Materials Today
Volume 29, October 2019, Pages 10-19
Journal home page for Materials Today

Short Communication
A chain is as strong as its weakest link – Stability study of MAPbI3 under light and temperature

https://doi.org/10.1016/j.mattod.2018.10.017Get rights and content

Abstract

The stability of perovskite solar cells is a key issue for industrial development. One reason for this is the volatile organic methylammonium (MA) cation, which is prone to degas under elevated temperatures from the perovskite. At the same time, small amounts of MA are used for practically all highest performing solar cells. These compositions have also shown relatively promising stabilities. This raises the question of MA stability with respect to different, application-dependent stability requirements. Interestingly, MA stability was mainly studied on thin films that differ from full devices or with architectures which are also prone to degrade. Therefore, the degradation behavior on complete MA containing devices with a relatively stable architecture is required to quantify the long-term stability of MA. This enables to determine at which timescales MA is unstable and which role it can play in future compositions. If MA is indeed unstable at much longer timescales than previously recorded, it also indicates that more severe degradation pathways are currently underappreciated. Here, “weakest link” MAPbI3 solar cells show stability where devices retained 100% of their initial efficiency over 1000 h of aging under constant illumination and maximum power point tracking at 20 °C. At elevated temperatures of 50 and 65 °C, the devices retained 100% and 90% of their initial efficiency after 500 h of illumination, respectively. Impressively, at 95 °C the MAPbI3 device retained 85% of its initial efficiency after 500 h under constant illumination, which is some of the best stability data reported to date for MA. Thus, MA-containing devices require further studying. Nevertheless to achieve the necessary industrial lifetimes of more than 25 years, the complete removal of MA is a sensible precaution to systematically avoid any long-term risk factors.

Introduction

In 1839, Gustav Rose discovered CaTiO3 and named it “perovskite” after the mineralogist L. A. Perovski [1]. Later, this mineral became the namesake for any ABX3 structure. In 2009, Kojima et al. [2] demonstrated that organic–inorganic perovskites could be used in solar cells. For these solar cells the different sites can be replaced or mixed with A being Cs+, methylammonium (MA) CH3NH3+, or formamidinium (FA) CH3(NH2)2+; B being Pb2+, or Sn2+; and X being Cl, Br, or I [3], [4], [5]. Perovskite solar cells (PSCs) have shown outstanding progress with a power conversion efficiency (PCE) of 3.8% in 2009 to a certified efficiency of 23.3% in 2018 [2], [6]. The first and still one of the most adopted perovskites is MAPbI3 because of its relative simple preparation from only two components, i.e. MAI and PbI2 dissolved in DMSO. This reduces potential error sources from contaminated chemicals or accidentally misadjusted cation ratios. For example, for the high-performance Rb5(Cs5(MA17FA83)95)95Pb(I83Br17)3 perovskite (from here on referred to as RbCsMAFA), requires six components, PbI2, PbBr, FAI, MABr, CsI, and RbI dissolved in DMF and DMSO [7]. Also, MAPbI3 has a beneficially red-shifted bandgap at 1.55 eV compared to 1.62 eV for RbCsMAFA [3]. Moreover, high PCE values of >20% were achieved with MAPbI3 [8], [9].

Unfortunately, even though the PCEs are highly attractive, MAPbI3, as all perovskites, suffer from severe long-term instability especially compared to silicon which is reported to have a mean degradation of 0.8% (relative) per year [10]. Significant efforts were made recently toward stability in PSCs. Looking at the main components for typical perovskite compositions, i.e. Cs, FA, MA, Pb, I, Br, it is especially the organic components MA and FA that are susceptible to light and heat. From earlier studies, FA was reported as thermally more stable than MA [11], [12], [13]. Thus, one way to improve stability under light and heat is to simply remove MA as previously suggested [11], [13], [15], [18], [19]. However, this appears to be challenging, and almost all highest efficient PSCs still use small amounts of MA (∼5% to 17%) as shown in Table 1. For example, MA is present in the double cation mixture (i.e. MA, FA, Pb, Br and I, referred to as MAFA), the triple cation mixture (i.e. Cs5(MA17FA83)95Pb(I83Br17)3, referred to as CsMAFA), and RbCsMAFA with efficiencies at 22.85%, 21.1%, and 21.6%, respectively [7], [20], [21]. Very recently, during the revision of this work, some of the highest efficiency for non-MA compositions, RbCsFAPbI3 and CsFA, were reported at 20.35% and 21.1% (both stabilized), respectively [17], [18]. Thus, at this stage, the presence of MA still correlates with the currently highest power outputs. This was recently connected to the smaller ionic radius of MA which leads to a smaller unit cell compared to pure FA resulting in a stabilized cubic perovskite phase which has better photovoltaic parameters [22]. Therefore, degradation behavior of full MA devices is crucial to understand and to quantify the long-term degradation incurred specifically by MA in mixtures. Only after a full study of MAPbI3, it is possible to determine at which timescales MA is unstable and which role it can play in future compositions.

