Stability of perovskite solar cells

Highlights

• Perovskite solar cells may provide efficient, low-cost energy generation.

• The stability of perovskite devices must be addressed to achieve commercialisation.

• The key factor is the perovskite material sensitivity to moisture.

• The stability of the device as a whole must be considered.

• Encapsulation and standard testing protocols are required to improve stability.

Abstract

The performance of perovskite solar cells has increased at an unprecedented rate, with efficiencies currently exceeding 20%. This technology is particularly promising, as it is compatible with cheap solution processing. For a thin-film solar product to be commercially viable, it must pass the IEC 61646 testing standards, regarding the environmental stability. Currently, the poor stability of perovskite solar cells is a barrier to commercialisation. The main issue causing this problem is the instability of the perovskite layer when in contact with moisture; however, it is important to explore stability problems with the other layers and interfaces within the device. The stability issues discussed in this review highlight the need to view the device as a whole system, due to the interdependent relationships between the layers, including: the perovskite absorber, electron transport layers, hole transport layers, other buffer layers and the electrodes. We also discuss other issues pertaining to device stability, such as measurement-induced hysteresis and the requirement for standard testing protocols. For perovskite solar cells to achieve the required stability, future research must focus on improving the intrinsic stability of the perovskite absorber layer, carefully designing the device geometry, and finding durable encapsulant materials, which seal the device from moisture.

Introduction

Projected increases in global energy demand, predicted to be as high as 1 GW/day, will place significant strains on current energy infrastructure [1]. This looming challenge, coupled with depleting traditional fossil-fuel based energy sources and the threat of climate change, requires the development of renewable energy technologies. Of the possible renewable energy approaches, photovoltaics (PV), the conversion of sunlight into electricity, represents a promising route. The current photovoltaics landscape is dominated by silicon solar cells, which have benefited from recent advances leading to reduced manufacturing costs [2]. However, this mature technology is constrained by some fundamental cost barriers, such as high temperature processing. A fundamental shift in thinking may allow for significantly reduced processing costs. One such alternative approach involves replacing the crystalline silicon with organic semiconductors. The key advantage of this approach is the possibility of manufacturing solar modules using solution processing techniques, which significantly reduces the cost and energy payback time [3], [4], [5]. In fact, Espinosa et al. proposed that a polymer solar cell module could be fabricated with an energy payback time of only one day, given advances in efficiency and stability [1]. Over a period of approximately 20 years, the power conversion efficiency (PCE) of polymer solar cells progressed to 12% [6]. Importantly, several demonstrations have shown that polymer solar cell modules may be fabricated using solution based printing and coating techniques and installed with an installation rate of >200 W_p/min [7], [8]. Whilst multiple small area devices, fabricated in ideal laboratory settings, have achieved PCEs above 10% [9], [10], [11], [12], the demonstrations of large scale polymer solar modules suffer from low efficiency. The system efficiency for the solar park demonstration was <2% [7]. Whilst these represent initial reports, much improvement is required in this area.

A new material class, organic–inorganic metal halide perovskite semiconductors, may be able to address this problem, such that truly low-cost photovoltaic modules may become a reality. The term “perovskites” is used to describe a group of compounds characterised by the general formula ABX₃, which have the crystalline structure of calcium titanium oxide (CaTiO₃). Fig. 1 displays a schematic figure of the generic ABX₃ perovskite crystal structure for a hybrid organic–inorganic metal halide perovskite. The A position contains an organic cation, B is a metal cation, and X is a halide anion. This structure was discovered in the 19th century; however, it has recently been applied to the field of photovoltaics, with unprecedented success [13]. The most commonly explored organic–inorganic metal halide perovskite, hereafter referred to as perovskite, applied in the field of photovoltaics, is CH₃NH₃PbI₃. This consists of a large organic cation, methylammonium ${(\text{CH}}_3{\mathrm{NH}}_3^{+})$, lead (Pb) as the smaller cation and iodine as the halogen anion. Variations on this compound have also been investigated, as described in this review. As absorber materials in photovoltaic devices, perovskites possess multiple desirable traits. Importantly, the perovskite layer may be deposited using low-cost coating and printing techniques, implying that advances made for polymer solar cells will be transferrable to this new material system. In fact, the fabrication processes available for perovskites are robust [2]. High efficiency devices have been demonstrated using sequential deposition [14], spin coating [15], vacuum evaporation [16], or CVD [17], [18] to deposit the perovskite film. Schmidt et al. have already demonstrated the fabrication of perovskite solar cells on flexible substrates, using only scalable printing and coating methods [19]. The absorption coefficient of perovskite absorbers is strong across a wide wavelength range [20], allowing for relatively thin absorber layers. The rate of non-radiative recombination in these materials is relatively low. This reduces the bandgap – voltage offset $\left(\frac{E_g}{q}-V_{oc}\right)$, allowing for high V_oc. Many reports have demonstrated V_oc values >1 V [14], [21], [22], which is higher than for efficient polymer solar cells. Using photothermal deflection spectroscopy (PDS), De Wolf et al. suggest this is due to a low density of deep trap states, characterised by a remarkably steep absorption onset, similar to the highly crystalline GaAs [20]. This is particularly remarkable given this is a new material deposited by solution processes, not expensive crystal growth methods. Additionally, these perovskite materials possess high carrier mobilities [23], [24] and large diffusion lengths [25], [26], allowing them to efficiently transport charge to be collected at the electrodes.

