Si bottom cell on perovskite/silicon multijunction solar cell have obtained PCE of 30%, 13.4%, and 13.7%, respectively. Due to the limitation of band gap mismatch and current matching constraints, a traditional tandem cell of Si and perovskites could only achieve PCE of 24% and 20%, respectively. Top cell absorbs high energy photons and bottom cell harness the remaining low-level energy photons. ${\mathrm{MAPbI}}_3$ makes an excellent top layer in c-Si solar cells. Hydrogenated indium oxide, having high mobility, is used as a rear electrode to fabricate semitransparent PSCs.⁷⁶ Fabrication of two-terminal tandem solar cells with perovskite as the top layer and kesterite copper zinc tin sulfide (CZTS) as bottom cells was reported to attain a PCE of only 4.6%. This reduction is primarily due to semitransparency of the top aluminum electrode, resulting in transmission loss.⁷⁷ Thus, the highly efficient and transparent top layer is a very critical factor. Further, they should be immune to thermal instability and hygroscopicity. Numerous transparent electrodes used here are ultrathin metal such as Au,⁷⁸ graphene,⁷⁹ carbon nanotubes,⁸⁰ and Ag nanowires.¹⁹ However, efficiency achieved was below 10%, and their contacts featured strong absorption in near-infrared (NIR). PSC that are individually efficient underperformed when used as tandem cells, due to sensitivity to a thickness of perovskite layer. Thus, bifacial tandem cell was fabricated, which utilizes scattered light back from the ground (Fig. 13) to achieve a PCE of $\sim 33\%$.⁷⁵ Further, they are immune to variations in perovskites layer thickness.

Fig. 12

Architecture of a four-terminal PSC.

Fig. 13

(a) Operational bifacial tandem solar panel, utilizing direct illumination of sunlight and a fraction of light that scatters from the ground onto back of the panel. (b) Traditional tandem structure and (c) a bifacial tandem structure. Reprinted with permission from Ref. 75. © 2015, AIP Publishing.

4.1.

Structure

Materials and structure are also essential factors that decides the stability of PSC. The mesoporous structure formed by ${\mathrm{TiO}}_2$-, ${\mathrm{Al}}_2{\mathrm{O}}_3$-, or ZrO-coated substrates, exhibited the pore-filling problem, low stability, and degraded performance.^17,64 Alteration of precursor influences the loss of organic material in PSC. This mass loss was reported to be 16%, 12%, and 5%, between temperatures of 250°C and 400°C¹³⁴ in three steps. Under sunlight and device operating temperature ($\sim 327°\mathrm{K}$), ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3$ undergoes a phase transition from tetragonal to cubic. A disordered character is shown by methylammonium ions during the cubic–tetragonal transition, whereas ${\mathrm{PbX}}_6$ octahedron displays a displacive character.^135,136 Various combinations of halogens, especially of Pb and Br, induce change and transformation in the crystal structure of ${\mathrm{CH}}_3{\mathrm{NH}}_3\mathrm{Pb}({\mathrm{I}}_{1-x}{\mathrm{Br}}_x)$. As $x$ changes from 0 to 0.13, crystal structure transforms from tetragonal to cubic form with a change in the band gap. Empherical representation by quadratic equation $E_g(x)=1.57+0.39x+0.33x^2$ describes the dependence of band gap on $x$.⁶¹ The low value of bowing parameter, 0.33, in the equation confirms high miscibility^137,138 of ${\mathrm{MAPbI}}_3$ and ${\mathrm{MAPbBr}}_3$. This low value shows that $\mathrm{MAPb}{({\mathrm{I}}_{1\text{-}x}{\mathrm{Br}}_x)}_3$ would constitute the entire composition of the compound and enables convenient band gap tailoring. For subservient Br fraction ($x<2.0$), PCE remained unchanged with improved stability factor when tested for 20 days at 55% RH. Diminished cross-section and transformation from tetragonal to cubic stage, due to higher Br content, has led to enhanced stability.