Several papers have shown that by changing the often used architecture with 2,2′,7,7′-tetrakis(N, N-di-p-methoxyphenylamine)-9,9′-spirobi-fluorene (spiro-OMeTAD) as hole transporting material (HTM) and TiO2 as electron transport material, can improve the stability of PSCs significantly [8], [23], [24], [25]. Especially the usage of polymeric HTMs has proven a very fruitful strategy to prevent detrimental metal migration into the perovskite layer [7], [26], [27]. It is the systematic reduction of these highly detrimental degradation pathways that enabled more stable device architectures. With this, a reevaluation on the long-term stability behavior at temperatures of 50 ± 10 °C and 60 ± 5 °C (as defined by the IEC 61215 protocol) of MA can be conducted [28]. Therefore, to study the degradation process in MAPbI3 devices, we use a mesoporous structure with poly-triarylamine (PTAA) as an HTM, which has been shown to result in stable devices for over 500 h even at high temperatures of 85 °C [7].

We measure MAPbI3 devices at −10, 50, 65, and 95 °C under constant illumination over 500 h. The PSCs retained 87%, 100%, 90%, and 85% of their initial PCE, respectively. Remarkably, at 20 °C, the MAPbI3 device retained its initial efficiency after 1000 h of aging. This aging study provides crucial clues to what extent MAPbI3 and MA-containing devices can be long-term stable even at higher temperatures. Furthermore, we show that maximum power point (MPP) tracking and additional layers on top of the perovskite significantly stabilize PSCs explaining some of the differences compared to previous thin-film studies [14], [16], [29]. Hence, while the removal of MA is a preferred industry choice in terms of long-term risk factors, a more rigorous testing is required that may unveil other (non-MA related), severe degradation pathways that are currently underappreciated.

Section snippets

Results and discussion

We studied the influence of elevated temperatures on MAPbI3 films. In Fig. 1a, we show black films of MAPbI3 and CsMAFA (left side). The X-ray diffraction (XRD) pattern in Fig. 1b shows peaks at 14° and 28°, corresponding to (110) and (220) of the tetragonal perovskite phase. The films are then heated at 150 °C on a hotplate in a nitrogen glove box. After two hours the MAPbI3 films turned yellow (see Fig. 1a right side) in contrast to CsMAFA, which stayed black and only turned yellow after

Discussion

First, the initial part of the stability tests up to 50 h in Fig. 2b is discussed. The PCE of the 65 and 95 °C aging trace increases while the 50 °C trace showed only a small increase as seen in Fig. 2b. It is consistent with Bush et al. and Sheikh et al. who also observed an efficiency increase after heating [12], [33]. The open-circuit voltage (Voc) stayed relatively constant for all temperatures, while the short-circuit current density (Jsc) and FF increased both for 65 and 95 °C as shown in

Conclusions

We studied MAPbI3 devices at −10, 50, and 65 °C under constant illumination over 500 h which retained 87%, 100%, and 90% of their original efficiency. Remarkably the degradation of the device aged at −10 °C was reversible. At 20 °C, the device retained 100% after 1000 h of aging which is one of the highest stability values reported in literature. The device, heated at nearly 100 °C, retained 85% of its initial PCE after 500 h of aging under constant illumination, which is one of the best

Substrate preparation, TiO2 compact layer, and Li-doped mesoporous TiO2

Nippon Sheet Glass 10 Ω/cm2 was etched with a 4% HCl-solution, then cleaned by sonication in 2% Hellmanex water solution for 15 min. After rinsing with deionized water, the step was repeated with ethanol and acetone solution. Then substrates were further cleaned with UV ozone treatment for 15 min. A precursor solution of 0.4 ml of acetylacetone and 0.6 ml of titanium diisopropoxide bis(acetylacetonate) stock solution (75% in 2-propanol) is diluted in 9 mL of ethanol. The precursor solution is

Acknowledgments

P. H. thanks the PROMOS scholarship for their support. M. S. acknowledges support from the co-funded Marie Skłodowska Curie fellowship, H2020 Grant agreement no. 665667. Furthermore, the authors thank S. Collavini for help with the device preparation. The authors thank Dr. A. Uhl for fruitful input.