The first reported solar cell employing CH₃NH₃PbI₃ achieved an efficiency of 3.81% [27]. Various alterations to the device geometry allowed for a meteoric rise in device performance, for which the highest reported value is 20.1% [14], [28]. A detailed overview of the historical evolution of perovskite solar cell performance, occurring over a short time period, can be found in multiple previous review articles [13], [23], [29], [30]. This result implies that perovskite solar cells could, for the first time in the history of PV, combine high efficiency with low cost and scalable processing [13]. In fact, there are multiple approaches to improve performance available to perovskite solar cells. As an alternative to low cost solar cells produced using techniques similar to those used for OPV, this material could be coupled with silicon to form high efficiency tandem devices [2], [31].

The architecture of perovskite solar cells was derived from the dye sensitized solar cell (DSSC) technology. The traditional architecture of DSSCs consisted of a porous TiO₂ scaffold, sensitized by a dye and infiltrated by a liquid electrolyte. Both CH₃NH₃PbBr₃ and CH₃NH₃PbI₃ where investigated as an alternative to replace the dye material, with moderate success [27], [32]. One key problem with this technology is the inherent instability of devices containing a liquid. Two key breakthroughs from the Snaith group fundamentally changed the approach to this type of solar cell, leading to significantly increased PCE and an unprecedented amount of research attention. The first of these breakthroughs was to replace the TiO₂ scaffold, used to transport electrons, with an insulator (Al₂O₃) [33]. This demonstrated, for the first time, that the perovskite material could effectively transport electrons without the underlying TiO₂ layer. This insight prompted the next advancement; the demonstration of a planar geometry solar cell with a perovskite thin film as the absorber layer [16]. The high efficiency achieved with this device structure demonstrates the ambipolar charge transport of the perovskite material, in that it can transport both positive and negative charge simultaneously. This is of great importance, as the advances made for polymer solar cells, regarding the low-cost fabrication methods, are now applicable to this material class.

For the goal of cheap and efficient solar modules to be realised by perovskites, manufacturers must be able to provide guarantees regarding the long term stability of the product. Current products on the market typically have a warranty of 20–25 years, which propose that the installation will retain 80% of its initial output after this time period, corresponding to a system loss of <1%/year [34]. As with polymer solar cells, perovskite solar cells have problems with environmental stability. This is caused by the intrinsic instabilities of the perovskite absorber, as well extrinsic factors which degrade the device as a whole. Research conducted over the past couple of years indicates that one key factor is decomposition of the perovskite after exposure to moisture; however, there are many considerations which must be addressed.

Initial understanding of the decomposition of the perovskite absorber has been reviewed by Niu et al. [35]. Perovskite solar cells are composed of many layers, as such; the purpose of this review is to take a holistic view of the mechanisms causing instability of the whole device. We review recent works exploring all of the layers, including the perovskite layer, electron transport layers, hole transport layers, buffer layers, and the influence of varied device geometry, with the aim of highlighting the interdependent nature of instability of the different layers. New insights regarding measurement-induced hysteresis and stability measurement protocols are also discussed. The primary focus of this analysis is on the prototypical perovskite absorber CH₃NH₃PbI₃; however, alternative materials, including lead-free and mixed halide perovskites, are also discussed.