New materials such as ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{SnI}}_3$ can resolve lead toxicity problems. Whole fabrication was conducted in ${\mathrm{N}}_2$ environment as it decolorizes if exposed to the room environment. This decolorization is primarily due to oxidation of ${\mathrm{Sn}}^{2+}$ ions, which lead to the formation of oxides of Sn and residual perovskites material. Sn-based PSC shows an enhanced PCE of more than 6% and has a short diffusion length. It restricts the fabrication of devices with planar architecture. A combination of Sn with other halides, namely ${\mathrm{CH}}_3{\mathrm{NH}}_3\mathrm{Sn}({\mathrm{I}}_{3\text{-}x}{\mathrm{Br}}_x)$ was developed to achieve PCE of 6%.¹³⁹ Schoonman has explained the degradation in the absence of moisture.¹⁴⁰ Crystal structure of 3-D perovskites incorporating corner-sharing ${\mathrm{PbI}}_6$ octahedra is broken into 1-D double chains consisting of two adjacent octahedral (${\mathrm{MAPbI}}_3{·\mathrm{H}}_2\mathrm{O}$, monohydrate form) and subsequently into 0D isolated octahedral (${\mathrm{MA}}_4{\mathrm{PbI}}_6{·2\mathrm{H}}_2\mathrm{O}$, dihydrate form) by intercalation of ${\mathrm{H}}_2\mathrm{O}[4]$. Structural degradation occurs due to ionization-simulated degradation; it happens when the hole of halide member, which the inner layers produce, migrates and reacts with negative iodine on the surface. The reports confirm that doping improves crystallization. ${\mathrm{GeO}}_2$ doping, in planar PSC forming $\mathrm{PEDOT}:\mathrm{PSS}\text{-}{\mathrm{GeO}}_2$ composite films, displayed active sites for crystal nucleus during annealing.¹⁴¹ Doping has an overwhelming influence on performance, due to various reasons. Primarily it is due to undesired band bending at ETL and HTL interfaces with perovskites, increase in Auger recombination, and traps/defects. For higher PCE, the Fermi level in perovskite should not be above (below) Fermi level of ETL (HTL) in equilibrium.²⁵

4.2.

Materials of PSC

Perovskite materials exhibit features such as direct band gap, high optical extinction coefficients (${10}^4$ to ${10}^5{\mathrm{cm}}^{-1}$) in the NIR-visible region, and high carrier mobility ($\mu \sim 1$ to $10{\mathrm{cm}}^2{\mathrm{V}}^{-1}{\mathrm{s}}^{-1}$) making it a prospective member for light harvesting. A high optical extinction coefficient of perovskite makes it a good absorption of light, at low mesoporous film thickness. Tuning the band gap of perovskite materials is achieved by modifying the chemical composition of perovskite⁶¹ or by alloying different perovskites.¹⁴² In addition to other properties, perovskite materials also exhibit ambipolar charge transport property (can be n- or p-type), which conducts electrons and holes through them, due to small effective mass of both electrons.^36,143–145 High dielectric constant leads to effective charge screening. By bandgap engineering through modification of the chemical composition of PSC, such as ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_{1\text{-}x}{\mathrm{Br}}_x$, the entire spectrum is utilized.¹⁴⁶ Enhancement in PCE happens due to reduction in binding energy. The same was confirmed by incorporating core-shell ${\mathrm{Au}}_2{\mathrm{SiO}}_2$ in PSC and achieved PCE of 11.4%, $J_{\mathrm{sc}}=16.91\mathrm{mA}/{\mathrm{cm}}^2$).¹⁴⁶ On addition of ${\mathrm{PbI}}_2$, Sn halide perovskite exhibits photovoltaic properties. It attained a maximum photon to current conversion of 4.18% with 1060-nm-thick Sn layer, with a redshift at Sn thickness of 260 nm.¹⁴² Variation in prominent layers and adaptation of methodology can increase stability and PCE. Addition of 5% by volume of water into MAI solution of isopropyl alcohol assists in preferential crystallization in (110) plane, achieving a large grain size of $\sim 600\mathrm{nm}$. This method achieved a PCE of 12.42%.¹⁴⁷

4.3.