References (64)

  • J. Kim

    Sol. Energy Mater. Sol. Cells

    (2017)
  • M.K. Klintenberg et al.

    J. Lumin.

    (2003)
  • G. Rose

    J. fur Prakt. Chemie

    (1840)
  • A. Kojima

    J. Am. Chem. Soc.

    (2009)
  • M. Saliba

    Angew. Chemie

    (2017)
  • J.-P. Correa-Baena

    Energy Environ. Sci.

    (2017)
  • J.-P. Correa-Baena

    Science (80-)

    (2017)
  • NREL Chart, Retrieved September 27, 2018, from...
  • M. Saliba

    Science (80-)

    (2016)
  • S.S. Shin

    Science (80-)

    (2017)
  • Q. Wang

    Adv. Mater.

    (2016)
  • D.C. Jordan et al.

    Prog. Photovoltaics Res. Appl.

    (2013)
  • Z. Wang

    Adv. Mater

    (2017)
  • K.A. Bush

    Adv. Mater.

    (2016)
  • T. Leijtens

    J. Mater. Chem. A

    (2017)
  • B. Conings

    Adv. Energy Mater.

    (2015)
  • M.A. Green et al.

    ACS Energy Lett.

    (2017)
  • H. Yuan

    J. Phys. Chem. Lett.

    (2016)
  • Z. Wang et al.

    Nat. Energy

    (2018)
  • S.-H. Turren-Cruz

    Science

    (2018)
  • X. Xu

    J. Mater. Chem. A

    (2018)
  • N.J. Jeon

    Nat. Energy

    (2018)
  • M. Saliba

    Energy Environ. Sci.

    (2016)
  • L.Q. Xie

    J. Am. Chem. Soc.

    (2017)
  • H. Tan

    Science (80-)

    (2017)
  • N. Arora

    Science (80-)

    (2017)
  • H.-S. Kim et al.

    J. Phys. Chem. C

    (2016)
  • T. Matsui

    ChemSusChem

    (2016)
  • K. Domanski

    ACS Nano

    (2016)
  • P. Holzhey et al.

    J. Mater. Chem. A

    (2018)
  • R.K. Misra

    J. Phys. Chem. Lett.

    (2015)
  • Y. Wang

    J. Am. Chem. Soc.

    (2018)
  • Cited by (55)

    • On current technology for light absorber materials used in highly efficient industrial solar cells

      2023, Renewable and Sustainable Energy Reviews
      Citation Excerpt :

      Studies also continue to be performed to distinguish between the intrinsic and extrinsic factors that affect the device stability. Crucial works include the two-fold investigations of performance degradation and their remediation that can be instigated by exposure to 1) visible or ultraviolet light triggering perovskite decomposition, defect formation, and increased non-radiative recombination rates [254–258], dissociation of organic cations [257,259,260], desorption of surface adsorbed oxygen molecules [257], pronounced ion and defect migration [261–264] or phase segregation [265]; 2) elevated ambient temperatures leading to phase transitions [266–270]; 3) contamination by oxygen and moisture inducing the formation of traps or mobile charge barriers [271–273], or field-mediated decomposition [274]; and 4) external biasing resulting in phase segregation, phase transition, or ion or defect migration, or perovskite decomposition [265,275–279]. Taken together, the main research efforts employed in the promotion of the industrialization of organic polymer or perovskite solar cells can be summarized into four categories as follows: (1) the ability to have large-area control of thin film crystallization; (2) robust encapsulation designs; (3) strategies encompassing the identification of failure modes to guarantee the long-term stability and reliability of the devices; and (4) the advancement of environmentally friendly production processes compatible with industrial-scale manufacturing.

    View all citing articles on Scopus
    View full text