Hysteresis

Around 2014, scientists observed hysteretic current–voltage behavior of PSC.^179,180 Hysteresis, which is a specific property, is formed likely due to slow transient effects, leading to an over or underestimation of solar cell efficiency. J–V hysteresis property of PSC is a function of dielectric response of perovskite material, mobile ions,¹⁸¹ and trap-state discharging.¹⁷⁹ The shape of measured current–voltage (IV) curves and efficiency are function of scan direction, light source, delay in measurement time, and voltage bias conditions before measurement. It was established by various modeling and experimentation that hysteresis is weak when surface recombination is low and diffusion length is long.¹⁸² Improved crystallinity, better fabrication process, and improved contacts are going to continue to improve hysteresis. HTMs influence hysteresis and J–V curve. Cu-based, CZTS HTM express less hysteresis than with spiro-MeOTAD HTM.¹⁸³ J–V hysteretic behavior of CuI-based devices with an efficiency of 7.5% is better than spiro-OMeTAD devices. Faster polarization relaxation was observed using EIS and OCVD for CuI-based devices.¹⁸⁴ Recently, in addition to normal-type of hysteresis, an inverted hysteresis that exhibits an opposite behavior to the normal is also reported.^185,186 The inverted hysteresis is due to interfacial charge extraction barriers formed by ionic defect accumulation and space charge build-up.¹⁸⁷ The inverted hysteresis is noted to be a function of bias range and sweep rate. Different bias scanning conditions lead to a varying ionic movement at the interface of the ETH or HTL/perovskite, which assists in the tuning of these hysteresis.¹⁸⁶

Deposition of fullerene layers onto perovskites, which assists in the elimination of photocurrent hysteresis, is reported.¹⁸⁸ Methodology and accuracy for evaluating performance characteristics of PSCs are crucial. Hysteresis influences the evaluation of performance parameters such as I–V curves and $P_{\max }$. Measurement conditions such as sweep direction, start voltage, end voltage, hold time, and sweep time have an overwhelming effect.¹⁸⁹ Inverted devices (p-i-n structured) are reported to have less hysteresis.¹⁸² Figure 18 shows that normalized time-dependent $J_{\mathrm{sc}}$ of the regular film has degraded to a higher rate than inverted structured PSC.¹⁹⁰ Though moving of electron contact to the rear side is a possible explanation, detailed reasoning requires further investigation. Incorporation of a [6,6]-phenyl-${\mathrm{C}}_{61}$-butyric acid methyl ester (PCBM) interlayer and high-quality perovskite absorber fabricated by thermal evaporation–spin coating method produces a hysteresis-free planar perovskites.⁷⁶

Fig. 18

I–V hysteresis and time-dependent photocurrent. (a) Scan direction-dependent I–V curves of the $\mathrm{cp}\text{-}{\mathrm{TiO}}_2/{\mathrm{MAPbI}}_3/\text{spiro}$-MeOTAD (normal) structure and (b) the $\mathrm{PEDOT}:\mathrm{PSS}/{\mathrm{MAPbI}}_3/\mathrm{PCBM}$ (inverted) structure. (c) Normalized time-dependent $J_{\mathrm{sc}}$ of normal and inverted structures. Open-circuit condition under one sun illumination was maintained before measuring $J_{\mathrm{sc}}$. Reprinted with permission from Ref. 190. © 2015 by the American Chemical Society.

Ions accumulate at the ion-blocking interface. The “p” and “n” doping induce ion migration at ion-blocking interface. These doping will screen the applied voltage and alter charge collection capability, resulting in hysteresis. In addition, ion migration is responsible for unusual behaviors in perovskites materials such as photo-induced phase separation,¹⁹¹ photoinduced giant dielectric constant,¹⁹² and absence of transistor behavior at room temperature.¹⁹³ In high efficient PSC cells, ion migration does not cause hysteresis.¹⁸² Ion migration within crystal structure,¹⁹⁴ which is also a major contributing factor for hysteresis, has a profound effect on device stability.¹⁹⁵ Experimental proof and test results show that ions migrate along the grain boundaries.¹⁹⁶

4.4.

Moisture

Moisture acts as a catalyst, and its presence tends to induce hydrolysis¹⁰⁷ in PSC, thereby converting PSC back to precursors, which is irreversible. Experiments to evaluate the stability of PSC against moisture were conducted in inert and other carrier gases to ascertain degradation exclusively due to moisture.¹² When exposed to moisture, perovskite structure tends to hydrolyze, undergoing irreversible degradation and decomposing back into precursors. Highly hygroscopic ${\mathrm{CH}}_3{\mathrm{NH}}_{3}X$ converts to $\mathrm{CH}{({\mathrm{NH}}_2)}_{2}X$ salts and ${\mathrm{PbX}}_2$, with $X$ = halide.¹⁹⁷ Heat, electric field, and ultraviolet exposure¹⁹⁸ accelerate this process. Figure 19 shows the mechanism of degradation:

Fig. 19

Proposed decomposition pathway of ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3$ in the presence of a water molecule.¹⁹⁷

Stability of PSC, due to moisture, requires evaluation in three modes. One is purely because of moisture, the second is under the influence of moisture along with light, and finally the third mode is a combination of moisture along with light and oxygen. Degradation of perovskites is reported to follow the mechanism of Grotthuss.¹⁹⁹ In this, one water molecule reacts with perovskite material and degrades to HI and ${\mathrm{CH}}_3{\mathrm{NH}}_2$. Further, HI reacts with ${\mathrm{O}}_2$ to form water, to initiate more deterioration. After exposing perovskite film ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3$ to 80% RH for 2.5 h, the possibility of intermediate crystalline phase such as a hydrated compound ${({\mathrm{CH}}_3{\mathrm{NH}}_3)}_4.{\mathrm{PbI}}_6.{\mathrm{H}}_2\mathrm{O}$ exists. Grazing incidence x-ray diffraction (GIXRD) confirms the formation of intermediate phase. XRD pattern shows strong peaks of hydrate at 8.42 deg and 10.46 deg, which confirm the existence of intermediate phase. Exposure of solar cell to RH of 0% to 50% has led to very little degradation. However, under 90% RH, it showed a rapid decrease in PV performance efficiency from 12% to less than 1% in just 3 days. Third day values are $J_{\mathrm{sc}}=3.1\mathrm{mA}{\mathrm{cm}}^{-2}$, $V_{\mathrm{oc}}=0.637$, and $\mathrm{FF}=0.36$, $\eta =0.7$.²⁰⁰ Increase in RH level beyond 55% resulted in a drastic increase in degradation rate. In addition to general characterization, a powerful method such as photothermal deflection spectroscopy evaluates the changes in absorption after exposing perovskites to moisture.²⁰¹ In Fig. 20, ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3$ shows absorption edge as $\sim 1.57\mathrm{eV}$, which shifts to 2.3 eV on exposure to moisture, this correspondence to ${\mathrm{PbI}}_2$ formation, and decomposition of ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3$. Figure 21 shows PCE versus heterojunction solar cells variation based on $\mathrm{MAPb}{({\mathrm{I}}_{1\text{-}x}{\mathrm{Br}}_x)}_3$($x=0$, 0.06, 0.20, 0.29) as per time, stored without encapsulation at room temperature with RH ranging from 35% to 55%.²⁰² It also shows variation in UV–vis spectra as per moisture content and variation in the morphology. ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbBr}}_3$-based perovskites are less sensitive to moisture than ${\mathrm{PbI}}_{\mathrm{X}}$-based Perovskites. However, the absorption capacity of pure ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3$ is not comparable to that of ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbBr}}_3$.^55,203 The first line of defense against the moisture is the top charge-transporting layer of the device. Resistance to moisture can be increased by inherently strengthening the bond between cation and metal halide.²⁰⁰ Few other approaches are inclusion of thin blocking layer between perovskites and HTL, such as ${\mathrm{Al}}_2{\mathrm{O}}_3$,^169,204 moisture-blocking hole transporters,²⁰⁵ and hydrophobic carbon electrodes.^{2,116,206–209} Incorporation of 2-D PSC with large aromatic phenyl ethylammonium (PEA) cations could achieve average PCE of 5.5% with excellent stability against moisture. After exposure of these 2-D PSCs for more than 90 h at $72\pm 2\%$ RH, they were stable and maintained 50% PCE of initial value.²¹⁰ Incorporating polyethylenimine cations in 2-D PSC helps in attaining better film quality and resistance to moisture degradation with PCE of 8.77%.²¹¹ In addition, near-single-crystalline-type thin films are fabricated, which have, preferential out-of-plane alignment of crystallographic planes, good charge transport, and moisture resistance.¹¹⁷ Thus, 2-D PSC has better moisture resistance capability than the 3-D PSC.^212,213

Fig. 20

PDS spectra of ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3$ films before (initial state) and after exposure to a relative humidity of 30% to 40% for varied time period. Reprinted with permission from Ref. 201. © 2013 by the American Chemical Society

Fig. 21

Effect of moisture: (a) in situ absorbance measurements of ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3$ exposed to 98% relative humidity. Measurements were taken at 15 min intervals, (b) normalized absorbance measurements (taken at 410 nm) for ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3$ films exposed to different relative humidity, (c) SEM images of uncoated ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3$, ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3/\text{spiro}$-OMeTAD, ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3/\mathrm{PTAA}$, and ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3/\mathrm{P}3\mathrm{HT}$ films after exposure to 98% RH for 24 h. (d) Powder x-ray diffraction patterns for a sample of ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3$ powder. Reprinted with permission from Ref. 202. © 2015 by the American Chemical Society. (e) PCE versus heterojunction variation on $\mathrm{MAPb}{(}_{1\text{-}x}{\mathrm{Br}}_x)3$($x=0$, 0.06, 0.20, 0.29) with time stored in air at room temperature without encapsulation. Reprinted with permission from Ref. 61. © 2013 by the American Chemical Society.

Solution-processed ${\mathrm{MAPbI}}_3$ was immune to changes when exposed to ${\mathrm{O}}_2$. The decomposition on exposure to moisture or ambient air is due to ${\mathrm{H}}_2\mathrm{O}$.²⁰² However, very low quantity ${\mathrm{H}}_2\mathrm{O}$ acts as an n-dopant for ${\mathrm{MAPbI}}_3$, affecting the energy level. Proposed decomposition of ${\mathrm{MAPbI}}_3$ eventually releases ${\mathrm{NH}}_3$ and HI as volatile species, leaving ${\mathrm{PbI}}_2$ and making decomposition reaction irreversible. Reports confirm that moisture coupled with light directly led to decomposition of ${\mathrm{MAPbI}}_3$ to ${\mathrm{PbI}}_2$, which was an irreversible reaction. Residuals such as ${\mathrm{PbI}}_2$ and ${\mathrm{I}}_2$ are formed easily under sunlight, instead of reversible hydrate in the dark, when ${\mathrm{MAPbI}}_3$ exposed to an environment of 18 h at 60% RH.¹⁰⁷ Though the analysis of stability in dark is a basic test, stability research under illumination is essential for evaluation of working efficiency and degradation factor. Unlike degradation in the dark, exposure to moisture in light leads to irreversible decomposition of ${\mathrm{MAPbI}}_3$ to ${\mathrm{PbI}}_2$.

4.5.

Light Spectrum (UV and Visible)

Mostly laboratory experiments are conducted under controlled light source such as LED. However, actual utilization involves sunlight with UV element, which accelerates degradation pattern. The ${\mathrm{TiO}}_2$ element in PSC, which is photocatalytic in the UV regime, assists degradation.¹⁴⁸ To further surprise, as shown in Fig. 22, degradation is higher in encapsulated devices than the unencapsulated devices for a conventional device.

Fig. 22

Light soaking tests of regular PSCs under A.M. 1.5 G 1 Sun light intensity (with and without UV filters). Reproduced with permission from Ref. 148. © 2013 by Springer Nature.

This phenomenon is due to electrons injected into ${\mathrm{TiO}}_2$ are entrapped in deep-lying unoccupied sites.¹⁴⁸ However, encapsulation along with UV-filter can enable attaining better results. Degradation is directly proportional to efficiency and follows a pattern of reduction in $J_{\mathrm{sc}}$. A ${\mathrm{TiO}}_2$-free mesoporous structured solar cell is comparatively less degradable than MSSC with ${\mathrm{TiO}}_2$, which is confirmed by transient absorption spectroscopy. MSSC has a property of degradation being high at maximum power and rather low at open circuit condition. However, for practical applications installation of a UV filter on PSC overcomes the degradation process. Degradation is attributed to the instability problem²¹⁴ and the catalytic role of metal oxide when exposed to the UV spectrum. A general degradation mechanism for UV exposure is shown in Fig. 23. On exposure to UV light, photogenerated holes that are generated by absorption of energy react with the oxygen radicals adsorbed at surface oxygen vacancies. Molecular oxygen is absorbed from these sites, leading to a formation of unoccupied deep surface traps sites and a free electron per site. These electrons will recombine with the excess of holes in doped-hole transporter. On photoexcitation of sensitizer, electrons can follow in two ways. In the first route, they get injected into conduction band by which they get deeply trapped and in other they directly enter into the deep surface traps. These deeply trapped electrons are not mobile and recombine readily with holes on hole transporter.¹⁴⁸ The primary bottleneck that exists is transporting the charge. In quantum dots, polydispersity strongly affects transport. In quantum dot film, if few dots are of different sizes, then trap states manifest within the bandgap and lead to recombination. Similar is the case with 1-D carbon nanotubes. Three schemes are formulated to nullify the effects of UV. First being the reduction of trap sites or pacifying it and second is to undertake measures to avoid UV light reaching ETL (especially ${\mathrm{TiO}}_2$). Finally, the third scheme is to replace ${\mathrm{TiO}}_2$ scaffold with another material or make it ${\mathrm{TiO}}_2$ free, which has been explored by Leijtens at el.¹⁴⁸ The efficiency of PSC device reduces to almost 0% after 12 h of exposure in a nonencapsulated format. However, with the introduction of ${\mathrm{Sb}}_2{\mathrm{S}}_3$ buffer layer in PSC, efficiency increased to $\sim 65\%$ of its initial value. This increase is due to the ${\mathrm{Sb}}_2{\mathrm{S}}_3$ blocking layer, which blocks the UV-induced photocatalysis in ${\mathrm{TiO}}_2$, by providing a form of passivation. In addition, this extra layer assists in maintaining the crystalline structure of ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3$.²¹⁵ A methodology such as usage of the downshifting molecule such as $N$-(2-(6-chloros-tetrazin-3-yloxy)ethyl)-naphthalimide (NITZ) with a strategy to suppress UV from the source is also explored, for UV protection in PSC.²¹⁶ They have a specialty to get excited in the UV spectrum and to emit into the visible spectrum, allowing UV photons utility.

Fig. 23

Proposed mechanism for UV-induced degradation Reprinted with permission from Ref. 148. © 2013 by Springer Nature.

4.6.

Thermal Stability

Variation in efficiency and increase in temperature will result in amplification of deformities. These anomalies are due to deterioration in perovskites materials and behavior of HTM. It is understood that when the heat of the device ingresses into cell materials, mechanical stress of the materials at interfaces increases, thereby decreasing the lifetime of the PSC device. Accelerated aging test results showed that ${\mathrm{PbI}}_2$ deteriorates much faster than ${\mathrm{PbBr}}_2$, approximately after 60 min. High stability of Br is due to shorter and stronger bonds of Pb-Br compared to Pb-I.²¹⁷ Heating of ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_{3\text{-}x}{\mathrm{Cl}}_x$ in ${\mathrm{N}}_2$ environment at 90°C not only leads to the formation of 3-D perovskites structure but also results in a high rate of degradation at 100°C.²¹⁸ Hybrid perovskite assists in maintaining stability, as inorganic part of hybrid perovskites increases thermal stability and its organic part increases the stability of device. Comparison of formaldinium lead trihalide $(\mathrm{HC}{({\mathrm{NH}}_2)}_2{\mathrm{PbI}}_3)$ and ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3$-based device reveals that the former is more stable as it deteriorates only after 290°C than the later, which decomposes at 230°C.²¹⁹ In PSC, deterioration rate due to temperature increases in the presence of moisture. Reports reveal that at 85°C time of decomposition is 283, 208, and 133 min when RH is 12%, 57%, and 85%, respectively.²²⁰ Period of exposure is also an influential factor for evaluation and characterization by hard x-ray photoelectron spectroscopy instead of conventional XRD, because assessment is done on chemical content regardless of crystallinity of the sample. Mechanism of degradation is as follows:

Films are made independent of parameters such as moisture and oxygen by subjecting it to ultrahigh vacuum heating chamber, to evaluate effects purely due to temperature.²²¹ I/Pb and N/Pb ratios define the characterization of films for degradation. Reduction of these ratios indicates degradation of perovskites into ${\mathrm{PbI}}_2$, as shown in Fig. 24. Variation in temperature induces a different degree of degradation in various layers of PSC. For example, perovskites made of scaffolded ZnO are found to be more stable than ${\mathrm{TiO}}_2$.^222,223

Fig. 24

I/Pb ratio (red circle) and N/Pb (black diamond) ratio plotted versus different temperatures. Reprinted with permission from Ref. 221. © 2015 by the American Chemical Society.

To understand degradation on perovskites purely due to thermal effects, semifabricated PSC, without HTL, ETL, and the top electrode are tested.¹⁰⁶ These semifinished planar structures undergo exposure to inert, ${\mathrm{O}}_2$, and ambient conditions for 24 h. Surprisingly, degradation in the form of delamination happens in an ambient environment rather than in ${\mathrm{O}}_2$. GIXRD and surface photovoltage measurements assisted in characterization and revealed that at temperature above 140°C and 160°C perovskite materials decomposed into ${\mathrm{PbI}}_2$ in vacuum.²²⁴ Up to 150°C, an increase in the Braggs peak with enhancement of perovskite film occurs. From 150°C to 200°C, these peaks reduced due to ${\mathrm{PbI}}_2$ formation and further heating to 330°C lead to complete decomposition into Pb.²²⁵ Empirical studies revealed that thermal annealing conducted for production of perovskite materials induces slight ${\mathrm{PbI}}_2$ production, causing drop in efficiency. Heat treatment of layers such as ZnO before deposition lead to an increase in efficiency and thermal stability, due to reduction in deportation of methylammonium at the interface.²²³ Thermal stability is a highly potent factor that affects different layers especially the HTM. Initiation of crystallization and formation of large crystals causing morphological hole traps affecting charge transportation and effecting PCE was reported for PSC with spiro-OMeTAD as HTM.²¹⁸ This crystalline process gets aggravated by additives or doping done in the layer to increase charge due to the hygroscopic nature of dopants.²²⁶ To achieve better stability, many new HTM materials were explored, such as PTAA and 5,10,15-trihexyl-3,8,13-tris(4-methoxyphenyl)-10,15-dihydro-5H-diindolo[3,2-a:3′,2′-c] carbazole (KR131).²²⁷ Introduction of bulky aromatic PEA cations was reported to produce thermally stable 2-D PSC, which maintained 65% of its initial PSC after exposure to 100°C for 70 h.²¹⁰ In addition, multidimensional 3-D/2-D perovskites with an intermediate dimension between them led to the formation of thermally stable PSC, due to reduced defect density.²²⁸

4.7.

Fabrication and Preparation

Fabrication is a vital factor for sustaining stability in PSC. It primarily constitutes two elements such as the synthesis method used and the substrate precursor. Synthesis temperature and time strongly influence the morphology, stability, and performance of PSC. In one-step deposition process, perovskite solution is dispensed on the substrate while turning at a fixed velocity. Though excess perovskite is spun off, perovskites crystallize on evaporation of the solvent. Uncontrolled precipitation yields nonuniform large- and small-size crystals, resulting in poor performance and solar cell efficiency. Adhikari et al. have proposed the usage of an antisolvent method for better crystal formation and growth. It involves annealing conditions and defines charge transport capability, as measured by nanoscale Kelvin probe force microscopy (KPFM).²²⁹ Annealing temperature effect on ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbI}}_3$ is analyzed from 80°C to 140°C as shown in Fig. 5. The peak of ${\mathrm{PbI}}_2$ at 12.7°C, in XRD, increases proportionally as per the annealing temperature. The annealing time is optimized to achieve correct crystal size, charge transport layer such as HTL formed CZTS nanoparticles (CZTS-30).²³⁰ Thermal annealing of ${\mathrm{CH}}_3{\mathrm{NH}}_3{\mathrm{PbBr}}_3$ films at 90°C, for short duration such as $\le 10\min$, would lead to increased crystallinity, eliminating residual solvent and achieving complete conversion of the precursor to crystal, in comparison to longer annealing time.²³¹ Greater annealing time would initiate sublimation, so accurate monitoring of time is essential. For ${\mathrm{TiO}}_2$ films, photonic curing technique is more effective than high-temperature annealing ($\sim 500°\mathrm{C}$), as it creates more conductive films.⁹⁶ Photonic curing is also called pulse-thermal processing, which utilizes high-intensity plasma arc lamp. It delivers a peak sintering power up to $\mathrm{20,000}\mathrm{W}/{\mathrm{cm}}^2$ for milliseconds and produces rapidly annealed thin films without damaging underlying plastic substrates. PEDOT:PSS-based PSC grown at room temperature without annealing was reported to exhibit a PCE of 14.27%.²³² They have high crystallinity when subjected to 40% RH at room temperature.²³²

4.8.

Device Encapsulation

Air sensitivity of PSC elements, in conjunction with moisture sensitivity of perovskite absorbers, emphasizes the requirement of encapsulation of high level to attain a stable outdoor lifetime of over 25 years.^3,64,61 Without encapsulation, highly efficient devices were reported to degrade rapidly, showed 80% degradation when subjected to ambient temperature.²³³ Usually, a thin glass coverslip is used to cover device, which is further sealed using a UV curable epoxy resin.^148,169,234 Ramos et al.²³⁵ reported that single step high-temperature encapsulation ($>150°\mathrm{C}$) resulted in delamination of active and metallization layers in devices. A two-step encapsulation approach minimizes the defect due to mechanical and thermal stress. The first step in a pre-encapsulation process consists of making electrical contact and later embedding it in a polymer matrix to ensure mechanical stability. The next step involves laminating the pre-encapsulated device using low temperature, to retain good photovoltaic characteristics of PSC. During the process and after, samples are monitored electrically (J–V curves). There are two types of configuration for encapsulation. Han et al. have compared these two types,¹⁶⁸ in first method, UV curable epoxy resin used to fill the area between silver contact and plain glass cover.^{131,236–238} The other configuration has an electrode, which has a top glass cover with a gap in between. Water absorbent, desiccant material was made to fill this gap, thus absorbing moisture that had penetrated through epoxy resin sealant.¹⁶⁸ For 2-D PSC films, encapsulation becomes very critical. Therefore, plastic barrier films are used. It is available in two formats, namely partial and complete. They laminate commercial encapsulate adhesive on to films at 100°C. This type of encapsulation could only retard the ingress of moisture penetration and can maintain the operational performance of PSC steady for more than 1 year.^{131,236,239–241} Conductive tapes made of commercial carbon assist in achieving a dual purpose such as terminal contact with perovskite and encapsulation. Encapsulated devices are leak tested with helium gas to assure moisture penetration. Investigations reveal that incorporation of impermeable barrier layers can be a long-term remedy to eliminate the ingress of moisture and oxygen.